1.DfES Recruitment and Retention Report Jul 2005

Recruitment and Retention of AcademicStaff in Higher Education

Hilary Metcalf, Heather Rolfe,Philip Stevens and Martin Weale

National Institute of Economic and Social Research

Research Report RR658

RESEARCH

Research Report

No 658

Recruitment and Retention of Academic Staff in Higher Education

Hilary Metcalf, Heather Rolfe,

Philip Stevens and Martin Weale

National Institute of Economic and Social Research

The views expressed in this report are the authors’ and do not necessarily reflect those of the Department for Education and Skills.

© National Institute of Economic and Social Research 2005

ISBN 1 84478 523 8

Acknowledgements

This study was only possible due to the generous support of a large number of people. We would like to thank them all. We will not name individuals, as most cannot be named for reasons of confidentiality.

In particular, our thanks go to the universities which participated in the study, those individuals who agreed and facilitated access and described their human resourcing. A special thanks goes to those individuals who helped provide us with the quantitative survey samples. In many cases, this was an onerous task and we greatly appreciate the tenacity and dedication shown. We would also like to thank the research students and academics who participated in the qualitative research for their time and trust to respond to highly personal questions. Finally, our thanks go to all the research students and academics who completed the detailed quantitative questionnaire.

CONTENTS DETAILED CONTENTS.....................................................................................................I

TABLES ....................................................................................................................... VII

FIGURES ...................................................................................................................... XII

REPORT SUMMARY ....................................................................................................XIII

1 INTRODUCTION ....................................................................................................1

2 PREVIOUS EVIDENCE ON RECRUITMENT AND RETENTION ISSUES IN ACADEMIA7

3 THE STRUCTURE OF ACADEMIC EMPLOYMENT IN HIGHER EDUCATION .........43

4 INTERSECTORAL AND INTERNATIONAL COMPARISONS OF PAY........................65

5 ENTRANTS TO ACADEMIC JOBS..........................................................................87

6 ACADEMIC EMPLOYMENT AND CAREERS........................................................131

7 LEAVERS...........................................................................................................157

8 THE ROLE OF HUMAN RESOURCE POLICIES AND PRACTICES..........................169

9 CONCLUSIONS ..................................................................................................203

BIBLIOGRAPHY ..........................................................................................................215

APPENDIX A : THE QUALITATIVE RESEARCH...........................................................223

APPENDIX B : THE STAFF AND STUDENT SURVEYS...................................................227

APPENDIX C : HESA STAFF AND STUDENT DATA ....................................................237

APPENDIX D : NATIONAL AND INTERNATIONAL COMPARISONS..............................243

APPENDIX E : INTRA-UK EARNINGS AND UK/US ACADEMIC EARNINGS ...............249

APPENDIX F : THE LIKELIHOOD OF STUDENTS ENTERING ACADEMIA ....................255

APPENDIX G : JOB SATISFACTION AND INTENTIONS TO LEAVE ACADEMIA ............263

APPENDIX H : RECENT AND PROPOSED POLICY DEVELOPMENTS ...........................289

i

Detailed Contents

ACKNOWLEDGEMENTS...................................................................................................I

DETAILED CONTENTS.....................................................................................................I

TABLES ....................................................................................................................... VII

FIGURES ...................................................................................................................... XII

REPORT SUMMARY ....................................................................................................XIII S.1 The study...................................................................................................xiii S.2 Recruitment and retention problems .....................................................xiv S.3 Pay .............................................................................................................xiv S.4 Recruitment to the sector ........................................................................xiv

S.4.1 Pattern of recruitment ............................................................................xiv S.4.2 Reasons for entering academia...............................................................xv S.4.3 Satisfaction of academic staff................................................................xvi

S.5 Retention ..................................................................................................xvii S.5.1 Factors affecting leaving the sector......................................................xvii

S.6 Discrimination and equal opportunities...............................................xviii S.7 Raising the supply to the sector ............................................................xviii S.8 Improving retention .................................................................................xix S.9 Management and human resource practices .........................................xix

S.9.1 Recruitment practice and recruitment difficulties ..................................xx S.9.2 Promotion and retention practices ..........................................................xx

1 INTRODUCTION ....................................................................................................1 1.1 Overview of the study .................................................................................2

1.1.1 HESA staff and student data.....................................................................3 1.1.2 The comparative analysis of pay ..............................................................4 1.1.3 The qualitative and quantitative survey research .....................................4

1.2 Nomenclature...............................................................................................5 1.3 Report layout ...............................................................................................5

2 PREVIOUS EVIDENCE ON RECRUITMENT AND RETENTION ISSUES IN ACADEMIA7 2.1 Introduction .................................................................................................7 2.2 Turnover ......................................................................................................7 2.3 Evidence on recruitment and retention problems..................................11

2.3.1 Current experience of the case study universities: recruitment..............13 2.3.2 Current experience of the case study universities: turnover...................14

2.4 Factors affecting recruitment and retention...........................................16 2.4.1 Introduction ............................................................................................16 2.4.2 Employers’ views on factors affecting recruitment and retention..........16

2.4.2.1 Pay....................................................................................................17

ii

2.4.2.2 Reputation and subject area .............................................................17 2.4.2.3 Redundancies ...................................................................................18 2.4.2.4 Promotion.........................................................................................18

2.4.3 The whole package .................................................................................18 2.4.4 Job satisfaction of academics .................................................................19 2.4.5 Factors affecting recruitment and retention: pay....................................20

2.4.5.1 Comparative pay levels within the UK............................................21 2.4.5.2 The importance of pay .....................................................................22 2.4.5.3 Pay systems......................................................................................23 2.4.5.4 Academic salaries in an international context .................................23 2.4.5.5 Pay discrimination ...........................................................................25

2.4.6 Factors affecting recruitment and retention: pensions............................26 2.4.7 Factors affecting recruitment and retention: job content........................26

2.4.7.1 Teaching...........................................................................................27 2.4.7.2 Research...........................................................................................27 2.4.7.3 Administration and management .....................................................27

2.4.8 Factors affecting recruitment and retention: promotion and progression 27

2.4.8.1 Changes in career path .....................................................................27 2.4.8.2 Internal promotion ...........................................................................28 2.4.8.3 Promotion criteria ............................................................................28 2.4.8.4 The RAE ..........................................................................................29 2.4.8.5 Gender..............................................................................................30 2.4.8.6 Ethnicity...........................................................................................31

2.4.9 Factors affecting recruitment and retention: workload and hours of work 31

2.4.10 Factors affecting recruitment and retention: security of employment....33 2.4.10.1 Staff on temporary contracts........................................................33 2.4.10.2 The pattern of use of temporary contracts ...................................34 2.4.10.3 Findings from the study of contract research staff in Scotland ...35

2.4.11 Factors affecting recruitment and retention: family-friendly practices..37 2.4.12 Factors affecting recruitment and retention: other .................................37

2.5 Summary and conclusions ........................................................................39

3 THE STRUCTURE OF ACADEMIC EMPLOYMENT IN HIGHER EDUCATION .........43 3.1 Introduction ...............................................................................................43 3.2 Institutional structure...............................................................................43 3.3 Job structure..............................................................................................44

3.3.1 Grade structure .......................................................................................44 3.3.2 Pay structure ...........................................................................................47

3.3.2.1 Additional earnings..........................................................................47 3.3.3 Contractual status ...................................................................................49 3.3.4 Primary employment function................................................................51 3.3.5 Mode of employment .............................................................................51

3.4 The characteristics of academic staff in Higher Education ..................53 3.4.1 Gender ....................................................................................................53 3.4.2 Age .........................................................................................................57 3.4.3 Ethnicity .................................................................................................58 3.4.4 Nationality ..............................................................................................58 3.4.5 Staff Qualifications.................................................................................59

iii

3.5 Summary....................................................................................................62 3.5.1 Job structure............................................................................................62 3.5.2 Characteristics of employees..................................................................63

4 INTERSECTORAL AND INTERNATIONAL COMPARISONS OF PAY........................65 4.1 Introduction ...............................................................................................65 4.2 Flows in and out of the UK Higher Education sector ............................65 4.3 A comparison of academic salaries across nine countries .....................68

4.3.1 Data.........................................................................................................68 4.3.2 Results ....................................................................................................70

4.3.2.1 Full-time Staff ..................................................................................72 4.3.2.2 The Distribution of Earnings ...........................................................73

4.4 An analysis of intra-UK earnings and UK/US academic Earnings ......77 4.4.1 Background.............................................................................................77 4.4.2 UK ..........................................................................................................78 4.4.3 US ...........................................................................................................79

4.5 Factors affecting international pay differences ......................................81 4.6 Summary and conclusions ........................................................................84

5 ENTRANTS TO ACADEMIC JOBS..........................................................................87 5.1 Introduction ...............................................................................................87

5.1.1 A note on the data...................................................................................87 5.1.2 Layout of the chapter..............................................................................88

5.2 Source of entrants to the sector................................................................88 5.3 Career routes into academia: existing staff ............................................93 5.4 Existing staff: decision process for entering academia..........................96 5.5 Students as a source of entrants.............................................................103

5.5.1 The supply of PhD students..................................................................106 5.6 Students’ career intentions.....................................................................107

5.6.1 What do research students want from a job?........................................112 5.7 Does academia offer what research students want? ............................115

5.7.1 Students’ perceptions of academia .......................................................115 5.7.1.1 Prior knowledge of academic employment....................................115

5.7.2 Research students’ perceptions of academic jobs ................................117 5.7.3 Do students think academia offers what they want? ............................118 5.7.4 Are students’ expectations met? ...........................................................120

5.8 The factors influencing the likelihood of students entering academia122 5.9 Summary and conclusions ......................................................................124

5.9.1 Source of academic recruits..................................................................124 5.9.2 Career decisions....................................................................................125 5.9.3 Students ................................................................................................126 5.9.4 What do research students want from a job?........................................127 5.9.5 Perceptions of academia .......................................................................127 5.9.6 Policy implications and further research needs ....................................128

6 ACADEMIC EMPLOYMENT AND CAREERS ........................................................131 6.1 Introduction .............................................................................................131 6.2 Career paths ............................................................................................131

iv

6.2.1 Expected progression............................................................................132 6.3 Academics’ satisfaction...........................................................................134

6.3.1 Introduction ..........................................................................................134 6.3.2 Are Female Academics Happier? .........................................................136 6.3.3 What Makes Staff Happy?....................................................................137 6.3.4 The determinants of satisfaction...........................................................148 6.3.5 Is there a single overall measure of satisfaction? .................................150

6.4 Summary..................................................................................................152 6.4.1 Progression within academia................................................................152 6.4.2 Overall satisfaction...............................................................................152 6.4.3 Satisfaction with individual aspects of the job .....................................152 6.4.4 Factors affecting satisfaction................................................................154

6.4.4.1 Personal characteristics ..................................................................154 6.4.4.2 Experience......................................................................................154 6.4.4.3 Grade..............................................................................................154 6.4.4.4 Contractual status...........................................................................154 6.4.4.5 Differences between universities ...................................................154 6.4.4.6 Hours..............................................................................................155 6.4.4.7 Research ranking............................................................................155

7 LEAVERS...........................................................................................................157 7.1 Introduction .............................................................................................157 7.2 Career plans.............................................................................................158

7.2.1 Career plans: changing universities......................................................158 7.2.2 Career plans: leaving the UK Higher Education sector........................160

7.2.2.1 Remaining in academia to retirement ............................................160 7.2.2.2 Expectations of leaving the sector in the following year ...............160

7.3 Factors affecting the likelihood of leaving UK Higher Education......164 7.3.1 Introduction ..........................................................................................164 7.3.2 Results ..................................................................................................165

7.4 Summary and conclusions ......................................................................166 7.4.1 Career expectations ..............................................................................166 7.4.2 Factors affecting leaving the sector......................................................167 7.4.3 Implications for policy .........................................................................168

8 THE ROLE OF HUMAN RESOURCE POLICIES AND PRACTICES..........................169 8.1 Introduction .............................................................................................169

8.1.1 A note on research staff ........................................................................169 8.2 Employment strategy ..............................................................................170 8.3 Recruitment practice ..............................................................................171

8.3.1 Recruitment strategy.............................................................................171 8.3.2 The recruitment process .......................................................................172 8.3.3 The pay offer ........................................................................................174 8.3.4 Other terms and conditions...................................................................176 8.3.5 Response to recruitment difficulties.....................................................176

8.4 Promotion practice..................................................................................177 8.4.1 Promotion systems................................................................................178 8.4.2 Procedures and criteria .........................................................................179 8.4.3 Response to retention difficulties/changes: promotion systems...........181

v

8.5 Pay system................................................................................................182 8.5.1 Pay level ...............................................................................................182 8.5.2 Retention incentives .............................................................................182

8.6 Appraisal, training and development....................................................184 8.6.1 Recruits.................................................................................................184 8.6.2 Appraisal...............................................................................................185 8.6.3 Training and development....................................................................186

8.7 Work demands and work allocation .....................................................187 8.8 Equal opportunities.................................................................................188

8.8.1 Gender ..................................................................................................190 8.8.2 Ethnicity ...............................................................................................194 8.8.3 Age .......................................................................................................195 8.8.4 Other .....................................................................................................196 8.8.5 Equality initiatives................................................................................196

8.9 Summary and conclusions ......................................................................197 8.9.1 Ethos and human resource structure.....................................................197 8.9.2 Employment strategy............................................................................198 8.9.3 Promotion practice................................................................................199 8.9.4 Pay system............................................................................................200 8.9.5 Work demands and work allocation .....................................................201 8.9.6 Equal opportunities...............................................................................201

9 CONCLUSIONS ..................................................................................................203 9.1 The labour pool .......................................................................................204 9.2 Pay ............................................................................................................204 9.3 Job content ...............................................................................................205 9.4 Workload .................................................................................................206 9.5 Other aspects of the job ..........................................................................207 9.6 Job security ..............................................................................................207 9.7 Career prospects......................................................................................208 9.8 Careers advice .........................................................................................209 9.9 Discrimination .........................................................................................210 9.10 Other issues..............................................................................................211 9.11 Management and human resource practices ........................................211 9.12 Further research......................................................................................211

BIBLIOGRAPHY ..........................................................................................................215

APPENDIX A : THE QUALITATIVE RESEARCH...........................................................223 A.1 The scope of the qualitative research ....................................................223 A.2 The sample ...............................................................................................223

APPENDIX B : THE STAFF AND STUDENT SURVEYS...................................................227 B.1 Aim ...........................................................................................................227 B.2 The university sample .............................................................................227

B.2.1 University sample population...............................................................227 B.2.2 University sample structuring...............................................................229

B.3 The staff sample.......................................................................................231

vi

B.4 The student sample .................................................................................232 B.5 Fieldwork .................................................................................................232 B.6 Response rates .........................................................................................233 B.7 Weighting .................................................................................................233 B.8 Bias ...........................................................................................................233

B.8.1 Staff ......................................................................................................233 B.8.2 Research students .................................................................................234

APPENDIX C : HESA STAFF AND STUDENT DATA ....................................................237 C.1 HESA staff data set. ................................................................................237 C.2 HESA Student Data ................................................................................240

APPENDIX D : NATIONAL AND INTERNATIONAL COMPARISONS..............................243 D.1 Data...........................................................................................................243

D.1.1 UK ........................................................................................................243 D.1.2 US.........................................................................................................244 D.1.3 Australia ...............................................................................................244 D.1.4 New Zealand.........................................................................................245 D.1.5 Canada ..................................................................................................245 D.1.6 Denmark ...............................................................................................245 D.1.7 France ...................................................................................................245 D.1.8 Sweden .................................................................................................246 D.1.9 Japan .....................................................................................................246 D.1.10 Additional tables...................................................................................246

APPENDIX E : INTRA-UK EARNINGS AND UK/US ACADEMIC EARNINGS ...............249 E.1 Background..............................................................................................249 E.2 Empirical Model......................................................................................249 E.3 Results ......................................................................................................251

E.3.1 UK ........................................................................................................251 E.3.2 US.........................................................................................................253

APPENDIX F : THE LIKELIHOOD OF STUDENTS ENTERING ACADEMIA ....................255 F.1 Model ........................................................................................................256 F.2 Results ......................................................................................................258

APPENDIX G : JOB SATISFACTION AND INTENTIONS TO LEAVE ACADEMIA ............263 G.1 The General Model .................................................................................264 G.2 The Job Satisfaction of Academics ........................................................265

G.2.1.1 The satisfaction of temporary staff ................................................266 G.2.1.2 Satisfaction in a dynamic framework ............................................267

G.2.2 Analysis ................................................................................................267 G.2.3 Is there a single overall measure of satisfaction? .................................278

G.3 Factors affecting the likelihood of leaving Higher Education.............280 G.3.1 Introduction ..........................................................................................280 G.3.2 Results ..................................................................................................281

APPENDIX H : RECENT AND PROPOSED POLICY DEVELOPMENTS ...........................289

vii

Tables

Table 2.1 Turnover in HE staff, 2001/2 (%)..................................................................8

Table 2.2 Turnover by broad subject (%) ......................................................................9

Table 2.3 Turnover by higher education institution type (%)......................................10

Table 2.4 Turnover by type of contract (%) ................................................................11

Table 2.5 Turnover as a percentage of total turnover by type of contract ...................11

Table 2.6 Comparison of job satisfaction of academics with whole economy............20

Table 2.7 Contract researchers: value of job attributes................................................36

Table 3.1 Employment by type of institution ..............................................................44

Table 3.2 Salary scales from 1 August 2004 ...............................................................45

Table 3.3 Staff grades (%) ...........................................................................................46

Table 3.4 Staff grades lecturing and senior researchers (%) .......................................47

Table 3.5 Academic Staff: Gross additional earnings in the previous 12 months (%) 48

Table 3.6 Academic staff: Gross additional earnings in the last year, by grade (%)...48

Table 3.7 Academic staff: Gross additional earnings by subject.................................49

Table 3.8 Employment activity by contractual terms (%) ...........................................50

Table 3.9 Breakdown of Staff by Primary Employment Function (%) .......................51

Table 3.10 Mode of Employment (%) .........................................................................52

Table 3.11 Mode of Employment by Terms of Employment (%)...............................52

Table 3.12 Mode of Employment by Employment Function (%) ...............................52

Table 3.13 Academic staff by gender ..........................................................................53

Table 3.14 Mode of Employment by Gender (%) .......................................................54

Table 3.15 Mode of Employment by Age and Gender (%).........................................54

Table 3.16 Breakdown of Staff by Subject and Gender ..............................................55

Table 3.17 Staff grades (%) .........................................................................................56

Table 3.18 Breakdown of Staff by Primary Employment Function (%) .....................56

Table 3.19 Breakdown of Staff by Terms of Employment (%)...................................56

Table 3.20 The Age of HE Staff (%) ...........................................................................57

Table 3.21 The Age Structure of HE Staff (%) ...........................................................58

Table 3.22 Breakdown of Staff by Nationality (%).....................................................59

Table 3.23 The Age of HE Staff by nationality (%) ....................................................59

viii

Table 3.24 Highest qualification of staff (%) ..............................................................59

Table 3.25 Qualifications by Terms and Mode of Employment (%) ..........................62

Table 4.1 Flows into and out of UK Higher Education sector, by nationality ............66

Table 4.2 Destination of academics within Higher Education sector, by nationality..67

Table 4.3 New academics from within Higher Education sector, by nationality ........67

Table 4.4 Academic salaries, in £ ................................................................................71

Table 4.5 Proportion of part-time academic staff ........................................................72

Table 4.6 Real earnings of full-time workers only ......................................................72

Table 4.7 Comparative actual academic earnings by age ............................................80

Table 4.8 Earnings of academics relative to average earnings ....................................82

Table 5.1 Origin of entrants to the HEI sector, 2001/02..............................................89

Table 5.2 Total staff movements by age group (%).....................................................90

Table 5.3 The Contracts of Entrants to HE by Age .....................................................91

Table 5.4 Academic staff: First academic job: type of job ..........................................92

Table 5.5 Academic staff: Contractual basis of first academic job .............................92

Table 5.6 Academic staff: Employment outside academia..........................................93

Table 5.7 Academic staff: Nature of employment outside academia ..........................93

Table 5.8 Academic staff: Entrants making career change by subject ........................94

Table 5.9 Academic staff: Whether made a career change to go into academia .........95

Table 5.10 Academic staff: Entrants making career change by type of university and by gender..............................................................................................................95

Table 5.11 Academic staff: Whether was seriously considering and comparing careers before got first academic job................................................................................97

Table 5.12 Academic staff: Career intentions prior to first academic job ...................98

Table 5.13 Academic staff: Career intentions prior to first academic job by type of university and by gender......................................................................................98

Table 5.14 Academic staff: Career intentions prior to first academic job by subject..99

Table 5.15 Academic staff: Reasons for entering academia (%)...............................100

Table 5.16 Academic staff: Reasons for entering academia by type of university and by gender............................................................................................................100

Table 5.17 Academic staff: Factors prompting career change to academia ..............102

Table 5.18 Academic staff: Factors prompting career change to academia by type of university............................................................................................................102

Table 5.19 First Destination, by degree (%) ..............................................................104

Table 5.20 Nature of study for those going into further study (%)............................104

Table 5.21 Breakdown of students entering employment in the Higher Education sector (by SIC code) (%)....................................................................................105

ix

Table 5.22 Occupation of graduates entering HE......................................................105

Table 5.23 First Destination, by degree and subject (%)...........................................106

Table 5.24 Research students: Importance of reasons for study................................108

Table 5.25 Research students: What best describes your career intentions? .............109

Table 5.26 Research students: Career intentions by degree subject (excerpt) (%)....110

Table 5.27 Research students: Desired and probable jobs.........................................111

Table 5.28 Research students: Desired jobs, by subject ............................................112

Table 5.29 Research students: Importance of factors in career choice, cumulative %............................................................................................................................113

Table 5.30 Research students: The importance of elements of the academic job (%)............................................................................................................................114

Table 5.31 Research students: The importance of a permanent job, cumulative % ..114

Table 5.32 Research students: Parental experience of teaching/research..................116

Table 5.33 Research students: HE Employment prior to current degree...................116

Table 5.34 Research students: HE Employment during current degree ....................117

Table 5.35 Research students: HE Employment whilst student ................................117

Table 5.36 Research students: How well does an academic career offer (%) ...........118

Table 5.37 Research students: Perceptions of academic employment ......................118

Table 5.38 Research students: Comparison of research students’ desired job attributes and expected attributes of an academic job .......................................................119

Table 5.39 Research students: Contract type.............................................................120

Table 5.40 Academic Staff: Hours of work in academic employment......................121

Table 5.41 Research students: Expected earnings in academia.................................122

Table 5.42 Research students: Actual salary .............................................................122

Table 6.1 Academic staff: Academic career expectations .........................................133

Table 6.2 Reported satisfaction in previous studies...................................................135

Table 6.3 Academic staff: Reported satisfaction (%) ................................................136

Table 6.4 Academic staff: Satisfaction by Gender ....................................................137

Table 6.5 Academic staff: The aspects of teaching and job satisfaction ...................138

Table 6.6 Academic staff: The aspects of research and job satisfaction ...................139

Table 6.7 Academic staff: The effect of administrative and organisational aspects of academic employment and job satisfaction .......................................................140

Table 6.8 Academic staff: Other aspects of academia and job satisfaction...............141

Table 6.9 Academic staff: Perceived fairness of decisions on individual pay, current university............................................................................................................142

Table 6.10 Academic staff: Perceived fairness of decisions on promotion and recruitment to senior posts, current university...................................................143

x

Table 6.11 Academic staff: Whether thinks academic jobs have improved since their first academic job in UK....................................................................................147

Table 6.12 Factor loadings for overall job satisfaction..............................................151

Table 7.1 Leavers: HEFCE-funded staff ...................................................................157

Table 7.2 Academic staff: Likelihood of moving to another UK university in the next year.....................................................................................................................159

Table 7.3 Academic staff: Reasons for moving to another UK university in the next year and for not moving.....................................................................................159

Table 7.4 Whether expects to stay in UK academia until retirement ........................160

Table 7.5 Academic staff: Likelihood of leaving employment in UK Higher Education in the next year...................................................................................................161

Table 7.6 Sector leavers: plans ..................................................................................162

Table 7.7 Academic staff: Sector leavers: reasons for leaving..................................163

Table 7.8 Academic staff: Whether has worked continuously in academia by subject............................................................................................................................164

Table 7.9 Academic staff: Total time spent working outside academia ....................164

Table 8.1 Equal opportunities, staff beliefs ...............................................................190

Table 8.2 Childcare provision, staff with children ....................................................193

Table 8.3 Adult care provision...................................................................................194

Table 8.4 Early retirement .........................................................................................195

Table A.1 Case studies: subject areas ........................................................................225

Table A.2 Specialist institutions excluded from the sample......................................228

Table A.3 Other institutions with fewer than 200 staff excluded from the sample ...229

Table A.4 University sample structuring ...................................................................229

Table A.5 Colleges of Higher Education...................................................................230

Table A.6 Universities in London..............................................................................230

Table A.7 Academic staff on clinical rates ................................................................231

Table A.8 Staff sample ..............................................................................................232

Table A.9 Student sample ..........................................................................................232

Table A.10 Weights ...................................................................................................233

Table A.11 Staff: comparison of characteristics: survey and HESA data .................234

Table A.12 Research students: comparison of characteristics: survey and HESA data............................................................................................................................235

Table A.13 Comparison of mean wages in LFS and HESA data ..............................244

Table A.14 Academic real earnings quartiles ............................................................247

Table A.15 Academic real earnings deciles...............................................................248

Table A.16 Variables used in empirical analysis.......................................................251

xi

Table A.17 UK Results ..............................................................................................252

Table A.18 US Results...............................................................................................254

Table A.19 Variables used in empirical analysis of academic career intentions.......257

Table A.20 Ordered probit results: Academic career intentions................................260

Table A.21 Variable used in the analysis of job satisfaction.....................................268

Table A.22 Results – satisfaction...............................................................................271

Table A.23 Including alternative salary.....................................................................276

Table A.24 Factor loadings for overall job satisfaction.............................................279

Table A.25 Variables used in analysis of likelihood of leaving ................................280

Table A.26 Results – Likelihood of leaving UK higher education............................284

xii

Figures

Figure 3.1 Age Profile of Qualifications (% of staff) ..................................................60

Figure 3.2 Age Profile of Staff with Doctorates (% of staff).......................................61

Figure 4.1 Real earnings quartiles of full-time academics ..........................................74

Figure 4.2 Real earnings quartiles of full-time academics, men..................................74

Figure 4.3 Real earnings quartiles of full-time academics, women.............................74

Figure 4.4 Real earnings deciles of full-time academics, total ....................................75

Figure 4.5 Real earnings deciles of full-time academics, Men....................................75

Figure 4.6 Real earnings deciles of full-time academics, women ...............................76

Figure 4.7 Net earnings deciles of academics for the UK and France.........................76

Figure 4.8 Predicted lifetime wage profiles, UK .........................................................79

Figure 4.9 Predicted lifetime wage profile, US ...........................................................80

Figure 4.10 Comparative predicted annual wage profiles ...........................................81

Figure 4.11 Predicted lifetime annual wage profile, US..............................................83

Figure 5.1 UK and overseas domiciled students obtaining PhDs..............................107

xiii

Report summary

S.1 The study The study seeks to identify the factors which lead to individuals entering and

leaving academic employment in the English Higher Education sector. Although the main focus was entry and exit from the sector, recruitment to and retention by individual institutions can shed light on this and was also investigated.

Academic employment was defined as jobs in higher education institutions (Universities and Colleges of Higher Education) whose main function was academic teaching or academic research, irrespective of the contractual terms of the job holder. Thus lecturing and research staff are included, but academic-related staff (e.g. technicians) are not. Full-time, part-time, permanent and temporary staff within these groups are included. However, the coverage of hourly paid staff is severely limited.

(The study excluded staff in Further Education Institutions and those on clinical grades.)

The study had five, inter-related, strands:

• a literature review conducted March to May 2003; the review covered the relevant academic and non-academic literatures in education, economics, management and human resources; it concentrated mainly on the last twenty years although it did include earlier work where relevant;

• analysis of HESA staff and student data, 2001/02, to provide a descriptive analysis of turnover in academia, to identify the basic characteristics of employment in the sector and to identify the student supply into academia;

• comparative analysis of pay both nationally and internationally, using 2001 data;

• qualitative research within universities exploring human resource policies and practices and factors affecting entry and exit from the sector; the fieldwork, involving face-to-face and telephone interviews, was conducted between July 2003 and July 2004;

• quantitative surveys of academic staff and of research students to identify factors which affect recruitment into academia and retention; the fieldwork was conducted between May and July 2004.

Thus evidence used in the research relates to a range of periods and some Higher Education policies and practices may have changed subsequently. To assist the reader, the Department for Education and Skills (DfES) has produced a list of relevant Higher Education initiatives. These are listed in Appendix H.

xiv

S.2 Recruitment and retention problems (Section 2.2, Section 2.3) The study did not identify severe recruitment and retention problems in the

sector. However, some problems were apparent: vacancies sometimes remained unfilled and there was some reported decline in the quality of applicants. The extent of difficulty varied by subject and fluctuated over time.

S.3 Pay (Chapter 4) Pay is one of the important factors in career and job choice, affecting both

recruitment and retention. The study compared the pay of UK academics with highly qualified people in the rest of the economy and also compared the pay of UK academics with academics in the USA, Canada, Australia, New Zealand, Denmark, Sweden, France and Japan. Differences in data definitions, differences in the nature of academic jobs and differences in cost of living between countries prevent exact cross country comparisons. However, we are confident that our findings identify the broad scale of differences.

Academic pay is low relative to that in other highly qualified jobs in the UK, which is likely to reduce entry to the sector (Section 4.4.2). (Retention is likely to be less responsive to pay differentials as careers progress, due, in many subjects, to a divergence in the skills developed in academia and those needed for senior jobs in other sectors.)

UK academic pay1 compares favourably with academic pay in Sweden, Japan, Australia and New Zealand (Section 4.3.2). UK academic pay is similar to that of Denmark, France and Canada. Pay in the US is higher for comparable academic staff and the difference is particularly marked at the top-end of the earnings distribution (Section 4.3.2; Section 4.4.3). Thus pay is likely to be a factor encouraging outflow of academics from the UK to the US, but also a factor easing recruitment from Sweden, Japan, Australia and New Zealand.

Whilst lower paying countries may prove a fruitful source of recruits, foreign recruitment should be viewed with some caution. Recruits from other EU (and EEA) countries, Australia, New Zealand and the US were identified as more likely to expect to leave UK academia and so reliance on foreign recruitment may lead to future retention problems (Section 7.3.2).

S.4 Recruitment to the sector

S.4.1 Pattern of recruitment The main sources of entrants to academia are students and employees in other

sectors (Section 5.2):

• employees in other sectors (UK: 42 per cent; abroad: 21 per cent, including 11 per cent in academia abroad)

1 This section refers to real (as opposed to nominal) pay, i.e. once cost of living differences have been taken into consideration.

xv

• students (34 per cent: 19 per cent UK domiciled students studying in the UK, 10 per cent foreign students studying in the UK; and five per cent students studying abroad)

Many make a career change to enter academia, normally from higher level occupations (managerial, professional, associate professional and technical), rather than from lower level jobs (via being a mature student) (Section 5.3). Career changers were particularly common in new universities and in Business and administrative studies, Computing, Subjects Allied to Medicine, Education, Social studies and Engineering.

Forty per cent of recruits are foreign (i.e. non-UK nationals). Amongst research students, non-British EU nationals were more likely to want to go into UK academia than other nationals (Section 5.2; Section 5.8).

Entrants to academia tend to be recruited to fixed-term contracts in research, although the percentage entering to permanent contracts rises with age, suggesting that previous skills are relevant and recognised (Section 5.2). It appeared that the tendency to recruit to research jobs and to fixed-term contracts had grown over time.

S.4.2 Reasons for entering academia Almost 40 per cent of research students were keen to have an academic career

and a further 21 per cent saw this of equal interest to some other career (Section 5.6). Amongst existing academics just over one half had been keen to have an academic career. However, many research students and academics did not seriously consider alternative careers and, for those progressing from a research degree, entering academia appears often to be a form of drift, to a job which is known. A significant minority of research students saw academia as providing a stepping stone to another career (12 per cent), which is likely to reduce retention.

The study examined what research students wanted from a career (Section 5.6.1) and which attributes they thought academia offered (Section 5.7). The main attraction of academic jobs to research students is doing research and setting one’s own research agenda. A career offering research was very important to 63 per cent of research students. Only 30 per cent were strongly attracted by teaching. Only one-third of research students saw a high salary as very important in their career choice. Non-pecuniary aspects were more often important: most often a good working environment, variety, freedom to use initiative and seeing tangible outcomes from their job. Close behind these factors are autonomy in the job, control of their research, career prospects, collaboration and flexibility of working hours. Also important were good physical work conditions, helping people and job security.

Broadly speaking, research students believed that an academic career offered the attributes they wanted from a career (Section 5.7). However, there are three important exceptions to this: career prospects, job security and high salary, none of which tended to be seen as good in academia. Slow career progression, lack of job security (particularly at the start of an academic career) and relatively low pay are likely to reduce the supply from this source. Research students appeared to be well-informed about the nature of academic employment. However, they tended to underestimate academic pay and not to be aware of the additional financial benefits of university pensions. This is likely to reduce entry to academia (but also reduce

xvi

turnover) and so more information on pay may be useful in promoting academic careers among research students.

For career changers, other motivations included a better lifestyle (36 per cent) or the desire for change (34 per cent) (Section 5.4). Few were driven through lack of further progression in their previous career outside of Higher Education (or being made redundant or retiring). Some of the factors prompting change to a career in Higher Education included stress in previous job, shift work, routine working and wanting variety at work, academia being ‘more relaxed and informal’, career progression requiring movement into management (and wanting to stay with one’s profession). For art and design lecturers one factor appears to be to fund artistic practice.

S.4.3 Satisfaction of academic staff (Section 6.3) Academic staff are somewhat less satisfied with their jobs than those in the

workforce as a whole (Section 6.3.1). Academics appear to be considering three separate sets of elements of their jobs, namely the pecuniary factors (both the salary and the ability to earn money from additional work), non-pecuniary factors relating to the qualitative dimensions of the job and longer-term factors such as promotions and job security. The factors affecting satisfaction are discussed below (Section 6.3.3; Section 6.3.4).

• Research is a major source of satisfaction for academics and many academic staff would prefer to spend more of their time on research, although self-determined research tends to be of interest, rather than that determined by others. The demand for research output and the RAE in particular are seen rather negatively. Hours spent on research increase staff satisfaction with the actual work itself.

• Whilst teaching is not the most important reason for becoming an academic, most would prefer a job that involves teaching. Teaching bright students and seeing their students develop are the positive aspects of teaching. The negative aspects relate to assessment, both of the amount required of the students and that of the staff themselves.

• Administrative tasks and organisational change tend to be viewed as negative aspects of the job by most academics. Hours of work spent on administration have a negative effect on satisfaction with almost all dimensions of academics’ job satisfaction.

• Academic staff tend to value more subtle elements of their jobs, such as the support of their peers and the ability to participate in the wider academic community.

• Being on a fixed-term contract significantly reduces satisfaction.

There was no difference between women and men in satisfaction (once one takes into account differences in other characteristics), except that women were more satisfied with salary (Section 6.3.4). This may be due to women having lower expectations of salary, due to discrimination in the labour market as a whole.

Academics from ethnic minorities tended to be less satisfied with the opportunity they have to use their own initiative, the hours they work and their relations with their colleagues than their white colleagues (Section 6.3.4). This is

xvii

likely to indicate that academics from ethnic minorities find themselves in less satisfactory jobs.

S.5 Retention About two-thirds of academics expected to remain in UK academia until

retirement (Section 7.2.2). This was higher in new universities. We would estimate that between three and six per cent would leave the sector in the year following interview. This comprised two to three per cent of those on permanent contracts and five to 11 per cent of those on fixed-term contracts.

Forty-three per cent of those who thought they might leave the sector in the following year expected to move to another job; this was most commonly to a UK job outside research and teaching, to an academic job abroad or for fixed-term contracts staff, to a UK research job. Contract staff could be seen as being driven out of the sector due to insecurity: they tended to leave due to their contract ending, the desire for a permanent job or pessimism about job opportunities in UK academia.

S.5.1 Factors affecting leaving the sector (Section 7.3) The following factors increased the likelihood of leaving the sector:

• dissatisfaction with non-pecuniary elements (the work itself, relations with manager, being able to use one’s own initiative, hours, relations with colleagues and physical work conditions)

• being a non-British EU (and EEA) national, Australian, New Zealander or US national

• having had a break in one’s academic career

• being on a non-permanent contract

• being closer to the end of a fixed-term contract

• hours worked

• hours spent on administrative tasks

• the fewer hours spent on research

• perception of excessive workload

• belief that decisions on either individual pay, recruitment to senior posts or promotion at their current university are not at all fair

• dissatisfaction with pay and the level of pay (but not estimated relative pay).

A number of aspects of academic employment that staff feel are important for their satisfaction did not affect their likelihood of leaving. Those who said that the RAE, QAA requirements and the general direction of higher education policy lowers their satisfaction by a lot are no more likely to expect to leave UK Higher Education than those who do not. The likelihood of leaving was no different between those who had changed career to enter academia and those who had not. Nor did it differ by gender, ethnicity, or having children. Part-timers were no more likely to expect to leave the sector than full-timers.

xviii

S.6 Discrimination and equal opportunities Discrimination, if worse relative to other employment, may exacerbate

recruitment and retention difficulties amongst the discriminated against groups. The sparse, previous, evidence found differences in pay between ethnic minorities and whites and in promotion between women and men (Section 2.4.5.5). This study found that, compared with whites, academics from ethnic minorities were less satisfied with a number of aspects of their job (Section 6.3.4 and summarised in Section S4.3 above), which may suggest that ethnic minority academics find themselves in less satisfactory jobs. Whilst a number of factors may be advanced for these differences, discrimination (whether direct or indirect) cannot be ruled out and further research is required in this area.

There was a commitment to equal opportunities in the case study universities at the level of policy, and many heads of department believed that men and women had the same opportunities for advancement, or at least that these were improving (Section 8.8). However, many of the staff believed that disability, age, religion and ethnicity as well as gender affected progression within Higher Education. Indeed, the percentage of staff who believed that these factors affected progression were 64 per cent (age), 44 per cent (gender), 38 per cent (disability) 26 per cent (ethnicity) and 10 per cent (religious affiliation). The qualitative research identified a number of possible reasons for women being disadvantaged, including unfair work allocation, sexism and a disproportionate impact on career progression of part-time working (Section 8.8.1). It was apparent, both from the survey and qualitative research that many staff had not considered the position and opportunities for other under-represented groups, which suggests that the diversity debate does not have a high profile amongst academics in UK universities.

S.7 Raising the supply to the sector (Chapter 9) The key messages from the research about how supply to the sector could be

increased are:

• increasing the supply of UK-domiciled students achieving a PhD (Section 5.2; Section 5.5.1);

• shifting the balance of academic job content towards research (Section 5.4; Section 5.6; Section 5.7);

• increasing pay (Section 5.6; Section 5.7); • targeting foreign recruitment (with the caveat that this may increase retention

problems) (Section 4.6; Section 5.2; Section 5.8; Section 7.3.2); • increasing the job security of researchers (Section 5.6; Section 5.7); • improving career progression for both lecturing and for research-only staff

(Section 5.6; Section 5.7).

These findings suggest there are a number of ways in which individual universities and the Government could raise retention of academics in the sector. Many, but not all of these would have financial implications.

xix

S.8 Improving retention (Section 7.3; Chapter 9) Throughout the study the message that academics were driven by the desire to

do research and neither enjoyed administrative tasks nor always could see benefits of these was repeated. Changing the relative time spent on research and administration (without increasing total hours) and reducing administrative demands would be likely to increase retention, as would reducing the total time spent working. Approaches to this might include increasing the staff/student ratio, reducing structural change and reducing change in administrative demands and reducing the demands of quality reviews and inspections.

Turnover is higher among academics who enter UK Higher Education as PhD students from overseas. Encouraging British students to undertake research degrees and so expanding the supply of UK-domiciled people with PhDs entering the sector is likely to increase retention. This might be achieved through more funding for PhD students or addressing the debt built up during earlier degrees.

Turnover is high amongst researchers. This is largely due to the use of temporary contracts for most and the lack of career opportunities (in research). Addressing these two issues is likely to have a major effect on the loss of researchers from the sector. It may require imaginative approaches to deployment of both researchers and lecturers.

Ensuring pay and promotion decisions are fair and seen to be fair would reduce loss of staff from the sector. Certainly, enhancing pay of recruits (above comparable levels for existing staff) and of valued staff expected to leave are regarded as ‘unfair’ and excite strong feelings (Section 6.3.3; Section 8.9.2). Market supplements and Golden Handshakes are also seen as unfair. Performance pay must be implemented well to avoid perceptions of unfairness. However, even so, the problem of the average employee tending to believe their performance is above average and so deserving a higher than average pay rise, means that more employees are likely to be dissatisfied by performance pay than satisfied.

Perceptions of the fairness of promotion decisions are likely to be influenced by the extent of opportunities for promotion: the more opportunity, the less important fairness becomes (and the more likely each individual will have experienced promotion). Thus expanding the opportunities for promotion is likely to increase retention, irrespective of other changes. It is also possible that the emphasis on promotion is, in part, a result of relatively low pay levels. If so, increasing the pay of academics is likely to increase retention also.

S.9 Management and human resource practices (Chapter 8) Management and human resource practices differ across universities and can

affect satisfaction. Most of the case study universities had mixed practices (good and bad) and seemed to be trying to address some of the problem areas.

There appears to be a lack of communication and understanding between the academics and the university administration (including the human resource specialists), which probably increases the burden of administrative demands on academics (Section 8.9.1).

xx

Implementation of human resource practices tends to be devolved. In old universities the post of head of department tends to be rotated every few years. Academics’ management and administrative responsibilities tend to be accorded little status. All these factors lead to lack of expertise and some poor practice.

S.9.1 Recruitment practice and recruitment difficulties (Section 8.3) Replacement of staff tended to be slow (sometimes taking more than a year),

due to the practice of reviewing the need for each vacant post (Section 8.3.1). This led to additional work demands on remaining staff.

It was common to down-grade vacated posts (Section 8.3.1).

Due to difficulties attracting good candidates, especially at senior levels and to some specialisms, active search methods had become common (using search committees and networks) (Section 8.3.2). Case study universities tended to overlook their own PhD students for recruitment.

In the research-intensive old case study universities, due to the RAE, there was an increased emphasis on recruiting ‘stars’ and poaching. This was expected to increase in the run up to the next RAE.

The case study universities were practising flexibility over the pay offered to new recruits, but were uncomfortable with the idea of market supplements because of the discrepancies they can create (Section 8.3.3). Pay was seen as only one means to attract new recruits (Section 8.3.4). Other incentives included facilities, equipment and start-up funding. There were indications that these did not always meet recruits’ expectations, and it is possible that this could result in early turnover.

S.9.2 Promotion and retention practices The criteria and transparency of promotion practices was found to vary

between the case study universities (Section 8.4). Universities were concerned that the criteria used should take account of the range of academic activity, including research, teaching, administration and enterprise. However, research continued to be the pre-eminent criterion, and in new universities was of greater importance than in the past. The allocation of teaching and administrative responsibilities within university schools or departments was therefore becoming more important than in the past, because individuals with heavy teaching or administration loads were less able to carry out research and therefore achieve promotion. This was found to lead to considerable resentment. In some cases, to retain a member of staff, the formal process was bypassed, leading to strong dissatisfaction amongst other staff.

Below professorial level, pay scales are set nationally (Section 8.5). However, case study universities were increasingly using pay incentives to keep valued staff, particularly those in hard to recruit subjects or senior research-productive staff. In some cases, this was formalised. In others the formal system was bypassed to award additional increments, for example, when a member of staff received another job offer. This was not normally seen (or found) to be effective and was thought to affect the morale of other staff.

Although the case study universities had appraisal systems, these were sometimes of recent origin and were not linked directly to the promotions process (Section 8.6). Neither were they linked to training and development. Many staff were

xxi

not aware of the procedures and criteria for promotion. There was evidence of bypassing the formal promotions process. Such practices are likely to be perceived as unfair and to lead to dissatisfaction. In response to retention difficulties, case study universities were looking at progression and promotion, to make criteria more transparent and to ease the process of progression. Their main concern was to identify and reward strong performers and to encourage them to stay. Transparent systems to enable such progression are likely to be more acceptable to the academic workforce than practising covert deals with individuals.

Although training is usually regarded as key to workforce retention, it was acknowledged that heads of department do not always have the skills needed to develop staff. The prevailing view was that academic staff are responsible for their own training and development (Section 8.6). However, there was evidence of an increase in attention to such issues, including management training for heads of department to assist them in identifying staff development needs (Section 8.4.3).

xxii

1

1 Introduction

Demand for academic staff in Higher Education has been increasing2 and may be expected to continue to increase given the Government’s intention that participation in Higher Education should increase substantially amongst those aged 18 to 30 years old.

At the same time, recruitment and retention problems have been growing in prominence (HEFCE, 2003) and there has been a long-standing concern that the sector faces a ‘retirement bulge’, as academics from the 1960s expansion reach retirement. Consequently, there is concern about the adequacy of the future supply of academics.

Other substantial changes in Higher Education in the past 10 to 20 years are likely to have contributed to the tightness of the academic labour market3. Polytechnics were granted university status in 1992, changing their funding regime, their focus and the demands on staff. The number of students has grown substantially, a growth which has not been matched by staff increases resulting in a large increase in the student:staff ratio. Changes in funding have led to much greater emphasis on research output (through the Research Assessment Exercise, the RAE), teaching quality (through the requirements of the Quality Assurance Agency, the QAA) and on academics raising research and consultancy funds. Other changes include tighter contractual terms (affecting holidays and hours worked), an increase in the use of short-term and hourly-paid contracts and the loss of tenure. Overall, these changes have tended to alter the nature of the job, reducing autonomy and increasing the workload, including that of administrative and teaching tasks. At the same time, both the salaries and status of academics are perceived to have deteriorated relative to alternative careers (Halsey, 1992; Keep et al., 1996).

Substantial change in the nature of any job is likely to increase turnover, as a mismatch develops between the nature of the job to which people were recruited and the actual job. If these changes tend to reduce the quality of the job, rather than just change it, and if the applicant requirements are not altered (and, probably, lowered), recruitment will also become more difficult. Both turnover and recruitment difficulties will be exacerbated by a relative decline in pay.

2 Between 1995 and 2000, the number of staff in HE (excluding staff in medical cost centres and those who were less than 40% FTE) increased by around 11% (HEFCE, 2002). 3 A market is said to be ‘tight’ if demand is high relative to supply. In competitive markets, prices will adjust to bring demand and supply into line. In markets such as that for academic labour this may not happen for at least two reasons: first, the price of academic labour does not float because of the existence of national pay scales; second, the final product (higher education) is not itself sold on a competitive market.

2

1.1 Overview of the study Against this background, the study was designed to identify the factors which

lead to individuals entering and leaving academic employment in the English Higher Education sector. Although the main focus was entry and exit from the sector, recruitment to and retention by individual institutions can shed light on this and was also investigated.

For the purposes of the study academic employment was defined as jobs in higher education institutions (Universities and Colleges of Higher Education) whose main function was academic teaching or academic research, irrespective of the contractual terms of the job holder. Thus lecturing (e.g. Professors, and Lecturers) and research staff (e.g. Research Assistants, post-docs and Senior Research Fellows) are included, but academic-related staff (e.g. technicians) are not. Full-time, part-time, permanent and temporary staff within these groups are included4.

Two main groups of academics were excluded from the study: those in Further Education Institutions and those on clinical rates of pay. The resources of the study precluded inclusion of these two groups5. Following discussion with the DfES, it was decided to focus on staff at English HEIs. This was done in order to prevent differences in the funding and structure of the HE sectors obscuring the analysis. Two exceptions to this rule are the analysis of the HESA data on research students and the chapter on international comparisons of pay (Chapter 4). These are discussed in more detail in sections 1.1.1 and 1.1.2 below (and in further detail in Appendix C and Appendix D).

The study had five, inter-related, strands:

• a literature review to establish the nature of the recruitment and retention problems and to identify previous evidence on the factors affecting recruitment and retention; this was conducted from March to May 2003;

• analysis of HESA staff and student data, 2001/02, to provide a descriptive analysis of turnover in academia, to identify the basic characteristics of employment in the sector and to identify the student supply into academia6;

• a comparative analysis of pay, using 2001 data, both for comparable employment nationally and for academics in Higher Education internationally, in order to establish the competitiveness of academic pay;

• qualitative research within universities exploring human resource policies and practices and factors affecting entry and exit from the sector; the fieldwork was conducted between July 2003 and July 2004;

4 However, the coverage of hourly paid staff is severely limited, owing to limitations in the HESA data (see below) and to practical difficulties of sampling for the survey (see below). 5 Inclusion of the former would have extended the study to Further Education Institutions. The latter have different terms, conditions and employment patterns from other academics in HEIs and so a larger sample would have been required to adequately cover this group. Staff on clinical rates accounted for five per cent of academic staff (HESA Individualised Staff Record 2001/2). 6 Analysis of more recent HESA data and of trends over time is contained in HEFCE (2005a) and HEFCE (2005b).

3

• quantitative surveys of academic staff and of research students to identify factors which affect recruitment into academia and retention; the fieldwork was conducted between May and July 2004.

The research findings are affected by the policies and practices in effect at the time. To assist the reader, the DfES has produced a list of relevant Higher Education initiatives since the survey fieldwork. These are listed in Appendix H.

Below, essential aspects of the methods are described. Further details of the methods appear in appendices. Appendix A describes the qualitative research. Appendix B discusses the quantitative surveys of academic staff and research students. Appendix C describes the HESA staff and student datasets. Appendix D describes the data used for the intra- and international pay comparisons. Appendix F describes the model used for the analysis of the likelihood of students entering academia. Appendix G describes the econometric analysis of job satisfaction and intentions to leave academia.

1.1.1 HESA staff and student data7 For staff, the HESA Individualised Staff Record for the academic year

2001/02 was used. The analysis was confined to institutions in England and to staff who were not on clinical grades. The sole exception to this is chapter 4, where the focus is expanded to the whole of the UK, to maintain consistency with the international comparisons analysis. For the study’s purposes, the Individualised Staff Record data has two important limitations. Firstly, they exclude employees8 whose total academic employment is below that of 25 per cent of a full-time academic (i.e. those with short hours or with substantial management and administrative responsibilities are excluded)9. One of the implications is that hourly-paid staff will be substantially under-reported and is unlikely to be representative of hourly-paid staff as a whole. Secondly, the data relating to leavers suffer from a high level of non-reporting: around 60 per cent of leavers’ destinations are missing10. Therefore the findings on movement out of the sector must be treated with caution.

For students, the Combined Student/Module Record for the academic year 2001/2 was used. This was combined with the First Destination Supplement (FDS), relating to those students who left in 2001/2. The analysis was not limited to students from English higher education institutions because the appropriate pool of domestic entrants into Higher Education academic post is the whole of the UK. It is important

7 Note that the University of North London was not included in either the staff or student data supplied by HESA because the university has asked that its individual data is not released. 8 Strictly, they exclude contracts whose total academic employment is below that of 25 per cent of a full-time academic, as the record reports contracts rather than individuals. For more information see Appendix C. 9 However, there are staff in the Individualised Staff Record with their FTE recorded as less than 25%; the majority of these records relate to staff who arrived or left during the year. 10 Internal work by HEFCE that matched the 2001-2 survey with that for the following year using staff code, data of birth and sex, found that 20% of those for whom the destination was not known remained at the same institution, 5% were found at a different institution and the remainder could not be matched with a record in the second year. This latter group are made up of those who left the sector and those who remained but whose record in the second year did not match with respect to one of the three criteria.

4

to note that the destination of postgraduate research students in the FDS has a particularly low level of response (38%).

Further details are given in Appendix C.

1.1.2 The comparative analysis of pay The comparative analysis of pay uses data from national labour force surveys

(and censuses in nine countries. These were chosen to illustrate the types of countries to and from which most international movement with UK academia occurs. They include the main English–speaking countries to which UK academics move (the USA and Australia), together with other English-speaking nations (New Zealand and Canada), three European countries (Denmark, France and Sweden) and Japan.

The analysis of the labour force survey data used in the international comparisons used data from the whole of the UK. This was to increase the sample size. However, we would not expect to find significant differences within the UK.

Identifying higher education academics was done using information on occupation and industry where available. In most countries we were able to obtain a sample group that matched the UK sample. Exceptions to this were the US, where the sample also included academic staff at state colleges, who also conduct teaching undertaken in the FE sector in the UK, and Australia and New Zealand, where it is possible that our sample excludes some researchers who have no teaching responsibilities. We discuss the implications of this in Chapter 4 and Appendix D.

Comparisons were made in both nominal and real terms. Earnings were converted using exchange rates to make nominal comparisons. In order to account for differences in the cost of living, purchasing power parity exchange rates developed by the OECD were used to make real earnings comparisons.

Further details are given in Appendix D.

1.1.3 The qualitative and quantitative survey research Qualitative research was conducted in thirteen English universities and

quantitative research conducted in a subset of these. A structured sample of universities was selected to ensure coverage of different types of universities (new, old and colleges of Higher Education), universities in London and elsewhere and universities with differing research ratings. Institutions with fewer than 200 academic staff and most specialist institutions11 were excluded. Small institutions were excluded because economies of scale in setting up the quantitative survey meant that their inclusion would have led to a smaller survey, as the project resources could not increase the sample through an increased number of institutions. Specialist institutions were excluded for similar reasons. (This did not reduce the subject coverage, as subjects taught in specialist institutions are also found in other HEIs.)

The purpose of the qualitative research was to identify factors which might affect recruitment and retention, including human resource practices and staff preferences. Qualitative interviews were held with senior staff with responsibility for

11 ‘Specialist institution’ is a classification developed for funding purposes and refers to institutions where 60 percent or more of funding is allocated to one or two cost centres.

5

human resourcing and, in eight of the universities, interviews were conducted with heads of two departments, and a sample of their academic staff and research students.

A survey of academic staff was conducted in ten12 of these universities. The survey covered both research and lecturing staff. Full-and part-time staff were included, but hourly paid staff were excluded (see Appendix B). The questionnaire collected data on personal characteristics, employment history, views on aspects of the job and career intentions. The survey was web-based. A total of 2805 staff responded, a response rate of 32 per cent. Survey data have been re-weighted to be representative of university academic staff in English HEIs. For more information on the weighting and other issues relating to the staff survey see section B.1 of Appendix B.

A survey of research students (full-time and part-time) was conducted in nine of the universities, where research students were those undertaking a Masters degree mainly by research or a doctorate. The questionnaire collected data on personal characteristics, employment history and career intentions. The survey was web-based. A total of 1330 research students responded, a response rate of 29 per cent. Survey data have been re-weighted to be representative of research students in English HEIs. For more information on the weighting and other issues relating to the staff survey see Section B.7.

Further details are given in Appendix B.

1.2 Nomenclature Throughout this report the following nomenclature is used:

• Student when referring to the student survey refers to research student.

• ‘Academic’, ‘academic staff’ refers to those employed in higher education institutions on either the research grade or the lecturing grade.

• University is used to refer to all higher education institutions, whether a university or a college.

• New and old universities. New13 universities are those that received university status in 1992 (when polytechnics and many colleges of Higher Education converted to university status) or later; old universities are those which had university status before this date.

1.3 Report layout The structure of the report is as follows. The next chapter sets the scene by

presenting evidence on turnover and recruitment and retention problems in higher

12 The aim had been to survey staff and students in twelve universities. Unfortunately, not all the universities were able to supply the sample, either due to data protection considerations or due to difficulties providing an email contact list. 13 This nomenclature is in common use now, but, previously, ‘new university’ was used to denote universities established in the 1960s and early 1970s. Perhaps the term was also used in the nineteenth century to refer to the redbrick universities when the sector was expanded in the Victorian era.

6

education. It also presents evidence on the factors affecting recruitment and retention of academics.

Chapter 3 then describes the structure of academic employment in Higher Education, including the grade structure and contractual status, and the main characteristics of academic staff. This description is used to raise some of the factors which might affect recruitment and retention.

Chapter 4 continues with the theme of structure, focusing on pay, and examines relative pay to investigate whether pay differences may be a cause of recruitment and retention difficulties. Both domestic and international comparisons are made. The former compares the relative pay of UK academics with that earned by other graduates in the UK; the latter compares academic salaries across eight countries to examine whether pay differences may encourage emigration to universities abroad (or immigration from universities abroad and new entrants to academia).

The following three chapters examine career paths and the nature of academic jobs. Chapter 5 describes entry into academic employment: the source of entrants and the factors affecting their career decisions. Students are one of the largest sources of entrants to academia and the chapter concentrates on the career choices of this group. However, some consideration is also given to other sources of entrants and the factors affecting career changes.

Chapter 6 turns to the nature of employment once in academia. It describes expectations about career progression and identifies factors which affect academics’ satisfaction in employment. This is examined in more detail using a model of the factors affecting satisfaction, as satisfaction is a good predictor of leaving, which is described in more detail in Appendix G.

Chapter 7 then examines movement out of the sector: expectations of leaving and destinations of leavers. A model of likelihood of leaving is developed, identifying the key factors which might affect retention.

Together, Chapters 3 to 7 identify the factors affecting recruitment and retention. Based on these findings, Chapter 8 considers how university human resource policies and practices may be influencing recruitment and retention. It describes practices in the case study universities and draws on common practice across universities.

Chapter 9 draws conclusions.

7

2 Previous evidence on recruitment and retention issues in academia

2.1 Introduction Using previous research and evidence from the case study universities, this

chapter examines the evidence on recruitment and retention issues in academia. Initially the chapter considers recruitment and retention problems, as evidenced through turnover and through reported recruitment difficulties. First, recent academic staff turnover in Higher Education is described. However, the level of turnover, per se, is not a direct indicator of problems. Whilst higher levels of turnover are normally seen as undesirable and lower as desirable, what is desirable depends on the specific employment circumstances. There is also an important distinction to be made between turnover within the sector and movement into and out of the sector. Moreover, irrespective of the level of turnover, recruitment may be difficult. The following section presents previous evidence on recruitment and retention problems and expands this with evidence of the case study universities’ recent experience. The evidence presented tends to rely on the view of senior human resource managers in higher education institutions. Defining recruitment and retention problems is highly subjective and will be influenced by circumstances, one’s view and one’s experience. Nevertheless, the approach does show whether universities are experiencing difficulties and allows changes in difficulties and differences between subjects to be identified.

Section 2.4 then discusses the factors which have been identified as affecting recruitment and retention in academia.

A summary is provided in the final section.

2.2 Turnover14 Eighty-seven per cent of academic staff in Higher Education in 2001/02 (for

whom we have data) were working in the same institution as they had in the previous year and the recruitment rate was 13 per cent (Table 2.1). However, these figures exclude the three per cent of staff for which data was missing. This is likely to include

14 This section uses HESA Individualised Staff Record data. As has been discussed in Chapter 1 (and Appendix C), certain caveats need to be borne in mind with respect to the reliability of data on the destination of leavers and the source of entrants because of the relatively high levels of non-reporting of these. Thus figures based on these data need to be treated with caution. Note also that the data are confined to academic staff who work 25% FTE and over and relate to contracts rather than individuals (for more on the precise definition of the data see Appendix C). Internal work by HEFCE comparing contract counts with counts of individuals at HEI’s (i.e. if an individual held a more than one post at a single HEI, they were counted as a single observation, but if they held posts at more than one HEI they were counted as more than one) found that the results are almost identical (See Table A.12 and Table A.13 in Appendix C)

8

a high percentage who were new to the institution. Thus, a recruitment rate of 13 per cent is likely to be a minimum estimate of recruitment.

Table 2.1 Turnover in HE staff, 2001/2 (%)

per cent na

New to the institution 13 New to the sector b 9 106,746

Left the institution 7 Left the sector c 5 102,558

Average tenure 8.2 years • a Note that these numbers are different because fewer people

report their state in the following year • b Included in those new to the institution • c Included in those leaving the sector • percentage relates to those whose status was known (i.e.

unknowns are excluded from the denominator) • Data refers to Staff at English HEIs and excludes those on

clinical grades • Note that the University of North London was not included

in either the staff or student data supplied by HESA because the university has asked that its individual data is not released.

• Source: HESA Individualised Staff Record 20001/2

Seven per cent of staff were recorded as having left. Whilst, with some growth in the sector, the number of staff recruited should be greater than the number leaving, the degree of disparity between these two figures suggests under-recording of leavers and that leaving would be closer to the level of recruitment. If so, the turnover rate in academic jobs was similar or slightly higher than that for all professional employees in the economy, 13 per cent in 2001 (CIPD, 2002).

Comparison with previous years suggests that turnover has grown substantially; between 1994-95 and 1998-99 turnover ranged between 2.8 per cent and 5.3 per cent p.a. (PREST, 2000, using HESA data)15. Further evidence of a growth in turnover comes from the findings of the Independent Review of Higher Education Pay and Conditions, which, using a survey approach, put the turnover rate for academic staff in 1998 at 6.4 per cent (Appendix E, Bett, 1999)16. Thus, although the level of turnover may be about average for this type of employee, some recruitment and retention difficulties may be caused because universities would not have been used to dealing with this rate of recruitment and retention. The rise may also contribute to universities’ perceptions of difficulty. 15 Whilst the data analysed are not strictly comparable, as the PREST data was restricted to HEFCE-funded staff (i.e. those who were funded through research council budgets, which would exclude many contract research staff), this difference is unlikely to account for the full increase in turnover. 16 The survey was of heads of personnel in all UK HEIs. The response rate for supplying turnover data was 40 per cent.

9

The above describes recruitment and retention to individual higher education institutions. Whilst this is important in understanding recruitment and retention problems, overall movement into and out of the sector is of greatest interest for this study. In 2001/02, HESA data recorded that a little over two-thirds of recruits to higher education institutions were new to the sector (accounting for nine per cent of staff in higher education institutions) and almost three quarters of leavers (for whom destinations were known) left the sector (accounting for five per cent of the sector) (Table 2.1). If accurate, the figure for new entrants seems high and likely to cause problems. Whilst the figure for leavers from the sector is not high17 (suggesting an average stay in academic employment of 20 years), it is almost certainly an under-estimate, as the destination of over 60 per cent of the staff who were recorded as leaving in the sector is ‘don’t know’.

Is this pattern consistent across subjects? Turnover is higher in science than non-science, and recruitment from outside the sector is higher for science posts (Table 2.2)18. Turnover appears to be higher in old universities, with 16 per cent (for whom we have data) recruited in 2001/02 compared with nine per cent in new universities (Table 2.3).

Table 2.2 Turnover by broad subject (%)

Total Non-

science Scienceper cent n

Inflows Same HEI 89 85 87 From other HEI 4 4 4 New to sector 7 11 9 Total known 100 100 100 106,746Not known 4 2 3 3,348 Total 110,094 Outflows Same HEI 95 93 94 Other UK HEI 1 2 2 Left sector 3 4 4 Retired or died 1 1 1 Total known 100 100 100 102,558Not known 7 7 7 7,536 Total 110,094

• Source: HESA Individualised Staff Record 2001/2 • Data refers to staff at English HEIs and excludes those on clinical grades

17 Although 75 per cent of leavers leaving the sector would be high. 18 Note that such comparisons are only valid if non-response does not vary by subject.

10

The differences between science and non-science and between old and new universities may be due to differences in the percentage of researchers (who are normally on fixed-term contracts) (see Chapter 3), as, unsurprisingly, turnover is higher amongst those on fixed-term contracts than permanent staff (Table 2.4). Employees on fixed-term contracts were also much more likely to have been recruited from outside the sector (18 per cent) and to leave the sector, compared with those on permanent contracts. None of this is surprising given the standard career path of entry to academic employment through research posts, almost all of which offer short-term contracts and the consequent drift from the sector for those who fail to secure a permanent post (see Section 2.4 and Chapters 5 and 7). Table 2.5 shows how these new entrants to the sector and leavers from the sector are concentrated amongst those on short-term contracts, with seven per cent of staff new to the sector and on short-term contracts and two per cent new and on permanent contracts.

Table 2.3 Turnover by higher education institution type (%)

Old University

New University Total

Inflows Same HEI 84 91 87 From other HEI 4 3 4 New to sector 11 6 9 Total known 100 100 100 Outflows Same HEI 92 96 94 Other UK HEI 2 1 2 Left sector 5 2 4 Retired or died 1 1 1 Total known 100 100 100


11

Table 2.4 Turnover by type of contract (%)

Permanent Fixed-term contract Total



Table 2.5 Turnover as a percentage of total turnover by type of contract

Permanent Fixed term contract

Total (including hourly paid)



2.3 Evidence on recruitment and retention problems Recruitment and retention difficulties for academic staff in higher education

appear to be a recent phenomenon. Certainly, the literature pointed to few concerns in

12

the UK about the ability of universities to recruit and retain academic staff during the 1970s and 1980s (HEFCE, 200319)20, with the exception of concern about a potential retirement crisis. The issue became more prominent during the 1990s and evidence, based on the difficulties reported by university human resource specialists, suggests that difficulties have grown over the past decade.

In respect of retirement, fear that the age profile of academics in many disciplines was dangerously skewed to those close to retirement was causing concern in the early 1980s and the 1990s (UGC, 1984; Keep et al., 1996). However, the age profile of academics is similar to that of the workforce as a whole (PREST, 2000), although analysis of the age structures of departments showed considerable variation, with the proportion of staff set to retire in the next five years ranging from under 10 per cent to over 25 per cent (UCEA, 2002). Analysis using a sophisticated personnel model suggested that the age profile does not present a particular problem in most subject areas, with the exception of mathematics, physics and engineering (HEFCE, 200221). We return to this issue in section 3.4.2 below.

The evidence for shortage is based on the reported experience of higher education institutions of difficulties recruiting and retaining academic staff (see, for example, Bett, 1999 and HEFCE, 2003). As would be expected, difficulties are not uniform and vary by institution, location, subject, grade and contract, and may also vary by type of individual. At the end of the 1980s the only area where there were major difficulties was engineering and technology (Pearson et al., 1990). Even by 1998, there were still not widespread problems with recruitment and retention but the number of areas in which difficulties existed had increased (Bett, 1999). Problematic areas included recruitment and retention in business subjects, IT, electronic engineering and some rarer specialisms, and in the recruitment of academics with professional experience (e.g. in law, health studies and teaching)22. Other areas suffering problems with retention included researchers and teaching staff on fixed-term contracts (Bett, 1999). Recruitment problems were sharpest in the old universities, in the South-East (excluding London) and the West Midlands.

However, over the next few years reported recruitment and retention problems intensified and became more widespread. Based on higher education institution human resource specialists’ reports of recruitment and retention, UCEA (2002)

19 This study included structured interviews to elicit the views of a number of stakeholders. These 15 organisations ranged from Association of University Teachers and Universities UK, to the Equal Opportunities Commission and the Economic Social Research Council. The study also collected evidence from various categories of academic staff employed in six HEIs. 20 The story is slightly different in the US, where predictions of staff shortages from the 1970s (e.g. Freeman, 1971, 1975; Stapleton, 1989) led to an increase in interest in academic labour markets (see Ehrenberg, 2003, for a summary). In the US, however, the main thrust of the enquiry was the shortages of PhD students rather than issues pertaining to academic staff themselves (Cartter, 1976; Ehrenberg, 1991, 1992) 21 Note that the HESA data supplied to HEFCE allowed aggregation over contracts held by the same individual. HEFCE analysis referred to in this report refers to staff and not contracts. HEFCE also have different cut-offs e.g. 40% FTE for some analysis. This means that our figures and HEFCE’s (2002, 2005a 2005b) will not necessarily tally. 22 These figures come from the staff survey conducted for the Independent Review of Higher Education Pay and Conditions (Bett, 1999). This was the result of surveys sent to heads of personnel of all 178 UK HEIs. The overall response rate was 77 per cent, giving information pertaining to 246,027 of the estimated 300,780 staff in the UK, including 110,070 of the estimated 133,977 academic, clinical academic and research staff in UK HEIs.

13

pointed to a considerable increase in recruitment and retention difficulties for both academic and support staff since 1998. Around one in five institutions reported experiencing difficulties filling academic positions in 2001 (18 per cent), compared to one in twenty in 1998 (six per cent)23. The subject areas causing the most problems were similar to those in the Bett Report, but had expanded. They now included computing/IT, business subjects (accountancy/finance, business/management, law and economics), engineering, science subjects (biological sciences, chemistry and physics), nursing/midwifery and professions allied to medicine, and education. The percentage increase in retention difficulties was of a similar magnitude, with 7.6 per cent reporting retention difficulties ‘most of the time’ or more in 2001, compared to 2.2 per cent in 1998. Departments most frequently mentioned as having a turnover problem were: computing and computer science; law; accountancy and finance; business, business management and information systems; engineering (including electrical); and education (Appendix 4, UCEA, 2002). Institutions reported that lecturers were the most difficult to recruit, with almost 60 of all institutions reporting difficulties. The percentage reporting difficulties in recruiting lecturers was similar in new and old universities, but lower in colleges. More old universities than new found it difficult to recruit professors and research assistants.

Recruitment and retention problems were particularly acute in areas which had to compete with the private sector, such as law, IT and engineering (UCEA, 2002). However, the same was true in areas competing with other public sector jobs with higher pay, such as education and subjects allied to medicine (ibid). The result of this was that HEIs reported difficulties attracting many candidates and those that they did attract were often not of the requisite quality. Moreover, they also reported that it was difficult to recruit good young academic staff as a result of low starting salaries (ibid).

2.3.1 Current experience of the case study universities: recruitment Further evidence on the nature of recruitment and retention problems was

gathered from the case study universities. This evidence is based on a much smaller number of universities than the UCEA, HEFCE and Bett studies (126, 35, and 137 respectively). However, within these universities, it drew on the views of Heads of department (two per university) as well as human resource specialists. This has the benefit of providing more detail from the ‘coal face’, but does mean that there is greater information about selected subjects.

The case study universities did not appear to have widespread recruitment problems for academic staff. Although they experienced difficulties in some areas of recruitment and difficulties attracting the quality of staff they would prefer, neither human resource managers nor heads of departments felt that this amounted to a recruitment ‘problem’ or an issue in urgent need of attention at institutional or Government level.

Contrary to the evidence presented above, respondents did not see difficulties as a recent phenomenon. They also did not see them as persistent, except in a few highly specialist areas. Difficulties in some subject areas were viewed as cyclical and linked to external factors including external markets.

23 Here a recruitment difficulty is defined as those who answered ‘always’, ‘usually’ or ‘more often than not’ to the question ‘how often do you have difficulty in retaining staff?’. The other responses were ‘sometimes’, ‘rarely and ‘never’.

14

Human resource managers and heads of department in the case study universities reported greater difficulty recruiting senior level staff, particularly to readerships and chairs. Junior staff were reported to be easier to recruit, although posts involving heavy teaching duties and little time for research were less sought after and the case study universities in London reported more difficulties than others in recruiting junior staff.

Recruitment difficulties in the case study universities were reported in subject areas with strong external labour markets. At the time of the fieldwork, these included education and health and, in the recent past, had included law, computing, business and engineering. Only new case study universities reported problems in psychology and economics. Difficulties recruiting to particular subject areas, for example computing, engineering and law, were seen as cyclical, varying with market conditions, by some human resource managers, a view which was confirmed by the experience of some academic staff.

The nature of the difficulties tended to stem more from the quality of applicants, rather than the number. Insufficient research experience and track record were the most commonly cited short-comings. In respect of recruitment in vocational subjects, problems were reported in relation to the competence, in higher education, of practitioners. A particular problem was identified in attracting those with sufficient energy or enthusiasm for academic work, and particularly for research. Universities were concerned not to recruit people who ‘couldn’t hack it’ in the private sector and who saw academic life as an easy ride. Again, this was a more common concern among case study new universities who were willing to recruit applicants without research experience, on the basis that they would acquire it. Case study old universities required all new staff to have research experience, with the exception of visiting lecturers, and therefore could avoid this potential problem.

Shortages due to lack of academic qualifications among some good applicants were exacerbated in a Health Studies department in a new university, where, until recently, practitioners were recruited (and could rise to senior positions) with a diploma. However, the university now required lecturers to have a degree and senior lecturers to have a Masters degree, which few applicants had. Moreover, the university did not recognise the professional qualifications in their pay decisions. This meant that many suitable practitioners could not be recruited, whilst others could only be offered jobs on the lecturing scale, at salaries substantially below those in the health service. Whilst Health Studies was the only example in the qualitative research where this issue was identified, we would expect the problem to be replicated in other subjects recruiting practitioners where practitioners were required to have relatively low level academic qualifications only.

2.3.2 Current experience of the case study universities: turnover Human resource managers said their universities did not have a problem with

overall turnover of academic staff. This was for two main reasons: turnover was objectively low, and some of the case study universities wished to reduce staff in some areas because of changes in student demand. Therefore, they welcomed movement of staff in some areas. At the same time, case study universities did see levels of turnover in some sections of the academic workforce, and in some subject areas, as problematic. Their principal concern was the loss of the more highly productive researchers. However, some were concerned that departures of research

15

active senior staff to other UK universities would increase in the lead up to the next RAE.

Most of the case study universities’ human resource managers reported turnover levels in the region of four to eight per cent. Only one case study university reported significantly higher levels, at 16 per cent. A number of human resource managers did not know the current rate of turnover in their institution. Many did not have an opinion on whether turnover in their institution could be considered high or not, but did not believe their university had a turnover problem, including the university where turnover was 16 per cent. Although many human resource managers seemed to lack an overall picture of turnover among academic staff, it was the configuration of staff that was of most concern to them. Their principal concern was that staffing levels matched income from students and research and that numbers of non-research active staff should be reduced. Consequently, they were aware of areas of greatest movement, both in terms of grade and subject area.

Far from wanting turnover to be lower, a number of human resource managers said they would even welcome an increase in staff turnover to more than eight per cent because this would allow for the staff profile of some departments and schools to be reviewed and make it easier to improve the calibre of staff. This view was expressed most strongly by those in case study new universities which had undertaken extensive redundancy programmes aimed largely at reducing the number of non-research active staff. In some cases, these programmes had not resulted in the re-balancing that universities had aimed for, but had reduced the age profile and teaching experience of lecturing staff. A preference for higher, rather than lower, turnover was also voiced by a small number of heads of department. One head of art and design in a new university stated,

‘There hasn’t been a turnover of staff and the staff profile has aged. I would like more of a cross-section and would like to appoint some of the visiting lecturers to lectureships, but I can’t do this unless current staff leave’.

Case study university heads of department expressing a preference for higher turnover were generally in new universities and under pressure to increase their research ratings. They reported that staff with shorter length of service were usually those with a stronger track record in research and who were more marketable to other institutions.

Turnover was reported to be higher in junior posts, particularly among researchers and junior lecturers leaving for senior positions or higher ranking universities. The exception to this was a new London university, which experienced higher turnover in the middle and senior grades, possibly for reasons of living costs, it was thought. Although the case study universities preferred not to lose junior staff to other institutions, particularly those who are research active, they appeared to be resigned to a certain level of loss on the grounds that moving around is part of the academic career and that they are powerless to prevent it. Movement was believed necessary to gain promotion and also ‘experience’ which could lead to promotion in time. Human resource managers in some old universities were more concerned at the loss of more senior research active staff to other universities in the UK and overseas than the loss of junior staff. However, turnover among this group was not thought to be excessive and, again, universities were resigned to it. A practice of particular

16

concern was of staff taking colleagues with them to their new department or research centre. Several departments gave recent examples of where this had happened to their own institution. It included the loss of almost an entire physics research team from one old university to another. This practice was reported to be very damaging for departments and faculties and highlighted the importance of retaining ‘star’ performers.

2.4 Factors affecting recruitment and retention

2.4.1 Introduction Recruitment and retention are affected by the whole employment package (the

rewards and disbenefits of the job) relative to other employment. These include pay and fringe benefits, intrinsic aspects of the job (e.g., for academics, teaching and research), job security, work organisation, autonomy, progression, family-friendly practices, congeniality of colleagues and the working environment etc. The more attractive the overall package, the more likely it will attract applicants and retain employees. The relative importance of these factors differs for recruitment and retention, due to informational differences between those in a job and potential recruits. Applicants (particularly those entering the sector) have less knowledge and the factors influencing recruitment tend to be those on which information is more easily available. This means that pay tends to loom larger for recruitment than retention. Moreover, the expected and the actual package may differ, leading to turnover.

This section presents the factors which have been identified in the literature as affecting recruitment and retention. It starts with a description of what human resource managers and, from the case study universities, Heads of department consider to affect recruitment and retention. It then turns to evidence based on employees’ views and behaviour. Much of the evidence relies on staff views on how certain factors affect their satisfaction. Whilst there is a strong link between overall job satisfaction and turnover (see below), it cannot be assumed that specific factors with which employees are satisfied or dissatisfied affect turnover. Without verification of the link between these factors and behaviour, this evidence should be treated with caution. We shall investigate the link between satisfaction with the various elements of the job and intentions to leave UK higher education. The model and results are presented in Appendix G and are discussed in chapters 6 and 7.

2.4.2 Employers’ views on factors affecting recruitment and retention With regard to specific elements of the package which affect recruitment and

retention, those considered by higher education institution human resources departments to be causing most problems were: uncompetitive salaries, location-specific issues (such as the HEI was in a rural or expensive area, or one with few job opportunities for partner); job insecurity from fixed-term contracts; workload; and poor promotion possibilities (Bett, 1999; UCEA 2002)24.

24 Whilst it is useful to know what factors universities considered to cause recruitment and retention difficulties, it is important to verify these. Indeed, if any causes have been incorrectly identified, this

17

Pay levels were the main reason cited by human resource departments for recruitment and retention difficulties in both the survey of HR departments and the case studies (at a sample of 14 higher education institutions) in UCEA (2002). Two thirds of all respondents mentioned pay as being a major factor behind recruitment and retention problems in the sector, particularly for staff whose expertise is valued in the private sector: IT and computing, law and accountancy were the areas where staff were most likely to leave the HE sector. It was also noted that institutions in London and other major UK cities reported that high housing and travel costs were exacerbating the problem of low salaries (ibid).

The case study universities interviewed as part of this project amplified some of the issues regarding pay and some mentioned other factors which they believed affected their recruitment and retention.

2.4.2.1 Pay

Particular problems in respect of pay were identified by the case study universities in a number of subjects with strong external labour markets. Health and education, where recruitment was primarily from the health service and from schools, respectively, were seen as highly problematic. These sectors have benefited from considerable pay increases in recent years and salaries have now risen above those offered in Higher Education. As a head of Education at a new university stated,

‘The difficulty is fundamentally that the largest base of recruitment is people coming from schools where salaries for people with positions of responsibility have outstripped the top of the senior lecturer scale’.

However, some heads of department referred to a steady decline in academic pay and conditions rather than an increase in salaries elsewhere. As a head of law at an old university explained,

‘A university career has become less attractive, with ever increasing levels of bureaucracy. Academics are no longer respected and pay now lags badly behind equivalent or lesser qualified jobs. Some colleagues consider they are grossly underpaid, compared with newly qualified solicitors. You couldn’t even begin to compare their salaries with their equivalents in private practice where they earn more than 5 times an academic salary’.

Although universities experienced problems with recruitment where external markets were strong, or pay and conditions superior, they also benefited from downturns. For example a number of new universities had recruited staff during downturns in the telecommunications and computing industries.

2.4.2.2 Reputation and subject area A department’s reputation and that of existing staff was seen as an important

influence on recruitment and retention by some of the case study old universities. itself might lead to recruitment and retention problems, as human resource strategies might be inappropriate and, possibly, counterproductive.

18

These said they were able to attract staff wishing to work with their well-known academics. Equally, some were concerned at the loss of research ‘stars’ since their departments then became less attractive to applicants. These issues were not raised by the case study new universities.

2.4.2.3 Redundancies

Some of the case study universities felt they faced greater recruitment difficulties due to recent redundancy programmes. Human resource managers in these universities reported difficulties recruiting high quality staff, required by the university to improve research ratings, because of knowledge of its staff cutting exercises in the wider academic community and because of potential employees’ concerns about job stability.

2.4.2.4 Promotion

In respect of retention, some human resource managers commented that junior staff often found it easier to leave their institution to gain promotion, and acknowledged that staff might be encouraged to stay if internal promotion was less slow and complicated. One head of human resources in a new university explained,

‘The deans say they are not able to keep pace with staff expectations of salary. Junior staff begin on a low salary and they can get experienced and very good in quite a short period of time. Often this is by having work dumped on them by the wily old lags. But progression is about time-serving and we don’t deviate from it. So the only way they can get a higher salary is to leave the university for other institutions. These are often younger and enthusiastic staff with strong academic ability who we don’t want to leave. At one point we (the human resources department) did propose an accelerated scheme, but it was strongly opposed by staff and by the unions’.

Not all heads of department felt powerless to prevent the loss of good junior staff. One head of engineering in an old university was developing ways in which lecturers could gain the experience necessary to receive promotion without having to move elsewhere.

‘The strength of academia is the movement of people as they develop their specialism, but this also means that keeping people can be difficult, especially in some areas where there is a real dearth of good people. Staff may need to leave their university to gain experience in another institution, but that shouldn’t mean losing them for good. We arrange industrial secondments and most of them come back, so it can work.’

2.4.3 The whole package Most previous research focuses on specific aspects affecting recruitment and

retention. However, as we have said, recruitment and retention will be influenced by the whole package and being able to rank different factors is useful.

19

Research in Australia may be instructive. Bellamy et al. (2003) conducted a survey of academic staff in business subjects at all 38 Australian universities to determine why they remain at universities ‘despite deteriorating working conditions and reduced job satisfaction’25. They found that the most important factors in becoming and remaining an academic were flexibility, autonomy, teaching, research, and the ‘community of scholars’26. Total income and university salary were rated tenth and eleventh out of fourteen factors.

A study of Nigerian HEIs found the most important factor affecting retention was pay (Mallam, 199427). The other factors (in declining order of importance) were: supervision, the work on the present job, the job in general, co-workers, and the commitment of the respondents to the institution at which they worked. These, and other, factors are examined below.

2.4.4 Job satisfaction of academics Previous UK research has not explored the overall attractiveness of the

academic job package for potential recruits (nor differences between expectations and the reality), but some studies have looked at overall satisfaction amongst academic employees. In the wider economy, reported job satisfaction measures have been found to be good predictors of future quits (Freeman, 1978; Akerlof et al., 1988; Clark et al., 1998; Clark, 2001), although (based on single institution studies) there is a little evidence to this effect from the US HE sector (Nicholson and Miljus, 1972; Hinsz and Nelson, 1990).

Most analyses of job satisfaction are based either explicitly or implicitly on the job descriptive index, first devised in the 1960s, e.g. Smith et al. (1969), and updated in Balzer et al (1997). This measures job satisfaction using 5 facets: work in present job, present pay, opportunities for promotion, supervision and co-workers. Subsequent work has tended to add to these facets and the elements that are generally considered to make up overall satisfaction with employment include the nature of the job itself, the hours of work, pay and promotion, co-workers’ and supervisors’ behaviour and physical working conditions. The most common method to measure workers satisfaction is using a 7-point scale running from extremely dissatisfied to extremely satisfied, with four on the scale representing indifference (e.g. Oshagbemi28 1996, 1998; Clark and Oswald, 1996; Ward and Sloane, 2000; Clark, 2001).

In his study of job satisfaction and turnover in the whole economy, Clark (2001) found that if one compares overall job satisfaction with satisfaction with the separate elements of work, one finds that it is overall job satisfaction that is the best predictor of quits, followed by pay and the work itself. However, the various elements 25 The study is based on 3161 respondents, a response rate of 42 per cent. 26 Between 61 per cent and 91 per cent of the respondents rated these aspects as ‘important’ or ‘very important’. The percentage rating these aspects as important for becoming an academic were almost identical to their importance for remaining an academic (although the latter figures are slightly lower (with the exception of research). 27 The study was based on a survey of 247 full-time faculty members from ten randomly selected HEIs, yielding a return of 208 respondents (a response rate of 84 per cent). 28 Oshagbemi (1996) investigated the satisfaction of HE staff from 23 universities in the UK with eight aspects of academic life (the survey achieved 554 respondents with a response rate of 50.27%), but does not give an overall measure of satisfaction. Although in later work (Oshagbemi, 1998), he reports a measure of overall satisfaction, but does not explain how this is derived from measures of satisfaction with individual elements.

20

of job satisfaction have been found to be very highly correlated both with each other and with overall satisfaction, particularly in academics (Ward and Sloane, 2000). Ward and Sloane (2000) suggest that academics place a lower emphasis on pecuniary relative to non-pecuniary aspects of work than other sectors of the workforce, although there is little discussion of this.

There have been no analyses comparing the satisfaction of academic staff in HE with the rest of the population using the same survey. However, three studies have examined very similar aspects of satisfaction and comparison of these may point to problems29. The levels of satisfaction reported by academic staff on the key aspects of job satisfaction are generally lower than those reported for the whole workforce (Table 2.6). Only on satisfaction with the opportunity to use their own initiative do academic staff not score below the general workforce. The greatest differences in satisfaction between academics and the population as a whole is with pay and promotions. Although these are the areas where satisfaction is the lowest in the economy as a whole, the difference between satisfaction with these two facets and the other is much larger for academics than for other workers. Academic staff are most satisfied with the work itself and teaching in particular30.

Table 2.6 Comparison of job satisfaction of academics with whole economy

Whole economy UK academics Scottish

academics Clark (2001) Oshagbemi (1996) Ward and

Sloane (2000) Overall job satisfaction 5.427 4.212* 5.04 Promotion 4.484 3.42 3.40 Pay 4.615 3.44 3.60 Hours 5.214 - 4.52

Work itself 5.562 Teaching = 5.09 Research = 4.66 Admin = 3.93

5.27

Job security 5.192 - 4.41 Opportunity to use initiative 5.745 - 5.81 Supervisors 5.529 4.18 5.09 Co-workers - 4.81 5.42 Physical work conditions - 4.33 -

• * Oshagbemi (1998) • All figures based on 7-point scale running from extremely dissatisfied to extremely satisfied,

with four on the scale representing indifference

2.4.5 Factors affecting recruitment and retention: pay One clear conclusion from the literature is that pay is a major issue (Bett,

1999; Machin and Oswald, 1999, 2000; UCEA, 2002; HEFCE; 2003). 29 Note that these comparisons are rather tenuous, as they refer to different surveys with slightly different questions. 30 Note that not all academic staff teach. Oshagbemi (1996, 1998) does not state whether these averages are averaged over all staff, or just those who perform these tasks.

21

Our comparison of elements of job satisfaction in Section 2.4.4 showed that academics were relatively less satisfied with their pay compared with the workforce as a whole. Pay also is a greater cause of stress amongst academics (Kinman, 1988; Kinman and Jones, 2004), particularly when compared with staff in a range of public sector employment (University of Plymouth, 2003).

2.4.5.1 Comparative pay levels within the UK

Certainly, starting salaries for postdocs have remained unchanged in real terms over the past 15 years, while the average figure for all graduates has risen substantially in this period (House of Commons Science and Technology Committee, 2002b, quoting evidence the Roberts Review, HM Treasury, 2002, para 5.29).

In their comparison of jobs in the HE sector with similar jobs in the public and private sector, the Hay study in Bett (1999) found that not only did academic jobs compare unfavourably with similarly ‘sized’ jobs in the private sector, but also with other public sector employment. For the Independent Review on Pay and Conditions, Hay Consulting (Appendix F, Bett, 1999) conducted a job evaluation of academic jobs and compared these jobs with other jobs in the public and private sectors of similar content or ‘size’. The conclusions of this study are that there were indeed large pay differentials between the HE sector and elsewhere. These were largest at the top and bottom of the scale (i.e. professors and Lecturer A, or equivalents) and in London and the South East31. In many cases, the upper quartile in the HE sector was lower than the median wage in the private sector. The most competitive wages were to be found at the Lecturer B level in the old university sector and at the Senior Lecturer level in new universities and colleges. The study also conducted an exercise in order to take into account other non-pecuniary factors (e.g. the length of the working week, holidays, cars, pensions and other fringe benefits) that might offset or amplify these differences in salary32. The results of this exercise were that the picture remains broadly the same, with tangible rewards lower for academic jobs. Indeed, for some more senior roles, rewards were even less competitive when the broader package is taken into account.

Machin and Oswald (1998, 2000) focused on one problem area (economics) and investigated the severe decline in UK PhD students, with the implicit assumption that this represents the primary source for economics academics33. They concluded that the main explanation for the decline in UK PhD students was low pay. They compared salaries and fringe benefits of academic economists to those of private sector economists, gleaned from the Society of Business Economists, between 1988 and 1998. The authors estimated that over this period the earnings of academic economists fell behind those of private-sector economists by approximately 20-30 per cent.

31 Note that there is a different grading system and nomenclature in old and new universities and colleges. We discussed this further in section 3.3.1. 32 Rees (1994) found that the cost of additional benefits were more significant in explaining faculty retention than pay for US academics. In this study, benefits were based on institutions’ assessment of the cost of their contributions to retirement and insurance plans, tuition waivers and payments, social security taxes, unemployment taxes, workers’ compensation taxes and benefits in kind with cash alternatives. 33 It is certainly true that the most commonly reported highest level of educational qualification of academics is the doctorate (Court, 1998).

22

2.4.5.2 The importance of pay

Whilst Clark (2001) found that overall satisfaction was the best predictor of turnover in the economy as a whole, pay alone was the second best predictor. Studies in the US have found that better paid faculty are less likely to leave their jobs (e.g. Ehrenberg et al., 1991; Rees, 1994). In a study of the factors influencing faculty turnover at 10 Nigerian colleges of technology/polytechnics, Mallam (1994) found that the most influential factor on voluntary turnover was dissatisfaction with pay. Oshagbemi (1996) found extensive dissatisfaction amongst UK university teachers with pay (54 per cent dissatisfied and 30 per cent satisfied)34. Pay was the factor with which the highest number of teaching staff expressed dissatisfaction in his survey, which examined satisfaction with seven factors.

It is not only aspects of the job itself that are likely to determine an individual’s satisfaction, but also those of alternative employment (Mobley, 1977; Clark and Oswald, 1996; Clark, 2001; Ward and Sloane, 2000). Using data from the British Household Panel Survey, Clark and Oswald (1996) found that workers’ reported satisfaction levels are inversely related to their comparison wage rates. This result has been replicated in the academic sector by Ward and Sloane (2000), who also found that subjective measures of comparison income35 were more important influences on academics’ overall job satisfaction than an objective one36. In our analysis we find that satisfaction with salary was strongly negatively related to perceptions of what staff felt they could earn elsewhere. For more on this see section 6.3.

It is helpful to consider work in other areas of the education sector. Dolton and van der Klaauw (1994, 1999) consider the effects of alternative wages in their competing risks analysis of the turnover of new UK school teachers. Dolton and van der Klaauw (1994, 1999) found that the higher the teachers’ salaries the less likely they were to leave and the higher the expected wage elsewhere37, the more likely they were to leave. However, expected wages only affected the propensity to leave teaching for a non-teaching job, they had no influence on the exit probability into non-employment. Female teachers were more likely than men to leave the labour market altogether, a difference that may be explained by family commitments. This work has recently been extended by Chevalier et al. (2002) to examine these effects in five cross-sections of university graduates covering the period 1960 to 1996. Chevalier et al. (2002) show how a time-series approach is a particularly powerful way to identify the effects of relative pay. They found cyclical differences in relative pay for teachers (either due to teachers’ wages not growing as fast or in periods when the economy is buoyant) and that the wage effects of alternative employment were stronger in periods when they were higher relative to teachers’ salaries.

34 The survey achieved 554 respondents (the response rate was 50.27 per cent). 35 Two subjective measures of comparative pay were used: the first was derived from a question asking respondents to say if they earned less or more than they deserve; the second being the difference between what they actually earned and what they felt they deserved. 36 This was obtained by estimating a human capital earnings regression on the wage data collected and a number of background characteristics such as gender, tenure, subject, grade, etc. The measure used was the difference between the predicted and actual value of earnings (i.e. the residual). As the authors note, there are a number of problems with this measure as the residual may capture many other reasons for deviations of actual from ‘deserved’ earnings, such as omitted variable bias or differences in unmeasureable factors such as ability. 37 Based on their predicted wage from an earnings function.

23

2.4.5.3 Pay systems

Another aspect of pay, which may have an impact on recruitment and retention in the HE sector, is the use of discretionary pay. Discretionary pay schemes have been extensively used in the private sector, particularly for middle and senior managers (Millward et al., 1992). During the 1980s and 1990s, a number of developments took place which introduced discretionary pay to the HE sector (Jackson, 1997). However, the implementation of discretionary pay in HEIs faced a number of problems: the difficulty of measuring individual performance (particularly in teaching); the lack of management and human resources structures found in the private sector (Keep and Sisson, 1992; Keep and Mayhew, 1995; Jackson, 1997; Bisset et al., 2000)38; and the frequent rotation of headship of departments and similar staff (Jackson, 1997). If staff are unhappy with the criteria for discretionary pay awards, these may do more harm than good (as might be suggested by the results of Court, 1999, for promotions criteria, discussed below). However, there is little or no evidence on staff’s satisfaction with the design and implementation of discretionary pay strategies.

2.4.5.4 Academic salaries in an international context

Relatively low academic pay is not unique to the UK (Enders, 2000). However, according to Enders (2000), salaries in the UK are among the lowest in Europe in absolute terms. Moreover, it is the countries with the lowest salaries where dissatisfaction with them is greatest. However, the author does not state on what evidence these conclusions are based. One problem with Enders’ comparison is that he compares wages in absolute terms and does not take into account differences in purchasing power.

There have been very few international studies of academic wages. This is unusual when one considers that the university sector labour market is one of the few that has a truly international dimension. One reason for the lack of international studies may be the difficulty in obtaining data from the Higher Education sector. The international higher education sector is a heterogeneous one, being made up of private and public sector organisations. In many countries it is a combination of the two. This has led those studies that have been done to sample a small number of higher education institutions in different countries to obtain the required data on earnings. In this study we bypass this problem using national survey data. These data are more readily available than information from education agencies, which differ between countries, or even within countries.

The Association of Commonwealth Universities (ACU) has undertaken a number of surveys of academic staff salaries and benefits in commonwealth countries. Provan (2001) compared academics in universities in six countries: Australia, Canada, New Zealand, the UK, South Africa and Singapore. The data were not collected in a systematic way, but rather a combination of finding information on the internet and a combination of web-based research and direct contact with university personnel officers. Provan (2001) converted nominal salaries into real salaries using World Bank purchasing power parity series. Provan (2001) found that the wages of UK

38 This situation is not unique to the UK. Crothall, Callin and Härtel (1997) suggest that few Australian universities have comprehensive guidelines for recruitment and retention and there is little systematic training of the personnel involved in the selection process38.

24

academics compared poorly with all other countries except for Canada. The problem with using Provan (2001) as a guide to policy on the influences of international disparities in academic salaries on recruitment and retention in UK higher education is that it excludes two major destinations of UK academics: the US and mainland Europe (Table 4.2).

The most recent ACU study (Maxwell and Murphy, 2003) looked at staff at a sample of forty-five institutions in seven countries: Australia, Canada, Malaysia, New Zealand, Singapore, South Africa and the United Kingdom. This study found that the UK still compared poorly with the other countries (except Malaysia) once one accounted for differences in purchasing power. The highest salaries at all levels were to be found in Singapore. However, the study found that academic salaries in the UK were more similar to those in other countries – in particular Australia, Canada, New Zealand, Malaysia and South Africa – than in earlier studies (e.g. Provan, 2001). The authors put this growing similarity of academic wages down to ‘growing awareness of competition and of developing salary standardisation for institutions that draw on the same academic pool for their teaching staff’ (p. 3). For UK policy purposes, there are two main problems with this study. First, as with Provan (2001), there are no institutions from the US or the rest of Europe, two of the most popular destinations for emigrating UK academics. Second, the method used to account for differences in purchasing power in the different countries is open to criticism. Maxwell and Murphy (2003) use the Economist’s ‘Big Mac Index’, an index of purchasing power created by comparing the prices of McDonalds’ hamburgers in different countries. This is appealing because it is a commodity that is identical across the world and thus differences in prices are likely to reflect differences in production between countries. Whilst in some respects it may be preferable to using exchange rates, which are volatile and are often dominated by short-run factors such as currency speculation, the Big Mac Index is a long way short of an appropriate measure of purchasing power. Indeed, in the case of the ACU studies, the use of the Big Mac Index by Maxwell and Murphy (2003) appears to be a step back from the association’s earlier work (Provan, 2001). The difference in their results with those of Provan (2001) may merely be due to the change in the conversion rates, as the authors acknowledge.

Ong and Mitchell (2000) provide a similar study of institutions from Australia, Canada, Hong Kong, New Zealand, Singapore, South Africa, the UK and USA. The data come from academic associations and unions and individual countries. The authors do not provide any information on how many universities were sampled, or what the coverage of the academic associations and unions is. Therefore, it is uncertain whether figures are for the whole of the sector, all of those in the sector who are members of a particular union or association, or whether they refer to a sample of individual universities. They find that the real salaries of UK academics are lower than staff in all of the other countries in the study, with the exception of South Africa. Although this study does contain information on the US, it still does not provide any European comparisons. Another major problem with Ong and Mitchell (2000) is their choice of conversion rates. Like Maxwell and Murphy (2003), Ong and Mitchell (2000) use the Big Mac Index. We compare salaries in a number of different countries using a more informative price index in Chapter 4 of this report.

In his comment on Machin and Oswald (2000), Freeman (2000) compares UK with US academic economists and finds that starting salaries in the UK are between half and two thirds of those in the US in purchasing power parity dollars. Machin and Oswald (2000) show how not only are the salaries of US academic economists higher

25

than their UK counterparts, but also that they are much higher than their colleagues in all other subjects, except computing and IT. The table (quoted from the US Faculty Salary Survey 1995-96) shows clearly that in the US, where wages are allowed to vary in order to equate supply and demand, the top salaries are paid to academics mainly in subjects that are suffering recruitment and retention difficulties in the UK, the top six being: computing and IT, economics, chemistry, astronomy, physics and biology. More evidence of the responsiveness of academic salaries to market forces in the US is provided by Graves, Marchand and Sexton (2002), who found that salaries of economics academics increased with the number of journal articles published. However, not all universities in the US appear to have the capacity to respond to market forces in their pay policies. Over the past twenty-five years, wages at public universities have grown much slower than those at private universities. Zoghi (2003) argues that this is due to the increased financial pressure in public universities. This has not been offset by a corresponding increase in non-pecuniary benefits; indeed higher-paying institutions (in particular private ones) also offer higher levels of benefits (Zoghi, 2003).

2.4.5.5 Pay discrimination

One particular aspect of pay within the HE sector that may have an impact upon recruitment and retention difficulties is that of pay discrimination. For economists, Blackaby and Frank (2000) found that, after controlling for individual and workplace effects, black and Indian economists suffer a 7.6 per cent earnings gap with whites39. The ethnic wage gap is similar to that reported in Blackaby et al. (2002) for the economy as a whole (10 per cent). Blackaby and Frank (2000) do, however, find an insignificant effect of ethnicity on grade. Thus not only can the lower wages of ethnic minorities not be explained by their personal or institutional characteristics but neither can it be put down to their lower grades.

For women economists, Blackaby and Frank (2000) found that, relative to unmarried men, married women suffered an earnings gap of 9.1 per cent and unmarried women a 14.1 per cent gap. The gap for women may in part be explainable in terms of their lower promotion probability, although this effect was only significant for unmarried women. This result is particularly interesting, as unmarried women are less likely to have time out of their career (Dearden et al., 2003). Work on a group of Scottish academics from a broader range of subjects (Ward, 2001) also points to a gender salary gap of 15 per cent40. However, when the sample is divided by job (researchers, lecturers, senior lecturers and professors), Ward (2001) found no within-job wage discrimination with respect to gender, after taking into account a range of personal and employment characteristics, except for researchers (for whom there was an unexplained 10 per cent wage gap between women and men). This suggests that the pay gap derives from gender differences in promotion.

However, whether pay discrimination exists or not, its effects on recruitment and retention would depend on comparative discrimination (or perceived comparative discrimination) and opportunities elsewhere in the labour market: what is important is whether academic employment is any different from alternative employment. If 39 The study of Blackaby and Frank (2000) is based on a survey of individuals in UK economics departments. The survey achieved 516 responses from 1,600 questionnaires. 40 Ward’s (2001) study is based on a survey of 900 academics in five Scottish universities, although the results refer to a maximum of 723 respondents. However, the response rate is not given.

26

academic employment is no more discriminatory than alternatives (or perceived to be no more discriminatory), there is no reason that discrimination should increase exit from the sector to other jobs. However, it may reduce labour market participation, which is likely to affect women with partners almost exclusively.

2.4.6 Factors affecting recruitment and retention: pensions It has been argued that the pension schemes enjoyed by academic staff are not

as good as those in the private sector because they involve higher employee contributions (Appendix F, Bett, 1999). However, in light of recent questions that have been raised about the solvency of many pension schemes this may have changed. This is particularly true given the high employer contributions into the USS which have kept it relatively solvent.

2.4.7 Factors affecting recruitment and retention: job content The nature of the job is an important influence on employee satisfaction and

hence the balance of the positive and negative factors will impact upon the ability of the HEI to retain current and to attract new staff. It is certainly believed that academics experience substantial intrinsic job satisfaction (Oshagbemi, 1996; Bryson and Barnes, 2000a; Ward and Sloane, 2000), although our comparison with satisfaction across all jobs (Section 4.2) suggests that intrinsic job satisfaction may actually be lower than average.

Academic work in the HE sector is made up of a blend of three elements: teaching, research and administration/management. Some academics may not have to undertake all of these responsibilities41, but most, to a greater or lesser extent, do. The balance of these three elements is important to the satisfaction of the workforce and hence turnover (Court, 1999). Some staff may not wish to undertake all three of these tasks and forcing all academic staff to undertake all three will push dissatisfied staff out of the sector. The Association of University Teachers has said that ‘There is no reason why staff should not, as an informed career choice, concentrate substantially in any one of the three major components of academic work’ (p.10, AUT, 1995). Because of this varying mix of teaching, research and administration, some factors will affect some members of staff more than others. For example, changes in student numbers may directly affect only those who teach (by increasing teaching hours and marking), but the effects may be felt indirectly by non-teaching staff as pressure is exerted on them to expand their role to include teaching. Likewise, those that do not undertake research may find that they are affected by things like the RAE if they find that their likelihood of promotion is limited by their lack of research output. Anything that places too heavy (or too small) an emphasis on one area of working may have detrimental effects on recruitment and retention, e.g. the perception that promotions depend primarily on research output (Court, 1999). Moreover, it appears that the positive aspects of academic life are being squeezed by burdens in areas which staff dislike, such as bureaucratic administration (Bryson and Barnes, 2000a). Ceteris paribus, these changes should result in higher turnover.

41 Those who only undertake administrative roles we exclude from this review.

27

2.4.7.1 Teaching

Oshagbemi (1996) found widespread satisfaction amongst UK academics with the teaching element of their job (80 per cent satisfied and only 13 per cent dissatisfied). The factors contributing to this satisfaction were course content (usually selected by the academic), whereas class size and teaching load could detract from satisfaction.

2.4.7.2 Research

Oshagbemi (1996) found that around two-thirds of UK university academics were satisfied with the research aspect of their job (65 per cent satisfied, with 27 per cent dissatisfied). This satisfaction was greater amongst respondents from pre-1992 universities. Gray et al. (2001) examined the importance to satisfaction not of research, but of its applicability. They conducted an analysis of industry-university research centres in the US and found that members’ satisfaction with the relevance of research is positively related to staff retention.

However, it has been suggested that the RAE has changed the job content, through changing the balance of teaching and research amongst academic staff, as teaching has been reallocated to non-research active staff, particularly in the RAE highest-rated departments (Heap, 1999, referred to in PREST, 2000). Moreover, the House of Commons Science and Technology Committee (2000a) found that the RAE could have a very damaging effect on morale, through the classification of some staff as ‘research inactive’.

If research does indeed represent a positive, compensating aspect of academic work, then if any group of staff has less opportunity to conduct research this is likely to lead to lower levels of job satisfaction and problems with staff retention. Blake and La Ville (2000) find that although men and women are equally successful in obtaining research grants, women are less likely to apply. This does not appear to be due to differences in inclination in women, but rather their under-representation at the top of the academic career ladder and over-representation in part-time and fixed term contracts.

2.4.7.3 Administration and management

Oshagbemi (1996) found relatively high levels of dissatisfaction amongst UK academics with administration and management (37 per cent dissatisfied and 40 per cent satisfied). The belief that too much emphasis is placed on administration appears to be much worse in new universities, whereas in the old universities the stress on research over teaching creates more dissatisfaction (Court, 1999).

2.4.8 Factors affecting recruitment and retention: promotion and progression The literature has identified a number of factors related to promotion and

career progression which may affect recruitment and retention.

2.4.8.1 Changes in career path

The increasing use of short-term contracts, particularly in research appears to have affected career progression, at least in science and technology, where the traditional path was from PhD to one or two post-doctoral fixed-term research posts

28

and thence to a permanent teaching post (House of Commons Science and Technology Committee, 2002b). The House of Commons Science and Technology Committee (2002b) found that, increasingly, the academics were getting stuck in fixed-term research posts and were unable to make the final step to a permanent post. Whilst the Committee identify a number of advantages of the use of contract research posts, they also conclude that it leads to much higher turnover and exit from the sector.

2.4.8.2 Internal promotion

There is evidence that promotion practice in HEIs may affect retention. Certainly, in a study of Nigerian HEIs, Mallam (1994) found that the second most influential factor on voluntary turnover was the opportunity for promotion42. Poor promotion opportunities for academics were mentioned as a reason for staff leaving institutions in the survey of HR departments in UCEA (2002). However, this does not mean that these staff left the sector; staff often left to join institutions that were larger, had higher reputations or better research opportunities. Part of the reason for poor promotions leading to dissatisfaction may be that many institutions have used the promotion of incumbent staff or the placement of incoming staff on too high a grade to overcome the low pay in the sector (Bett, 1999). The Bett report concluded that this may be creating a time bomb for the future as staff hit the upper limits of pay scales at later stages of their career.

Court (1999) found widespread dissatisfaction with promotion prospects amongst university academics: two-thirds of respondents in a survey of teaching staff disagreed with the statement ‘I have satisfactory promotion prospects at my current institution’, with disagreement highest in the pre-1992 universities and amongst senior staff. Given possible biases in the sample43, this suggests, at minimum, a substantial minority of teaching staff were not satisfied with their promotion prospects within their institution. Oshagbemi (1996) also found substantial dissatisfaction amongst UK university teachers with promotion (50 per cent dissatisfied and 26 per cent satisfied). More recently, the interviews conducted as part of HEFCE (2003) also noted dissatisfaction with promotion criteria, particularly the lack of transparency and the fact that part-time and fixed-contract staff were excluded from staff development and promotion opportunities.

2.4.8.3 Promotion criteria

It is not merely the lack of promotions themselves which are seen as a problem by academics, but the criteria on which they are based (Oshagbemi, 1996). There appears to be a widespread view that research is weighted too heavily as a criterion by which promotion is awarded (Court, 1999; Coe and Boddington, 2003). Although it is widely believed that research is one of the positive aspects of working as an academic in higher education, many staff feel that, partly as a result of the RAE, HEIs concentrate too heavily on research as a criterion for recruitment and internal 42 The most important factor was pay. The other factors (in declining order of importance) were: supervision, the work on the present job, the job in general, co-workers, and the commitment of the respondents to the institution at which they worked. 43 The survey was of teaching staff who were AUT members with a response rate (by our estimation) of about 35 per cent. The profile of respondents, with a few exceptions, was similar to that of the population.

29

promotion (Court, 1999). This will dissuade individuals who do not wish to pursue a research-led career, or who miss out on research opportunities because they are working part time (Coe and Boddington, 2003). Court (1999), based on a survey of teaching staff, and drawing on Halsey (1995), Henkel and Kogan (1996) and Fulton (1993), concluded that academics’ careers were strongly framed from the beginning by the RAE and that research excellence strongly affected recruitment and progression. This marked some change in emphasis (from teaching to research), particularly for the post-1992 universities (Henkel and Kogan, 1996).

In his survey of academics from 23 universities in the UK, Oshagbemi (1996) notes that the reason respondents were dissatisfied with promotions policy were the bias in favour of quantity instead of quality of publications, the relative neglect of teaching and administrative responsibilities, the lack of vacancies at professorial level and the lack of clearly stated promotions policies. In his study of the views of UK academics on their career opportunities44, Court (1999) finds that over half of the respondents believe that promotion at their institution places too much emphasis on research and a quarter too much emphasis on administration, whilst few thought that too much emphasis was placed on teaching45. However, his findings suggest that the emphasis on research excellence caused dissatisfaction amongst a substantial percentage of teaching staff. The dissatisfaction with the research bias of promotions policies was stronger in the old universities than in the new ones, where dissatisfaction was rather with the emphasis on administration46.

2.4.8.4 The RAE

As well as affecting promotion criteria, the RAE has been thought to have affected recruitment and retention, through its effect on satisfaction and morale and through increased competition for academics who will add to the RAE score.

With respect to morale, the House of Commons Science and Technology Committee (2002a) found that:

‘41. The RAE can have a very damaging effect on staff morale. Being labelled 'research inactive' for tactical reasons can blight research careers, and even bring them to an end.’ … It is clear that the RAE has had a negative effect on university staff morale.’

Henkel and Kogan (1996) found that the RAE affected departmental pecking orders, which is likely to lead to a change in satisfaction, although whether to a net increase or decrease in satisfaction is not possible to predict, a priori. Court (1999) also found evidence of a minority of university teachers believing that normal recruitment and promotion procedures were circumvented in order to ensure that ‘stars’ were recruited or retained to boost ratings for the RAE. Whilst Court did not investigate the extent to which this caused dissatisfaction amongst existing staff, it may be a cause of dissatisfaction amongst some.

44 The survey is based on replies from 586 of 2000 staff contacted from AUT records and so may not be representative of the population as a whole. 45 The phrasing of his questions does not allow the opposite to be deduced, that those saying there was not too much emphasis believed the emphasis was correct. 46 Both of these differences were statistically significant using the Mann-Whitney test.

30

PREST (2000) also concluded that the RAE had increased turnover, but with movement highest amongst those who were research active (PREST, 2000). PREST (2000) concluded that the RAE had led to institutions buying staff to boost their RAE score. However, such movement was rare and the significance of the ‘transfer market’ had been exaggerated (and the ‘one-year rule’, whereby staff who move in the year prior to the RAE may be cited by both institutions, should have reduced this). Analysis of the RAE database by HEFCE (2000) confirmed the fact that there was an increase in staff moving in the year immediately before the RAE, but that such movement was rare. HEFCE (2000) also noted that the percentage of research active staff moving to another institution in the year after the RAE was ‘significantly higher than the two years before it’ (Annex J, p. 4). This suggests that any incentives to retain research active staff for the RAE only had a temporary effect on their retention.

2.4.8.5 Gender

Women are particularly under-represented at the higher levels in the HE sector. It is widely believed that this is not just a cohort effect (Dearden et al., 2003), but that women face particular difficulty in obtaining promotion. Women are certainly less likely to be promoted than men in some areas, such as economics (Booth et al., 2000). Blackaby and Frank (2000) identified a lower promotion probability for unmarried women (compared with unmarried men), although surprisingly, did not find this for married women.

A number of potential causes of this disparity have been identified.

First, the importance of research output as a criterion for progression may reduce women’s career progression. This is, in part, because women are more likely to undertake more teaching and pastoral work at the expense of the research (House of Commons Science and Technology Committee, 2002a) and, in part, because of career breaks (PREST, 2000), particularly as childbearing usually coincides with the most productive period of an academic’s life in terms of research output (Dearden et al, 2003)47. The RAE is thought to have exacerbated this as a problem (House of Commons Science and Technology Committee, 2002a).

Second, the lower mobility of women reduces their opportunity to obtain a higher ranked job at another HEI (Booth et al., 2003; Coe and Boddington, 2003). This lack of mobility relates to what is called the ‘loyal servant hypothesis’ (Booth et al., 2003) and is suggested by Ward (2001) as a potential explanation, although it has not been rigorously tested. According to this hypothesis, women are less mobile than men because in married or cohabiting households where both individuals work, the woman tends to apply for jobs in the region where the husband works. This will cause women to apply for fewer jobs than men and to be less likely to leave their current employment.

Third, some subjects and/or grades are not (or have not been considered) amenable to part-time employment or flexible working hours (Coe and Boddington, 2003). The age structures of staff may inhibit promotion and so be linked with discrimination against women who have had career breaks for child rearing (HEFCE,

47 The prevalence of career breaks is not known. However, in a survey of HEI academics in the US, 4 per cent of men reported interrupting their career for health or family reasons compared to 25 per cent of women (Hagedorn and Sax, 1998).

31

2003). Fourth, male senior staff may implicitly or explicitly wish to select staff in their own image (ibid).

2.4.8.6 Ethnicity

Compared with academics as a whole, British ethnic minorities are concentrated in research and fixed-term contract jobs and are under-represented in lecturing posts and at senior levels (Carter et al., 1999). However, British ethnic minorities in academia have a younger age profile than British white staff (with 73 per cent aged 45 and under, compared with 55 per cent of British white staff) and it is unclear the extent to which this accounts for their under-representation in more senior jobs48. (Note that Blackaby and Frank, 2000, did not find ethnicity had an effect on grade for economists, although this study included non-British staff.) Certainly, in a survey of discrimination and racism in academia, 18 per cent of British ethnic minority staff considered they had experienced racism in promotion (Carter et al., 1999).

2.4.9 Factors affecting recruitment and retention: workload and hours of work One benefit of working in academia has been the degree of autonomy of the

hours and pattern of working. Recent evidence suggests that, at least with regard to the number of hours worked, this is no longer (if it were ever) the case. Indeed, workload causes greater stress amongst academics than amongst people in a range of other public sector jobs (University of Plymouth, 2003). Stress is particularly high in this respect in the new universities.

The Hay study reported in Bett (1999) found that, on average, academic staff spent 47-48 hours a week working49. This is more than the national average of 40.5 hours a week (p. F80, Bett, 1999), but similar to that of managers. (The Hay study notes that over 78 per cent of managers, nationally, worked more than 40 hours a week and 34 per cent more than 50 hours; source: Worrall and Cooper, 1998). Moreover, whilst over half of managers, nationally, worked in the evening and a third at weekends, academics had flexibility over their work patterns and this leads Hay to conclude that there was ‘no reason to regard higher education work as radically more or less demanding than work in the rest of the economy’ (p. F80, Bett, 1999). Indeed, they argued that the flexibility academics had in organising their work patterns compensated for the extra hours worked. Note however that there are negative aspects to this increased ‘flexibility’. In a study of stress and work-life balance in academic staff, Kinman and Jones (2004) found some worrying issues concerning work-life balance50. Their results suggest that the boundaries between home and work in the life of the academic are wafer-thin, particularly for the 20% of those polled who lived with another academic. On average, a quarter of academics’ work is done at home and around 10% of academics check their email five times a day at home. Whilst many people appreciate this flexibility in working, the blurring of the lines between home

48 Carter et al. (1999) report grade differences as remaining after adjusting for age. However, the adjustment is made using simple cross-tabulations. It would be useful to conduct a fuller multivariate analysis. 49 The Hay study includes both full-time and part-time staff, although less than 5 per cent of the jobs were part-time. 50 The study was based on 1,108 replies to 5,000 questionnaires sent to academic and academic related staff at UK HEIs who were AUT members, a response rate of 22%.

32

and work may lead to stress particularly when those hours are in excess of a full working week in the university.

The pattern of hours varies by activity. Those with teaching and research responsibilities had the longest hours, averaging 49.9 hours, whereas those with teaching only or research only averaged 45.3 hours and 44.4 hours (Bryson and Barnes, 2000a). Academics in new universities worked on average two hours fewer than those in old universities, in teaching and teaching and research posts, although slightly more in research only posts (ibid).

These long hours may be a factor in the under-representation of women in HE, although this must be offset against the greater flexibility of hours (Appendix F, Bett, 1999). At the same time, it is not clear the extent to which women may reduce these hours through working part-time. Only 13 per cent of HESA recorded jobs were part-time. Although this is low compared with employees across all employment (of whom 24 per cent are part-time), part-time working for academics in HE is slightly higher than for all employees with a higher degree (11 per cent) (LFS Spring 2000). Moreover, the percentage of academics working part-time is under-reported due to the exclusion from the HESA data of those working less than 25 per cent full-time. At the same time, men comprised a much higher percentage of part-time employees than is the norm amongst employees with higher degrees (43 per cent and 28 per cent respectively). It is unclear whether the low level of part-time employment is due to lack of availability or lack of demand; if the former, then those wanting to reduce their hours seem more likely either to continue full-time or leave the labour force, rather than move to a different sector. Certainly, Dearden et al. (2003) suggest that although women may have the opportunity to reduce the hours of an existing job if required by family circumstances, this does not necessarily mean that they will be able to find a new part-time opportunity. It would be worthwhile to check the availability of part-time working for those who want it (and also to explore further the role of part-time work in male academics’ careers).

Moreover, the variation in the proportion of women across departments may be due to differing working cultures (Coe and Boddington, 1999). Their qualitative research with academic chemists highlighted concern about the ‘24 hour culture’ in chemistry and laboratory-based subjects in general51.

It is widely held that the workload of academic staff has been increasing. Certainly stress levels do appear to have risen between 1998 and 2004 (Kinman, 1998; Kinman and Jones, 2004). If so then we would expect this to increase staff exit from the sector. Certainly evidence from Kinman (1998) and Kinman and Jones (2004) suggest an increase in staff who had considered leaving higher education, from 44% in 1998 to 47% in 2004. However, evidence from Australia runs counter to this. The university system in Australia has undergone similar change to that of the UK. There has been a dramatic increase in student numbers, a reduction in per student funding and increased demand for accountability (Bellamy et al., 2003). Coaldrake and Stedman (1999) concluded that as a result of these changes there was growing pressure on time, workload and morale. McInnes (1999) noted an increase in stress and a drop in the general level of job satisfaction for Australian academics, but that 51 As part of their study Coe and Boddington (1999) conducted four discussion groups with staff in HE and two with graduates working outside HE. However, the total number of participants was only 42. Coe and Boddington (2003) conducted individual interviews with 35 women in seven chemistry departments.

33

this had not led to an increase in academics seeking employment outside the sector. The reasons for this were explored by Bellamy et al. (2003). These seem to suggest a very strong attachment to academic employment.

2.4.10 Factors affecting recruitment and retention: security of employment Job security has declined in HEIs, with the abolition of tenure and an

increasing use of temporary contracts. The abolition of tenure (by the Education Act 1988) affected permanent lecturing staff in the old universities. The growth in fixed-term contracts has been largely in research (and, therefore, mainly affecting old universities where most research posts are located). In the old universities, fixed-term contract research staff comprised 15 per cent of academic staff in 1978, rising to 28 per cent in 1988 and to 40 per cent in 1997 (Bryson, 1999). Furthermore, Husbands and Davies (2000) have noted the growing resort to using part-time adjunct teachers for the performance of conventional teaching.

Although job security is an issue for permanent staff (indeed, it leads to much higher levels of stress amongst academics than amongst staff in a range of other public sector employment, University of Plymouth, 2003), the literature has concentrated on this issue for temporary staff, particularly, for contract staff52, and so temporary staff have to be the focus of the rest of the section. As two-thirds of academic staff on fixed-term contracts are research staff (Table 3.8), it may be assumed that, where literature does not explicitly state otherwise, that the findings either refer to or are dominated by the experience of fixed-term contract researchers.

2.4.10.1 Staff on temporary contracts

The increased use of temporary contracts53 has been a major concern in studies of recruitment and retention and in studies of academics’ job satisfaction. It has often been cited as a reason for retention problems (Bett, 1999; Bryson and Barnes 2000a, b; UCEA, 2002). One in five personnel departments asked in the UCEA (2002) survey mentioned that fixed-term contracts were causing recruitment and retention problems in their institution. There may also be an impact on recruitment to permanent posts via the inflow of research staff (94 per cent of whom were on fixed-term contracts in 2001/02, Table 3.8), where this is often seen as a first step into an academic career, as in IT and computer sciences (EPSRC, 1999) or clinical science (Academy of Medical Sciences, 2000).

Certainly the House of Commons Science and Technology Committee (2002b, Summary) considered there was a major problem:

52 Other temporary employment includes hourly paid employment (mainly used for teaching). 53 The increase in the proportion of total staff on temporary contracts has been driven largely (but not solely) by the increase in numbers of temporary research staff. The percentage of what HEFCE (2002) call ‘academic grades (lecturer and above)’ on temporary contracts rose from 14% in 1995 every year until 1999, when it was 16%, and fell to 15.6% in 2001 (note that all these figures relate to census counts of staff with FTE of 40% or more as of 1 December of the respective year and so will not be the same as those published by HESA; for more information see Appendix A of HEFCE, 2002). The percentage of ‘assistant academics (below lecturer)’ rose from 95.5% in 1995 to 96.2% in 1997, before it fell to 95.3% in 2000. Over this period, the proportion of staff who were academic assistants rose from 28.2% to 29.8%.

34

‘We found widespread dissatisfaction and demoralisation among contract researchers, some of whom have been employed on 20 different contracts in as many years. For many researchers there is no career structure and little hope of obtaining a permanent position…. Many researchers are either new in position or searching for their next contract. Research is left unfinished or unpublished…..’

Staff on temporary contracts have suffered a double disadvantage in respect of security. As well as their employment only being guaranteed until the end of their contract, they have also had fewer employment rights (HEFCE, 2003). Most HEIs have required them to waive their rights to redundancy pay (Bryson, 1999). Protection against unfair dismissal has also been waived. However, the EU Fixed-term Directive is likely to affect the use of such employment by HEIs (HEFCE, 2003)54.

2.4.10.2 The pattern of use of temporary contracts

The staff survey conducted as part of Bett (1999) found that 56 per cent of clinical academics and 34 per cent of non-clinical academics were on fixed-term contracts. (These figures do not include researchers55.) The use of such contracts was much higher in old as opposed to new universities (18 per cent compared to 8 per cent). It is in research, however, where fixed-term contracts were in most use, with some 93 per cent of research staff being on fixed-term contracts. One result of the use of fixed-term contracts has been that such staff leave before the end of the contract (Bett, 1999). This is entirely rational behaviour if individuals have any doubts as to whether their contract will be renewed and something we investigate further in Chapter 7 and Appendix G.

Women are more likely to be on fixed-term contracts than are men (House of Commons Science and Technology Committee, 2002b). It is unclear whether this is due to the increasing number of female entrants to the sector (since fixed-term contracts tend to be held at an early stage of the academic career) or difficulties in securing permanent jobs (House of Commons Science and Technology Committee, 2002b). However, the House of Commons Science and Technology Committee (2002b) identified other factors which might result in women being more highly represented on fixed-term contracts (and with the additional consequence that women would be more likely to leave academia). These include direct discrimination in recruitment to permanent posts, difficulties of return after maternity leave or a career break, less control over job mobility, leading to less choice over job, women’s academic achievement being rated less highly.

Husbands and Davies (2000) have noted the growing resort to using part-time temporary teachers for the performance of conventional teaching. It is difficult to obtain a true figure for this because HESA do not collect data on individuals who are less than 25 per cent FTE (full-time equivalent) at some point during the academic year, or on a casual basis. However, according to the Bett report, a third of academics work part-time, with the figure for new universities (43.9 per cent) being roughly

54 The implementation of the Directive was delayed in the UK (Planned implementation was July 2001), but has now introduced the Fixed-term Employees (Prevention of Less Favourable Treatment) Regulations 2002. 55 They therefore indicate a much higher use of contracts amongst lecturing staff than do the HESA data (see Table 2.1). It is unclear which is more accurate.

35

twice that for old (21.7 per cent), a finding mirrored in the survey of Bryson and Barnes (2000a). Note that the use of part-time staff does have positive aspects as it may allow women who would otherwise leave the labour market to care for children to remain in employment.

2.4.10.3 Findings from the study of contract research staff in Scotland

A recent study of the careers of academic contract research staff in Scotland was conducted by Hasluck et al. (2001). Its most pertinent findings to the current study are highlighted below.

In respect of the pattern of employment and careers, the survey found that 44 per cent of academic contract research staff had had more than five years experience in research employment. Most were concentrated in the lower grades: 79 per cent on Research grades IA and IB, with a further 10 per cent on Grade II. Interestingly, those on the higher grades tended to have had shorter research careers, and Hasluck et al. (2001) suggested they had probably taken fixed-term research contracts at a senior grade from employment outside higher education. This suggests that promotion within academic contract research is rare. Seventy-eight per cent of academic contract research staff were aged 25 to 39.

Research interest is the main factor affecting the decision to enter academic contract research and remains of supreme importance. The main reason that people entered academic contract research staff was interest in research. The career intentions prior to entering were, primarily a research career: 66 per cent entered because they wished to pursue a research career and only 23 per cent had entered seeing it as a stepping stone to academic lecturing. However, 24 per cent had seen their initial contract research job as a stepping stone to a professional career.

Once employed, research interest remains overwhelmingly important, with interesting and challenging work very important to 76 per cent of academic contract research staff, (along with other aspects of research interest, such as innovative work, time to pursue research interests) (Table 2.7). However, although only two per cent of academic contract research staff said their job did not provide this at all, fewer than half (44 per cent) said that their jobs were extremely interesting and challenging, suggesting potential dissatisfaction56. The next most important issue (amongst the attributes examined) was security of employment, very important to 38 per cent of academic contract research staff. At the same time, 36 per cent said this was not provided by their job at all and, as only one per cent said security was not at all important, it is likely that this was a cause of dissatisfaction amongst many.

56 Hasluck et al. (2001) identified the importance to contract researchers of a range of job attributes, together with respondents’ assessment of how much these attributes were provided in their job. Unfortunately the findings are not reported fully, reducing the conclusions which may be drawn by the report reader.

36

Table 2.7 Contract researchers: value of job attributes

very important not provided by job at all*, **

interesting and challenging work 76 2 the chance to do innovative work 42 4 job security 38 36

continual skill development 34 7 work making a positive contribution to society 32 4 time to pursue my own research interests 31 16 time to pursue leisure interests 26* 12 opportunities to take managerial responsibility 25* 33 autonomy 24* 5

competitive salary 24 16 opportunities for promotion 24 37 opportunities for international career 18* 20 working with people with whom I enjoy socialising 16* 9 competitive work environment 11 13 * Percentage is read from a graph and so may not be quite accurate. ** whether the job provided the attribute was ranked on a 5 point scale. Source: Hasluck et al. (2001)

Other possible areas of dissatisfaction are the opportunities for career progression (very important to 24 per cent and of no importance at all to only two per cent) and the opportunity to take managerial responsibility (very important to 25 per cent). Hasluck et al. (2001) also add pay to this list.

Flexibility of the working environment and the scope for moving between jobs or projects, broadening experience and developing and absorbing new ideas were also identified, in qualitative research, as important attractive attributes of contract research.

Despite some apparent dissatisfaction with the research interest of the job, the main reasons for leaving contract research tended to relate to the contractual status. Of most importance was insecurity. Seventy per cent saw obtaining a permanent job as important and academic contract research was seen as unsustainable as a career in the long term. Despite this, 34 per cent planned to remain in academic research. The results of insecurity can be seen in job search activity and career plans, with 43 per cent actively seeking their next post. The other main reasons for leaving contract research were limited scope for progression, the need for geographical mobility and difficulties developing more specialised knowledge.

Although the long-term career plans of most academic contract research staff were to continue in academia (59 per cent), only 25 per cent saw this as in a traditional academic career, rather than contract research (34 per cent). The main planned careers leaving the sector were professional practice or management (12 per cent) and research in industry or commerce (11 per cent), with independent, public sector or voluntary sector research 6 per cent.

37

In the short term though, 40 per cent sought a further research post in HE as their next job and only 18 per cent of the contract researchers were aiming for a lecturing post in HE. However, 37 per cent wanted a job outside HE.

2.4.11 Factors affecting recruitment and retention: family-friendly practices There is almost no direct evidence relating to the effects of family friendly

policies57. It is generally believed that conditions of work in HE are relatively family-friendly, due to flexibility in working hours and relatively high holiday entitlement. Certainly the Hay Report found that over half of academics spent more than 20 per cent of their time ‘working on other sites, away from the university’ (Appendix F, Bett, 1999). On the other hand, there is little part-time employment, 13 per cent (see Section 4.7), and an ongoing study has identified juggling work/life balance as being more stressful for academics than amongst staff in a range of other public sector employment (University of Plymouth, 2003).

It is more difficult to assess issues such as maternity leave, although in its 2003 letter to local representatives regarding the most recent regulations on maternity and paternity leave (the Employment Act 2002), the AUT noted that ‘it should be noted that some universities provide maternity and paternity leave arrangements equal to, or an improvement on, the new statutory rights set out in regulation’ (AUT, 2003). One advantage is that, when calculating entitlement to maternity pay, years of continuous service often begin when an individual enters service in HE, rather than the institution itself (NATFHE, 2001).

Some interesting findings on the role of family and family-friendly practices come from the United States, where Hagedorn and Sax (1998) found that having dependent children had a negative influence on satisfaction, although marital status and having career interruptions for family or health reasons had no influence58.

2.4.12 Factors affecting recruitment and retention: other The literature provides some elucidation of a number of other factors which

affect job satisfaction, stress at work or might affect recruitment and retention. These include discrimination, work relationships, physical conditions, resources, control and commitment.

Discrimination may reduce recruitment from those groups discriminated against and increase their turnover. Discrimination in relation to pay and promotion has been discussed above, but discrimination may surface in other ways. Certainly, in a survey of discrimination in Higher Education, 23 per cent of British ethnic minority staff reported encountering race discrimination in recruitment and 20 per cent reported experiencing harassment in academia (Carter et al., 1999). Furthermore 32 per cent considered their institution not very or not at all committed to Equal Opportunities. Bagihole (2002) discusses a range of forms which gender discrimination may take and how the high degree of autonomy of staff in universities allows such discrimination to continue.

57 One respondent to the recruitment and retention survey reported in Bett (1999) reported the ‘lack of focus on family-friendly policies’ as an issue. 58 Based on a survey of 59,933 staff from 446 HEIs in the US, with a response rate of 42 per cent.

38

There is mixed evidence on the role of working relationships. Oshagbemi (1996) found that the majority of staff were satisfied with their supervision (52 per cent satisfied and 34 per cent dissatisfied) and with co-workers’ behaviour (70 per cent satisfied and only 17 per cent dissatisfied). However, work relations lead to higher stress amongst academics than amongst staff in a range of other public sector employment (University of Plymouth, 2003).

Oshagbemi (1996) and the University of Plymouth (2003) also provide mixed evidence on physical conditions and working facilities. Oshagbemi (1996) found 57 per cent satisfied and 31 per cent dissatisfied with these, whilst resources and communication lead to higher stress amongst academics than amongst staff in a range of other public sector employment (University of Plymouth, 2003).

In Australia, Bellamy et al. (2003) found autonomy to be an important retention factor (80 per cent of academics said it was an important reason for remaining). However, in the UK, academics appeared to feel similar degrees of control over their job as staff in a range of other public sector employment , although those in new universities experienced more stress in this regard (University of Plymouth, 2003). However, Bryson and Barnes (2000a) found that autonomy, a positive aspect of working in higher education, had been declining. Retention, in particular, is affected by the accuracy of prior expectations of the job. Doing a PhD may provide better knowledge of the job, particularly given the extent to which PhD students get involved in teaching. It would be useful to know the extent to which doing a PhD provides an appropriate induction into academic life and creates useful expectations of such work (and whether other employment does this to a lesser extent). It would also be useful to know whether this differs by discipline and in respect of new and old universities, particularly as most PhD experience will be at old universities. Little evidence exists on this subject, although Golde and Dore (2000) found in their study of US graduate students that graduate study is not enough to prepare students for an academic career59.

One aspect which might encourage turnover is the degree of commitment that staff feel the university has towards them. This is relatively low, compared with that felt by staff in a range of other public sector employment (University of Plymouth, 2003). This is reciprocated by a relatively low commitment by academics to their organisation. At the same time, in Australia, Bellamy et al. (2003) concluded that academic staff remained in the sector (despite deteriorating working conditions and reduced job satisfaction) for the same reasons they join it, because being an academic is a ‘calling’ rather than just a job. Stress may also affect retention. An ongoing study is examining stress amongst HEI staff. This has found that, compared with a range of jobs in the public sector, academic jobs were, overall, less stressful (University of Plymouth, 2003). The study examined specific aspects affecting stress and these have been discussed at appropriate points above. Another study of stress in academia has been done in Australia (Winefield et al., 2002). This study found that around half of the Australian university staff studied were ‘at risk of psychological illness, compared with only 19% of the Australian population overall’. This level is even higher than that found in a study of Australian correctional officers (Dollard et al., 1992). Stress was found to affect middle-ranking academic staff, those with teaching duties and

59 Their study was based on a survey of 9,645 students in eleven disciplines at 28 major US research universities.

39

those in the humanities and social studies the hardest. The strongest influences on psychological strain were found to be job insecurity and work pressure.

Bellamy et al. (2003) also found no evidence that staff (in Australia) remained in academia due to inertia (for those who have been in academia all their life) or that those well established are poorly skilled/suited (or believe themselves to be) to adapt to better paid jobs in other sectors at their stage of career, i.e. they cannot command a high wage outside HE. The paper finds that ‘poor job opportunities outside academia’ is the second least important reason (out of thirteen) for remaining in academia. Sixteen per cent of people cite it as an important reason for remaining.

2.5 Summary and conclusions The extent of recruitment and retention problems faced by universities for

academic staff is not entirely clear. Whilst there has been a growth in reported problems, the nature of difficulties seems to fluctuate. Moreover, turnover is similar to or only slightly higher than that for professional occupations generally. Difficulties appear to be concentrated in subjects where subject-specific knowledge is transferable to other sectors. This means that difficulties fluctuate with changes in economic activity in the relevant sectors. However, there was evidence that recruitment problems are being masked by universities being able to recruit but to a lower quality, i.e. there is a gradual decline in quality of academic staff. Thus, we would conclude that there are recruitment problems and that these vary by subject and seniority, but that problems are not severe. Whatever the current degree of difficulty, expansion of the sector (unless accompanied by a downturn in the economy) would increase any difficulties.

Previous research suggests that pay is a major factor affecting the recruitment of academic staff and that pay and job satisfaction are major factors affecting retention. For contract researchers, interest in research is the main reason for entering academic employment. The evidence points to a decline in pay for UK academics relative to pay in other sectors in the UK (including in comparable jobs) and that, internationally, UK pay is relatively low. Using a standard measure of job satisfaction, academics appear to have relatively low levels of satisfaction, with comparatively low levels of satisfaction in respect of promotion and pay, in particular, and also with hours, the work itself (especially administration), job security and relationships with supervisors. The only aspect on which academics compare favourably is in respect of satisfaction with being able to use their own initiative. In terms of job content, academics were most satisfied with teaching.

Discrimination (whether direct or indirect), if greater (or perceived to be greater) in academic jobs than in alternative employment, is likely to exacerbate recruitment and retention problems for the discriminated against groups. There is little robust evidence on discrimination in academic employment (i.e. evidence which takes into account differences in personal and academic characteristics). There is some evidence of both an ethnic and a gender pay gap, with the latter deriving from gender differences in promotion. The reasons advanced for the latter include men focussing more strongly on the career-enhancing aspects of the job (i.e. research), childbearing, married or cohabiting women’s lesser mobility, resistance in some jobs to part-time employment or flexible working hours and direct discrimination.

40

The use of fixed-term contracts, primarily for research staff, has been found to decrease retention in the sector. Whilst many contract researchers remain in the sector for several contracts, many of those who do not gain a permanent job leave due to insecurity.

Other attractions of the job include working-time flexibility and pensions. Detractions of the job include physical conditions and working facilities.

These, and other, issues were investigated further in the qualitative research at case study universities, through the staff and research student surveys and through analysis of relative pay, nationally and internationally and are discussed in subsequent chapters.

41

43

3 The structure of academic employment in Higher Education

3.1 Introduction In 2001/02, 110,094 people worked in academic posts (of at least 25% full-

time equivalent, FTE) in English higher education institutions (excluding those on clinical grades)60, or 88,102 full-time equivalent staff (HESA Individualised Staff Record 2001/02). This chapter provides an overview of the nature of their jobs and who they are.

The next section describes the institutional structure of employment (across new and old universities and Colleges of Higher Education). Key contractual aspects of the job (grading, pay, contractual status, primary employment function and mode of employment) are described in Section 3.3. The chapter then turns to the characteristics of academic employees, describing their gender, age, ethnicity, nationality and qualifications. The final section provides a summary.

Nearly all the evidence in this Chapter is based on analysis of the HESA Individualised Staff Record 2001/02. Throughout the data refer to English HEIs and exclude those on clinical grades. Caveats to the Individualised Staff Record data have been described in Chapter 1, i.e. data actually relate to contracts, rather than individuals (which leads to double-counting of those with more than one post) and contracts which are for less than a 25 per cent full-time equivalent academic job are excluded. Due to this exclusion, the total number of staff is under-reported. (For further information see Appendix C.)

The chapter also makes use of the staff survey data. This has been re-weighted to be representative of non-clinical grade academics, excluding hourly-paid staff in English non-specialist institutions with over 200 academic staff. (For further information see Chapter 1 and Appendix C.)

3.2 Institutional structure

In 1992, Higher Education in the UK underwent a major change, with polytechnics (and some other higher education institutions) converting to university status. Pre-1992 universities are commonly referred to as ‘old universities’ and those converting since 1992 as ‘new universities’. This altered their governance and financing. Despite continuing change in the finance and nature of universities, differences between old and new universities remain (e.g. in terms of research 60 To be precise, there were 110,094 contracts of employment with staff. For reasons of privacy the data supplied by HESA did not include staff identifiers and so, in common with most uses of the staff record (exception to this is HEFCE, 2002, who have access to the identifiers), we cannot exclude multiple contracts. HESA suggest that around three per cent of staff hold multiple contracts. Note that the University of North London was not included in either the staff or student data supplied by HESA because the university has asked that its individual data is not released.

44

intensity, academic status, size and internal governance) and so the term remains useful.

Employment is concentrated in the old universities, which account for almost two thirds of total staff numbers (Table 3.1). The majority of the remainder is made up of employees at new universities. By comparison, Colleges of Higher Education and specialist institutions together employ only around 10 per cent of higher education institution staff61.

Table 3.1 Employment by type of institution

Staff numbers FTE Part-

time Total 110,094 88,102 14,453 HEI Type Old Universities 64,780 52,333 7,086 New Universities 34,239 26,995 5,495 Colleges 4,352 3,585 630 Specialist institutions 6,723 5,188 1,242

• Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

3.3 Job structure In the previous chapter a number of factors were identified which had been

found to affect recruitment (pay) and retention (pay and satisfaction, with satisfaction influenced by progression, job security and the nature of the job, amongst other factors). This section describes some of the key job structures which affect these: the grading structure (affecting progression), the pay structure, contractual status (affecting job security) and primary employment function (main job content). In addition, the mode of employment is described.

3.3.1 Grade structure Universities have a well-defined grade structure, which, below professor, is

closely linked to pay. The structure differs between new and old universities.

In old universities, the teaching grades are lecturer, senior lecturer and professor (with readers on the same scale as senior lecturers) (Table 3.2). In addition, some universities have ‘teaching fellows’, who are not expected to do research. In the case studies teaching fellows were found only in the old universities and could not rise beyond senior lecturer. Research grades run roughly parallel to teaching grades.

61 Colleges of Higher Education, 4% and Specialist Institutions, 6%. In terms of factors affecting recruitment and retention, colleges of Higher Education are likely to be similar to new universities. Specialist Institutions are so classified for financial purposes. They are mainly composed of old universities, research institutions, art, drama and music colleges. They also include colleges of Higher Education and agricultural colleges. About 3.6% of all staff are in Specialist Institutions which are similar to old universities (i.e. where research plays a major role e.g. constituent institutions of London University, Research Institutions).

45

Table 3.2 Salary scales from 1 August 2004

old universities new universities

Non-clinical grade academic staff* Research staff Academic staff * Research staff

Grade Salary Grade Salary Grade Salary Grade Salary

12,887 13,953 14,751 15,699 16,773

A

17,601

18,777 19,460 19,614 20,540 20,540

IB

21,640 21,640

22,507

22,507 23,643 23,643 23,643 23,643 24,820 24,820 24,450 24,450 25,699 25,699 25,432 25,432

26,671 26,671

L E C T

A 27,116 27,116 27,390 27,390

27,989 27,989 28,360

B

28,360

29,128

IA

29,128

L E C T U R E R 29,479

30,363 30,363 30,36331,544 31,544 31,54432,666 32,666 32,364

33,26034,227 34,227 34,22735,883 35,883 35,208

S E N I O R

36,42837,558a 37,558c 37,226

38,14239,114a 39,114c 39,114

L E C T U R E R

B

40,091a

II

40,091c 40,09141,212 41,212 41,212 42,573

42,573 42,059

43,513b 43,513d

P R I N C I P A L

43,037

44,616b 44,616d

P R O F

min

S E N I O R

45,885b

IV min

III

45,885d

* Those on clinical grades in old universities have a separate grade scale. a discretionary points, Lecturer B b discretionary points, Senior Lecturer c discretionary points, Research Grade II d discretionary points, Research Grade III Source: AUT

In new universities, the teaching grades are lecturer, senior lecturer, principal lecturer and professor (with readers on the same scale as principal lecturer). Senior lecturers at new universities are fairly equivalent to lecturers at old universities in terms of pay. Research grades are limited to two, with the highest point similar to the top of the Senior Lecturer scale (or the bottom of the Principal Lecturer scale). Thus, for researchers, the progression opportunities (without moving into lecturing) in new universities are severely curtailed.

The pattern of employment across these grades differs substantially between old and new universities (Table 3.3). ‘Researchers’ (all research grades except UAP

46

research grades III and IV, the senior research grades in old universities) are highly concentrated in old universities and rare in new universities and colleges. Although 29 per cent of all academic staff are ‘researchers’ (all research grades except UAP research grades III and IV), there are relatively few in new universities (nine per cent) and colleges of higher education (three per cent) whereas in old universities they represent the largest group of staff (43 per cent) (Table 3.3)62. Around nine per cent of staff are professors or UAP research grade IV, but this figure is much higher in old universities than other higher education institutions. Other than in old universities, lecturers (including senior lecturers on the PCEF scale) form the largest group of staff.

Table 3.3 Staff grades (%)

Total HEI type n % Old New College Spec

inst Professors; researchers Grade IV (old universities) 9,669 9 12 4 3 6

Senior lecturers & researchers Grade III (old universities); Principal lecturers (new universities)

17,955 16 16 17 12 15

Lecturers (old and new universities); Senior Lecturers (new universities) 39,389 36 23 55 69 42

Other researchers 32,234 29 43 9 3 22 Other grades 10,847 10 6 15 14 15 110,094 100 100 100 100 100

• Table shows the percentage of staff in each grade • Breakdown based on Earl (2001) • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

The grade structure of employment suggests that progression opportunities differ between old and new universities. Focusing on teaching staff and senior researchers (i.e. excluding ‘Researchers’ and ‘other grades’ from Table 3.3), almost one quarter of posts in old universities are professorial (24 per cent), compared with five per cent at new universities (Table 3.4). Similarly, 31 per cent of posts at old universities are Senior lectureships or equivalent, compared with 22 per cent at new universities. Indeed, 72 per cent of teaching and senior research posts at new universities are at the lectureship (and equivalent grade), compared with 45 per cent in old universities.

62 In order to obtain a useful grouping of equivalent grades, we follow the categorisation suggested by HESA (Janet Earl, 2001, ‘Further Guidance on the Staff Individualised Record 1999/00 (C99021)’, KB Article 763, HESA). This classifies the PCEF, UAP, CSCFC, clinical and locally-determined scales into the five groups listed in Table 3.3.

47

Table 3.4 Staff grades lecturing and senior researchers (%)

Total HEI type n % Old New College Spec

inst Professors; researchers Grade IV (old universities)

9,669 14 24 5 4 10

Senior lecturers & researchers Grade III (old universities); Principal lecturers (new universities)

17,955 27 31 22 14 24

Lecturers (old and new universities); Senior Lecturers (new universities)

39,389 59 45 72 82 67

67,013 100 100 100 100 100 • Table shows the percentage of staff in each grade • Breakdown based on Earl (2001) • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

3.3.2 Pay structure Pay has been identified as a major influence on recruitment and retention. To

provide an indication of the pay rates that academics face and the speed of increase, the pay scales for academics (from 1 August 2004) are given in Table 3.2.

Academic staff can join a contributory final salary pension scheme. Old and new universities offer different schemes. However, pension scheme members who move between universities offering the other scheme have the option of remaining in their original scheme.

3.3.2.1 Additional earnings

In addition to their salary, academic staff may have additional earnings from, for example, external work or royalties. Each university has its own rules on the amount of additional work which may be undertaken and, normally, for work related to university work, some of the earnings will go to the university.

In the NIESR/DfES staff survey, 42 per cent of academic staff reported supplementing their salary with additional earnings in the previous year, although a further 13 per cent did not answer this question (Table 3.5).

The likelihood of receiving additional earnings rose with grade, from 26 per cent of researchers to 69 per cent of professors (Table 3.6). It also varied by subject (Table 3.7). Staff in business and administrative studies were most likely to have additional earnings (61 per cent). Otherwise, generally, staff in other social sciences, arts and humanities and education more often had additional earnings (45 per cent to 54 per cent) than those in maths and sciences. In particular, additional earnings were least common amongst those in the physical sciences, engineering, medicine and dentistry and biological sciences (27 per cent to 34 per cent). This pattern is likely to reflect differences in the grade structure across subjects.

48

Table 3.5 Academic Staff: Gross additional earnings in the previous 12 months (%)

All staff Excluding

missing no additional earnings 50 58 additional earnings 37 42 missing 13 Total 100 100 n (weighted) 2,783 2,422 n (unweighted) 2,779 2,426

• Source: NIESR/DfES Staff survey, 2004

For most staff, additional earnings were, at most, a small addition to their salary: 81 per cent had no additional earnings or earnings of £2,000 or less (Table 3.6)63. Only six per cent reported additional earnings of £5,000 or more in the previous year. The amount rose with grade, but, still, only nine per cent of professors reported receiving more than £5,000 in the previous year.

Table 3.6 Academic staff: Gross additional earnings in the last year, by grade (%)

Professor

Senior Lecturer/

Senior Researcher

Lecturer Researcher Total

No additional earnings 31 49 62 74 58 Additional earnings 69 51 38 26 42 <£1,000 6 9 6 6 7 £1,000 - £2,000 31 22 12 8 16 £2,001 - £5,000 12 9 7 4 7 £5,001 - £10,000 5 4 3 1 3 £10,001 - £20,000 3 2 2 1 2 £20,001 - £30,000 1 0 1 1 1 £30,001 + 0 0 0 1 0 Earnings not known 11 6 7 4 7 Total 100 100 100 100 100 n weighted 293 497 897 625 2,422 n unweighted 276 485 897 653 2,426

• Total includes those whose grade was not fully described. • Table excludes those who do not state whether they have other earnings; these account for 13% of

the total population.. • Source: NIESR/DfES Staff survey, 2004

63 The exact figures should be treated with caution, owing to the 13 per cent who did not report whether they had additional earnings and the further six per cent (of all staff) who did not report the amount of additional earnings. (We do not think it should be assumed that non-responses indicated having additional earnings, nor that those not giving the amount had high earnings.)

49

The amount of additional earnings varied by subject (Table 3.7). However, there were no clear differences between groups of subjects (sciences and humanities etc) nor did the amount follow the same pattern as likelihood of having any additional earnings.

Table 3.7 Academic staff: Gross additional earnings by subject % of those with

additional earnings, earning:

% of total staff earning:

% with other

earnings < £1,000

> £5,000

< £1,000

> £5,000

n (weighted)

n (un-

weighted)

business and administrative studies 61 9 23 138 14 138 130 modern languages 54 29 10 61 5 61 52 education 53 18 11 172 6 172 161 other humanities 52 20 16 175 8 175 142 art and design 52 - - 48 10 48 45 combined studies across subject groups 51 25 13 94 6 94 80 English literature and classics 50 - - 46 4 46 36 architecture, building and planning 49 - - 37 10 37 31 social studies, including economics and social/economic geography 45 9 22 326 10 326 281

mathematical sciences (excluding computing) 42 19 10 85 4 85 91 computing sciences 42 12 29 132 12 132 140 subjects allied to medicine 40 32 16 313 7 313 364 engineering 34 7 17 147 6 147 165 physical sciences 34 19 12 213 4 213 236 medicine and dentistry 27 - - 121 5 121 126 biological sciences 27 21 10 253 3 253 284 Total 42 19 17 2404 7 2404 2409

• Table excludes those who do not state whether they have other earnings; these account for 13% of the total population.

• Table also excludes 8 individuals who do not report their subject area. • Cells with a base of less than 30 (un-weighted) observations are marked with a ‘-‘;agriculture and

related subjects, librarianship and information science and other technology have been excluded due to small cell size for all data. These are included in the total.


3.3.3 Contractual status Job security is an important influence on job satisfaction and the use of fixed-

term contracts has been identified as an important factor causing academics to leave the sector (see above). This section describes the pattern of use of different forms of contract.

50

There are essentially three types of contract: permanent, fixed-term and hourly paid64. Below, the contractual pattern of employment is described using HESA Individualised Staff Record data. It should be remembered that this data excludes jobs which entail less than a 25 per cent full-time teaching or research job (i.e. those whose jobs include a substantial non-teaching and research element and/or work low part-time hours are excluded). Thus jobs with few hours are excluded. This will tend to exclude disproportionately the hourly paid. It is useful to describe briefly the types of jobs this might include. In the case study universities, in addition to lecturing grade staff and contract researchers, universities have ‘visiting’ lecturers (who include guest lecturers making one or two presentations a year to those with more regular work, sometimes covering for vacant posts) and other teaching staff (for example those solely taking tutorials, seminars or demonstrating). They also include stop-gap teaching roles to cover for illness, sabbatical or temporary shortages. The contracts such staff were on differed between case study universities. Hourly-paid or fixed-term contracts tended to be used for small amounts of teaching. Such services were provided by PhD students, researchers and people outside academia, amongst others. For this work, paid hours tended to be based on contact time (i.e. preparation and marking time was unpaid) and, in some cases, the contact time paid was adjusted downwards for teaching which required little preparation or no marking.

Excluding staff on clinical rates, in 2001/02, 41 per cent of English academic staff were on fixed-term contracts, with a further four per cent on hourly rates, Table 3.8. There are distinct differences between the terms of employment for staff whose function is teaching only, research only or both teaching and research. Those who perform both teaching and research are most likely to be on permanent contracts, with 84% on permanent contracts. Research only staff are almost exclusively on fixed-term contracts, with 94% on fixed-term contracts. Those whose only function is teaching appear fairly evenly distributed across contract types, although as many hourly paid/casual staff are teaching only staff and hourly-paid are under-reported in the HESA data, the real distribution may be more skewed towards hourly-paid. The number of fixed-term contract staff had increased by 34 per cent between 1994/95 and 2000/01 (House of Commons Science and Technology Committee, 2002b, Appendix 53).

Table 3.8 Employment activity by contractual terms (%)

Contract Teaching &

research Teaching

only Research

only Total n

Permanent 84 38 6 55 61094 Fixed-term contract 15 28 94 41 45073 Hourly paid/Casual staff 1 34 0 4 3927 Total 100 100 100 100 110,094 % by activity 60 9 31 100 n 65,567 10,329 34,198 110,094

• Source: HESA Individualised Staff Record • Data refer to staff at English HEIs and exclude those on clinical grades

64 In the Individualised Staff Record, ‘fixed-term contracts’ includes rolling fixed-term contracts, but term-time only staff who are on open ended contracts are considered to be permanent.

51

Old universities use more staff on fixed-term contracts. This is likely to be due to the greater use of research-only contract staff (see Table 3.9 below). In new universities and colleges three-quarters of staff are on permanent contracts, whereas almost one-half are on fixed-term contracts in old universities and specialist institutions.

3.3.4 Primary employment function Job content is an important influence on job satisfaction. The Individualised

Staff Record records the primary employment function of staff, which shows whether a member of staff chiefly teaches, conducts research or does both.

The majority of staff perform both teaching and research duties (Table 3.9). This figure is much lower in old universities than it is in the other types of HEI. This is because a large portion of academic staff in old universities – where the majority of research takes place – perform only research. The majority of single-function staff are research only. Specialist institutions have the most even spread of staff. This is partly because this group is the most heterogeneous group of institutions – some are primarily focused on teaching, such as schools of music and art, whereas others conduct a great deal of research, such as medical institutes.

Table 3.9 Breakdown of Staff by Primary Employment Function (%)

HEI Type Total Old New College Spec inst

Teaching & research 60 49 78 87 43 Research only 31 45 10 3 25 Teaching only 9 6 12 10 32 100 100 100 100 100


3.3.5 Mode of employment The availability of part-time employment can be a key factor in recruiting and

retaining staff. Whilst this is most commonly important for mothers with young children, it can be important for all staff, enabling them to choose their work/life balance or to combine more than one job.

Thirteen per cent of academic staff are working part-time65 (Table 3.10). Part-time working is least common in old universities, as is the use of hourly paid/casual staff. Part-time working is also twice as common among staff on fixed term contracts as it is among permanent staff.

65 In the Individualised Staff Record, term-time only staff who work full time during term time are considered to be full time. Therefore, an individual could be 30/52=58% FTE, but still be recorded as full time. Likewise, someone who works two days per week during term time would be 23% FTE.

52

Table 3.10 Mode of Employment (%)

Total HEI Type n % Old New College Spec inst

Full-time 91,202 83 88 76 77 71 Part-time 14,453 13 11 16 14 18 Hourly paid/casual 4,433 4 1 8 9 10 110,094* 100 100 100 100 100

• * Note that this total does not tally as 6 staff are entered as ‘other’. • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

Table 3.11 Mode of Employment by Terms of Employment (%)

Permanent Fixed term contract Total

Full-time 91 79 83 Part-time 9 19 13 Hourly paid/casual 0 2 4 Total 100 100 100


Part-time working is most common among teaching only staff. This, allied to the fact that almost all hourly paid/casual staff are employed to conduct teaching only, suggests that such employment is undertaken to fill in gaps in teaching. Similar proportions of staff whose primary employment is either both teaching and research or research only work part time and almost no staff are employed on an hourly paid/casual basis for these types of jobs.

Table 3.12 Mode of Employment by Employment Function (%)

Teaching

& research

Research only

Teaching only All

Full-time 88 87 40 83 Part-time 12 13 21 13 Hourly paid/casual 1 0 40 4 100 100 100 100


53

3.4 The characteristics of academic staff in Higher Education The personal characteristics of employees may provide pointers towards

factors which have affected recruitment and retention. For example, a high age profile may indicate future problems of a retirement bulge; a low percentage of women may indicate historical discrimination in recruitment, discrimination prompting leaving or a lack of family-friendly working. As these examples suggest, the pattern of employment by characteristics cannot identify factors affecting recruitment and retention, but can suggest issues for further examination. The structure of characteristics of employees can also indicate the types of practices which might be used to improve recruitment and retention, e.g. for retention, family-friendly practices are more important the higher percentage of women employed.

This section describes selected characteristics of academics in higher education institutions, namely, gender, age, ethnicity, nationality and their qualifications.

3.4.1 Gender Women comprise 38 per cent of the academic workforce in higher education

institutions (Table 3.13). This falls to 36 per cent of full-time equivalent (FTE) staff, because of the greater number of women working part time. However, the HESA data may underestimate the percentage of females, due to the exclusion of those working less than 25 per cent FTE.

Table 3.13 Academic staff by gender

per cent FTE

per cent Part-time

per cent Total

n

Men 64 42 62 67,741 Women 36 58 38 42,353 Total 100 100 100 110,094 n 88,102 14,453 110,094


Although women are substantially under-represented in academia (38 per cent) compared with women in the labour force as a whole (46 per cent female), comparison with employees with higher degrees, of whom 41 per cent are women, suggests only a slight under-representation. Women’s under-representation varies across departments66.

Part-time working is twice as common among women as it is men (Table 3.14). One-fifth of women work part time. (This rises to one quarter if one includes hourly paid/casual staff). Indeed, even though women represent 38 per cent of total staff numbers, they represent 58 per cent of part-time academic staff. 66 Figures in Coe and Boddington (1999) show that even within science subjects, the proportion of female employees by cost centre varies from 8% (in electrical, electronic and computer engineering) to 73% (in Nursing and Paramedical Studies) in the old university sector and by a similar amount among new universities.

54

Table 3.14 Mode of Employment by Gender (%)

Men Women Total Full-time 87 76 83 Part-time 9 20 13 Hourly paid/casual 4 5 4 Total 100 100 100


For men, the proportion of staff in part-time employment increases slightly with age until they reach 60, when it more than doubles (Table 3.15). For women, the increase happens when they enter their thirties, and remains relatively constant until they reach their sixties. This is consistent with the view that women use part-time employment to balance work and family commitments.

Table 3.15 Mode of Employment by Age and Gender (%)

<30 30-39 40-49 50-59 60+ Total Men Full-time 90 91 89 87 70 88 Part-time 7 7 8 10 24 9 Hourly paid/casual 3 3 3 3 6 3 100 100 100 100 100 100 Women Full-time 87 76 72 73 58 76 Part-time 10 20 23 22 30 20 Hourly paid/casual 3 4 5 6 12 5 100 100 100 100 100 100

• Table shows the percentage of each age group in each mode of employment • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

Although 38% of staff are female, this proportion varies by subject (Table 3.16). Women represent almost half of the staff in subjects like biological sciences and languages and are even in the majority in subjects allied to medicine. There are many fewer women in subjects like the physical, mathematical and computer sciences, engineering or architecture.

55

Table 3.16 Breakdown of Staff by Subject and Gender

Staff % Female % PhD

Medicine and dentistry 1,249 43 49 Subjects allied to medicine 8,669 62 34 Biological sciences 14,972 46 60 Agriculture and related (in vet) 1,147 37 44 Physical sciences 11,918 19 69 Mathematical sciences 3,645 22 55 Computer science 3,911 24 35 Engineering and technology 8,885 13 50 Architecture, building and planning 1,473 25 22 Social studies 11,024 45 41 Law 4,597 41 33 Business and admin studies 5,254 36 21 Mass comm. & documentation 1,124 49 16 Languages 6,329 50 46 Historical and philosophical studies 5,511 35 60 Creative arts and design 5,267 39 13 Education 5,156 53 18 Combined or general science 1,932 45 27 DK subject 8,031 45 0 Total 110,094 38 41

• Subject relates to cost centre staff returned in • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

Women tend to be concentrated in lower level jobs than men. Men tend to be

in higher grade posts than women, with 31 per cent of men being either professors or senior lecturers and researchers, compared to 15 per cent of women (Table 3.17). Women are much more likely to be on lower lecturer or research grades. Although almost two-thirds of male academics perform both teaching and research, the figure for women is much closer to one-half (Table 3.18). Higher proportions of women are in single-function roles, this is true for both teaching only and research only posts. We have seen above (Table 3.12) that single-function staff are more likely to be on non-permanent contracts and men are also much more likely to be on a permanent contract than women (Table 3.19).

56

Table 3.17 Staff grades (%)

Men Women Total Professors 12 3 9 Senior lecturers & researchers 19 12 16 Lecturers 33 40 36 Researchers 26 34 29 Other grades 9 11 10 100 100 100

• Table shows the percentage of staff in each grade • Breakdown based on Earl (2001) • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

Table 3.18 Breakdown of Staff by Primary Employment Function (%)

Men Women Total

Teaching & research 64 53 60 Research only 28 35 31 Teaching only 8 12 9 100 100 100


Table 3.19 Breakdown of Staff by Terms of Employment (%)

Total HEI Type Men Women Old New College Spec inst

Permanent 60 49 55 45 73 77 55 Fixed-term contract 37 47 41 54 21 14 35 Hourly paid/Casual 3 4 4 1 7 9 10 100 100 100 100 100 100 100


This pattern raises the question of discrimination. Women’s under-representation increases as grades increase (EC, 2000, Laafia and Larsson, 2001). In their work on women in economics, the Committee for Women in Economics (CWE) of the Royal Economics Society found that this was not a cohort effect, as there are fewer women than men even at the lowest grades, suggesting that there may be other barriers to entry and/or problems in HE employment (Dearden et al., 2003). Some of these may be due to discrimination, or Human Resourcing policies, others may be due to the nature of the job itself.

57

3.4.2 Age The age structure of staff in Higher Education is important for a number of

reasons. Some have voiced concerns that there exists a demographic time-bomb, with a large number of staff approaching retirement. It is imperative for policy design to know whether this is indeed the case and, if so, which staff are affected. We can see from Table 3.20 that there does not appear to be a bulge in staff approaching retirement. Male staff are older on average than women (the mean age of men is around 43 years, whereas for women it is around 40). This may be due to a higher leaving rate amongst women. It is almost certainly also a cohort effect, with fewer women entering academia in earlier cohorts. Staff tend to be younger at old universities than the other types of higher education institution. A more sophisticated analysis of the issue was conducted in HEFCE (2002), which concluded that the current age profile ‘will not lead to a marked increase in the overall leaving rates’ (p. 3). Exceptions to this conclusion were mathematics, physics and engineering, where the report said that what was required was an ‘increase in recruitment rates to maintain current numbers’ (ibid.).

Table 3.20 The Age of HE Staff (%)

<30 30-39 40-49 50-59 60+ Men 12 30 25 26 6 100 Women 17 33 29 19 2 100 Part-time 9 29 29 24 9 100 HEI Type Old University 19 35 23 19 5 100 New University 7 25 33 31 4 100 College 4 25 34 32 4 100 Specialist institutions 11 30 29 26 5 100 Total 14 31 27 23 5 100 Total (n) 15,392 33,906 29,135 25,411 5,115 108,959*

• Table shows percentage of staff type in each age group • * Note that 1,135 staff did not report their age • Data refer to staff at English HEIs and exclude those on clinical grades • Source: HESA Individualised Staff Record 2001/2

It is useful to consider the age structure of different groups of staff in order to

gain an understanding of who these staff are. Table 3.21 presents the percentage of each staff type in each age group. From this a potential career path of academics emerges. Fixed-term contracts are generally taken at the beginning of an academic career, with 70% of staff on fixed-term contracts being under forty years of age67.

67 Note that an alternative explanation of the large number of younger staff on fixed-term contracts is that they are a relatively recent phenomenon and staff enter HE on a fixed-term contract career path and will continue to take fixed-term employment throughout their career. However, we do not think this likely.

58

Conversely, three-quarters of permanent staff are forty or over. Given what we already know about the numbers of research staff on fixed-term contracts, it is little surprising to learn that there is a similar pattern to the age structure of research only staff. A third of research only staff are under thirty years of age and almost four-fifths under forty.

Table 3.21 The Age Structure of HE Staff (%)

<30 30-39 40-49 50-59 60+ Total Terms of employment Permanent 2 14 19 19 3 56 Fixed-term contract 12 17 7 4 1 41 Hourly paid/Casual 0 1 1 1 0 3 Total 14 31 27 23 5 100 Primary employment function Teaching only 1 2 3 2 1 9 Research only 11 13 5 2 0 31 Teaching & research 2 16 19 19 4 60 Total 14 31 27 23 5 100

• Table shows the percentage across all staff of each staff type within each age group • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

3.4.3 Ethnicity Fewer than 7% of academic staff were from ethnic minorities, with under half

of the latter British nationals (Carter et al. 1999). As 7.5% of employees with a higher degree are from ethnic minorities, this means that academic employment may mirror that of its labour market. However, only one quarter of people from ethnic minorities who have a higher degree are British nationals (Labour Force Survey, Spring, 2000), suggesting that British ethnic minorities, as a group, are unlikely to be under-represented in academic jobs.

3.4.4 Nationality The ratio of UK to foreign staff is around three to one (Table 3.22). This ratio

is almost identical across gender and so this does not provide any support for the thesis that women are less mobile than men, at least internationally. More foreign staff are employed in old universities and specialist institutions than new universities and colleges. Foreign staff tend to be younger than UK staff (Table 3.23).

59

Table 3.22 Breakdown of Staff by Nationality (%)

HEI Type Men Women Total Old New College Spec inst

UK Staff 77 76 76 73 81 92 77 Foreign Staff 23 24 24 27 19 8 23 Total 100 100 100 100 100 100 100


Table 3.23 The Age of HE Staff by nationality (%)

<30 30-39 40-49 50-59 60+ Total UK Staff 13 27 28 27 5 100 Foreign Staff 19 45 23 11 2 100 Total 14 31 27 23 5 100 Total (n) 15,392 33,906 29,135 25,411 5,115 108,959*

• Table shows percentage of staff type in each age group • * Note that 1,135 staff did not report their age • Data refer to staff at English HEIs and exclude those on clinical grades • Source: HESA Individualised Staff Record 2001/2

3.4.5 Staff Qualifications The most common qualification held by academic staff is the doctorate,

although it is far more common among men than women (Table 3.24). Women are more likely to hold a first or higher degree than men. The proportion of staff with doctorates is the same for the under forties as it is for older staff68, although considering all higher degrees, older staff tend to be more qualified: 67 per cent of staff aged forty years and over have at least a higher degree compared to 62 per cent of younger staff. However, this simple dichotomy may mask the underlying pattern of qualifications in the data. Therefore, we examine this in more detail below.

Table 3.24 Highest qualification of staff (%)

Men Women Under 40 40 plus Total Doctorate 48 31 41 41 41 Higher Degree 21 27 20 26 23 Postgrad diploma 3 5 3 5 4 First Degree 14 21 18 15 16 Other qualification 8 8 9 7 8 No formal academic qualification 1 2 2 1 2 Not known 5 7 7 5 6 Total 100 100 100 100 100


68 Actually, it is a little higher for younger staff, with 41.42% of staff under the age of 40 having a doctorate compared to 41.16% of over forties.

60

We can perhaps gain a little more insight into the qualifications gained by staff of different ages by considering Figure 3.1. This shows a dip in the percentage of staff whose highest qualification was a doctorate and a corresponding increase in the percentage whose highest qualification is a higher degree. It is uncertain whether this is due to average qualifications increasing over time – i.e. older staff did not need a PhD when they started working in the sector – or because staff with doctorates are more likely to leave Higher Education in middle age69. It is possible that the qualifications demanded of staff have risen but that over fifties (particularly in their late fifties) with doctorates are less likely to leave. Another contributing factor may be that entrants at the younger ages require higher qualifications than those entering with other skills and experience (e.g. in vocational subjects). Certainly, the turnover data suggest that staff with doctorates are less likely to leave an institution (4.4%) than those whose highest qualification is a first degree (4.9%) and only slightly more likely to leave than those with masters degrees (4.2%), although we cannot tell how or whether these have changed over time. The staff data are not of good enough quality to calculate the departure rate by both age and highest qualification, but the departure rate for all staff in their forties is in fact the lowest of all age groups70. This suggests that by this time, staff have made their decision whether they wish to remain in the sector. Departure rates are highest among the young and those approaching retirement.

Figure 3.1 Age Profile of Qualifications (% of staff)

0

10

20

30

40

50

60

25 30 35 40 45 50 55 60age

%DoctorateHigher degreeFirst degree

• Data apply to both men and women • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and exclude those on clinical grades

69 There are of course other potential explanations, such as the potential that the age profile varies by subject, as does the qualification profile. 70 Individualised Staff Record 2001/2.

61

Although the percentage of female staff with doctorates is lower than that of men at all ages, the disparity appears to increase over time. Either more women have entered HE with doctorates in recent years or women with doctorates are more likely to leave than similarly qualified men. It may the case that women are more highly represented in subjects where it is less important to have a doctorate71. Staff at new universities are much less likely to have a doctorate than at other institutions.

Figure 3.2 Age Profile of Staff with Doctorates (% of staff)

0

10

20

30

40

50

60

25 30 35 40 45 50 55 60age

%AllWomenNew Universities


Doctorates are much more common among permanent and fixed-term staff

than among hourly paid/casual staff (Table 3.25). It is possible that many of the latter are still studying for their doctorates. However, note that the proportion of staff for whom we do not have information on qualifications is significantly higher in hourly paid/ casual staff than for permanent and fixed-term staff. This is likely to be because records are less rigorously kept for such adjunct staff. Doctorates are also more common among full-time than part-time staff.

71 34% of staff in subjects allied to medicine, for example, hold a doctorate, compared to the average of around 41%.

62

Table 3.25 Qualifications by Terms and Mode of Employment (%)

Terms of employment Mode of employment

Permanent Fixed-term contract

Hourly paid/casual

Full-time

Part-time

Hourly paid/

casual Doctorate 45 39 6 45 26 6 Higher degree 26 20 19 23 29 20 Postgrad diploma 5 2 5 3 6 4 First degree 14 19 19 15 21 20 Other 6 10 6 8 7 6 No academic qualifications 1 2 5 1 2 5

Not known 4 7 41 4 9 39 Total 100 100 100 100 100 100


3.5 Summary

3.5.1 Job structure The university sector has a well-defined grading structure, linked to pay. The

structure differs between new and old universities and there are much greater opportunities for promotion in the latter for those who have reached lecturer status: 25 per cent of teaching and senior research posts are at the professorial grade in old universities (five per cent in new universities) and 31 per cent are at the senior lectureship grade (or equivalent) (22 per cent in new universities).

Posts are divided into teaching and research (the standard lecturing post), research only and teaching only. There is a strong link between role and contract status. 84 per cent of teaching and research posts are permanent, 94 per cent of research only are fixed-term contract posts. The use of casual/hourly paid contracts is less possible to determine, due to the HESA data only recording those working at least 25 per cent full-time equivalent. However, it appears that these contracts are mainly used for teaching only posts (which would include visiting lectureships, people undertaking tutorials and demonstrators, much of which is done by PhD students).

Research posts (and hence fixed-term contracts) are concentrated in old universities (with 45 per cent of academic staff in old universities in research only posts, compared with 10 per cent of those in new universities).

With the caveats on HESA data above, overall just over half (55 per cent) of academic staff are on permanent contracts, 41 per cent on fixed-term contracts and four per cent on hourly/casual.

Thirteen per cent of academic staff work part-time. Part-time working is less common in old universities (11 per cent compared with 16 per cent in new

63

universities). However, it is more common for those on fixed-term contracts (19 per cent compared with nine per cent of permanent posts).

3.5.2 Characteristics of employees Women comprise 38 per cent of academic staff, only a slight under-

representation compared with women with higher degrees in the workforce. However, women are concentrated in lower level jobs: the lower grades, single function (research only or teaching only) and on fixed-term contracts. Previous research (on women in economics) has found that this is not a cohort effect, suggesting possibilities of discrimination.

Part-time working is twice as common amongst female academics (20 per cent) than male (nine per cent). The pattern by age, not surprisingly, suggests that women academics move to part-time working due to motherhood. However, they do not appear to return to full-time academic employment as their children age. Part-time working amongst female academics also rises from the age of 60. The factors leading to part-time employment amongst male academics seem to differ, as men largely move to part-time over the age of 50 and, particularly over the age of 60.

The age structure of academic staff is similar to that of the national workforce and there does not appear to be a retirement bulge. Male staff are slightly older than female, almost certainly a cohort effect, but possibly also due to a differential leaving rate. The age pattern by type of job and tenure suggests the career pattern of young staff entering on fixed-term contracts in research posts, prior to moving into permanent lecturing posts.

Fewer than seven per cent of academic staff were from ethnic minorities, with under half of the latter coming from the UK (Carter et al., 1999). As 7.5% of employees with a higher degree are from ethnic minorities, this means that academic employment may mirror that of its labour market. However, only one quarter of people from ethnic minorities who have a higher degree come from the UK (Labour Force Survey, Spring, 2000), suggesting that UK-born people from ethnic minorities are unlikely to be under-represented in academic jobs.

About three-quarters of academic staff are UK nationals, with similar percentages for women and men. Foreign staff tend to be younger than UK staff. More foreign staff are employed in old universities and specialist institutions than new universities and colleges.

Forty-one per cent of academic staff have doctorates. The percentage is lower for those aged between 40 and 55. The most likely explanation of this is that, over time, there has been an increase in the demand for recruits to have doctorates, that those older recruits are also judged on skills and experience and that those over the age of 55 without a doctorate are more likely to leave. Women are less likely to have a doctorate than men (31 per cent and 41 per cent respectively). Staff at new universities are much less likely to have a doctorate.

64

65

4 Intersectoral and international comparisons of pay

4.1 Introduction Pay has been identified as a major factor affecting the recruitment and

retention of academics (Chapter 2). This chapter explores the relative pay of UK academics (relative to other UK employment and relative to academic employment abroad) to identify whether pay differentials may be a serious driver of shortages. Previous research into international comparisons of academic salaries has been scant and tended to adjust for differences in the cost of living between countries inadequately (see Chapter 2). Our approach adjusts for differences using the OECD purchasing power parity.

In order to get a flavour for the issues, we first consider information on staff flows, using the HESA Individualised Staff Record 2001/02. We expand the focus of our analysis to the whole of the UK in order to retain consistency with the remainder of the chapter72. In section 4.3 we use data from labour force surveys to compare a number of indicators of earnings across nine countries in 2001. These countries have been selected to include the main English–speaking countries to which UK academics move (the USA and Australia), together with other English-speaking nations (New Zealand and Canada) and three European countries (Denmark, France and Sweden). We also have a figure for mean academic wages in Japan, which we include for comparison. Together, these illustrate the types of countries to and from which most international movement with UK academia occurs.

In section 4.4 we take a more sophisticated approach to the comparison of earnings, using multivariate analysis to take into account differences in individual characteristics. The analysis examines comparative earnings between academia and the rest of the UK economy for people with a degree and compares UK and US academic pay. Section 4.5 starts to explore why academic pay may be higher in the US than the UK.

4.2 Flows in and out of the UK Higher Education sector This section considers academic staff flows into and out of the sector using the

HESA Individualised Staff Record. However, before we continue, we must make it clear that this particular part of the Record may not be very reliable. The number of staff reporting their destination on leaving a post and their previous post when

72 Because the LFS is a sample of the population, we increase the focus to the UK as a whole to increase the sample size. We would not expect to find significant differences within the UK. This is because of the similarities in the salary scales and distribution of staff within the UK, particularly in comparison to the other countries.

66

entering is small relative to the number of non-responses73. Any inferences that one draws from these data will only be affected if there exist systematic differences in non-reporting across the responses. Therefore these tables are only intended to provide a pointer to the issues and so we only discuss the previous posts of incoming staff and destinations of leavers relative to each other.

Over half of the staff inflows into the UK Higher Education sector for whom we know their previous activity come from outside the Higher Education sector (Table 4.1)74. The UK student body supplies just over one quarter of new UK Higher Education academic staff. Flows from other countries’ Higher Education sectors represent 11%. The outflows from the sector are broadly similar. Just over 10% go to a higher education institution in another country. One quarter of academics leaving UK Higher Education retire (or die). Almost two-thirds of UK academics leaving UK Higher Education leave the sector entirely. However, the picture is not the same for academics from all nationalities. Few foreign nationals retire from the UK Higher Education sector, suggesting that it is the early part of their career that they spend in the UK. Indeed 68% of foreign academics are under the age of forty, compared to 40% of UK nationals75.

Table 4.1 Flows into and out of UK Higher Education sector, by nationality

UK Europe US Australasia Other Total Inflows Outside of sector 41 5 1 1 11 58 Student at UK HEI 18 3 0 0 6 27 Student at non-UK HEI 0 1 0 0 2 5 EU HEI 0 2 0 0 1 4 US HEI 1 0 1 0 0 2 Other foreign HEI 1 0 0 1 3 5 Total 70 11 2 2 16 100 Outflows Outside of sector 43 8 1 1 12 65 Retirement/death 23 1 0 0 1 25 EU HEI 1 2 0 0 1 4 US HEI 1 0 0 0 1 3 Other foreign HEI 2 0 0 0 2 4 Total 70 10 1 2 16 100

• Source: HESA Individualised Staff Record, 2001/02 • Data refer to staff at UK HEIs and exclude those on clinical grades

73 Around 15.6% of the population left the HESA record in 2001, the destination of nearly half (7.5%) of these is not known. 12.4% were new to the record and it is not known what around a quarter (3.1%) were doing previously. 74 Around 40% of UK staff come from other parts of the UK public sector, and 17% from the private sector. 75 Source: HESA Individualised Staff Record, 2001/2.

67

If we consider academics for whom we have data that leave their institution but not the sector, we find that one quarter of these move overseas (Table 4.2). However, this is mainly made up of non-UK nationals. The US appears to be by far the most popular destination for UK academic migrants, representing the destination of over one third of staff who move to a higher education institution abroad.

Table 4.2 Destination of academics within Higher Education sector, by nationality

Destination UK

national

Non-UK

national Total

Another HEI in UK 87 54 76 HEI in EU 3 20 8 HEI in US 5 9 6 HEI in other overseas country 6 17 10 100 100 100

• Data for staff who report destination as within HE sector outside institution • Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at UK HEIs and exclude those on clinical grades

There does not appear to be a reciprocal movement in the other direction, as most of the recent employees at UK higher education institutions for whom we have data that have not moved from a post at another UK higher education institution are students from UK higher education institutions (Table 4.3). More than two-thirds of the new foreign academics in the UK studied here previously. The figures in these two tables are consistent with the story that foreign students take up academic appointments at UK higher education institutions after their study, but return eventually to their home countries.

Table 4.3 New academics from within Higher Education sector, by nationality

Source UK national

Non-UK national Total

Another HEI in UK 97 0 95 Student in UK 3 69 4 Student from elsewhere 0 31 1 HEI in EU 0 0 0 HEI in US 0 0 0 HEI in other overseas country 0 0 0 100 100 100

• Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at UK HEIs and exclude those on clinical grades

This suggests that there are flows into and out of the UK higher education sector, but there are quite different reasons for the movements of UK and non-UK nationals. The majority of foreign staff studied here and may return to their home countries. Thirteen per cent of British academics who move higher education

68

institution move abroad. More than one-third of these academics move to a single country: the United States. Less than one-quarter migrate to other EU countries, our nearest neighbours. It is likely, therefore, that proximity is not the deciding factor in international movements of academics in the Higher Education sector.

Obviously, pay is only one of many factors which influence the decision to enter and remain in academia and to move between academic institutions. These have been reviewed in Chapter 2 and the relative role of pay and other factors on leaving the sector is assessed in Chapter 7. It may be expected that international moves may have similar determinants to intra-national moves: academics may move, for example, because they seek promotion, higher pay, a more highly ranked institution or a better quality of life. However, international moves may be more problematic because they entail more unknowns (both about the employment market and about non-work aspects). At the same time, there are well-trod career paths, entailing permanent and temporary moves abroad.

Comparison of earnings is complex, entailing decisions about the range of benefits to be included (for example, pension contributions, sports facilities, company cars). International comparisons are particularly difficult, as earnings may be compared in a number of ways. Comparisons may be made using nominal salaries (i.e. adjusted across countries using exchange rates) or using real salaries (i.e. adjusted for differences in the cost of living). However, there are a range of exchange rates and different ways of assessing real salaries. Ideally, the choice should be based on the behaviour of academics: how they compare salaries internationally. But this is not known. Therefore we present analyses using both nominal and real salaries, with the latter adjusted using the purchasing power parity (PPP) exchange rates developed by the OECD. However, it should be noted that any adjustment can only be an approximation and that whilst movement abroad may be affected by official indicators of relative cost of living, remaining abroad will depend on actual cost of living. Moreover, nominal, rather than real, differences are important if individuals plan to return to their own country: items that will be converted back into domestic currency, such as savings and house purchases, should ideally be excluded from such calculations.

4.3 A comparison of academic salaries across nine countries In this section we compare academic wages in nine countries using national

survey data. The countries that we study are the UK, US, Canada, Australia, New Zealand, Denmark, France and Sweden. We also have a figure for mean academic wages in Japan, which we include for comparison. For all countries, we examine average earnings for all academics and, where data allows, the distribution of earnings. We also consider differences in the extent of part-time working.

4.3.1 Data The data come from a mixture of surveys and censuses. The advantage of

labour force surveys over samples of individual institutions is that we can be far more certain that the data are representative of the total population. The national statistical agencies that undertake these surveys have spent a lot of time and effort to ensure that samples are truly random and representative of the economy as a whole. For example, samples of academics within a sample of institutions are not totally random as each

69

individual in the country’s whole Higher Education sector does not have an equal chance of being sampled; most academics have a zero probability of being sampled. Since there is a great deal of heterogeneity between higher education institutions, a large amount of the variation in wages may be explained by the institutions themselves rather than the individual academics. Whilst using only data from those institutions that post data on the internet (as some studies mentioned in the previous section have done) may appear random, there is likely to be a systematic bias towards particular types of institution doing this76. Of course, it would be preferable to have datasets on the actual staff population in each country as we do with the HESA Individualised Staff Record in the UK, but this is not possible. Therefore, as with any analysis that is based on the use of samples it is important to ensure that the sampling procedures are of good quality. National Labour Force surveys and censuses meet this requirement, whereas ad hoc methods, such as scanning for data available on the internet, do not. In this section we discuss the main data issues and outline the data used. For further, detailed discussion of the data, see Appendix D at the end of this document.

The primary data issue is how to identify higher education academics. Unfortunately, it was not possible to identify precisely all of the appropriate staff, research staff in particular for all countries. Identifying higher education academics was done using information on occupation and industry where available. In the UK, France, New Zealand and the US, where it was possible to classify data by both these dimensions, individuals were considered to be higher education academics if they were in an appropriate occupation (e.g. lecturer or researcher) and working in the higher education sector (although in the US it is not possible to separate teachers and researchers at universities from those at state colleges77, who also conduct teaching undertaken in the FE sector in the UK). Otherwise, individuals were identified by their occupation only. In Australia and Canada, the data refer to university teachers and lecturers78, and exclude individuals from the FE sector. However, for Sweden and Denmark, the occupational classification could only be obtained at a level that precluded the separation of higher education academics from individuals who worked more broadly in the tertiary sector. The data issues are dealt with in more detail in Appendix D. One issue to note is that it is possible that pro-vice chancellors and deans may be excluded from our analysis if they were coded as managerial occupations. This will only take place if the majority of their time is spent in managerial tasks. Moreover, the number of such staff is very small.

Another issue to consider is whether to examine earnings net or gross of taxation. One would prefer all of the data to be net of taxation so we could compare take-home pay. Unfortunately, the data for the majority of countries was only 76 It may be the case that only certain universities post information. For example universities paying low salaries may not wish to advertise this. 77 In order to give some idea of the order of magnitude of the extent to which our data may understate US academics’ wages, Clery and Christopher (2004) report average salaries in the academic year 2002/03 of $52,377 and $63,524 for faculty in two-year and four-year public institutions respectively, with the number of faculty in each being 332,665 and 438,459 respectively; this would have yielded a mean salary of $58,715, an understatement of $4,808, or 8%. Note however that this is for ‘faculty’ and excludes teaching and research assistants. Therefore, since these account for around a third of academics in public four-year colleges and less than 1% in two-year colleges, the actual degree of overstatement will be rather smaller. 78 An attempt was made to include other staff in higher education (e.g. research staff), but this was not possible. For more on this, refer to the data appendix (Appendix D).

70

available on a gross basis and so in what follows we compare gross annual salaries. The wage data for France was only available net of taxation and so we also include figures for net UK earnings to enable us to compare the two.

Unfortunately, we could not study some of the European countries that we would wish to because of data considerations. The EU legislation instructing member states to run labour force surveys does not stipulate income as information that is required. Therefore, although all EU member states have labour force surveys, their coverage of income is patchy. The Italian and Spanish labour force surveys do not collect information on earnings and so are of no use to our study. The German Mikrozensus does include information on income, but it is not considered to be reliable (see Appendix D).

4.3.2 Results In order to make useful international comparisons we need to convert salaries

into a common metric. The simplest is the current exchange rate. This is the easiest conversion rate to obtain and seems the most likely one that individuals considering employment in another country will use to compare wages, at least as a first approximation.

When we consider nominal academic salaries in pounds sterling, we see that academics in the UK are situated approximately in the middle of our sample79. Overall, UK academics earn less than those in Japan, Denmark and the US, but more than those in Sweden, Australia, New Zealand, Canada and France (columns 2-4 of Table 4.4). The pattern is somewhat different by gender. For example, compared with the US, male academics in the UK earn considerably less in nominal terms, whereas female academics in the UK earn slightly more than their US counterparts.

Although converting foreign currencies into pounds sterling is a simple calculation and provides a useful indicator of international comparisons of salaries, it is only a nominal comparison. There are a number of drawbacks with using exchange rate data. Exchange rates are highly volatile. Comparisons will therefore be sensitive to the date of comparison, or the time scale over which an average is calculated. They are also the product of many other factors, including currency speculation. Moreover, relative real earnings, i.e. in terms of its purchasing power, are obviously important. In order to account for differences in purchasing power across countries, we convert the salaries using the OECD purchasing power parity (PPP) rates. These rates are also much more stable than exchange rates and represent a more stable and representative basket of goods than a single-good measure like the Big Mac Index.

79 Note that the data for the UK, Canada, France and US relate to both teaching and research staff in higher education. In the other countries they refer to university lecturers and teaching staff only. Since research staff without teaching responsibilities tend on the whole to be lower paid than their counterparts, the figures in these countries may be overstated slightly, although the effect will be small.

71

Table 4.4 Academic salaries, in £

Nominal Real Total Men Women Total Men Women e

a

WW

Gross Salary US 31,640 35,775 25,886 31,959 36,136 26,147 1.30 UK 29,183 31,210 25,964 29,183 31,210 25,964 1.60 Denmark 34,518 37,932 29,871 29,123 32,004 25,203 1.20 Canada80 24,458 26,319 21,485 28,860 31,056 25,352 1.22 Japan 34,358 † † 25,079 † † 1.31 Sweden 25,015 26,677 22,708 22,756 24,268 20,657 1.13 Australia 18,068 19,655 15,847 24,751 26,924 21,708 1.44 NZ 17,925 20,393 15,279 22,777 25,912 19,414 1.32

Net Salary UK 19,605 21,088 17,475 19,605 21,088 17,475 1.59 France 16,749 18,237 14,542 19,475 21,206 16,909 1.44

• Nominal salaries are converted using the current exchange rates • Data for Canada, France, UK and US relate to both teaching and research staff in higher

education. In the other countries they refer to university lecturers and teaching staff only. • Real salaries are converted using PPP rates from OECD Main Economic Indicators • Final column (Wa/We) refers to wages of academics relative to rest of economy. • † Data not available.

The final three columns of Table 4.4 present the real wages of academics, deflated using OECD PPP rates. These present a different picture from the comparison of nominal wages. Once differences in the cost of living have been taken into account it is only in the US where academics earn more on average than in the UK. The premium paid to academics in Denmark and Japan disappears when earnings are measured in real terms because these are relatively expensive countries. Conversely, some or all of the nominal premium UK academics enjoy over those in Canada, Australia, New Zealand and France disappears. Indeed, the real wages of Canadian academics are similar to those in the UK even though the nominal wages are much lower. The premium that UK academics enjoy over Swedish academics increases because of the relative price levels in the two countries. This difference may be greater in post-tax income, because taxation is much higher in Sweden than in the UK. However, these taxes are reflected in higher state provision of services in Sweden that are not reflected in the OECD PPP figures.

Perhaps the most important facts that arise from the international comparisons of nominal and real salaries are the figures for the US and Canada. As we noted above, 6% of staff who move to another higher education institution, go to one in the US (Table 4.2). The similar price levels in the UK and US mean that the differential in real earnings is almost identical to the nominal one. The effect of cheaper prices in Canada is to increase the wages of academics working there relative to those in the UK. The lower prices in Canada mean that the real wages of Canadian academics are similar to those in the UK.

80 The figure for Canada is the weighted average of university professors and teaching and research assistants.

72

4.3.2.1 Full-time Staff

The above analysis refers to all staff, irrespective of hours of work. Differences in annual earnings could be entirely due to differences in hours. Owing to data limitations, we were unable to examine hours of work directly, but, for many of the countries, we can improve comparability by removing part-time employees from the analysis.

As we can see from Table 4.5, the UK has a smaller proportion of part-time staff than the US and so this cannot explain the lower wages seen in the UK81. The UK has a greater proportion of part-time staff than in France.

Table 4.5 Proportion of part-time academic staff

UK France US Men 16 10 28 Women 31 16 41 Total 23 12 33

• Sources: Enquête Emploi (France), LFS (UK) and CPS (US)

Table 4.6 presents the earnings of full-time workers in academia in the UK, Canada, Sweden, Australia, New Zealand, France and the US. We can clearly see that the differences in the proportions of part-time staff in higher education do not explain the differences in earnings between the two countries, indeed they exacerbate it. Once we account for the different structure of employment the gap between UK and US male academics actually increases.

Table 4.6 Real earnings of full-time workers only

Total Men Women US 38,706 42,157 32,745 UK (gross) 32,589 34,341 29,532 Australia 31,092 32,192 Canada82 28,860 31,056 25,352 NZ 26,126 28,428 23,227 Sweden 23,184 UK (net) 21,908 22,973 20,132 France (net) 21,142 22,641 18,635

• Real salaries are converted using PPP rates from OECD Main Economic Indicators

81 The proportion of part-time staff is much higher than one would expect given the figures in HESA Individualised Staff Record (13% of staff were classified as part time, 9% of men and 20% of women), although part of the explanation may be due to contract staff being coded as part time in the LFS (by this definition, 17% of staff would be part-time, 24% of women and 12% of men. Staff who are studying for a degree may not appear in the HESA records if they are less than 25% FTE. There are both part-time and full-time staff in the LFS who are reported to be studying for a higher degree. 82 Again, this is a weighted average of university professors and teaching and research assistants.

73

4.3.2.2 The Distribution of Earnings

Comparisons of the mean value of wages are simple to undertake and easy to understand. However, they are a narrow indicator of the earnings among academics in the higher education sector. Although a low average wage may reflect the general level of wages, it may also reflect relatively low starting wages or, conversely, low earnings at the pinnacle of the career. We do not have access to wages by grade in different countries, but we can gain some more insight into how earnings in different countries compare by examining the dispersion of wages. Figure 4.1 to Figure 4.3 show the earnings quartiles for the UK, the US, Australia, New Zealand, Canada, Sweden and France, the countries for which data were available83. The figures underlying these graphs are presented in the Appendix D (Table A.16). Note that for Canada we do not have data for all staff in Higher Education, but rather university professorial staff (i.e. teaching staff and those who undertake teaching and research) and teaching and research assistants separately. It may be the case that individuals enter the latter posts early in their career and progress to the former. This is borne out by the fact that the upper quartile point of teaching and research assistants’ wage distribution is below the lower quartile point of university professors84.

The countries in our sample have quite different earnings profiles. The earnings distribution in the US is rather dispersed, whereas in the UK, Sweden and France it is quite compact85. The earners at the top quartile point in the US earn twice that of the bottom quartile point. Because of this greater dispersion in the US, although the lower quartile point of academics in the UK and the US earn similar real wages86, the median wage in the US is similar to that of the UK’s upper quartile point. By the time US academics reach the upper quartile, they are earning far more than their UK counterparts. The distribution of earnings in Australia is more similar to the UK, although there is less dispersion in the earnings of male Australian academics than those in the UK. The distribution of Canadian university professors’ wages is broadly similar to that of the UK at all points, although the inclusion of teaching and research assistants would undoubtedly increase the dispersion of wages observed in Canada, depending on the relative size of these groups.

83 For those not familiar with the concept of quartiles and deciles, note that since a quartile is the value of the boundary when the data are divided into four parts, each containing a quarter of the population, there are only three quartiles. The lower quartile or first quartile is another name for the 25th percentile. It is the location that cuts off the smallest quarter of the data. The upper quartile or third quartile is another name for the 75th percentile. It is the location that cuts off the largest quarter of the data. The same is true for deciles, i.e. because the data is divided into ten parts there are only nine deciles. 84 Note that this classification of university professors is similar to that of the US rather than that of the UK. 85 Wages for all academics in Denmark also display a similar pattern, although we do not have them broken down by part-time/full-time staff. 86 the quartile point is higher for the UK than in the US for men and lower for women.

74

Figure 4.1 Real earnings quartiles of full-time academics

0

10

20

30

40

50

60

UK(Gross)

US Australia NZ Canada(prof )

Canada(assist)

Sweden France(net)

UK (net)

Ann

ual S

alar

y, (£

000) 1st quartile

2nd quartile

3rd quartile

Figure 4.2 Real earnings quartiles of full-time academics, men

0

10

20

30

40

50

60

UK(Gross)


Canada(assist)

Sweden France(net)

UK (net)

Ann

ual S

alar

y, (£

000) 1st quartile

2nd quartile

3rd quartile

Figure 4.3 Real earnings quartiles of full-time academics, women

0

10

20

30

40

50

60

UK(Gross)


Canada(assist)

Sweden France(net)

UK (net)

Ann

ual S

alar

y, (£

000) 1st quartile

2nd quartile

3rd quartile

• Real salaries converted using PPP rates from OECD Main Economic Indicators • Data for UK and US relate to both teaching and research staff in higher education. In the other

countries they refer to university lecturers and teaching staff only. • Figures for female Australian academics are unavailable once broken down by gender and

quartile

75

We can look at the earnings distribution in greater detail for six of the countries (UK, US, Australia, New Zealand, Canada and France) by examining the wage deciles in Figure 4.4 to Figure 4.7. Again, the figures underlying these are presented in the Appendix D (Table A.17). What is immediately apparent from the figures is that differences in earnings between UK and US academics are very small at the lower end of the distribution, but increase as we move up. The top 10% of US academics in particular are paid considerably more than their UK counterparts. In the US there appears to exist a premier league of academic high-fliers that does not exist in the UK. The UK is much more similar to Australia and New Zealand.

Figure 4.4 Real earnings deciles of full-time academics, total

0

10

20

30

40

50

60

70

80

1stdecile

2nddecile

3rddecile

4thdecile

5thdecile

6thdecile

7thdecile

8thdecile

9thdecile

Ann

ual s

alar

y (£

000s

)

UK

US

Australia

Canada (prof)

Canada (assist)

New Zealand

• Because the wages for Australia are based on a banded variable, they have an upper limit of

£38,354 and thus the top two deciles may be equal.

Figure 4.5 Real earnings deciles of full-time academics, Men

0

10

20

30

40

50

60

70

80

1stdecile

2nddecile

3rddecile

4thdecile

5thdecile

6thdecile

7thdecile

8thdecile

9thdecile

Ann

ual s

alar

y (£

000s

)

UK

US

Australia

Canada (prof)

Canada (assist)

New Zealand


£38,354 and thus the top two deciles may be equal.

76

Figure 4.6 Real earnings deciles of full-time academics, women

0

10

20

30

40

50

60

70

80

1stdecile

2nddecile

3rddecile

4thdecile

5thdecile

6thdecile

7thdecile

8thdecile

9thdecile

Ann

ual s

alar

y (£

000s

) UK

US

Canada (prof)

Canada (assist)

New Zealand


£38,354 and thus the top two deciles may be equal. • Figures for female Australian academics are unavailable once we break them down by gender

and decile

When we compare the net earnings distributions of UK and French academics (Figure 4.7), we see that there is very little difference between them. The difference between UK and French academics’ earnings remains constant across most of the distribution; UK net earnings are always slightly higher than those in the France apart from in the highest deciles.

Figure 4.7 Net earnings deciles of academics for the UK and France

0

5

10

15

20

25

30

35

1stdecile

2nddecile

3rddecile

4thdecile

5thdecile

6thdecile

7thdecile

8thdecile

9thdecile

Ann

ual s

alar

y (£

000s

)

Total, UK

Total, France

Men, UK

Men, France

Women, UK

Women, France

• French salaries converted using PPP rates from OECD Main Economic Indicators • Data for UK relate to both teaching and research staff in higher education. In France they

refer to ‘Professeurs arégés et certifiés’ (aggregate and certified Professors), ‘Enseignants de l’enseignement supérieur’ (teachers of higher education) and ‘Chercheurs de la recherché publique’ (Researchers of public research) in ‘enseignement supérieur’ (Higher education).

77

To sum up, the earnings distribution of academics in higher education in the English-speaking nations is generally more dispersed than those elsewhere, particularly when compared with the two Nordic countries, Denmark and Sweden. The earnings distributions for women are slightly flatter than those for men. This is particularly true at the top of the earnings distribution. In the US, the top 10% of academics earn almost a third more than the next 10% and twice as much as the median worker. Although, the UK does appear to have a small group of relatively high earners, the earnings premium they earn is small in comparison to their US counterparts. Moreover, male US academics enjoy an increasing premium over their UK peers throughout almost the entire distribution. Female academics in the US and UK earn similar real wages in the lower half of the earnings distribution, but above the median female US academics also enjoy an increasing premium over their UK (and Antipodean) peers. At the upper echelons, therefore, US academics earn half as much again as those in the UK. This suggests that, for men, earnings differentials are likely to encourage out-migration of academics to the US (except, possibly in the earliest years of an academic career) and to discourage in-migration. However, for women the incentives towards out-migration are concentrated in the upper half of the career structure.

These figures suggest that UK academics tend to be similarly or better paid than their counterparts in the selected European counties and Australia and New Zealand. Thus the level of UK academics’ pay is more likely to prove an incentive to in-migration from these countries, rather than an incentive to out-migration.

4.4 An analysis of intra-UK earnings and UK/US academic Earnings

4.4.1 Background In the previous section we saw how the earnings of academics in higher

education across the world vary. However, the analysis has taken little account of differences in the composition of academic staff between countries, which may affect earnings. It is important to take into account many of the other factors that may vary in order truly to compare like with like. We have already seen, for example, how the proportions of part-time staff vary between the UK and the US. If we rely solely on the figure for the whole of academia we would obtain an altogether too sanguine view of the wages of UK academics relative to their US counterparts, particularly for males.

In this study we employ an econometric method to explore the extent to which the earnings differential between the UK and the US can be explained by differences in characteristics of the individuals in the labour market. The analysis addresses two questions. It pursues the international theme, to explore the extent of the earnings differential between the US and the UK, once differences in characteristics have been taken into account. It also turns to the domestic labour market to explore academic pay relative to that in the rest of the UK economy.

The model and results of our econometric estimation are discussed in detail in Appendix E. In this section we discuss the implications of the results for these two questions.

78

4.4.2 UK The results of the estimation of our model for the UK are presented in Table

A.19 of Appendix E. From our results we can see that people remaining in academia have a different lifetime earnings profile to graduates working outside academia. Academic earnings start from a lower point, but increase faster over the span of an academic career.

Once we account for age, qualifications and ethnicity, UK academics earn less than non-academics, ceteris paribus. In fact, the wage an individual can earn as an academic will be below that which they could earn elsewhere until they reach their mid-fifties when academics earn slightly more than graduates who remain in employment87. The experiences of men and women working in higher education are not significantly different.

We can get a clearer picture of the pattern of earnings by considering the lifetime wage profiles of our average workers by age, presented in Figure 4.8. This figure shows the annual wage of average academics and non-academics alike88. Academics earn less than non-academics over almost all ages. At the age of thirty, the average academic earns 80% of the average graduate outside academia. At forty this figure is 87% and at fifty it is 95%. Academics achieve parity with non-academics in their mid-fifties and by the time they are sixty they are earning slightly more. It is important to note that this difference is not due to differences in personal characteristics. If we plotted a line to say what an average academic would earn outside academia, it would be indistinguishable from the lifetime wage profile for non-academics. Therefore, unless there are other, unobserved, differences between academic and non-academic graduates89, this difference represents the pay cut a new academic would have to suffer entering from outside the Higher Education sector and the annual pay premium an academic would be expected to achieve if they left the sector.

Although, over their career, academics’ pay gradually creeps towards that of other graduates, it is only late in their career that it reaches and then surpasses that of graduates in other jobs. It has long been known that people discount future events relative to more immediate ones when making decisions90. Therefore, the difference between academic and non-academic earnings may be even greater than these figures suggest. The implication of this is that these differences in earnings may have a greater impact upon recruitment and retention of young staff. The staff remaining in higher education are likely to be either those for whom relative earnings are less 87 Note that it is possible that high wage earners outside academia invest some of the wage premium that they earn earlier in their career and so retire early. This is supported by the fact that participation rates of graduates drop fairly swiftly in their fifties. If it is the case that the higher wage earners retire early, those who remain in the labour force will earn less on average ceteris paribus and this reversal of fortunes is illusory. 88 That is, the values of the variables in Table A.19 are set to their mean values for academics and non-academics and the predicted value of the wage is calculated at each age. 89 This would need to be some difference between graduates who become academics and those who do not that is negatively associated with income. The most likely contender is unmeasured ability, but it is difficult to sustain the hypothesis that academics are, on average, of a lower ability than other graduates. Certainly, academics are likely to have higher degree results than the average non-academic graduate since many posts as well as graduate programmes require a minimum of an upper second. It is also possible that there may be differences in other intangible difference between graduates in academia and those outside, such as motivation, drive and personability. 90 i.e. jam today is worth more to most people than the same amount of jam in thirty years time.

79

important, or who are less able to earn a high wage elsewhere (because of some unmeasured ability, for example).

Figure 4.8 Predicted lifetime wage profiles, UK

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

25 30 35 40 45 50 55 60age

Ann

ual e

arni

ngs

Non-academics

Academics

• Chart shows the predicted annual wages of the average academic and non-academic graduate

based on the results reported in Table A.19 and averages calculated from the sample.

4.4.3 US The results of our estimation for the US are presented in Table A.20 of

Appendix E. The earnings profile of US academics is similar to that in the UK, although it is flatter (Figure 4.9). The earnings of US academics increase over their working lives. However, unlike in the UK, US academic wages reach a plateau at the end of their career. The average US academic continues to earn more over their entire working life, although the amount their earnings increase slows over their career.

A question of particular interest is what UK academics believe that they could earn in the US. Academics in the US earn more at all ages (Table 4.7 and Figure 4.10). It is important to note that the US data include individuals working in the US state colleges, who undertake teaching similar to that in UK FE colleges and so this comparison may actually understate the differences in wages. This will provide an incentive to UK academics if the US figure does indeed represent what they could earn in the US. However, first we must discount the possibility that the difference in earnings represents differences in the personal characteristics of UK and US academics. We can do this by considering what the average UK academic would earn in the US based on the results of our estimation91, taking into account differences in personal characteristics92. We see that the lifetime wage profiles of the average UK academic (were they in the US) and the average US academic are almost identical. This suggests that the compositions of UK and US academics are very similar. The 91 i.e. given the coefficients in the US earnings regression. 92 Note that we are implicitly assuming that our UK academic works in an area of US HE where they obtained their degree.

80

difference in wages therefore is due to how the attributes of academics (their age, gender etc.) are rewarded. This leads us to conclude that the differences in UK and US academic wages highlighted in Table 4.7 are unlikely to be due to differences in the academics themselves, but rather in the labour markets generally and in systems of higher education between the two countries.

Figure 4.9 Predicted lifetime wage profile, US

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

25 30 35 40 45 50 55 60age

Ann

ual e

arni

ngs


Table 4.7 Comparative actual academic earnings by age

age UK US Under 30 £16,197 £19,738 30-39 £24,124 £25,780 40-49 £29,724 £32,530 50-59 £29,526 £35,690 Over 60 £27,920 £39,888 Total £25,832 £33,039

• Real salaries are converted using PPP rates from OECD Main Economic Indicators • Includes FT and PT staff

81

Figure 4.10 Comparative predicted annual wage profiles

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

25 30 35 40 45 50 55 60age

Ann

ual e

arni

ngs

US Academic

UK Academic in the UK

UK academic in the US

4.5 Factors affecting international pay differences We have seen that the earnings of academics vary across countries, both in

nominal and real terms. There are a number of factors that affect the levels of academic pay in different countries. One is the wealth of the economy. Wealthier economies may be able to afford to pay higher salaries to their academic staff, ceteris paribus. Another is the general state of the labour market and that of the academic labour market in particular. Regulations and taxation policy are also likely to affect the wages of academics. Where the higher education sector is in the public sector, Government policy will come into play. Even if institutions have autonomy over the setting of salaries, these decisions are heavily influenced by the total amount of funding available.

The academic labour market may operate in a different manner from other labour markets. It could be argued that academics are more peripatetic in nature. If academics are more mobile than other types of labour, one would expect them to respond more to international differences than other sectors. As well as international differences in earnings, this might also include the relative size of countries’ higher education sectors or the reputation of their institutions. Therefore, if academic institutions have to compete in an international market for academic staff, they will have to set salaries, at least in part, with reference to their competitors. The degree of competition between domestic and foreign higher education institutions will differ across countries. Higher education institutions from countries who share a common language are likely to have a relatively high degree of competition compared to those that do not, ceteris paribus. Likewise, it may be the case that the degree of competition depends on geographical proximity, particularly in the EU where the single European labour market removes many of the regulatory obstacles to migration.

It is likely that the ‘other’ influences on academic pay outlined above will affect other employees similarly. We can get a picture of the effect of countries’ economies on differences in academic earnings by considering their salaries relative to those paid to the economy as a whole. This will remove (or at least attenuate) the

82

effect of other factors affecting pay. Table 4.8 compares academics’ wages to those paid in the rest of the economy. Academics in the UK are paid relatively well relative to the whole economy when compared to the other countries. Note that these relative wages may reflect many other factors apart from the wealth of the economy, such as differences in the characteristics of academics relative to the rest of the labour force, the demand and supply of academic staff, earnings inequality etc. For example, the relatively high wages of UK academics when compared to the richer US may merely reflect the greater earnings inequality generally in the US. Nevertheless, we can see from Table 4.8 that the wealth of nations is not the prime driver of variations in academic earnings.

Table 4.8 Earnings of academics relative to average earnings

Relative earnings UK (gross) 1.60 UK (net) 1.59 Australia 1.44 France (net) 1.44 NZ 1.32 Japan 1.31 US 1.30 Canada 1.22 Denmark 1.20 Sweden 1.13

We can focus more clearly on the reasons behind international differences in academic salaries if we extend our UK-US comparison. Both countries share a common language. Moreover, we have seen that the US represents the largest single destination of UK academics who move within the Higher Education sector, for whom we have data.

Pay in any sector will be a product not only of pay within that sector – domestic and foreign – but also in other sectors that compete for the same (or similar) workers. In order to compare the UK and US Higher Education sectors, we also need to know what academics in these countries would earn if they were not in academia. In our empirical analysis we chose to compare academics to all other graduates (as opposed to just those with higher degrees)93.

Using our results we can see that, as with the UK, US academics earn less than their compatriots for the first two-thirds of their working life (Figure 4.11). Academics have different lifetime earnings profiles from non-academic graduates. US graduates outside higher education earn much more in their early life compared to academics. This difference is larger than in the UK. One might expect, therefore, shortages of academic staff in higher education to be an even bigger problem in the 93 There are two reasons for this: first, although the majority of Higher Education academics have a higher degree, many only have a bachelor’s degree. Second, it is not clear precisely when individuals make their decision to work inside or outside of academia. If the choice of whether to undertake a higher degree is based on the relative earnings the graduate could earn in Higher Education or elsewhere, it is important to make the comparison with all graduates, even if some do not make their final decision not to enter Higher Education after their higher degree.

83

US than in the UK, but this is not the case. Indeed there has been very little discussion of this issue in recent years in the US by comparison to the UK94. Non-academics’ earnings do tail off quicker in the US than they do in the UK and so the average US academic earns more than the average non-academic graduate from their late forties onwards. However, it must be pointed out that it is dangerous to place too much emphasis on comparisons at the end of the working life because of differences in retirement patterns.

Figure 4.11 Predicted lifetime annual wage profile, US

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

25 30 35 40 45 50 55 60age

Hou

rly e

arni

ngs

Non-academics

Academics


There are a number of potential explanations for the difference between US and UK earnings. It probably reflects the US labour market generally. It may be that the freer academic labour market allows US institutions to reward excellence in a way that the more-structured UK does not. However, we must be careful not to confuse two reasons for wages being distributed thus. The first is that earnings are more widely distributed within US academic institutions. The second is that earnings are more widely distributed between as well as within US higher education institutions. It may be the case that there is a greater heterogeneity of institutions in the US Higher Education sector, from the Ivy League to some of the smaller State Universities and Colleges. Higher wages in some institutions may merely reflect the social make-up of their student bodies. US higher education institutions are much more reliant on alumni contributions than those in the UK. Answers to questions of this nature cannot be answered without information about individual’s earnings at institutions, to which we are not privy for the United States95.

94 See our earlier literature review. 95 There is a survey of earnings of academics staff at US HEIs. However, access is not available to the data outside the US.

84

4.6 Summary and conclusions In this chapter we have sought to place the salaries of UK academics in higher

education in their national and international context. We have done this by comparing academic wages in the UK to those of other highly qualified employees and comparing UK academic wages with academic pay in eight other countries: four English-speaking, the US, Canada, Australia and New Zealand; three European, Denmark, France and Sweden; and Japan.

Compared with other highly qualified employees in the UK, academic starting salaries are low. Academics continue to earn less than comparable employees until they are in their mid-50s, at which stage academic salaries are slightly higher than for comparable employees. However, the gain at this age is much smaller than the earlier loss.

Compared with academics in the eight comparator countries, UK academics are in the middle of our sample (once differences in purchasing power have been taken into account). Overall, UK academics earn less in real terms than those in the US, they earn a similar amount to those in Canada, Denmark and France, and more than those in Australia, New Zealand, Japan and Sweden.

In terms of dispersion of earnings, the UK earnings distribution lies somewhat in the middle of our sample. The UK distribution is more dispersed than those in the Nordic countries, but less so than in the US. This may reflect differences in the higher education sectors in these countries, or it may merely reflect the characteristics of the different labour markets as a whole.

It is the comparison with the US that is perhaps the most important. Of the 13% of UK academics who left the UK to move to another higher education institution in the academic year 2001/2, over one third moved to the US. We have found that it is not merely the higher real salaries that set academic staff in the US apart, the distribution is much more dispersed. Although academics at the lower end of the scale in the UK and US earn similar amounts, the median worker96 in the US earns more than three-quarters of UK academics. The upper reaches of US academics earn far more than their UK counterparts. From this we can conclude that in the US there is a premier league of academic high flyers that does not exist in the UK.

Because of the importance of the comparison in terms of staff movement and the differences in both the level and structure of earnings, we analysed the UK and US graduate labour markets in greater detail using multivariate analysis. This enabled us to net out differences in earnings that may be due to other factors, such as differences in the age structure of the academic workforces in the UK and US, or differences in ethnicity or gender. By comparison, Higher Education academics in the UK earn less than those in the US at all ages, particularly in the latter half of their working lives. The difference in real wages between academics in the UK and US cannot be explained by differences in their personal characteristics. An explanation of the difference in earnings of UK and US academics lies elsewhere, such as differences in the higher education funding system and labour market.

There are number of implications of our results for the recruitment and retention of academics in UK higher education. First, the difference between

96 The median worker is the one who is situated precisely halfway up the distribution of earnings, i.e. they earn more than half of the workers and less than the other half.

85

academics and graduates in the rest of the economy is significant and, importantly, is greater at the earlier stages of individuals’ careers. Thus, academic salaries are liable to discourage entry to the sector and, if people discount their future earnings, the difference is likely to weigh even more heavily on their minds when they are making their career decisions. A second implication of this result relates to the staff who chose to enter and remain in academic employment. Since their continued presence in academia is contingent on past career decisions, it is likely that those who choose to remain are either those who cannot earn a higher wage in an alternative career or for whom wages are less important.

The fact that the earnings of UK academics level off in their mid to late forties, but those in the US continue to increase almost to the end of their career, suggests that institutions of higher education in the US are less constrained in their pay structures and are able to reward academic high-fliers in order to attract and retain them. If it is also the case that US universities have more flexibility at the lower end of the scale, the UK risks losing its most able academics at both ends of their careers.

86

87

5 Entrants to academic jobs

5.1 Introduction The literature on the nature of the supply of potential academic staff is very

limited (see Chapter 2). Thus whilst we know that the most common highest qualification of entrants is a PhD (Bryson and Barnes 2000a)97, the literature does not address the stages at which students tend to take decisions about entering an academic career, nor the factors which affect their decisions about entering academia. Nor does it investigate the relative importance of home and foreign students studying in the UK. With respect to other sources of entrants (foreign academics, UK employed and non-students outside the labour market) there is a dearth of information about the nature of supply, the pattern of employment and the factors affecting movement into UK academia.

Pathways into academia can be long and varied. They include the traditional route (progression from school through higher education to a PhD to academia) and the career change route (from other employment, or other activities, to academia). There are many variations on these routes: entrants may have moved a number of times between employment (or economic inactivity) and education prior to entering academia; stop-gap jobs or unemployment may punctuate the pathway prior to entering academia; academic employment may be interspersed with other activities. These pathways are difficult to identify without a longitudinal dataset.

5.1.1 A note on the data This chapter draws on a range of data (HESA Individual Staff Records, the

HESA combined student/module record, the quantitative surveys of staff and of research students and the qualitative research) to provide more information on these pathways. A number of caveats on the data should be borne in mind.

• HESA Individual Staff Records show the previous year’s activity for staff in the dataset. This enables the identification of those entering the sector in the previous year. However, this will not always provide a good indication of pathways. In particular, it cannot distinguish between stop-gap activities (for example, take up of another activity whilst completing a PhD or whilst searching for an academic job) and longer-term activities (e.g. employment in a previous career). Nor can it distinguish between those entering academia for the first time and those who have previously moved between academia and another activity. Therefore, in our analysis of HESA staff records, ‘entrants’ refers to those who were not recorded as being employed in an academic job in higher education in the previous year. Moreover, three per cent of data are

97 Note this is not the same as saying that the largest pool comes directly from studying.

88

missing and it is likely that a high percentage of these were new to the institution, and probably sector.

• HESA combined student/module record provides useful data on progression to higher degrees but does not record entry into academic employment (as opposed to employment in any higher education jobs).

• The student survey allowed analysis of the factors affecting the career decisions of research students. However, as the sample was confined to research students the career decisions leading to being a research student could not be fully explored; in particular, it could not identify why others had not become research students.

• The staff survey allowed exploration of the range of pathways into academia, including the career change route. However, as the sample was confined to those who were in academia, it could not identify why others had not entered academia.

Despite these limitations, the data provide a greater understanding of sources of supply, routes into academia and the factors affecting entry.

5.1.2 Layout of the chapter The next section describes the sources of academic recruits. The following

sections then describe the career routes that existing academic staff have pursued prior to entering academic employment (Section 5.3) and the factors which affected entry (Section 5.4). The rest of the chapter focuses on students. In Section 5.5, the route from first degree to academia is examined. Section 5.6 examines research students’ career intentions and the factors which are important in their career decisions. Finally Section 5.8 identifies factors which affect the likelihood of seeking an academic career.

5.2 Source of entrants98 to the sector There is little previous evidence on the source of entrants to the sector. The

assumption in the literature is that staff follow ‘traditional’ routes, i.e. PhD student to permanent academic post (possibly via a post-doctoral post). The HESA data (relating to those working 25% or more FTE) do not appear to support this99.

Although students were a large source of entrants to higher education institutions, they accounted for only one-third of entrants (34%) (Table 5.1)100. Most had been studying in the UK, with universities abroad accounting for only 5% of total

98 ‘Entrants’ to the sector are those who took up an academic job in a higher education institution having not been employed in an academic job in a higher education institution the previous year. However, in the US, Anderson (1998) and Austin (2002) consider that it is not the first appointment but the PhD that should be thought of as the first stage of the academic career as PhDs provide cheap teaching assistants. 99 As has already been mentioned, it is important to note that the HESA data on leavers does not record the previous situation of all entrants. 100 Survey evidence put the percentage of students higher, at 45% (Bryson and Barnes, 2000a). However, the representativeness of the survey is open to doubt having achieved a very low response rate: 1,451 respondents from 17,234 academic staff invited to participate, to which was added 34 pilot responses (Bryson and Barnes, 2001a,b).

89

entrants (or 15% of student entrants). However, many student recruits were foreign nationals (about 40%), many of whom had been studying in the UK. In total, 10% of entrants were foreign nationals who had been studying in the UK, whilst 5% had been studying elsewhere.

Table 5.1 Origin of entrants to the HEI sector, 2001/02

UK Foreign Total Student 19 14 34 Student in UK 19 10 29 Student abroad 0 5 5 UK Public Sector 21 4 25 Public sector research institute in UK 3 1 3 Other public sector in UK 11 2 13 Other education inst in UK 8 2 9 UK non-public sector employment 13 4 17 Private industry/commerce in UK 7 3 10 Self-employed and other UK employment 6 1 7 Other abroad 5 16 21 HEI in other country 2 9 11 Research institute in other country 1 5 6 Other employment abroad 1 3 4 Not in regular employment 2 1 3 Total entrants with known destination (%) 60 40 100

(number) 6,051 3,993 10,044 Employment last year not known, n 1,900 1,448 3,348 N 84,112 25,982 110,094

• Source: HESA Individualised Staff Record, 2001/02. • Data refer to staff at English HEIs and exclude those on clinical grades

UK employment was the largest source of academic entrants, 42%, with 60% of these coming from the public sector (including research institutes and other educational institutions). Foreign employment accounted for 21% of entrants, with the majority coming from higher education institutions abroad, 11% of the total.

The above data cover both research and teaching academic staff. For research staff alone, a lower percentage enter from employment. A survey of contract research staff in Scotland found that for about two-thirds, contract research was their first job (Hasluck et al, 2001). Those who had been employed tended to have been in professional practice, training (outside higher education), other research work, technical and support jobs and also in lecturing. In 2001/02 the European Union was

90

the main geographical source of entrants (source: HESA Individualised Staff Record, 2001/02).

Evidence from the qualitative research suggested diversity in recruitment markets. The case study new universities recruited more academic staff locally and regionally, while the case study old universities recruited from further afield, including overseas.

In the case study new universities, jobs in vocational subjects were particularly likely to attract local applicants. For subjects such as business and computing, this was particularly true for universities located in large cities. Health and education departments attracted local applicants because they recruited experienced practitioners with strong local connections. These included staff joining on secondment who are likely to be unwilling to relocate themselves and their families for a short period. The more academic subjects and stronger departments in new universities were able to attract candidates from a wider area, including overseas.

Some case study departments in old universities were becoming increasingly reliant on overseas recruits, from countries including Russia and China. Case study new universities attracted fewer overseas applicants, although some departments reported an increasing number of applicants from within the UK with English as a second language. Postgraduate students were also increasingly recruited from overseas, particularly by the case study old universities, which offer the majority of PhD studentships.

Just under one-third of the entrants into academic employment are under 30 (Table 5.2). These young staff enter Higher Education employment on a fixed-term contract, 92% of new entrants aged under thirty years of age are on fixed-term contracts (Table 5.3). (In total, 90 per cent of academic staff aged under 30 are on fixed-term contracts and the job function of 77% is solely research.) A further two-fifths of entrants into academic employment are in their thirties (Table 5.2). These are more likely to enter on permanent contracts than younger entrants (Table 5.3).

Table 5.2 Total staff movements by age group (%)

<30 30-39 40-49 50-59 60+ All ages Inflows In HEI last year 9 25 24 22 5 84 New to sector 4 5 2 1 0 13 From other HEI 1 2 1 0 0 4 Total known 14 31 27 23 5 100 Not known 1 1 1 0 0 3

• All figures are percentages of the known total for all ages, with the exception of Not known, which is the percentage of staff whose employment in the previous year is not known

• Source: HESA Individualised Staff Record 2001/2 • Data refer to staff at English HEIs and excluds those on clinical grades

91

Table 5.3 The Contracts of Entrants to HE by Age

Permanent Fixed term contract

Hourly paid/Casual Total

<30 Same HEI 12 86 3 100 From other HEI 20 79 1 100 New to sector 7 92 1 100

30-39 Same HEI 47 50 3 100 From other HEI 47 52 1 100 New to sector 26 72 2 100



60+ Same HEI 69 25 6 100 From other HEI 52 45 3 100 New to sector 31 56 14 100

All ages 55 41 4 100 Not known 1 4 1 5

• All figures are percentages of each group with the respective contract type, with the exception of the final row, which are the percentage of each contract type for whom last year’s employment is not known.


The proportion of current staff whose first job had been a research job increased consistently since the 1970s, with the exception of the early 1990s, and a tiny dip in 2001. The proportion of staff whose first academic job was in the past five years and was a research job is significantly higher than that for any five year period (Table 5.4). Similarly, there is a decline in the proportion of first jobs that were permanent for every period up until 2001 except 1990-4 (Table 5.5). Either entry to research posts and the use of contracts for entrants appears to have increased over time or those recruited to fixed-term contracts and to research posts have a much higher leaving rate. Whilst the latter is true, it seems unlikely that differential leaving rates would persist over a long period and so it does not explain differences for those recruited.

92

Table 5.4 Academic staff: First academic job: type of job

When first had research or lecturing job at UK university Lecturing Research Total n

Prior to 1970 70 31 100 101 1970-1979 65 34 100 343 1980-1984 62 38 100 215 1985-1989 59 41 100 300 1990-1994 64 36 100 501 1995-1999 56 43 100 539 2000 48 52 100 158 2001 49 51 100 174 2002 41 59 100 210 2003/2004 40 59 100 220 Totala 56 43 100 2787

• a Includes 27 who did not give date of entry to academia • Source: NIESR/DfES Staff Survey, 2004

Table 5.5 Academic staff: Contractual basis of first academic job

When first had research or lecturing job at UK university Permanent Not

Permanent Total n

Prior to 1970 50 50 100 101 1970-1979 48 52 100 343 1980-1984 44 56 100 215 1985-1989 35 65 100 300 1990-1994 41 59 100 501 1995-1999 31 68 100 539 2000 30 70 100 158 2001 28 72 100 174 2002 30 70 100 210 2003/2004 39 60 100 220 Total 37 62 100 2787

• a Includes 27 who did not give date of entry to academia • Source: NIESR/DfES Staff Survey, 2004

The literature provides little further elucidation on other sources of academic staff, such as international recruitment, the relative importance of home and foreign students studying in the UK and the nature of the supply from other employment (particularly in respect of vocational subjects). Nor does it indicate the source of non-mainstream teaching (e.g. hourly paid, termly contracts or other short periods). These might be expected to be filled by post-graduate students, researchers, those in other employment and those outside the labour market. In the US, the former are very important as teaching assistants (Anderson, 1998; Austin, 2002), but the pattern in the UK has not been explored.

93

5.3 Career routes into academia: existing staff The ‘standard’ career path for an academic might be thought of as going from

school to university to further degree(s) to temporary research posts to a lecturing post and then to progress up the academic career ladder. However, as we have seen, according to the HESA data, only one-third of academic staff for whom there is data were recruited direct from studying.

The NIESR/DfES survey of academic staff found that 61 per cent of academic staff had had a non-academic job (other than as a student) and nearly one third had been in a non-academic job for at least six years (Table 5.6). Over one quarter had held professional jobs outside academia and been employed for at least two years (Table 5.7).

Table 5.6 Academic staff: Employment outside academia

Percent had a non-academic job, other than as a student 61 only had a non-academic job as a student 27 never had a non-academic job 12 Period of employment outside academia other than as a student none 39 1 year or less 11 2 - 5 years 19 6 - 10 years 12 11 - 15 years 9 16 years or more 10 Total 100 n 2720 Missing 7 o Source: NIESR/DfES Staff Survey, 2004

Table 5.7 Academic staff: Nature of employment outside academia

Percent Those with two or more years employment outside academia 50 manager or senior official 6 professional 27 associate professional or technical 5 administrative or secretarial 3 some other job 7 not stated 2 less than two years outside academia 50 Total 100 n 2720 Missing 7 o Source: NIESR/DfES Staff Survey, 2004

94

Indeed, 36 per cent considered they had made a career change to enter academia (Table 5.8)101.

Table 5.8 Academic staff: Entrants making career change by subject

Main subject specialism per cent of subject specialists

making a career change n

Business and administrative studies 81 161 Computing sciences 47 146 Subjects allied to medicine 46 371 Education 43 214 Social studies, including economics 42 374 Engineering 40 169 Combined studies across above subject groups 35 106 Medicine and dentistry 28 142 Art and design 26 53 Other humanities 25 209 Mathematical sciences (excluding computing) 22 94 English literature and classics 19 52 Biological sciences 15 292 Modern languages 15 71 Physical sciences 13 242 Total 36 2785 o Subjects with fewer than 50 survey respondents are not reported. o Source: NIESR/DfES Staff Survey, 2004

Fewer recent entrants to academia than longer-established staff made a career change to enter academia (Table 5.9). There are two explanations for this, depending on whether career changes into academia tend to be permanent or temporary in their change of career. The first explanation is that fewer people are changing careers to enter academia than was the case previously. The second explanation is that those who change careers to enter academia are more likely than other entrants to leave academia, i.e. career changers are individuals who are more peripatetic in nature (either through personal preference or the existence of past contacts outside academia). If this is the case, then greater percentage of recent entrants who have changed careers to enter academia is due solely to the fact that they haven’t made their next move yet. Given the fact that retention problems have increased rather than decreased over the past two decades (e.g. Bett, 1999; UCEA, 2002; HEFCE, 2003), it seems that the latter is the more likely explanation. Certainly this ‘unobserved heterogeneity’ explanation has been used as a reason for the observed negative relationship between tenure and turnover (e.g. Gibbons and Waldman, 1999).

Although the incidence of career change varied by subject, for most subjects between one in five and one half of staff had made a career change to enter academic 101 Note that the data relate to current employees. Hence if the turnover rate differs between career changers and non-changers (and by subject), the pattern of career changing to enter Higher Education amongst entrants will differ.

95

employment. Career changers were most common in Business and administrative studies (81 per cent) and in Computing (47 per cent), Subjects Allied to Medicine (46 per cent), Education (43 per cent), Social studies (42 per cent) and Engineering (40 per cent). They were least common in Modern languages (15 per cent) and English literature and classics (19 per cent). Thus, the degree to which a subject might be regarded as vocational or demanding similar skills outside academia appears related to the likelihood of career change for entrants.

Table 5.9 Academic staff: Whether made a career change to go into academia

Less than 2 Years

More than 2 years Total

made a career change 57 61 61 did not make a career change 43 38 39 Total 100 100 100

• Source: NIESR/DfES Staff Survey, 2004

Career change was much more common amongst academics in new universities, with about one half reporting a career change to enter academia, compared with about one quarter of those in old universities (Table 5.10). It was slightly more common for women than men.

Table 5.10 Academic staff: Entrants making career change by type of university and by gender

old new men women total Made a career change 26*** 51*** 33*** 39*** 36 n 1726 1059 1790 996 2785

*** Significantly different at the 1% level • Source: NIESR/DfES Staff Survey, 2004

The qualitative research provided more details on careers of career changers. All the academics in health and in education interviewed in new universities had followed careers as practitioners and had moved into Higher Education to pursue their interest in teaching students. Time spent in practice varied from a few years to 24, with many spending about 10 years working as a health or education professional before deciding on an academic career. Some had their first ‘taste’ of an academic job through secondment or a visiting (part-time) lecturer post. Entry points varied according to academic qualifications gained as a student or while on the job. Those with an MA had been able to join the institution (a new university) as a senior lecturer, while others had joined on the lecturer scale. Art and design lecturers in the qualitative research had carried out a range of full-time and part-time jobs in Higher Education and outside, producing and exhibiting their work. Other examples of routes taken by staff in new universities are presented below.

96

Examples of academic career paths: career changers

A senior lecturer in Education had graduated in politics in his mid-twenties and began a career in social work. He then took a teacher training course and taught in secondary schools for 22 years. As a teacher he lectured part-time at an old university and took a masters degree at a new university. He was then seconded to teach one day a week to his current institution and, after a year, was recruited to a full-time senior lecturer post at the age of 48. Senior lecturer, new university

A lecturer in occupational therapy had started her career in arts administration. She then trained to be an occupational therapist and worked in the health service, acting as a visiting lecturer at another university. Although a non-graduate, she was recruited at the age of 45 to her current job as a lecturer in occupational therapy. Lecturer, new university

A now-41 year old senior lecturer in History had left school at 17 and worked in factory jobs in Canada before coming to the UK aged 20 in the early 1980s. After a series of temporary jobs he worked for a local authority. He then studied part-time for a degree and then combined part-time study for a PhD with temporary lecturing. This included two part-time jobs at universities more than 200 miles apart. He was recruited to his current post in 2003. Senior lecturer, new university

A now-45 year old psychology lecturer had worked as a hairdresser for almost 20 years before taking a degree, then a PhD, before gaining a research job, then lectureship, all at the same university. Lecturer, new university

In the qualitative research, examples were also found of career changers in old universities. They included academics in law who took up employment in Higher Education after a career in legal practice, and a small number of others who were attracted to Higher Education as a ‘career change’. The qualitative research also included a few academics who entered the sector as mature students and followed a traditional academic route after graduation.

5.4 Existing staff: decision process for entering academia

Career decisions are often thought of in terms of choosing from a wide range of options. This approach may not match reality if information about different careers is poor. However, evidence from the staff survey suggests a further problem, irrespective of access to information, that potential entrants may not make comparisons. Indeed, almost half of academics claimed that they had not seriously considered and compared careers prior to gaining their first academic job (Table

97

5.11). This may be because, for those entering after a research degree, the decision to enter an academic career was taken at an earlier stage. This is investigated below. Men were slightly more likely than women to have made this comparison. There was no difference between type of university, which is surprising, given the disparity between types of universities in those making a career change.

Table 5.11 Academic staff: Whether was seriously considering and comparing careers before got first academic job

men women total seriously considering and comparing careers 58*** 55*** 56 not seriously considering and comparing careers 42*** 45*** 43 Total 100 100 100 n 1790 996 2785

• ***sig different at the 1% level (not significantly different by type of university) • Source: NIESR/DfES Staff Survey, 2004

The qualitative research found many examples of lack of thought given to careers where existing academic staff had simply continued in Higher Education having performed well at undergraduate and then postgraduate level and had never investigated alternative career opportunities. A professor of Law described how he had ‘drifted’ into teaching law rather than practice because of his academic interests. The following explanation given by a reader in Sociology is not untypical:

‘Doing a PhD seemed a good idea at the time. I wanted to continue in academic sociology when I finished my degree because I enjoyed it and had been successful at doing it’.

As one might expect, a large proportion of staff had very much wanted an academic career (Table 5.12). This proportion is higher among longer-established staff than it is for recent entrants (i.e. who had been employed in academia for two years or less), which suggests that these people are more likely to remain in academia. This hypothesis is supported by the fact that a greater proportion of recent entrants is only in academia because they thought that it would help them with a career outside academia.

More academics in old universities, compared with academics in new universities, had very much wanted an academic career (Table 5.13). More men than women had also very much wanted an academic career.

The percentage of academic staff who had wanted an academic career varied substantially by subject (Table 5.14). Eighty-one per cent of academic staff in English literature and classics had very much wanted an academic career. The percentage was also high in other humanities, maths, modern languages and physical sciences. At the other extreme were staff in Art and design (only 36 per cent of whom had very much wanted an academic career), Business and administrative studies (38 per cent) and engineering (44 per cent). However, for most of those who had not been very keen on an academic career, the reason was that other careers also appealed, rather than academia not appealing. Those who thought they would use academia as a

98

springboard into another career were concentrated in Art and design (11 per cent), Engineering (eight per cent) and Computing Science (eight per cent).

Table 5.12 Academic staff: Career intentions prior to first academic job

Less than 2 Years

More than 2 years

Total

You very much wanted an academic career 54 58 57 You were aiming for a career outside academia and you thought an academic job would help in this career

8 3 4

You preferred some other career 4 2 2 A number of careers were of equal interest with academia 21 21 21

You were thinking about careers but you did not know what you wanted to do 10 9 9

You were not thinking about careers 3 8 7 Total 100 100 100


Table 5.13 Academic staff: Career intentions prior to first academic job by type of university and by gender

type of university gender

Old New men women Total You very much wanted an academic career 60 52 60 53 57 You were aiming for a career outside academia and you thought an academic job would help in chosen career 4 3 4 3 4

You preferred some other career 2 3 2 2 2 A number of careers were of equal interest with academia 20 22 21 20 21

You were thinking about careers but you did not know what you wanted to do 10 8 7 12 9

You were not thinking about careers 4 9 5 8 7 Total 100 100 100 100 100 n 1726 1059 1790 996 2785

• Difference by gender and by type of university significant at the 1 per cent level • Source: NIESR/DfES Staff Survey, 2004

99

Table 5.14 Academic staff: Career intentions prior to first academic job by subject

Main subject specialism

Very much wanted an academic

career

Wanted other career thought an academic

job would help

Preferred some other career

Academia and other careers of

equal interest

Thinking about careers did not know what to do

Not thinking

about careers

Total n

English literature and classics 81 10 8 2 100 52 Other humanities 73 0 2 15 4 4 100 209 Mathematical sciences (excluding computing) 68 3 16 6 4 100 94

Modern languages 68 3 10 13 6 100 71 Physical sciences 67 2 2 19 7 3 100 242 Biological sciences 61 3 1 20 10 4 100 292 Medicine and dentistry 59 6 3 13 11 5 100 142 Social studies, including economics 57 3 2 20 13 6 100 375 Subjects allied to medicine 55 4 2 19 9 9 100 370 Computing sciences 55 8 4 20 6 8 100 145 Education 54 1 2 21 11 10 100 214 Combined studies across above subject groups 54 3 1 24 11 7 100 107

Engineering 44 8 3 31 7 5 100 169 Business and administrative studies 38 5 3 36 8 9 100 160 Art and design 36 11 34 4 11 100 53 Total 57 4 2 21 9 6 100 2785

o Source: NIESR/DfES Staff Survey, 2004

Table 5.15 shows individuals’ reasons for taking up the first academic job. A greater percentage of recent entrants (i.e. those who had been employed in academia for two years or less) had entered academia because they wanted to continue doing research for a while than had longer-established academics. The reverse is true for those who had wanted to teach for a while. This suggests that, unless the types of staff entering academia has changed over time, those who enter academia in order to undertake research either intend to leave after this desire is sated or because they discover that academic careers offer less in the way of research than they had hoped. We return to this subject when we compare the research postgraduates’ perceptions of what the academic job involves with the experience of recent entrants in Section 5.7 below.

The reasons for entering academia differed between old and new universities (Table 5.16). Those in old universities were more likely to have very much wanted an academic career or to have wanted to continue with research. More staff in new universities had wanted to teach or had been lured in through a particularly attractive job.

Compared with men, women too were less likely to have very much wanted an academic career. Instead, a higher percentage had entered academia to take up a job they had found particularly attractive.

100

Table 5.15 Academic staff: Reasons for entering academia (%)

Less than 2 Years


You very much wanted an academic careera 54 58 57 You were aiming for a career outside academia and you thought an academic job would help in chosen careera 8 3 4

Other reasonb 38 40 39 The job was particularly attractive to you 20 17 18 You wanted to continue doing research for a while 17 14 14

You wanted to teach for a while 6 9 8 The job was better than others you could get 6 4 5 It was a job 6 4 5

You hadn't been able to get the non-academic jobs you preferred 4 3 3

None of these 1 3 3 Total 100 100 100

• a those giving these responses were not asked for other reasons. • b respondents were allowed to give more than one response. • Source: NIESR/DfES Staff Survey, 2004

Table 5.16 Academic staff: Reasons for entering academia by type of university and by gender

old new men women Total You very much wanted an academic careera 60** 52** 60** 53** 57 You were aiming for a career outside academia and you thought an academic job would help in chosen careera

4 3 4 3 4

Other reasonb The job was particularly attractive to you 15** 22** 16** 21** 18 You wanted to continue doing research for a while 18** 9** 15 14 14

You wanted to teach for a while 5** 13** 7 9 8 The job was better than others you could get 4 6 4 5 5 It was a job 5 4 5 4 5 You hadn't been able to get the non-academic jobs you prefer 3 3 4 2 3 None of these 2** 4** 2 3 3 Total 100 100 100 100 100 n 1726 1059 1790 996 2785 • a those giving these responses were not asked for other reasons. • b respondents were allowed to give more than one response. • ** significantly different at the 5 per cent level • Source: NIESR/DfES Staff Survey, 2004

The qualitative research provided more details of this attraction to an academic career. Interest in the subject was often expressed in emotional rather than intellectual terms. Therefore, academics spoke of ‘love’, ‘fascination’, ‘passion’ and ‘excitement’ for their subject. It encompassed the academic subject itself, but went beyond it into ‘ideas’ in general. As a reader in Sociology expressed, ‘It was something I found I enjoyed doing. It intrigued me and excited me’. This interest tended to manifest itself in a desire to do research. Many had become interested in

101

research as undergraduates, which continued in their postgraduate studies, leading to entering temporary research posts and then lectureships. Although some were aware of opportunities to carry out research outside Higher Education, they felt that nowhere offered the same freedom to pursue ideas as did the university sector. This freedom was available in old universities through the availability of funding, but more importantly of time. With regard to the teaching role, some academics in the new universities described a desire to pass their knowledge and interest on to students, principally undergraduates. Respondents in the old universities who also described having been motivated by teaching tended to mention interest in postgraduate teaching, but for many this was a secondary rather than primary motivation for entering academia. Moreover enjoyment of teaching stemmed from interest in the subject rather than teaching per se.

As well as the career aspects and interest in teaching or research, the qualitative research identified more personal reasons for seeking an academic career. A few lecturers were motivated to enter academia by an aspiration to equal the achievements of an academic parent. These lecturers were familiar with the demands of an academic job, but also the attractions of an academic lifestyle, for example good holiday entitlement, flexible working hours and travel. A small number of academics in new universities were motivated by a desire to achieve academic success, where they had previously fallen short of their abilities. One education lecturer explained that she learned how to be a ‘good student’ while doing a part-time postgraduate course and wanted to encourage others to experience the same satisfaction. A similar motivation was found among some postgraduate students who had underachieved at school and discovered an interest and ability in academic study as a mature student. Thus it appeared some were motivated by a desire to achieve success, and progression in their studies and career was tangible evidence of this.

For career changers, additional reasons may be relevant and these are reported in Table 5.17. The main reason given for changing career is that academia appeared to be more interesting than respondents’ previous work. This was a factor prompting almost two thirds of those who changed career to academia. For just over a third, it was the lifestyle of academics that appealed. In the qualitative research, head hunting was found to be a further prompt to entry, with some engineers reporting having been targeted for recruitment by the university when posts were proving difficult to fill. The hypothesis that some of those who changed career were intrinsically peripatetic is supported by the fact that 40 per cent of recent entrants felt they ‘wanted a change’ compared to 33 per cent of longer-established staff.

There were only small differences between staff in new and in old universities for changing careers (Table 5.18), with more changers employed in old universities having believed academia would be more interesting and fewer employed in old universities wanting a change. However, given the much greater numbers of career changers employed in new universities, new universities had a high percentage of staff who had entered academia because it appeared more interesting than their previous work (30 per cent) and because they wanted a change (19 per cent). One fifth had also changed careers to academia because they saw it as offering a better lifestyle, compared with only nine per cent in old universities.

102

Table 5.17 Academic staff: Factors prompting career change to academia

Per cent of career changers Less than 2 Years


Academia appeared more interesting 64 60 61 Academia appeared to offer a better lifestyle (e.g. hours, autonomy) 35 37 36

You wanted a change 40 33 34 No further progression was possible in your career 12 11 11 You were made redundant 2 6 5 You retired (including early retirement) 1 1 1 None of these 10 10 10 Total 100 100 100 • Source: NIESR/DfES Staff Survey, 2004

Table 5.18 Academic staff: Factors prompting career change to academia by type of university

per cent of career changers

per cent of all academic staff

Why changed career to academia old new old new total Academia appeared more interesting 65 59 17 30 22 Academia appeared to offer a better lifestyle 35 39 9 20 13 You wanted a change 31 37 8 19 12 No further progression was possible in your career 12 10 3 5 4 You were made redundant 4 6 1 3 2 You retired (including early retirement) 2 2 0 1 0 None of these 12 10 3 5 3 Total 100 100 26 51 36 Did not change career 74 49 64 Total 100 100 100 n 1726 1059 2785 • Source: NIESR/DfES Staff Survey, 2004

The types of considerations for career changers were explored in more detail in the qualitative research.

A reader in physics at an old university compared both the career structure and pay in Higher Education favourably with opportunities for physicists in industry. Higher Education was seen by academics in some fields, for example computing, engineering and law, as less pressurised than industry. Therefore, some said the more relaxed and informal academic ‘life-style’ had attracted them. A small number of academics from overseas also compared pay, research opportunities and career prospects as better in the UK, particularly compared to Eastern Europe and some other European countries, for example Germany.

103

In the areas of Health and Education, where all respondents had made a career change, the reasons for changing to an academic career reflected experience of working as practitioners. For health practitioners, enjoyment of teaching other health professionals and trainees was a prime motivation to enter an academic career. Another strong motivation was to escape the stress, shifts and routine of working in the health service. In a similar vein, teachers had felt the need for more variety and challenge, although some were also driven by an interest in research. A number of health and education lecturers had considered the option of moving to a management position in their field, for example head teacher or NHS management, but had wanted to remain more closely connected with practice than they felt management positions allowed.

While movement into Higher Education was a positive choice for lecturers in health and education, the engineering lecturers in the new universities had been prompted by redundancy from engineering jobs and, in one case, difficulty finding another job in telecommunications. As stated above, these lecturers had been targeted for recruitment by their university.

The reasons for entering an academic career were different again for the academics in art and design, with the principal reason being to fund artistic practice. However, teaching and the contact with students was a further attraction, which was seen as both inherently enjoyable and benefiting their own artistic practice.

5.5 Students as a source of entrants As we have seen, students in the UK provide the largest source of entrants to

academic employment in England (students in UK institutions represent 29% of the known entrants to academic employment in the Higher Education sector). Little evidence exists on whether research degrees such as PhDs do provide an appropriate induction into academic life, although Golde and Dore (2000) found in their study of US graduate students that graduate study is not enough to prepare students for an academic career102. Problems with the recruitment of future academics from research postgraduates are more apparent. There is evidence that the supply of potential academics from PhD programmes has become very tight in recent years in areas such as economics (Machin and Oswald, 1999, 2000) and chemistry Coe and Boddington (1999). This has led to worries being expressed about the quality of the students that progress from doctoral studies or post-doctoral posts into academia, particularly given the increasing reliance on non-UK postgraduates103 (ibid). The EPSRC (1999) survey of postgraduate students and permanent staff in IT and computer science found that attracting strong British candidates to PhDs and the fact that PhD students were not moving on to academic careers were bigger problems than retaining staff. One of the problems may be the increasing indebtedness of undergraduates (due to changes in funding), leading to a reluctance of undergraduates to continue to higher degrees.

102 Golde and Dore (2001) were concerned that doctorates do not provide enough experience of the practical teaching and administration activities of academic staff. Even in research, they concluded that doctoral students were too focused on one narrow area of research, both in terms of subject and method of research. 103 Both HEFCE (2003) and the case studies undertaken in UCEA (2002) suggest that staff shortages are encouraging HEIs to look overseas.

104

This section provides more information about this route into Higher Education employment. Of particular interest is identifying the stages at which students enter academic employment or leave the route into academic employment. The First Destination Supplement (FDS) is a useful indicator of the movement between study and employment, which can be matched to the Student Record in order to analyse the data by student characteristics. The first destinations of students leaving Higher Education in the 2001/2 academic year are tabulated in Table 5.19.

The most common destination for all types of student is into some type of employment. One-fifth of students graduating with a bachelors degree go on to further study or training. Two-fifths of these graduates go on to some kind of taught course, roughly equal numbers studying for a taught higher degree or a diploma or certificate, and 13% proceed directly on to a research degree (Table 5.20). Students studying for a masters degree are less likely to enter into UK employment, although they are more likely to find employment overseas (Table 5.19). Of the 17% of masters graduates who continue with their studies, more than two-thirds undertake a research degree.

Table 5.19 First Destination, by degree (%)

Bachelors Other undergrad Masters Other

postgrad Doctorate

UK paid full-time employment 53 48 48 85 62 UK paid part-time employment 7 6 4 6 4 UK self-employed 1 1 3 0 1 UK unpaid employment 1 0 0 0 0 Other employment in UK* 1 1 2 2 7 Employment overseas 3 1 12 1 15 Undertaking study or training 20 37 17 1 5 Not available for employment or training 6 2 5 2 3

Assumed to be unemployed 7 3 5 2 2 Others 2 1 4 0 1 Total 100 100 100 100 100

• * category and/or mode not reported • Source: HESA Combined Student/Module Record 2001/2

Table 5.20 Nature of study for those going into further study (%)

Bachelors Other undergrad Masters Other

postgrad Higher degree - research 13 0 68 21 Higher degree - taught 40 3 11 44 Diploma or certificate (including PGCE) or professional training course 38 4 15 27

First degree course 5 90 2 3 Private study 1 0 1 1 Other study or training 3 3 3 5 Total 100 100 100 100

• Source: HESA Combined Student/Module Record 2001/2

105

These figures show that dropout from pursuing the qualifications which enable one to enter academia (as a first career) (i.e. from a bachelors degree to a higher degree and from a Masters degree to a research degree) are high at each stage. Therefore, at whatever stage people decide to become an academic, they either are pursuing a qualifications route similar to many who have not decided on this route or many are changing their minds.

It would be useful to know the percentages entering academia from each type of Higher Education qualification. Unfortunately, the FDS does not provide a fine enough occupational breakdown to identify this. Nearly half of the students in the FDS (who report the industry of their employment) who go into the Higher Education sector to work (SIC 8030) (in any occupation) are graduates with a first degree (Table 5.21). There are two reasons for this. First, there are large numbers of students graduating with a first degree compared to those with post-graduate qualifications. (Only two per cent of students graduating with a bachelors degree enter the Higher Education sector.) Second, first degree graduates often enter into non-academic jobs in the sector, such as clerical and administrative staff. Only one-quarter of first degree graduates enter teaching or some form of scientific or technical role (Table 5.22). They are primarily other occupations.

Table 5.21 Breakdown of students entering employment in the Higher Education

sector (by SIC code) (%)

Breakdown of students entering HE

% of qualification group entering HE

Doctorate 34 47 Masters 16 7 Other postgrad 2 1 Bachelors 47 2 Other undergrad 1 0 Total 100 • Source: HESA Combined Student/Module Record 2001/2 Table 5.22 Occupation of graduates entering HE

Scientists & engineers Teachers Technicians Other Total

Doctorate 38 47 2 14 100 Masters 12 39 3 45 100 Other postgrad 1 29 1 69 100 Bachelors 8 14 4 73 100 Other undergrad 2 14 6 78 100 Total 18 29 3 49 100 • Scientists & engineers = SOC groups 2000 (chemists) to 2243 (veterinary research workers) • Teachers = SOC groups 2300 (university teaching professionals) to 2302 University and HE

lecturers • Technicians = SOC groups 3000 (Laboratory technicians) to 3090 (Other scientific technicians

n.e.c.) • Source: HESA Combined Student/Module Record 2001/2

106

Table 5.23 First Destination, by degree and subject (%)

Clinical & nursing Science Engineering Computing Business &

economics Law Others

First degree UK Employment 16 49 9 2 3 1 20 Employment overseas 14 55 9 2 4 1 16 Undertaking study or training 15 53 5 0 3 1 24 Not available for employment or training 13 49 5 1 1 2 30

Assumed to be unemployed 8 55 16 2 1 0 18 Others 7 45 13 4 5 0 27

Total 15 50 9 2 3 1 20 Masters UK Employment 4 17 6 14 18 3 38 Employment overseas 2 13 14 6 24 9 32 Undertaking study or training 7 22 7 6 7 6 45 Not available for employment or training 3 11 14 8 22 4 38


Total 4 17 8 11 18 4 38 Doctorate UK Employment 13 14 6 5 15 2 46 Employment overseas 4 15 6 3 15 2 55 Undertaking study or training 5 24 5 3 8 11 44 Not available for employment or training 4 19 6 4 17 4 47


Total 10 17 6 5 13 4 46 • Source: HESA Combined Student/Module Record 2001/2

5.5.1 The supply of PhD students There is evidence on one important factor affecting the supply: the supply of

UK-domiciled PhD students.104 The number of doctorates awarded annually to UK domiciled students nearly doubled between 1994/5 and 2003/4 (Figure 5.1). This was over a period when the number of academic staff in Higher Education increased by around 11% (HEFCE, 2002)105. Over a similar period, the percentage of new entrants to teaching posts with a PhD rose from 35% to 41% between 1994/95 and 1998/99 (PREST 2000). As more demands have been made on potential applicants, worries have been expressed about the quality of applicants to academic jobs, particularly as higher education institutions have increasingly drawn on non-UK postgraduates.

104 Note that, contrary to this, figures provided by Alan Howarth MP in the debate on the Second Report from the Science and Technology Committee on the Research Assessment Exercise (Hansard, 27th June 2002) suggested that between 1995/6 and 1999/2000, the number of doctorates awarded to UK-domiciled students fell by 9%. Obviously, the supply would be a lagged effect of the decision to pursue a PhD three or more years previously. 105 Note that this excludes staff in medical cost centres and those who are less than 40% FTE.

107

Figure 5.1 UK and overseas domiciled students obtaining PhDs

0

2

4

6

8

10

12

1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04

Stud

ents

obt

aini

ng P

hDs

('000

s)

UK Domiciled

Overseas Domiciled

• Source: Table 14a HESA On-line Information Service HESA

(http://www.hesa.ac.uk/holisdocs/pubinfo/stud.htm)

5.6 Students’ career intentions When directly asked, 58 per cent of students surveyed felt that they had

seriously considered and compared different careers106, i.e. 42 per cent had not. The qualitative research suggested that lack of knowledge and not just lack of thought may have led to this:

‘My first choice is a university post but plan B would be a research-based job in the voluntary or public sector, but my thinking or knowledge about job opportunities is not well developed at present’. (End of second year in Arts subject, old university)

‘I haven’t really thought about careers. It’s possible I don’t know about much else at this stage in my life.’ (End of third year, Social Science, old university).’

Fifty-eight per cent having seriously considered and compared careers might seem rather low, considering that research degrees are often seen as the entrance 106 This was in response to the question ‘Would you say that you have been seriously considering and comparing careers?’ The responses to an alternative question are tabulated in Table 5.25, and show that, even if students feel they have not given serious consideration to a range of careers, they felt able to state what type of career they intended to pursue.

108

qualification to work in academia. Why then do students undertake postgraduate research if not for their career? In contrast to the above figure, when asked about the importance of a number of factors in deciding to do their current degree, 63 per cent felt that they thought that it would help them get the type of job they wanted was either ‘very’ or ‘extremely’ important (Table 5.24). However, only 44 per cent said that wanting an academic career was a very or extremely important reason for doing a research degree. By far the most important reasons for postgraduate study were interest in the subject of study and research in particular. Few are doing their degree to change careers, but more in order to further their existing career.

Table 5.24 Research students: Importance of reasons for study

important, cumulative % not important, cumulative %

extreme-ely

at least very

at least quite

not very /not at all/NA

not at all/NA

I was interested in the subject 53 87 98 3 1 I was interested in research 50 82 95 5 1 It will help me to get the type of job I want 33 63 84 16 6

I wanted an academic career 24 45 71 28 8 I wanted to improve my prospects in my existing career 19 43 58 42 32

I like being a student 10 26 47 53 25 My previous tutors/lecturers encouraged me 8 26 54 46 31

I wanted to change careers 7 14 26 74 57 I could not think what else to do 3 7 16 85 68 It was better than getting a job 2 6 17 83 68 I wanted to come to the UK 1 5 13 88 79 I needed a European qualification 1 3 8 92 86

• Source: NIESR/DfES Research Student Survey, 2004

This interest in research and the subject was explained further in the qualitative research, where some respondents appeared to see a research degree just as an extension of their studies, often in tandem with not having thought about careers. It seemed that these respondents felt they had not yet finished their education, until they had done a PhD. Stimuli to this interest or to doing a PhD were their undergraduate research project or having further study suggested to them by a tutor or supervisor.

In addition to the above reasons, the qualitative research found that a desire to achieve could lead to doing a research degree. A mature Social Science student explained,

‘I left school with no qualifications and worked briefly in a factory. This was enough to know it wasn’t for me so I went to FE college and then to university at 20. I had a personal challenge to see how far I could go and I wanted to prove to others, teachers and people in my

109

village, who thought I was good for nothing and I’d never make anything of myself’.

Academic careers are not unpopular among research postgraduates (Table 5.25). Almost two-fifths are keen on having an academic career and a further one-fifth feel that one is of equal interest with a number of other careers. This is consistent with the view that many of these students made a decision to enter, or at least aim for, a career in academia when deciding to undertake postgraduate research. It is interesting to note that 12 per cent of the respondents are aiming for a career outside academia, but felt that an academic job would help this career. This result may help to explain the number of academics who leave early in their careers. Indeed a survey by IRS (2002) of new entrants found that only just over half felt that they were likely still to be in the HE sector five years after the survey took place. It seems that a sizable minority of those undertaking an academic career sees it as providing some kind of credential required for another career or something to be experienced prior to another, more permanent, career. This is particularly true in engineering, where only one-quarter of research postgraduates are keen on an academic career and a further one-quarter see academia as a stepping stone to another career (Table 5.26). In contrast to this, 60 per cent of those in ‘other’ humanities (such as history) are keen on an academic career and the majority of the remainder feels that a number of careers are of equal interest with academia.

Table 5.25 Research students: What best describes your career intentions?

% You are keen to have an academic career 38 A number of careers are of equal interest with academia 21 You'd prefer a career outside academia (and an academic job first is not important) 19

You are aiming for a career outside academia but an academic job would help in this career 12

You may go in for academia because you cannot think of other careers you want to have 2

You don't know what you want to do 5 You are not thinking about careers 3 Total 100 • Source: NIESR/DfES Research Student Survey, 2004

110

Table 5.26 Research students: Career intentions by degree subject (excerpt) (%)

Keen to have an

academic career

Number of careers of equal

interest with academia

Aiming for career outside

academia but an academic would

help in this career

Prefer a career outside

academia

Business and administrative studies 69 17 6 9 English literature and classics 60 21 8 8 Other humanities 60 20 3 14 Architecture, building and planning 58 11 16 16 Art and Design 54 8 8 21 Medicine and dentistry 53 15 15 14 Mathematical sciences (excluding computing) 53 29 12 6

Modern languages 48 22 11 19 Computing science 47 19 8 19 Social studies, including economics and social/economic geography 47 20 9 23

Education 42 16 16 26 Combined studies across subject groups 37 36 11 14

Subjects allied to medicine 36 29 16 19 Biological sciences 35 27 16 21 Agriculture and related subjects 33 33 17 17 Physical sciences 29 23 13 33 Engineering 26 20 24 29 Other technology 0 18 64 18 Total 41 22 13 21 • Percentages based on an un-weighted total of less than fifty are shown in italics • Note that rows do not add up to 100 as we have excluded the three smallest groups in Table 5.25 • Source: NIESR/DfES Research Student Survey, 2004

The qualitative research suggested that, for some, staying in academia was a conservative approach to careers. As has already been described, a large minority of students had not seriously considered alternative careers and, in the qualitative research, settling on academia seemed to be to opt for ‘what one knew’ or because of lack of confidence to go into the unknown. Postgraduates described being ‘comfortable’ in the academic environment. This combined with an interest in research and their subject area led to wanting an academic career. In some cases alternatives were quickly dismissed: one postgraduate in computing said she could not picture herself in industry, suggesting lack of confidence in a non-academic environment.

Aspirations and expectations can often differ markedly, however. We can look at the question of what research postgraduates feel the future holds for them from a different angle by considering what type of job they would most like and what they think it is probable that they will actually get. The job that most would like to have after completing their degree is a university research post, followed by a university lectureship (Table 5.27). Similar numbers (around one-third) would like a non-university research job in the public and private sectors. Again similar numbers (just under one-quarter) would like another type of job in the public and private sectors.

111

This ordering of university research post, lecturer, research job outside universities and other job is fairly consistent across subjects (see Table 5.28). Students of the sciences state a slightly stronger preference for research than the social sciences and humanities, although in medicine and dentistry it appears that the preference is rather for lecturing over research.

Table 5.27 Research students: Desired and probable jobs

% would like job

% think will probably get

job University research post (e.g. Research Assistant, Research Fellow) 55 53

University lecturer 42 26 A research job outside universities but in the public sector 33 23 A research job outside universities in the private sector 32 24 Another type of job in the public sector 23 25 Another type of job in the private sector 23 26 • Note that columns do not add up to 100 as respondents could answer yes to more than one type of

job • The ‘% would like job’ column depicts the percentage who answered ‘yes’ to whether they would

like a job after completing their degree • The ‘% will probably get job’ column depicts the percentage who answered ‘yes’ to whether they

think it is probable that they will get that job after completing their degree • Source: NIESR/DfES Research Student Survey, 2004

The respondents appear to be rather pessimistic about their probability of actually getting these jobs. With the exception of a university research post, fewer think they will actually become university lecturers or get other research jobs than would like them. This is not the case for other types of job in either the public or private sector. Indeed, our results suggest that the opposite may be the case, with more people thinking they will get these jobs than would like it, although the difference is not great. At the same time, given the percentage of entrants to Higher Education who enter on a research post (see above), research students may be over-estimating their chances of gaining a lecturing job immediately after their degree. Even those who expect to get an academic job, may not see this as an easy career. Certainly, in the qualitative research, some expected to find the passage to a permanent job in Higher Education a difficult one and were expecting a number of years in temporary research posts and lectureships before securing a permanent job. Competition was seen as particularly strong in old university departments and in popular subject areas, such as English.

112

Table 5.28 Research students: Desired jobs, by subject

University research

post

University lecturer

Research job outside university

Another type of

job Architecture, building & planning 77 54 35 38 Combined studies across subject groups 66 50 34 21 Other humanities 65 63 18 21 Modern languages 65 60 15 18 Computing science 64 48 37 24 Agriculture & related subjects 64 9 50 . Mathematical sciences (excluding computing) 62 48 36 14

Biological sciences 61 29 41 21 Art & Design 57 43 29 25 English literature & classics 57 71 17 17 Physical sciences 54 26 39 27 Engineering 52 35 43 27 Other technology 50 25 46 33 Social studies, including economics & social/economic geography 48 51 30 22

Subjects allied to medicine 46 37 26 25 Education 38 38 22 22 Medicine & dentistry 37 49 25 19 Business & administrative studies 35 78 22 28

• Note that columns do not add up to 100 as respondents could answer yes to more than one type of job

• Table depicts the percentage who answered ‘yes’ to whether they would like a job after completing their degree

• Final two columns are averages of public and private • Percentages based on an un-weighted total of less than fifty are shown in italics • Source: NIESR/DfES Research Student Survey, 2004

5.6.1 What do research students want from a job? There are many factors that influence career choice. A job can be thought of as

a set of characteristics; similarly goods can be seen as sets of characteristics (Deaton and Muellbauer, 1980, Ch. 10; Lancaster, 1971). This is the approach taken in analyses of job satisfaction (e.g. Clarke, 1997, 2001; Clark and Oswald, 1996; Oshagbemi, 1996, 1998; Ward and Sloane, 2000). Whilst pay might be one of the more obviously desirable characteristics of a job, it is certainly not the only one. Individuals also value other aspects, such as autonomy, flexibility, stability, and working conditions. Research students appear to look more for the non-pecuniary aspects of their job, such as a good working environment, variety, freedom to use initiative and seeing tangible outcomes from their job than a high salary and a good pension (Table 5.29). Close behind these desiderata are autonomy in the job, control of their research, career prospects collaboration and flexibility of working hours. Also important are good physical work conditions, helping people and job security. This is consistent with the literature that it is the content and non-pecuniary elements of the work itself (e.g. IRS, 2002) that is important. This does not mean that financial reward

113

is unimportant and not a potential source of recruitment and retention difficulties, but rather that the current crop of potential applicants are willing to accept lower levels of pay (at least in the short run) if a job offers other, non-pecuniary benefits.

Table 5.29 Research students: Importance of factors in career choice, cumulative %

extremely important

at least very

important

at least quite

important

at least slightly

important total good working environment 34 79 96 99 100 freedom to use your own initiative 37 77 96 99 100 variety in the job 26 72 94 99 100 seeing tangible outcomes from the job 27 70 92 98 100 good career prospects 26 65 88 96 100 flexibility over working hours 28 64 91 99 100 control over what research you do 21 62 90 96 100 substantial degree of autonomy in the job 20 61 91 97 100 having a collaborative/ team working approach 24 61 85 95 100 job security 21 56 85 96 100 good physical work conditions 17 54 84 96 100 helping people 19 52 81 95 100 good pension 14 44 75 91 100 a socially useful job 15 41 70 88 100 clear progression route 11 38 70 89 100 high salary 9 34 76 94 100 high status job 9 30 61 82 100 foreign travel 10 30 60 82 100 contact with customers/ clients 10 29 56 79 100 managing people 5 18 45 74 100 a public profile 4 14 36 64 100 being the boss 3 12 34 62 100 high pressure 2 9 32 64 100 good working environment 34 79 96 99 100

• Figures depict responses to question ‘The following lists factors which influence career choice. How important is each in your career choice?


Turning to the elements specific to the academic job, we find that it is research, and in particular that specified by oneself that is most important to potential recruits (Table 5.30). This re-enforces the result that respondents desire autonomy in their jobs and that changes in policy either at the department/university or the sector level that change the balance between individual initiated and department specified research are likely to create problems of staff retention caused by staff dissatisfaction.

114

Table 5.30 Research students: The importance of elements of the academic job (%)

Strongly

prefer included

Strongly or

slightly prefer

included

Do not prefer either way

Strongly or

slightly prefer

not included

Strongly prefer

not included

Total

Teaching 30 75 21 13 5 100 Research 63 90 8 2 1 100 Own research 58 88 10 3 1 100 Research specified by others 10 47 30 22 8 100

• Note that the total is the sum of columns 3, 4 and 5 (as column 2 is already included in column 3 and column 6 is already included in column 5).

• Figures depict responses to question ‘How much would you prefer a job which included or excluded teaching/research/doing your own research/doing research specified by others?


One issue that creates problems of retention, highlighted by a number of earlier studies, is the preponderance of fixed-term jobs in academia, particularly in research and early-career posts (e.g. Bett, 1999; UCEA, 2002). If the job security provided by a permanent or open-ended contract is important for current and potential academic staff, the increased use of fixed-term contracts is likely to be the cause of dissatisfaction and possibly difficulties in staff retention. Just over half of the respondents felt that it was very important for them to have a permanent job within two years of getting their degree (Table 5.31). Over 80 per cent felt that it was at least quite important. This is in sharp contrast with the number of fixed-term contracts offered to individuals in their early career (see Table 5.3) If we extend the time frame to five years, the proportion of respondents who feel it is very important to obtain a permanent job rises to over 70 per cent. Again, these figures support earlier work based on studies of personnel departments that fixed-term contracts create a lack of permanence that is not popular with staff and is the cause of staff retention problems. We return to the issue of job security in Chapter 6.

Table 5.31 Research students: The importance of a permanent job, cumulative %

very important

at least quite

important

at least slightly

important total

Permanent job within two years of getting degree 53 81 94 100

Permanent job within five years of getting degree 71 89 95 100

• Figures depict responses to question ‘How important is it to have a permanent (and not a temporary) job within two/five years of getting your degree?’


What attracts research postgraduates to academia, and those who have changed career in particular, is the nature of the job itself. This reflects not only a desire to continue to pursue research, but also the importance placed on aspects such as autonomy and flexibility. What is likely to dissuade them is the lack of stability and

115

security, particularly at early stages in the academic career. It appears that these potential academics demand higher levels of job security than is offered.

5.7 Does academia offer what research students want? It is important to understand both what research students want from their

career and whether they consider academia offers this. If they do not believe academia offers what is important to them, then they will not apply. This section describes research students’ perceptions of academia and the foundations of their knowledge and then compares their aims of a career with their perceptions of academic careers.

The study of research students is important in aiding our understanding not only of academic recruitment, but will also shed light on issues relating to the retention of academic staff. This is because these courses provide an induction into academic life and create expectations of what such work involves. Considered like this, it is not the first appointment but the PhD that should be thought of as the first stage of the academic career (Anderson, 1998; Austin, 2002). If the expectations formed during postgraduate research prove to be incorrect or unrealistic, they are likely to cause staff to become dissatisfied with their jobs and leave academia early in their careers. We therefore also consider how well students’ perceptions match reality.

5.7.1 Students’ perceptions of academia An important element of recruitment is the perceptions that prospective

academics have of academic employment. Incorrect perceptions can create recruitment difficulties in two ways: First, if potential academics underestimate the benefits or have generally negative views of academic employment, this will make them less likely to apply for and enter jobs in higher education. Second, if new entrants discover that their expectations of academic employment are incorrect, this is likely to lead to dissatisfaction and retention difficulties. This dissatisfaction and turnover of academic staff is the subject of another chapter. Here we describe the perceptions that prospective academics have of academic work. First, we describe the extent to which research students have direct (and second-hand) knowledge of working in academia.

5.7.1.1 Prior knowledge of academic employment

Decisions that potential recruits to academia make are likely to be informed by prior experience of academic employment. This not only relates to actual experience of research or teaching itself that individuals have themselves, but also knowledge derived from their own tutors and lecturers and from others, including their parents. The study gathered data on previous research and teaching experience and parental occupation, but did not examine knowledge gained from others. However, entrants to academia will almost certainly have gleaned some information or views on academic employment from their own tutors and lecturers.

The occupation and income of parents have consistently been found to be an important influence on occupational choice, education level and earnings (Dutta et al., 1999).There is evidence of quite a high level of research students’ parents having experience of teaching and research, both within higher education and elsewhere. Around 12 per cent of the research students in our survey have at least one parent who

116

has worked as an academic and a further four per cent have at least one parent who has had some other job in higher education (Table 5.32). Furthermore, almost a quarter of the respondents have one or more parent who has been a teacher or lecturer outside higher education. Since the whole of the education sector accounts for only eight per cent of the population107, research postgraduates are much more likely to come from families with parents who have some form of experience of the education system. This does of course reflect the class make-up of students in higher education and it is not possible to disentangle these two influences with the data at hand. Nevertheless, it suggests that an important influence on the supply of potential recruits into academia is parental experience of academia and the education sector generally.

Table 5.32 Research students: Parental experience of teaching/research

% As an academic (lecturing or research) in higher education 12 In another job in higher education 4 As a researcher outside higher education 4 As a teacher/lecturer outside higher education 23 • Source: NIESR/DfES Research Student Survey, 2004

Turning to the actual work experience of research postgraduates themselves, we can see that a sizeable minority of them has some experience of employment prior to their degree (Table 5.33). This may reflect a body of students who have returned to education to ‘top-up’ their education credentials. This may be because a PhD is increasingly being seen as a pre-requisite for either work in academia generally or higher ranking posts in particular. A large proportion of research postgraduates has been employed in some form of teaching or research role during their current degree (Table 5.34). This suggests that many new entrants to academic employment already have some practical experience of what the job entails.

Table 5.33 Research students: HE Employment prior to current degree

% In a research post (paid on the standard research pay scales) 15 In any other research post 10 As a lecturer (paid on standard lecturing pay scales) 8 As a lecturer (seasonal) 5 To do occasional lecturing 5


Many of the respondents have had experience of the elements of academic employment whilst they have been students (Table 5.35). The majority have undertaken some form of teaching or lecturing, almost all have done some marking and just under two-fifths have conducted additional paid research whilst a student.

107 Quarterly Labour Force Survey.

117

Table 5.34 Research students: HE Employment during current degree

% Demonstrator 42 Tutor/seminar leader 26 Teaching assistant 13 Research Assistant/Fellow (appointed on the standard research pay scales) 13 Other research 8 Lecturer (appointed on the standard lecturing pay scales) 6


Table 5.35 Research students: HE Employment whilst student

% Teaching other than lecturing 72 Marking 49 Research 39 Lecturing 29


5.7.2 Research students’ perceptions of academic jobs Academic jobs were seen as scoring highly in terms of autonomy and

independence: control over one’s research, freedom to use one’s initiative, flexibility over working hours, autonomy and variety in the job (Table 5.36). Over two thirds also saw them as being socially useful, enabling one to help people, providing foreign travel, a good working environment, providing tangible outcomes from one’s work, allowing collaborative and team working and providing long holidays. Somewhat fewer, but still over half saw the jobs as high pressure, providing good physical working conditions, high status, providing a public profile, having clear progression routes, providing contact with customers or clients and the opportunity to manage people. However, fewer than half saw academic jobs as having good career prospects, providing job security, allowing one to be the boss, providing a good pension or having a high salary.

Certainly, the respondents to our survey suffer no illusions about academic careers offering a high salary. Eighty per cent of respondents feel that an academic career does not offer a high salary; one-third feel that it does not at all well in this respect. Only 41 per cent are aware that pensions, an important part of the remuneration package for academics, are relatively good. Research postgraduates are more sanguine about the possibility of offering various elements of autonomy and flexibility. In particular, a sizeable majority feel that an academic career does quite or very well in allowing control of what research they do.

118

Table 5.36 Research students: How well does an academic career offer (%)

Very well

Quite or very

well

Don't Know

Not very or not at

all well

Not at all well Total

control over what research you do 25 86 2 12 1 100 freedom to use your own initiative 32 84 3 13 2 100 flexibility over working hours 35 82 2 16 4 100 substantial degree of autonomy in the job 19 81 3 16 3 100 variety in the job 18 74 2 24 3 100 helping people 22 73 2 26 5 100 a socially useful job 23 73 3 24 5 100 foreign travel 21 72 3 26 6 100 good working environment 15 71 3 26 6 100 seeing tangible outcomes from the job 16 70 2 27 5 100 long holidays 23 67 3 30 7 100 having a collaborative/ team working approach 18 67 3 31 8 100 high pressure 17 63 2 35 7 100 good physical work conditions 12 61 3 37 9 100 high status job 13 55 2 44 11 100 a public profile 10 55 3 43 10 100 clear progression route 11 53 3 43 12 100 managing people 8 52 2 46 11 100 contact with customers/ clients 10 52 3 44 15 100 good career prospects 9 48 3 49 13 100 job security 11 47 3 49 22 100 being the boss 6 44 2 55 15 100 good pension 7 40 17 41 13 100 high salary 3 18 2 80 33 100

• Figures depict responses to question ‘How well do you think an academic career offers …?’ • Source: NIESR/DfES Research Student Survey, 2004

Turning to what precisely research postgraduates believe academics do, just over half of research postgraduates believe that academic employment will include fairly similar amounts of research and teaching (Table 5.37) and a further 29 per cent feel that it will consist mainly of research.

Table 5.37 Research students: Perceptions of academic employment

% Fairly similar amounts of research and teaching 53 Mainly research 29 Mainly teaching 18 Total 100 • Source: NIESR/DfES Research Student Survey, 2004

5.7.3 Do students think academia offers what they want? We saw in Section 5.6.1 above the range of factors that research students

wanted from a job. An important question for the supply of new academics to the higher education sector is whether potential recruits feel that an academic career offers what they desire from a career. If potential recruits do not feel that an academic

119

career offers these desiderata, this will reduce the numbers entering academia. Broadly speaking, potential recruits do appear to feel that an academic career does offer them what they want from a career (Table 5.38).

Table 5.38 Research students: Comparison of research students’ desired job attributes and expected attributes of an academic job

very important job attribute

quite or very well provided in academic

jobs good working environment 79 71 freedom to use your own initiative 77 84 variety in the job 72 74 seeing tangible outcomes from the job 70 70 good career prospects 65 48 flexibility over working hours 64 82 control over what research you do 62 86 substantial degree of autonomy in the job 61 81 having a collaborative/ team working approach 61 67 job security 56 47 good physical work conditions 54 61 helping people 52 73 good pension 44 40 a socially useful job 41 73 clear progression route 38 53 high salary 34 18 foreign travel 30 72 high status job 30 55 contact with customers/ clients 29 52 long holidays 21 67 managing people 18 52 a public profile 14 55 being the boss 12 44 high pressure 9 63


The important exceptions are good career prospects (very important to 65 per cent and seen as being at least quite well provided by 48 per cent), job security (very important to 56 per cent and seen as being at least quite well provided by 47 per cent), and high salary (very important to 34 per cent and seen as being at least quite well provided by 18 per cent). Since the early part of many academic careers consists of temporary or fixed-term employment, this may put off a number of potential recruits. As we saw above, the desire for job security, as measured by the importance of a permanent job, increases over a career (Table 5.31). This suggests that although academic staff may put up with some insecurity at the early stages of a career, they are less willing to as their career progresses. This may be, for example, due to increasing family commitments.

120

5.7.4 Are students’ expectations met? In order to gain some understanding of whether the expectations of potential

entrants to higher education are met, it is useful to consider the experience of recent entrants, rather than all academic staff. In what follows we use the term ‘recent entrants’ to describe those staff whose first job in academic employment was in 2002 or later. This represents around 15 per cent of the sample. We shall refer to the remainder of staff as ‘established’. It has long been known that many new job matches end early (see Farber, 1999, for a summary of work in this area). This is in part because jobs are inspection goods and potential workers are uncertain about match quality, i.e. whether the job is right for the worker and the worker is right for the job (Jovanovic, 1979).

We have noted that an important factor that research postgraduates look for in a job is that it be permanent (Table 5.31). Just over half thought it was very important that they had a permanent job within two years of getting their degree, and 71 per cent thought it important to obtain one within five years. Over two-thirds of recent entrants are on some form of temporary contract (Table 5.39). This figure falls to just over one-half for those who have worked in the higher education sector for between two and five years. This suggests that there is a dissonance between what research postgraduates want from jobs and what academic employment can offer them in terms of their contract type.

Table 5.39 Research students: Contract type

Recent entrants

Between 2 and 5 years

Five years or more Total

Permanent (or open-ended) 30 47 82 68 Fixed-term contract 67 51 17 30 Termly/Seasonal 1 0 0 0 Other non-permanent 1 1 1 1 Total 100 100 100 100


We saw above that just over half of research postgraduates feel that academic employment will include fairly similar amounts of research and teaching and the majority of the remainder feel that it will consist mainly of research (Table 5.37). In reality, recent entrants spend around half of their time on research and around a quarter on teaching (Table 5.40). The amount of research time falls with experience with the hours being more likely to be spent on teaching and administrative tasks. Note that it is not only the time spent on teaching and administration relative to research that increases but total hours of work per week. If research is seen as one of the positive elements of academic life and administration one of the negative ones, one can clearly see how satisfaction with work might fall with experience. Nevertheless, research postgraduates do appear to have a fairly good idea of what academic work consists of, at least in the early stages of their careers.

121

Table 5.40 Academic Staff: Hours of work in academic employment

Hours of work Mean Percentile

25 Median Percentile 75

Recent entrants Total per week 43 40 42 50 Teaching 10 0 4 18 Research 23 5 20 38 Admin tasks 6 0 5 10 Other activities 5 0 2 6

More than 2 years experience Total per week 47 40 47 55 Teaching 15 5 15 20 Research 15 3 10 25 Admin tasks 12 5 10 16 Other activities 5 0 4 7

Whole sample Total per week 46 40 45 52 Teaching 14 4 14 20 Research 16 4 10 28 Admin tasks 11 4 10 15 Other activities 5 0 3 7

• Number of hours spent teaching includes time with students, preparation and marking • Number of hours spent doing admin tasks includes student recruitment • Source: NIESR/DfES Staff Survey, 2004

Although a high salary was not expected to be one of the more desirable aspects of employment, a dissonance between the expectations of potential recruits and the salary they in fact earn is likely to cause dissatisfaction and possibly retention problems. Research postgraduates are, if anything, a little too pessimistic about the earnings they could expect in academia. Slightly less than three-quarters of them expect to earn between £15,000 and £25,000 in their first job (Table 5.41). The mean salary of the academics in the staff survey in the first year of academic employment was at the top end of this band (Table 5.42). As one might expect, the variation in predicted earnings increases the further our respondents look into the future. Around two-thirds of them expect to earn between £20,000 and £30,000 after five years. The mean salary of the academics in the staff survey in the first year of academic employment was £28,579. Again, this is at the upper end of the band predicted by the prospective academics. The median and modal prediction of expected earnings after ten years is in the £30,000 to under £35,000 band. Once more mean salary of the academics in the staff survey in the first year of academic employment was at the upper end of this band (£34,082).

122

Table 5.41 Research students: Expected earnings in academia

In first job, after degree.

After five years

After ten years

Under £ 10,000 1 0 0 £10,000 to under £15,000 11 1 0 £15,000 to under £20,000 41 8 1 £20,000 to under £25,000 33 31 9 £25,000 to under £30,000 9 31 21 £30,000 to under £35,000 3 15 24 £35,000 to under £40,000 1 6 16 £40,000 to under £45,000 0 3 10 £45,000 to under £50,000 0 2 5 £50,000 to under £55,000 0 1 4 £55,000 to under £60,000 0 0 1 £60,000 or more 1 1 6


Table 5.42 Research students: Actual salary

Mean Percentile 25 Median Percentile 75 First year £24,932 £20,515 £24,500 £28,500 Five years £28,579 £24,323 £28,000 £32,375 Ten years £34,082 £29,478 £34,000 £38,000


The results of our survey suggest that research postgraduates have a fairly good idea of what to expect from an academic career. They are certainly under no illusions of it offering a high salary; indeed their expectations of earnings rather understate what they might in fact earn. There are of course many difficulties with making such comparisons, since those who choose to enter and to remain in academic employment may be different from those who do not enter or who leave, and those who leave of their own accord may be different from those who leave because their contract is not renewed. For example, one well documented finding is that wage changes are higher on average for those who leave of their own accord than for those who are laid off (Bartel and Borjas, 1981; Antel, 1985; Mincer, 1986). This may be because staff of high ability may be able to find better-paid employment elsewhere, whilst those of lower ability remain and only leave when pushed (Gibbons and Waldman, 1999). These two types of individual are likely to have different expectations of what wages they could earn.

5.8 The factors influencing the likelihood of students entering academia In this section we will investigate the factors influencing the likelihood of our

research postgraduates seeking academic employment. We do this by using their responses to the question ‘What best describes your career intentions?’ From the

123

responses, set out in Table 5.25 on page 109, we estimate a model of the strength of the preference for an academic career.

We also include variables to account for whether the respondent’s parents have worked in teaching or research jobs either in or out of higher education. However, it may be the case that these factors will have a stronger influence on the decision to undertake higher education and to progress to a higher degree than the decision to continue into an academic career. Nevertheless, since the respondents have already chosen to undertake postgraduate education, if parental occupation has an additional influence on whether an individual becomes an academic this will have important policy implications.

Perhaps more important than parental experience of academia at this stage of the decision process is the individual’s own prior experience of academic employment. We therefore include variables to account for whether the individual worked as a lecturer or researcher prior to undertaking their current degree and whether they have done so during it. One would expect a priori the former group to be more likely to continue into academic employment as they may be undertaking a research degree in order to top up their academic credentials. For those currently teaching or undertaking paid research, its effect on their entering into academia may go either way, it may give them a taste of the academic life or put them off it.

One important dimension of the recruitment difficulties facing institutions of higher education is the variation across subject areas. Therefore, we include subject dummies in order to determine whether there are differences that could lead to recruitment difficulties in the future (the baseline subject area is social studies). Finally we include university dummies in order to pick up any unobserved heterogeneity at the institution level.

The results are presented in Table A.22 on page 260 of Appendix F.

Most of the personal characteristics included in the model do not appear to have a significant effect on the desire for entering academia.

There are no significant differences between males and females, ceteris paribus. Any differences between men and women therefore are likely to be the result of differential rates of entry to higher degrees by research. Since there are similar numbers of men and women in our sample overall, this is likely to be due to the differential rates across subject area and so the direct effect on the likelihood of research postgraduates pursuing an academic career will be picked up in the subject dummies. The effect of having children appears to be negligible. Students who are over 30 years of age have a slightly stronger intention to enter academia, although this effect is only significant at the 10% level in the first specification in the table. Again students from non-white ethnic backgrounds are slightly more likely to enter academia108, but this effect is only significant at the 10% level in the second specification.

Students with an EU passport from outside the UK express a stronger desire to progress from their course into academic employment. This result is consistent across specifications.

108 Note non-white is defined by ethnicity and not nationality, i.e. it includes foreign non-white students.

124

Turning to prior experience of working in higher education, those who have worked as a lecturer in HE prior to current degree appear more likely to enter academia. This result is statistically significant at the 1% level and is consistent across specifications. Since the survey is of research postgraduates and not entrants to academic employment, it is likely that this effect represents academics returning to the sector and those already working in the sector who feel that they need to top-up their qualifications in order to get an academic job, or a better one. Doing paid work as lecturer or researcher during their current degree appears to have a positive effect on students’ desire for an academic career. Indeed those who do so are more likely to enter academia. This result is also statistically significant at the 1% level and is consistent across specifications. Having one or more parents who have worked in higher education appears not to have any effect on the likelihood of entering academia. Any such influence may have its impact earlier on in the process, affecting the likelihood of attending university or choosing to undertake postgraduate study. The only significant effect of parental occupation appears to be that of having a parent that has worked as a researcher outside higher education. The cause of this is uncertain, but it may reflect an increased probability of becoming a researcher outside HE.

Once more, the effect of previous qualifications may operate through the decision to undertake a postgraduate research degree rather than the likelihood of entering academia. Certainly, the 40 per cent of research postgraduates who have first class degrees is much higher than the rest of the graduate population. Research postgraduates with first class undergraduate degrees have a stronger desire to enter academia than those with upper second class degrees (the baseline in the estimated model), as do those with lower second class degrees or lower. The reasons for these two effects are most likely different. Only ten per cent of our respondents have a lower second class degree or lower, whereas forty per cent have firsts109. It is likely that the former had to overcome considerable difficulty in order to study for a higher degree.

Students whose research is in physics and engineering and technology are less likely to wish to enter into academia (the baseline is social studies). Students in languages and other humanities are more so, although the latter effect is not significant if we include the ‘don’t knows’.

5.9 Summary and conclusions

5.9.1 Source of academic recruits The above has shown that the supply pool of academics is wide: the main pool

should not be seen in terms of research students, but of research students and employees in other sectors making a career change. This varies by subject, with career change employees concentrated in vocational subjects. Other important, but much smaller, sources are foreign higher education institutions, foreign research institutes and students abroad.

109 Note that for four per cent of the sample, we do not know the degree classification of their first degree.

125

The qualitative research suggested that the labour market for some posts could be local. (Local recruitment was found in some vocational subjects at new universities.)

About one third of existing academics had made a career change to come to academia. This tended to be from higher level occupations (managerial, professional, associate professional and technical), rather than from lower level jobs (via being a mature student). Career changing is more common in those subjects which are regarded as vocational and demand similar skills inside and outside academia (namely Business and administrative studies (81 per cent) and in Computing (47 per cent), Subjects Allied to Medicine (46 per cent), Education (43 per cent), Social studies (42 per cent) and Engineering (40 per cent). It was least common in Physical Sciences (13 per cent), Biological Sciences (15 per cent), Modern languages (15 per cent) and English literature and classics (19 per cent). Half of those in new universities had made a career change, compared with about one quarter in old universities.

Entrants tended to be recruited to fixed-term contracts in research, although the percentage going on to permanent contracts rises with age, suggesting that previous skills are relevant and recognised. (92 per cent of those recruited to the sector under 30 were recruited to a fixed-term contract, falling to 47 per cent for those aged 50-59). Retrospective data from the survey showed that the tendency to recruit to research jobs and to fixed-term contracts had grown over time.

5.9.2 Career decisions Only just over half of academics and research students seemed to have made

career decisions based on serious consideration of alternatives (56 per cent and 58 per cent respectively).

Academia, as opposed to another career, had been the prime aim for just over half of existing academics. In addition, 21 per cent had been equally interested in other careers and 10 per cent had not known what they wanted to do. Interestingly eight per cent had seen academia as a launching pad for some other career. More academics in old universities and men had been set on an academic career, compared with academics in new universities and women. The percentage of academic staff who had wanted an academic career varied substantially by subject. Eighty-one per cent of academic staff in English literature and classics had very much wanted an academic career. The percentage was also high in other humanities, maths, modern languages and physical sciences. At the other extreme were staff in Art and design (only 36 per cent of whom had very much wanted an academic career), Business and administrative studies (38 per cent) and engineering (44 per cent).

Most commonly, the reasons for entering academia for those who were not committed to academia, either as a sole aim or as a stepping stone to another career, were because a specific academic job was particularly attractive (45 per cent) and because of being able to continue doing research (36 per cent). Only 21 per cent entered in order to continue teaching. From the qualitative research, it seemed that the attraction of research in academia was the freedom it offered to pursue one’s own interests and not to conduct research to someone else’s agenda.

A greater percentage of new recruits (compared with established academics) entered academia because they wanted to continue doing research for a while. The reverse is true for those who wanted to teach for a while. This suggests that, unless the

126

types of staff entering academia has changed over time, those who enter academia in order to undertake research either intend to leave after this desire is sated or because they discover that academic careers offer less in the way of research than they had hoped.

Further motivations for career changers were explored. Career change into academia was more often motivated because academia appeared more interesting (61 per cent of career changers), because academia appeared to offer a better lifestyle (36 per cent) or because a change was desired (34 per cent), than because of lack of further progression in the existing career, being made redundant or retiring. There were slight differences in motivations between career changers in old and new universities. However, owing to the much higher percentage of career changers in new universities, this translated into substantial differences in the motivations when we consider all staff: new universities had a much higher percentage of staff who had changed career because academia appeared more interesting than their previous work (30 per cent, compared with 17 per cent in old universities), because they had wanted a change (19 per cent, compared with eight per cent in old universities). One fifth had also changed careers to academia because they saw it as offering a better lifestyle, compared with nine per cent of staff in old universities.

In detail some of the factors prompting change included stress in previous job, shift work, routine working and wanting variety at work, academia being ‘more relaxed and informal’, career progression requiring movement into management (and wanting to stay with one’s profession). For art and design lecturers one factor appears to be to fund artistic practice.

5.9.3 Students It is unclear at what stage students tend to decide to pursue an academic

career. Dropout from pursuing the qualifications to enter academia (as first career) (i.e. from a bachelors degree to a higher degree and from a Masters degree to a research degree) are high at each stage. Therefore, at whatever stage people decide to be an academic, they either are pursuing a qualifications route similar to many who have not decided on this route or many have changed their minds.

Interest in the subject and doing research are the most common motivators to doing a research degree (very important for 87 per cent and 82 per cent respectively). The qualitative research found that, with these interests, some people had drifted into their research degree with little thought or felt that their education was not yet complete. Stimuli to this interest or to doing a PhD included doing an undergraduate research project or having further study suggested to them by a tutor or supervisor. Wanting an academic career was a very important reason for starting a research degree for 45 per cent of research students (of some importance to 71 per cent). Thus an academic career was not uppermost in the decision to do a research degree.

By the time of interview, 38 per cent of research students were keen to have an academic career and a further 21 per cent saw this of equal interest to some other career. Thus it appears that doing their research degrees may dissuade people from pursuing an academic career.

The student survey identified a sizeable minority of research students who expected to go into academia as a stepping stone to the career of their choice (12 per cent). This may help explain early turnover in academia. This is particularly true in

127

engineering, where only one-quarter of research postgraduates are keen on an academic career and a further one-quarter see academia as a stepping stone to another career. In contrast to this, 60 per cent of those in ‘other’ humanities (such as history) are keen on an academic career and the majority of the remainder feels that a number of careers are of equal interest with academia.

Lack of knowledge of alternative careers and lack of confidence in going into unfamiliar employment situations seemed to encourage some research students to enter academia. This has interesting implications for the desirability of providing good careers information, as it implies that better information may reduce entry to academia.

Research students seemed to see getting a university research post as easier than getting a research post elsewhere. Although they were pessimistic about gaining a lecturing job straight from their degree, one quarter thought they would. This seems overly optimistic given the high percentage of entrants to the sector who have a research post.

5.9.4 What do research students want from a job? One third of research students said that a high salary was at least very

important to them in their choice of career. Instead, they looked more for the non-pecuniary aspects of their job and, most often, to a good working environment, variety, freedom to use initiative and seeing tangible outcomes from their job. Close behind these factors are autonomy in the job, control of their research, career prospects, collaboration and flexibility of working hours. Also important are good physical work conditions, helping people and job security.

Academic jobs can have at least two major advantages in attracting research students: they include research and they may enable one to set one’s own research agenda. These were very important to 63 per cent and 58 per cent of research students, respectively, whilst 22 per cent did not want to do research specified by others. (In addition 27 per cent and 30 per cent respectively were slightly attracted by these elements.) Only 30 per cent were strongly attracted by teaching, but a further 35 per cent saw this of some interest, whilst only 13 per cent would prefer not to teach.

Fixed-term contracts appeared to be a disincentive to enter academia (or to remain in academia): 81 per cent saw it as at least very important to have a permanent job within two years of gaining their degree.

5.9.5 Perceptions of academia The above has described what students wanted of a job. Whether they apply to

academic jobs then depends, in part, on whether they think academia offers them what is important.

Research students should have good knowledge of the nature of academic jobs, having been employed in a research or teaching role during their current degree (including: demonstrator, 42 per cent; tutor/seminar leader, 26 per cent; teaching assistant, 13 per cent, Research Assistant/Fellow 13 per cent) or previously. At the same time, 12 per cent had a parent who had been a university academic and more than a quarter in some other university job or in teaching or research outside Higher Education. Indeed, the only dissonance identified between beliefs about academic jobs and reality is that research students underestimated academic pay. This may lead

128

to somewhat few recruits to the sector (although high pay tended to be a major consideration for a small minority of students), but is likely to reduce turnover once in academia.

Broadly speaking, potential recruits do appear to believe that an academic career does offer them what they want from a career. However, there are three important exceptions to this for career prospects, job security and high salary. At the same time, research students tend to underestimate the pay of academics. For those for whom pay is a major influence on career choice, correcting this misapprehension might lead to an increased interest in an academic career.

Personal characteristics had little effect on whether one sought an academic job, with the exception of non-British EU nationals, who exhibited a stronger desire to seek academic employment. Doing some teaching or paid research in Higher Education whilst doing their research degree (e.g. as a demonstrator or tutor) does not reduce the likelihood of seeking an academic job and may encourage it. Those with first class degrees were more likely to seek an academic job. Students whose research is in physics and engineering and technology are less likely to seek academic jobs and those in languages and other humanities are more so. If this lower expression of desire to pursue an academic career translates into a lower inflow of young staff in these areas, this is likely to have a detrimental effect on the ability of HEIs to maintain or expand staff levels, unless they can be found from outside the sector.

5.9.6 Policy implications and further research needs What are the policy implications of the findings on entrants? First, policy

decisions should be informed by the importance of career changers to the supply of academics and not focus solely on increasing the supply from research students. At the same time, more research is necessary on career changers (and potential changers) to understand how this source could be exploited more fully. Our study was unable to investigate properly the factors which affected career change, as it only interviewed those who had entered academia and not those who had not (whereas we were able to look at both groups for research students). Research should identify the factors which lead to career change into academia (and which deter it). At the same time, it would be useful to examine in more depth whether the route (career change or traditional research student path) has any implications for the effectiveness of academics.

It is clear that research students and career changers are attracted by many positive aspects of the academic job (e.g. working environment, research, flexibility, autonomy). To maintain the supply, it is important that these aspects are not eroded and, as we shall see in the next Chapter, changes in Higher Education may reduce these attractive elements. The job attraction of research amongst research students (and much less common interest in teaching) should be borne in mind in future changes to higher education. Academic jobs tended not to meet research students’ (and, probably career changers’) aspirations for pay, career prospects and job security. In respect of pay, this is not surprising as within the UK, academic pay is relatively low and remains relatively low throughout an academic’s career (see Chapter 4). A relative improvement in pay (and its progression over the career) might increase the supply of academics. At the same time, research students tended not to be aware of the sizeable addition to pay provided by the pension package (nor to place priority on this). Ensuring that this is seen as part of the pay package might be useful. In addition, the supply of academics might be improved through offering greater job security early

129

in the career. This may need imaginative approaches to how job security (with career development) might be provided (such as combining research and teaching, transfers between research posts, combining research posts and developing a formal career route from research to lecturing)110.

A further response is to encourage more recruitment from abroad. Pay was relatively good compared with a range of other countries (see Chapter 4) and it may be that concentrating on recruiting from selected countries could be expanded (from higher education institutions, research institutes and students). However, this approach is likely to have poor implications for retention (and long-term workforce development), as we shall see in the next chapter.

Research students who had some experience of paid research or lecturing were more likely to want an academic career. Whilst this could be a selection effect (those interested in an academic career do relevant paid work), it is likely also to affect career choice. Therefore ensuring a higher percentage of research students have this experience may increase the supply of academics.

For resource reasons, the study was only able to examine the career decisions of research students, not career decisions throughout higher education. It would be useful to understand how interest in research and an academic career develops at earlier stages, to see if policy changes could be used to increase the supply of research students interested in an academic career.

110 Whilst the Fixed-Term Employees Regulations 2002 provide for equal terms and conditions for contract and permanent staff and also the right to a permanent contract after four years (with some exceptions) it is not clear how far this is going to alter the use of fixed-term contracts.

130

131

6 Academic employment and careers

6.1 Introduction This chapter considers two main aspects of employment for academics: career

progression within academia and job satisfaction. It starts by describing career paths in terms of movement from fixed-term contract and research jobs to permanent, lecturing jobs. It then examines career expectations. The following section turns to job satisfaction, describing overall satisfaction and satisfaction with specific aspects of the job.

6.2 Career paths We have already seen that just under one-third of the new entrants into

academic employment are under 30 and they enter into Higher Education employment on a fixed-term contract (usually a research only post) (Table 5.3). Although the data are not longitudinal, it appears that those who are still on fixed-term contracts in their thirties and beyond either progress into permanent contracts or leave the sector as 70% of fixed-term staff are under forty years of age. From the lower half of Table 5.3, it seems that the progression from fixed-term to permanent contracts as staff get older is mirrored in the progression from research only to teaching and research roles. Other than identifying that recruitment, on entry to the sector, on to permanent contracts increases with age, the data are not able to suggest the extent of movement between fixed-term contracts and permanent contracts for older entrants.

The qualitative research found that, in some cases, the route to a senior, or even a permanent, university post could be long, particularly in old universities. Routes in new universities were found to be much more varied, often including a period outside Higher Education in professional practice (for example in health, education, law or computing). Three examples of career paths taken by academic staff in old universities are given below. Once they had decided on an academic career, some individuals demonstrated considerable determination to reach a permanent position at a university.

132

Examples of academic career paths: staff in old universities

A now-35 year old lecturer in Biological Sciences had completed his first degree, had had a gap year (for financial reasons) and then returned to university to study for a PhD. He combined postgraduate study with demonstrating work in the department. On completing his PhD he took a three year post-doctoral job in research at a Scottish University, which was extended to four years. He then took a three year research post at an English university. When this finished, in 2003, he moved universities again to take up his current lectureship. Lecturer, old university

A reader in Computer Science had done his first degree in Scotland and his PhD at an English university. He had worked part-time in his own university and the local new university (then polytechnic) while doing his PhD and writing up. He then took a lectureship at his current university at the age of 26, was promoted to senior lecturer at the age of 40 and to reader at the age of 44. Reader, old university

A senior lecturer in Engineering Sciences had taken an engineering apprenticeship on leaving school and then taken an engineering degree course, followed by post-graduate research. He worked as a research assistant while studying for his PhD. He then moved to a research post at another university where he was appointed to a lectureship after six years. He moved to a senior Lectureship post at his current university in 2002, aged 36. Senior Lecturer, old university

6.2.1 Expected progression

Most staff interviewed expected to be promoted to a more senior post in the future, either within their own institution or by moving. About one third expected to reach a chair (and a small percentage a senior management role, including Vice Chancellor) (Table 6.1). Sixty-three per cent, in total, expected to reach Senior Lecturer (old university) or Principal lecturer (new university). Very few expected to reach higher grade research posts, a realistic expectation given the low number of these.

Irrespective of the type of university in which staff were currently employed, around two-thirds expected to reach at least Senior lecturer (old university)/Research Grade III/Principal Lecturer (new university). However, expectations of further promotion were much lower for those currently employed in new universities: 49 per cent did not expect to progress further, compared with 28 per cent in old universities. In old universities, 44 per cent expected to reach a chair/Research Grade IV/pro-Vice Chancellor/Vice Chancellor (compared with 25 per cent in new universities). Given that 12 per cent of old universities’ posts are chairs and four per cent of new

133

universities’ (Table 3.3), those in new universities may be over-estimating their likelihood of reaching a chair, unless they are likely to gain chairs in old universities.

Table 6.1 Academic staff: Academic career expectations

Highest level job expects to reach Old New Male Female TotalResearch grade IA/Research B or lower 3 1 2 3 2 Research grade II 3 0 1 3 2 Research grade III 2 0 1 1 1 Research grade IV 2 0 1 1 1 Lecturer (new university)/Assistant Lecturer 3 1 2 3 2 Senior Lecturer (new university)/Lecturer (old university) 8 18 9 17 12

Principal Lecturer (new university)/Senior Lecturer (old university) 11 29 16 21 18

Reader 10 9 9 9 9 Professor 38 22 42 20 32 Pro-vice Chancellor (or equivalent) 3 3 4 2 3 Vice Chancellor (or equivalent) 1 0 1 1 1 Don't Know 13 12 9 16 13 No Answer 2 5 3 3 3 Total 100 100 100 100 100 n 1727 1060 1575 1181 2788

• Significantly different by university type and by gender at the 1% level • Source: NIESR/DfES Staff Survey, 2004

Men had higher expectations than women: more than twice as many expected to reach a chair (42 per cent and 20 per cent respectively). Instead women are much more likely to expect to reach only Principal lecturer (new university)/Senior Lecturer (old university) or even to plateau at Senior lecturer (new university)/lecturer (old university).

The qualitative research suggested that one of the reasons for lack of expectations of senior posts may have been lack of desire; almost without exception, academic staff in old universities wanted promotion and, in most cases, expected movement to the next grade within a three to five year period. However, some respondents in the new universities did not want promotion because of the amount of administrative duties and bureaucracy a more senior post entailed. This reflects the difference in senior posts in new and old universities, where, in the former, the main promotion posts (Heads of department and Principal lectureships) entail substantial administration. Some staff who were already in senior posts wanted to be dean or head of faculty, but, as with their counterparts in new universities, some academic staff did not want to be a head of department or hold any other management post because they disliked administration and bureaucracy.

Other reasons for low expectations identified in the qualitative research were lack of research expertise or output, snobbery about their academic subject (for example Health) and sexism. The practice in some universities of ‘freezing’ promotion or capping numbers was also seen by some to limit their chances of promotion in the short or medium term.

134

Further light on views of the promotion process, in some cases, was that promotion was age related: unless they achieved a chair by the age of 50, they were unlikely to gain one.

6.3 Academics’ satisfaction

6.3.1 Introduction In this section we examine the determinants of academics’ job satisfaction.

Overall job satisfaction is a product of satisfaction with a number of different aspects of the job. We will refer to these ten elements of a job as its dimensions. In order to understand why the HE sector is suffering from a retention problem, we need to understand the determinants of overall job satisfaction as well as satisfaction with these dimensions. It may also be the case that differences in retention difficulties across subjects and gender are due to differences in overall job satisfaction or in satisfaction with particular job elements. Job satisfaction is an important factor influencing decisions to leave a job and hence retention difficulties. Reported levels of job satisfaction have been found to be good predictors of quit behaviour (Freeman, 1978; Akerlof et al., 1988; Clark et al., 1998; Clark, 2001). Overall job satisfaction is a product of satisfaction with a number of different aspects of the job. In this section, we consider a number of elements of job satisfaction of English academics.

The academic labour market is one where individuals appear to earn less than similarly qualified individuals elsewhere (Machin and Oswald, 2000; NIESR, 2004). This suggests that there are elements of academic employment that compensate for these disparities (Rosen, 1986). This also means that staff who have already decided to join academia may have already discounted the opportunity to earn a high salary and so one might expect the pecuniary aspects of the job to affect academics less than more highly-paid graduates in other sectors, ceteris paribus. Of course, circumstances change over an individual’s career and the influence of factors such as families may change their attitudes toward these and other elements of the job.

The literature on the job satisfaction of academics is limited111. Oshagbemi (1996, 1998), investigated job satisfaction of UK academics, but the analysis does not go beyond cross-tabulation. More recently, Ward and Sloane (2000) studied the job satisfaction of academics in five Scottish higher education institutions (HEIs) in a more systematic framework, using ordered probit analysis. Our dataset allows us to consider the satisfaction of a much larger group of academics from ten English HEIs.

One important point to note about studies of satisfaction is that reported levels of satisfaction in the economy as a whole tend to be fairly high (consider the first column of Table 6.2). Workers tend to be fairly satisfied with most aspects of their jobs. The levels of satisfaction reported by academic staff (Oshagbemi, 1996, 1998; Ward and Sloane, 2000) are generally slightly lower than those reported for the economy as a whole (e.g. Clark, 2001). Academic staff are most satisfied with the work itself and teaching in particular (Oshagbemi, 1996). It must be noted, however, that such comparisons are only illustrations at best, for they are tenuous because they

111 In an extensive review of job satisfaction, Oshagbemi (1996) found no studies published between 1981 and 1995 relating to the job satisfaction of university staff and only seven from an earlier date. Only one of the latter covered more than one university.

135

refer to different surveys, which have different sample methodologies and slightly different questions. There have been no analyses comparing the satisfaction of academic staff in Higher Education with the rest of the population using the same survey112. With this caveat in mind, the single area where academic staff compare well with the population as a whole is in the opportunity to use initiative (Ward and Sloane, 2000)113. The greatest differences in satisfaction between academics and the population as a whole are with pay and promotions. Although these are the dimensions where satisfaction is the lowest in the economy as a whole (Clark, 2001), the difference between satisfaction with these two facets and the other is much larger for academics than for other workers.

Table 6.2 Reported satisfaction in previous studies

Whole economy UK academics Scottish academics

Clark (2001) Oshagbemi (1996) Ward and Sloane (2000)

Overall job satisfaction 5.427 4.212* 5.04 Promotion 4.484 3.42 3.40 Pay 4.615 3.44 3.60 Hours 5.214 - 4.52

Work itself 5.562 Teaching = 5.09 Research = 4.66 Admin = 3.93

5.27

Job security 5.192 - 4.41 Opportunity to use initiative 5.745 - 5.81 Supervisors 5.529 4.18 5.09 Co-workers - 4.81 5.42 Physical work conditions - 4.33 -

• * This comes from Oshagbemi (1998)

The respondents in our staff survey were asked to rate their satisfaction with ten dimensions of their jobs on a similar seven-point scale to those of the studies summarised in Table 6.2 (from completely dissatisfied to completely satisfied). Their responses are summarised in Table 6.3. Our results echo those of Ward and Sloane (2000) in that we find that academics rate the work itself, the opportunity to use their own initiative and their relations with their colleagues most highly. The dimensions with which they are least satisfied are their promotion prospects, their salaries and their total earnings.

Factors leading to feelings of job insecurity were identified in the qualitative research. Lecturers in some of the case study universities and departments were concerned at the decline in student numbers in their own area and at the implications of this for their job security. Others expected their departments to be closed if its achievement within the RAE was not high. Some had concerns largely as a result of earlier redundancy programmes in their institutions. 112 Note, however, that this difference is consistent with the fact that satisfaction is generally found to be negatively correlated with education (Clark and Oswald, 1996). 113 Note that Oshagbemi (1996) does not ask a question relating to the use of initiative.

136

Table 6.3 Academic staff: Reported satisfaction (%) Level of Satisfaction

satisfied neither dissatisfied 7 6+7 5+6+7 4 1+2+3 1+2 1

Mean

Relations with colleagues 18 67 85 7 8 3 1 5.59 The actual work itself 10 65 86 3 10 3 1 5.48 Being able to use own initiative 21 63 84 6 10 5 2 5.54 Relations with manager 16 52 66 12 22 12 5 4.96 Job security 15 44 60 12 28 18 11 4.63 The hours you work 13 43 59 12 29 13 5 4.68 Physical work conditions 9 42 61 10 28 13 5 4.66 Total earnings 4 24 40 15 44 22 9 3.95 Salary 4 23 40 13 48 25 11 3.85 Promotion prospects 4 19 34 21 45 27 15 3.69

• Table shows the percentage of respondents reporting a given level of satisfaction • Satisfaction is coded as follows: 1 = Completely dissatisfied, 2 = Mostly dissatisfied, 3 =

Somewhat dissatisfied, 4 = Neither satisfied nor dissatisfied, 5 = Somewhat satisfied, 6 = Mostly satisfied, 7 = Completely satisfied


6.3.2 Are Female Academics Happier? One common finding from analyses of satisfaction is that females tend to

report higher levels of satisfaction than men (e.g. Clark, 1997). Ward and Sloane (2000) found that for Scottish academics, this was not the case. The responses to our survey support this finding for some aspects of job satisfaction but not others (Table 6.4 ). The aspects with which there is no significant difference in reported job satisfaction were what one might call the non-pecuniary aspects of the job: i.e. the actual work itself, being able to use their own initiative, the hours, relations with colleagues and the physical working conditions. In a compensating variation or equalising differences framework (e.g. Rosen, 1986), these are the aspects of academic work that one might think of as compensating for low wages.

It is, however, with the pecuniary aspects of academic work that there are significant differences between the reported satisfaction of men and women. In particular, women are even less satisfied with their promotion prospects than men. This is not surprising, given the lower promotion probability of women and the possible barriers to promotion they experience (see Section 2.4.8.5).

They are, however, more satisfied with their earnings. One reason for this is that the inequality within academia, although present, is less than that elsewhere (see Chapter 4).

137

Table 6.4 Academic staff: Satisfaction by Gender

Total Male Female Difference (F-M)

Mann-Whitney

Relations with colleagues 5.59 5.58 5.61 0.03 0.912 The actual work itself 5.48 5.48 5.48 0 0.593 Being able to use own initiative 5.54 5.55 5.53 -0.02 0.058 Relations with manager 4.96 5.01 4.89 -0.12 1.766* Job security 4.63 4.69 4.55 -0.15 1.817* The hours you work 4.68 4.70 4.65 -0.05 0.686 Physical work conditions 4.66 4.66 4.66 0 0.106 Total earnings 3.95 3.83 4.11 0.28 4.323*** Salary 3.85 3.72 4.02 0.30 4.323*** Promotion prospects 3.69 3.80 3.54 -0.27 3.870***

• * significant at 10%; ** significant at 5%; *** significant at 1% • Source: NIESR/DfES Staff Survey, 2004

6.3.3 What Makes Staff Happy? We have seen what factors influenced staff’s decision to enter academia and

how satisfied they are with the dimensions of their job. In the next section we will see how staff’s satisfaction with these elements of the job are influenced by factors such as their personal characteristics as well as those of their job and institution. Before we do so we shall consider in greater detail what it is about the elements of the academic job that affects job satisfaction.

Academics typically spend their time doing one of three things: teaching (including preparation and marking), research and administration. These roles involve a number of activities, some of which may be pleasurable and others less so. For example, whilst conducting research in an academic’s area of interest may in itself be intellectually satisfying, not all staff have complete freedom in determining their research agenda; conducting research often also involves raising funds and submitting oneself to external assessment. We saw above that although most postgraduates would prefer jobs that involved research, when we look in greater detail, we find that it is research specified by oneself rather than by others that they would like.

Whilst teaching was not the most important reason why postgraduates want to enter academia (Table 5.30 in section 5.6.1), most would prefer a job which involves some teaching. Academics, too, see teaching in general as one of the positive elements of the academic job – almost half of the academic staff surveyed say that teaching raises their satisfaction a lot (Table 6.5). However, not all of these aspects of teaching are viewed as being equally gratifying. The positive aspects of teaching involve teaching very bright students and seeing their students develop. Whilst the latter might be a reward one gains from all students, irrespective of ability, the former may only be available to staff in universities with a student intake of higher ability. Nevertheless, a sizeable minority feels that teaching non-traditional students does in itself raise their satisfaction. These elements of teaching are satisfying to a great many staff. The amount of student assessment is not viewed so positively. The amount of teaching has relatively little effect on job satisfaction. Whilst the amount raises satisfaction either a lot or a little for one third of staff, only 10 per cent says it raises

138

satisfaction a lot. The majority of staff (78 per cent) say that it has no effect or only raises or lowers satisfaction a little.

The survey could only examine a few elements of teaching. The qualitative research suggested that other aspects of the nature of teaching affected satisfaction. For example, some academics said they liked teaching, but usually qualified this by saying they liked giving one to one or small group tuition, conveying ‘ideas’ rather than facts, or said they enjoyed teaching if they had adequate preparation time.

Table 6.5 Academic staff: The aspects of teaching and job satisfaction raises satisfaction lowers satisfaction

a lot a little or a

lot no effect a little or a

lot a lot Total

Seeing students learn and develop 73 96 4 0 0 100 Teaching very bright students 58 86 13 0 0 100 The specialisms you teach 54 89 10 1 0 100 Teaching 48 85 9 6 2 100 Academic level of your teaching 26 65 20 15 5 100 Teaching non-traditional students 21 49 39 12 4 100 The amount of student assessment 2 12 28 60 26 100 Amount of teaching 10 33 27 41 12 100


On the other hand, heavy teaching loads, an increase in student numbers and in students’ expectations had led to greater demands on academics’ time from students and this had reduced satisfaction. Students were seen as increasingly demanding on lecturers’ time and to expect more contact and direct instruction (‘spoon-feeding’) in place of independent study. Some had concerns about the quality of provision in their institution. These included what one described as a ‘production-line approach’ to students and the poor standard of some first year entrants. As a senior lecturer in History (new university) stated,

‘Although I get along well with the students, they aren’t very good. Half don’t want to be there and come because their prospects of doing anything else are so limited. They aren’t interested in the subject and lower the standards of the whole group.’

The fact that research is so important for so many postgraduates considering entering academia and was one of the more important reasons for incumbent academics for doing so leads one to expect that it will be one of the more important positive influences on academics’ satisfaction with their jobs, unless they discover that it is very different from what they expected. Research is indeed seen as a more positive element of academic employment than teaching (Table 6.6), with almost two-thirds of staff stating that their research raises their satisfaction a lot. However, staff are less sanguine about the time that they have for research and the access they have to research funding. Moreover, the qualitative research found that some staff (all in the lower-research ranked universities) considered little support was given for research, in terms of time and other facilities. In one department, staff were set targets for publishing and if this was not achieved their research time would be reduced. (Although staff had always retained this with the argument that they needed it to meet

139

the following targets.) This was thought to be finance driven. The head of department said ‘people are stick driven; they whack you over the knuckles’.

Table 6.6 Academic staff: The aspects of research and job satisfaction raises satisfaction lowers satisfaction

a lot a little or a

lot no effect a little or a

lot a lot Total

Your research 65 86 10 3 2 100 Time for research 37 56 13 31 19 100 Access to funding for research 27 42 22 36 22 100 Demands for research output 11 28 34 39 16 100 The RAE 3 12 39 49 31 100


The demand for research output is seen on balance rather negatively. The RAE is seen as the worst aspect of research, with just under a third stating that it lowers their satisfaction a lot and a little under half feeling that it lowers their satisfaction at least by a little. This suggests that the implementation of such a scheme of research assessment must be careful not to have a negative effect on the research conducted in higher education by discouraging the workers that it seeks to assess. The qualitative research suggested that dissatisfaction with the RAE did not stem only from pressure for outputs or dislike of targets and accountability. The RAE was also seen as having a detrimental effect on the quality and nature of research itself. In particular it was seen to lead to the devaluing of research with long-term outcomes and which could not be published within a short time-scale. The effect of dissatisfaction with the research assessment exercise on academics’ likelihood of leaving UK higher education is examined below.

Whilst teaching and research are generally seen as positive aspects of academic employment, the administrative and organisational side is seen rather more negatively (Table 6.7). In particular, the amounts of administration that academic staff have to undertake, the changes (both organisational and in the administrative demands placed on staff) are detrimental to the satisfaction of many staff. QAA requirements are no more popular than the RAE is; indeed, they seem slightly less so.

The type of issues identified in the qualitative research were the volume of administration, which was seen as excessive, and increasing demands for ‘accountability’ through detailed records of course planning and delivery. Administration and bureaucracy were seen as increasing problems, which some explained with reference to reviews and inspections, particularly by Ofsted and the requirements of the Quality Assurance Agency (QAA). These demands were seen as having a negative rather than positive impact on quality of provision. Therefore as one principal lecturer in Education (new university) remarked,

‘There is too much emphasis on accountability and not enough on development. Quality reviews concentrate too much on the form and not the content, whether the right headings are there and not what the report is saying’.

140

Table 6.7 Academic staff: The effect of administrative and organisational aspects of academic employment and job satisfaction

raises satisfaction

lowers satisfaction

a lot a little or a lot

no effect a little

or a lot a lot Total

Administrative support 7 29 23 47 28 100 The value attached to teaching by the university 6 26 36 37 19 100 Your ability to influence major university decisions 5 17 32 52 34 100 Relations with the university administration 3 17 41 42 21 100 The value attached to admin by the university 2 7 42 51 30 100 The value attached to research by the university 12 41 28 31 16 100 The QAA 1 6 40 54 34 100 The amount of organisational change 1 5 32 63 35 100 The amounts of administration 0 5 24 71 40 100 Changes in the administrative demands placed on you 0 3 30 66 35 100


Complaints included new initiatives, target-driven exercises, constant inspections and corporate reorganisation. Staff who complied with university requests for information, including by taking part in committees, complained at the lack of recognition, either in terms of pay or career progression, of taking on administrative roles. As one senior lecturer in a new university explained,

‘I’m not happy with my situation. In my non-teaching semester, it is all admin. I could have said no but I didn’t want to appear churlish. There’s little to no reward for all this extra effort’.

Others complained about its effects on teaching and on research activity, as illustrated by the following statements,

‘There is more and more burdensome administrative work which I don’t enjoy having to do. Some is necessary, but some unnecessary and gets in the way of what we’re doing. We have endless meetings about teaching quality, when we could be getting on with teaching.’ (Senior lecturer in Sociology, old university)

‘I hate the crushing burden of bureaucracy, which is threatening the only good aspect of the job - the opportunity to do research. There are enormous administrative overheads which to my belief are mostly unnecessary’. (Senior lecturer in Law, old university)

Some felt that the level of trust, which had previously governed relationships between staff and management, had been eroded. Bureaucratic systems were seen to substitute for face-to-face contact and reducing the autonomy which academics had traditionally enjoyed. Some staff said they felt they were being constantly monitored and evaluated.

The pressure to become more involved in the administrative function of their department, faculty or university through joining committees was an issue and

141

examples were found of this leading to stepping down from positions of responsibility and avoidance of participation in committees if they possibly could.

A further source of dissatisfaction for some staff was the under-valuing of their subject area by other academics in their institution. This was raised by academics in less traditional areas for universities, including health and art and design.

A number of academic staff in new universities referred to poor management in their institution, including promotion practices and slow decision-making. Some senior staff complained at interference by the faculty into the work of their own department.

Overall we have seen academics tend to be less satisfied with their workload than employees nationally. The qualitative research suggested this was not just in terms of the amount of work, but the effect of excessive demands. Some staff in the lower research-intensive old university and the new universities reported that when they could not meet all the work demands, research activity suffered. Although in many of the case study universities there was substantial emphasis on research, it was still the first activity to be sacrificed if staff found the demands of the job too great.

Academic staff value more subtle elements of their jobs, such as the support of their peers and the ability to participate in the wider academic community; these are things that raise satisfaction for a lot of academics and have a negative impact for very few. In the qualitative research academics expanded on this (Table 6.8). Some mentioned appreciating working with colleagues with similar interests to their own and with a similar approach to their work and working in departments with research expertise in particular academic areas. A senior lecturer in Sociology (old university) described his colleagues as ‘very stimulating and challenging people to work with’ and a senior lecturer in Law (old university) spoke of working in a ‘team of world-recognised academics’. This experience was seen as a very positive feature of working in departments which were highly-rated in terms of research. With regard to participating in the wider academic community, some enjoyed collaborating with colleagues at other institutions.

Table 6.8 Academic staff: Other aspects of academia and job satisfaction raises

satisfaction lowers

satisfaction a lot a little

or a lot

no effect a little

or a lot a lot Total

Ability to participate in the wider academic community outside your university 36 74 19 7 3 100

Peer group support within the department or university 32 70 16 15 6 100 Ability to participate in other activities, connected with your academic job, outside the university 31 68 26 6 2 100

The general direction of higher education policy 3 14 21 66 38 100 Career progression 19 41 18 41 19 100 Your working environment 17 54 16 30 11 100 Pay 16 47 14 39 17 100


Pay and career progression are aspects of academia which raise the satisfaction of fewer academics and lower the satisfaction of rather more (Table 6.8).

142

More than half of the staff surveyed felt that the general direction of higher education policy had a negative impact on their job satisfaction. Whether this policy dissatisfaction actually translates into a desire to leave UK higher education is something we tackle in Section 7.3.

Some of the issues around pay, identified in the qualitative research, were pay erosion (for those who had worked in academia for a number of years), inadequacy in high-cost locations and being ‘grossly underpaid’ compared with people at similar levels of seniority in other sectors, such as law and accountancy. A number of staff also felt that the status of an academic job had fallen in recent years.

Exacerbating dissatisfaction in respect of pay was not only the level, but the fairness with which individual pay was set. Whilst the majority of staff felt this was done quite (or very) fairly, 39 per cent did not think it was (Table 6.9). The types of issues raised in the qualitative research were pay being set in relation to demand for the individual rather than in relation to their skills and attributes, and negotiation being used. (For example, some academics might negotiate a better package at recruitment or gain a pay rise through threatening to leave.)

Table 6.9 Academic staff: Perceived fairness of decisions on individual pay, current university

% Very fair 4 Quite fair 44 Not very fair 27 Not at all fair 12 Refused 13 Total 100 n 2785


Some criticism was made of the professorial system. One professor at an old university said, ‘It would be interesting to know where I am on the scale’, since he believed that ‘transparency is woefully lacking’ when it comes to professorial pay scales.

‘It is not possible to find out where you are on the ladder, or indeed, how many rungs there are on the ladder, since professorial salaries are shrouded in secrecy and there is no information about them. The setting of professorial pay is an invisible, secret process’.

Similar problems of fairness were apparent for promotion, including recruitment to senior posts. Sixty-four per cent of academics saw promotion decisions as unfair (Table 6.10). For internal promotion, the issue stemmed from restricted opportunities for promotion. Within these restrictions, promotion was carefully determined through competition with strict criteria. However, it was felt that such criteria were not always applied, especially where valued individuals appeared likely to leave.

143

A lecturer in an old university said, ‘The system is fine, but it’s not the system that’s important. It’s how the system works’. His experience at his current university was too short to comment on its implementation, but his previous experience of academic promotion systems was that they were ‘shrouded in secrecy. Not a transparent process’.

Table 6.10 Academic staff: Perceived fairness of decisions on promotion and recruitment to senior posts, current university

Fairness of decisions on promotion, current university

Fairness of decisions on recruitment to senior posts,

current university Very fair 15 16 Quite fair 3 4 Not very fair 37 39 Not at all fair 27 25 Refused 18 16 Total 100 100 n 2785 2785


Other criticisms were levelled at the formal system. A professor in a new

university commented,

‘It’s terrible here. No promotion and the process is extremely opaque. It was better at [university] where they had an annual promotion round.’

At the same time, certainly some individuals who felt the system was not fair did not understand the system or know the criteria.

Dissatisfaction was particularly strong where it was thought that the formal system was being circumvented to retain leavers. ‘Backdoor’ promotion tended to cause strong resentment amongst academic staff. A reader in an old university commented,

‘Promotion is atrocious. I’ve no problem with the formal scheme: it’s open, generally fair. There’s increasing irritation and concern at the backdoor arrangements, driven by the RAE, whereby people say they are leaving and so get promotion. This is done irrespective of the resentment it causes. It leaves a very bad taste and ultimately it will be counterproductive.’

However, staff views of the system may be affected by the availability of promotion. In the higher ranking old universities staff expected to achieve promotion and more seemed to be more satisfied with the system. However, even in these universities, the actual opportunity for promotion depended on how these were distributed by the university, with some concentrating senior positions according to RAE rating. A head of department in an old university considered that

144

‘The career path is unattractive. There is a battle at each level. It is very slow. The real problem is that very few are promoted each year. [In my faculty] this year 40 applied. Thirty were worthy of promotion. Only five or six were promoted’.

Employment at the new universities was much more concentrated in the senior lecturer grade (more or less equivalent to lecturer in the old universities) and there were substantially fewer promotion opportunities.

However, responses varied. For example, reports were very mixed in one old university, where within each department, views varied from, for example, the promotion system being considered ‘generous and fair’ to ‘ill-focused and not incentivised’.

Recruitment to senior posts was an alternative to internal promotion and 64 per cent of academics saw the decisions on this as unfair. The criticism here was that different criteria were used internally and externally, so that it was easier for external candidates to be appointed to more senior posts. For example a head of department in an old university commented,

‘An academic is recruited from outside to a certain position on the basis of promise and is promoted internally on the basis of performance. This tends to result in much stricter criteria for internal rather than external applicants.

Similarly, a professor in an old university who was ‘very unhappy’ about the system of promotion said,

‘There is a bias against internal candidates. This is antiquated, a disincentive, and it is de-motivating. It is ‘bizarre’ that the employers cannot see the merit in internal promotion, since it works well in many businesses.’

For some, it was not only certain elements of their job which raised satisfaction but the combination of those elements. As a senior lecturer in History said of his job,

‘It’s the best job I’ve ever done. No-one is looking over your shoulder and I’m paid for reading books. I really enjoy the teaching which means I have to learn about different areas and aspects of history. I do very much enjoy my job’.

Whilst peer group support was valued, some respondents in the qualitative research mentioned factors which were reducing this, including:

• heavy teaching loads

• the practice of working at home

• contact through email rather than face to face

• a ‘commuter’ ethos which makes after-work socialising difficult

145

• lack of a departmental common room

• distribution of staff across a number of buildings

The flexibility and autonomy which was an attractive feature to potential recruits, was also mentioned in the qualitative research by existing staff as increasing satisfaction. Some referred to ‘flexibility and freedom’ as twin benefits of their job, suggesting that flexibility over one’s working time confers a sense of freedom. Some described the freedom they enjoyed in their job and some compared this to previous jobs in industry where their time and activities had been more tightly controlled. For example, the head of a school of engineering (new university) referred to less pressure to meet targets in Higher Education than in industry and a higher degree of certainty over the future of the organisation. A senior lecturer colleague, who had been employed as an academic for only six months compared the flexibility in his academic job with his previous jobs in the telecommunications industry,

‘The flexibility is helpful. You have to fulfil your commitments but no-one is checking up on you, so it is good in terms of how you can run your life. You have a lot more personal responsibility in the job than in industry, you can do the teaching how you want. In industry it is all team working and a collaborative approach and there’s less scope for doing things as you’d like them done.’

However, others who had only worked in academia, made unfavourable comparisons between their current freedom and autonomy and their experience earlier in their career.

Autonomy and flexibility was also gained through being able to organise one’s own time, and this was reported as giving a sense of freedom. Related to this, some mentioned the control they had over their weekly time-table and even their daily activities. As a lecturer in Computing (old university) stated,

‘I can do research and not spend too much time on other activities. I like the lifestyle, I can work 10am to 6pm or 6pm to 5am. There are no specific times when I have to work, and I work more than eight hours a day anyway. No-one will complain about the hours you work’.

However, for some, there were problems. Some lecturers felt so pressured in their jobs that they found it difficult to relax when away from work, including when on holiday.

More detailed issues about the working environment were explored in the qualitative research. Academics were asked if the ethos of academia and that of their university suited them. This referred to academic and university values, culture and attitudes. Good aspects of the ethos included the level of debate and openness to discussion and diversity of opinion. Some academics also referred to ‘individualism’ as part of the academic ethos. This was seen to allow for the development of one’s own approach. As one principal lecturer in education stated, ‘there is no fixed party line’. Openness within universities was seen by some to extend to faculty management rather than to be confined to intellectual discussion within departments.

146

Some staff who had worked outside the Higher Education sector described this openness as a welcome surprise. However, not all academics valued the ethos of their institution: some staff (who were in less traditional academic areas such as health and art and design) felt that their subject did not fit easily with an intellectual ethos and that they were sometimes affected by academic ‘snobbery’.

Some staff who had previously worked in the private sector said they liked working for a public sector organisation, and others with experience of the private sector remarked favourably on their university’s ethos. At the same time, some academics complained of an increasingly ‘corporate’ culture in their institution. This included universities operating like commercial companies and the increasing emphasis placed on employment as an outcome of academic study. Some staff who had valued their institution’s orientation towards mature students were disheartened that courses which attracted them had been cut. Similarly, some staff regretted the increasing emphasis on research within their institution and felt that this had led to problems of ‘identity’, with some universities not knowing whether they were teaching or research institutions.

Others valued the freedom, flexibility and opportunity for creative thought. However, a number complained that, while this ethos still exists within departments and faculties, it is threatened by an increasingly financial ethos in university management. Therefore, departments were seen to exercise freedom within tightening constraints, principally governed by financial and bureaucratic considerations. A professor of Law complained that increasing emphasis on teaching and bureaucratic procedures in his university meant that his department was no more than a ‘glorified secondary school’.

As discussed above, some academics were highly critical of their universities’ increasingly commercial outlook, including links with industry. Some lecturers welcomed these links, but they were seen by many to threaten universities’ independence and values. A reader objected in particular to his university’s new practice of referring to students as ‘customers’, which he felt resulted from this change in ethos. At the same time, students’ attitudes to staff and to their studies in general were seen to have changed as a result of the introduction of tuition fees, which gave them both the status and mentality of customers of the university.

In the qualitative research, a number of issues directly relating to changes in policy on Higher Education were raised, including the impact of tuition fees on students’ attitudes and the effect of Government targets on universities and staff. Some staff also complained of under-funding of universities and changes, including increasing ‘commercialisation’. As a senior lecturer in Sociology stated,

‘Universities are run more and more as commercial enterprises. Research is organised around end-users and consumers rather than the production of ideas’.

Indeed, overall, three-quarters of academic staff thought that academic jobs had worsened since they had joined academia (Table 6.11). The percentage tended to rise the longer the period in academia.

147

Table 6.11 Academic staff: Whether thinks academic jobs have improved since their first academic job in UK

Improved Worsened When first had research or lecturing job at UK university

Greatly Greatly or slightly

Neither improved

nor worsened

Greatly or slightly Greatly

Refused Total

Prior to 1970 3 3 91 67 3 100 1970-1979 1 5 5 89 69 1 100 1980-1984 2 4 5 90 58 1 100 1985-1989 3 8 89 53 1 100 1990-1994 1 6 17 75 38 2 100 1995-1999 1 8 31 59 17 3 100 Total 1 6 15 77 43 2 100


Aspects of change were identified through the qualitative research. These included the increase in bureaucracy and management, covering the requirements of planning, monitoring and evaluation of teaching, requirements of the Quality Assurance Agency (QAA) and Research Assessment Exercise (RAE). While some staff believed that more scrutiny over teaching and research quality was to some extent useful in improving quality, negative features were identified. Staff in new universities remarked on the increase in management layers and hierarchy associated with the growing bureaucracy associated with accountability, while staff in old universities complained bitterly of an erosion in trust and academic autonomy. In some departments of new universities, the faculty was seen as having more involvement in departmental affairs than in the past, and this was not generally welcomed.

Staff in both old and new universities complained of an increasing workload, which some found stressful. Academics in old universities remarked that, although it had always been valued, research was now given an even higher profile. Similarly, academics in new universities felt increasing pressure to carry out research, along with heavy teaching and administration loads. Staff across all universities commented on the increase in student numbers which brought with it a higher workload. Some also commented on the growth in size in their institution, department or faculty, referring to increases in staff numbers. However, it was the increased ratio of students to staff which received most widespread comment. Students’ demands and expectations had increased and took up considerably more time than in the past. Some academics felt the calibre of students had declined alongside the expansion of Higher Education, while others believed they were simply more needy and less willing to work independently. Email and voicemail contact were seen as both a blessing and a curse, making escape from teaching difficult. Such communication was also seen to substitute for face-to-face contact with other staff, resulting in feelings of isolation for some academics.

Other developments identified by staff in old universities included chronic under-funding of Higher Education, which was seen to stem from the Thatcher administration. Institutions were also seen to make increasing use of short-term contracts, for lecturers as well as research staff. Contractual issues were also raised by lecturers in new universities which have changed terms and conditions for academic staff, including the reduction of holiday entitlement. Some staff in both old and new

148

universities felt under pressure to retire early as a result of their university’s cost-cutting programmes. This was most strongly felt by older staff with little involvement in research. Finally, staff in both new and old universities complained of low pay, and particularly of erosion in academic salaries.

6.3.4 The determinants of satisfaction In this section we discuss the results of our econometric analysis of the factors

influencing the job satisfaction of academics. Our model of job satisfaction and intentions to leave UK higher education is set out in Appendix G along with the results of its application to our survey of academic staff. Our model of job satisfaction is a generalisation of Clark (1997), Ward and Sloane (2000) and Lydon and Chevalier (2002), although it differs slightly in both motivation and application. In particular, we investigate satisfaction with ten elements of the job and in the next chapter we will use our measures of job satisfaction to help us explain the influences on the likelihood of academics leaving the UK higher education sector.

The results of our estimation on the full sample are set out in Table A.24 and Table A.25 of Appendix G. The results are discussed in more detail in Appendix G; to summarise the main results:

• There are no significant differences between the satisfaction of men and women with many of the elements of their academic employment.

! It is only with respect to their salary and total earnings that women report significantly higher levels of satisfaction.

• Members of non-white ethnic minorities are less satisfied with the opportunity they have to use their own initiative, the hours they have to work and their relations with their colleagues than their white colleagues.

• There is weak evidence that members of non-white ethnic minorities are less satisfied with their salary and total earnings.

• There is generally a non-linear relationship between experience and job-satisfaction.

! Satisfaction with the non-pecuniary aspects of the job tends to decrease with experience for the first half of academics’ careers but then increases. Whether this can be seen as a pure learning effect as noted in the introduction is unclear, as the results imply that it can take ten or twenty years to learn whether one likes the job or not.

! Two elements where experience does not have a significant effect are satisfaction with salary and total earnings.

! The one influence on earnings that the individual can predict less well is promotions; satisfaction with promotions is decreasing over the early career and then increasing is supported by the data.

• Professors and, to a lesser extent, senior lecturers are generally happier in their jobs than lower grades – particularly, and unsurprisingly, with promotions.

! One explanation for this result is that if all staff have similar levels of ability, but jobs are rationed so that only a lucky few get promoted, then those that do not get promoted will exhibit higher levels of dissatisfaction.

149

! Another explanation is that the ability of staff varies and those that are promoted are the ones of the highest ability. In this case, if lower ability staff feel they are of the same ability as those who are promoted (or higher), they will exhibit higher levels of dissatisfaction.

! The truth may lie somewhere between these two extremes. • Staff on non-permanent contracts are significantly less satisfied with their

promotion prospects and their job security.

! There is some weak evidence that they are more satisfied with the actual work itself than permanent staff.

! It is interesting to note that they are more satisfied with their earnings than permanent staff. This is consistent with the idea that they are willing to sacrifice earnings for other aspects of the job at this early stage of their career.114

• There are few clear patterns in the variation in satisfaction across subject areas.

! Academic staff working in medicine and dentistry, biological and physical sciences are more dissatisfied with their job security than those working in other areas, although those in physical sciences appear to be more satisfied with their salary and other earnings.

• The universities where satisfaction with salary is lower are located in the south of England where living costs are higher.

! However, this dissatisfaction is not present at all of the southern and London universities.

! Indeed the satisfaction at the first of the highly-rated southern old universities (and the London one) is significantly higher than the other.

• There is a negative relationship between total hours worked and satisfaction with all elements except the actual work itself.

! This reflects the general premise that people generally prefer fewer hours of work than more.

! However, for those who do it, hours spent on research have a positive effect on satisfaction with the actual work itself, the ability to use one’s own initiative and with hours generally115.

! Hours of work spent on administration has a negative effect on satisfaction with all dimensions of academics’ job satisfaction, with the exception of their promotion prospects and relations with their manager.

! The fact that the more time staff spend on research the more satisfied they are with the actual work itself is consistent with the idea that research is a non-pecuniary benefit of academic work. That is, research is one of the factors that offsets the low salaries in academic relative to alternative employment.

! Conversely, hours spent on administration appear to reduce satisfaction almost across the board.

114 Note that staff on non-permanent contracts are, on average, ten years younger than permanent staff. 115 Note that this is the effect of an hour of research, leaving the total number of hours worked unchanged. The net effect of an additional hour of research on satisfaction is only positive with respect to the actual work itself.

150

• Staff in five star departments tend to enjoy the work itself more than other staff (as, to a lesser extent, do staff at five-rated departments) and tend to be more satisfied with their promotion prospects, their relations with their manager and their ability to use their own initiative.

! This is consistent with the idea that staff wish to work at institutions of academic excellence, and the stability that the recognition of this excellence imparts is also appreciated by staff.

! Staff at five star departments are, however, less satisfied with their salaries (although not their total earnings).

o This may reflect the fact that the national pay scales in academia constrain institutions’ ability to reward the most productive workers.

o This conclusion is supported by the results of the inclusion of a measure of individual staff’s own assessment of their abilities – the wage they would expect to command outside of academia – is included.

o In this case, the negative effect of RAE5* on satisfaction with salary and total earnings disappears.

• Earnings are not significantly correlated with satisfaction with any of the non-pecuniary aspects of the job.

! This suggests that staff are able to consider the pecuniary and non-pecuniary aspects of their job independently.

6.3.5 Is there a single overall measure of satisfaction? Many studies utilise a single measure for overall job satisfaction, with the

assumption (either implicit or explicit) that the factors mentioned above are intermediate determinants. This is sometimes done by extracting the first principal component of reported satisfaction with these elements. Beyond the statistical objection to performing factor analysis on categorical variables, our discussion above suggests that this may be an overly simplistic view of job satisfaction and that there may be a number of separate and possibly orthogonal elements that need to be considered. A more appropriate methodology is to perform a factor analysis on the predicted values of latent variables assumed to underlie these categorical reported measures of satisfaction and examine how much of the total variation in satisfaction with these ten elements can be explained by the extracted factors116.

Because of the reduction in sample size imposed by the inclusion of w* in our analysis, we restrict our analysis here to the predicted values obtained from the analysis outlined in Table A.24. The results of performing principal-components factor analysis on our predicted values of the ten elements of job satisfaction are reported in Table 6.12. This table reports the factor loadings, along with the eigenvalues and the proportion explained by the factors extracted with eigenvalues of more than one117.

116 For more on factor analysis, see for example Harman (1976). 117 The so called ‘Kaiser-Guttman rule’ after Guttman (1954) and Kaiser (1970).

151

Table 6.12 Factor loadings for overall job satisfaction

Element of satisfaction Factor 1 Factor 2 Factor 3 The actual work itself 0.815 -0.044 -0.201 Promotion prospects 0.431 0.120 0.800 Salary 0.011 0.986 -0.167 Total earnings 0.015 0.986 -0.161 Relations with manager 0.874 0.051 0.139 Job security -0.045 0.181 0.917 Being able to use own initiative 0.892 0.037 0.107 The hours you work 0.801 -0.124 -0.336 Relations with colleagues 0.736 -0.014 -0.066 Physical work conditions 0.751 0.000 -0.045 Eigenvalue 4.158 2.013 1.724 Proportion explained 0.416 0.201 0.172 Cumulative 0.416 0.617 0.790

There are three principal factors and together these explain 79 per cent of the variation in the ten latent satisfaction variables118. The first factor explains just over half of the common variance of the predicted satisfaction variables. The factor loadings suggest that this factor explains much of the variation of six of the factors: respondents’ relations with their manager, being able to use their own initiative, the hours they work, relations with colleagues and physical work conditions119. We label this factor ‘satisfaction with non-pecuniary elements of the current job’. The second factor merely reiterates what we saw in the analysis above, namely that the explanation for both salary and total earnings are highly correlated. Moreover, the analysis on the whole sample suggests that the pecuniary factors are quite separate from the non-pecuniary ones. The final factor in the analysis of the whole sample includes satisfaction with promotion prospects and job security. We call this factor ‘longer term prospects’.120

These results suggest that there are in fact three separate sets of factors which determine the job satisfaction of academics. The most important from the viewpoint of most economists, namely earnings, is distinct from the other dimensions of the job. Satisfaction with their longer-term prospects explains almost as much of the total variation. The majority of the variation in six of the job dimensions is explained by a single common factor, which we call ‘satisfaction with non-pecuniary elements of the current job’. It is important to note that these proportions of variance explained are not the same as weights – they do not rank the relative importance of these factors. However, the influence of these factors on the likelihood of leaving can be assessed and this is explored in the next chapter (see Section 7.3).

118 In what follows, we will refer to the ten predicted latent satisfaction variables as ‘predicted satisfaction variables’ for brevity. 119 It also explains some of the satisfaction with promotion prospects. 120 Note that promotion prospects also enter into the principal ‘non-pecuniary’ factor, although with a lower factor loading than the other non-pecuniary factors.

152

6.4 Summary This chapter has covered a wide range of issues. In this summary, we first

summarise the findings within four sections: progression routes; overall satisfaction; individual factors affecting satisfaction; and the main factors affecting satisfaction (based on the multivariate model).

6.4.1 Progression within academia Career paths for young academics tend to be from fixed-term contract,

research posts into permanent lecturing posts or out of the sector. For some, achieving a permanent post requires persistence, as it can take a long time and involve many moves or multiple job holding. The pattern for older entrants is not clear and further research on this would be useful.

One third of academics expected to reach a chair, with expectations higher in old universities than new. Given the pattern of promotion posts, without substantial movement from new to old universities, the expectations of academics in new universities seemed rather high and, in old universities, somewhat low. Men had much higher career expectations than women. It was not clear if this was realistic or if women’s low expectations contributed to their under-achievement in academia. In new universities, some people were put off seeking promotion because the main promotion posts entailed substantial administration and bureaucracy. In old universities, because of the different nature of promotion posts, this was not a barrier for chairs, but could be to taking on the role of dean or head of faculty.

6.4.2 Overall satisfaction The standard measures of job satisfaction suggest academics are less satisfied

than other employees. They score lower on all of the main aspects of job satisfaction (the work itself, promotion, pay, hours, job security, relationships with supervisors, relationships with colleagues and physical work conditions) except for the opportunity to use their own initiative.

6.4.3 Satisfaction with individual aspects of the job Research is a major source of satisfaction for academics (65 per cent saying it

raised their satisfaction a lot). On the whole access to research funding and the time available for research raised satisfaction, but lowered it for a substantial minority. However, a number of aspects of research tended to reduce satisfaction: notably, the RAE and the demands for research output. Criticisms of the RAE included the devaluing of research with long-term outcomes and which could not be published within a short time-scale.

Teaching is a major source of satisfaction (48 per cent saying it raised their satisfaction a lot). Teaching bright students, seeing students learn and develop, teaching one’s specialism all substantially contribute to satisfaction. On balance teaching non-traditional students and the level of teaching also contributed to satisfaction. However, the amount of teaching and, especially, the amount of student assessment, tended to reduce satisfaction. From the qualitative research, other factors increasing satisfaction were identified: one to one or small group tuition, conveying ‘ideas’ rather than facts, having adequate preparation time. Those reducing satisfaction included: the ‘production-line’ approach to students, the quality of

153

students, spoon feeding and treating students as customers. For some, students were seen as increasingly demanding of lecturers’ time and expecting more contact and direct instruction in place of independent study.

The amount of administration greatly reduced academics’ satisfaction (greatly reducing it for 40 per cent). Contributing factors were the QAA, the amount of organisational change and changes in administration. Complaints included new initiatives, target-driven exercises, constant inspections, monitoring and evaluation and corporate reorganisation. Targets (e.g. the QAA and the RAE, as well as university initiatives) were seen as replacing trust and reducing academics’ autonomy. Administration detracted from research, reducing time on more interesting work and affecting career progression. It also detracted from teaching and was considered by some to have a negative impact on the quality of provision.

Corporate governance and its values. Relations with the university administration and academics’ ability to influence major university decisions tended to reduce satisfaction. The value placed by the university on research had a positive effect on satisfaction, but the value placed on teaching and on administration tended to have a negative effect. From the qualitative research, it appeared that the academic ethos tended to suit academics: a questioning environment, diversity of opinion, intellectual discussion, creativity. But some found snobbery and that the ethos was being undermined by an increasingly commercial outlook.

Pay and career progression. On balance pay raised satisfaction amongst academics. However, a substantial minority (39 per cent) said it lowered their satisfaction, compared with a larger minority (47 per cent) who said it raised their satisfaction. Equal numbers had their satisfaction raised and lowered by career progression. Whilst the level of pay was an issue, it appeared that the fairness of pay setting for individuals was also a problem, including ‘back-door’ pay rises to retain valued members of staff. In respect of promotion, the slowness and lack of career progression appeared to be a problem, combined with concerns about the fairness of promotion and belief that external candidates were advantaged. Systems were not always clear, and they were not always applied. As with pay, dissatisfaction was reported with ‘backdoor’ promotion to retain valued members of staff.

Overall there was an issue about workload and stress, with competing demands from research, teaching and administration.

From the qualitative research autonomy, freedom and flexibility were greatly valued and seen as greater than in other industries, although there was some concern about decline in these. Aspects include timing of working day (and week), how work is organised and the content of work.

Peer group support tended to raise satisfaction, as did ability to participate in the wider academic community and in other activities.

Job insecurity was a problem, with insecurity affected by demands on student numbers, meeting RAE targets and previous redundancies.

The general direction of Higher Education policy reduced satisfaction. Much of this may have been due to perceived changes in higher education which had worsened jobs. Three-quarters of staff thought academic jobs had worsened since they joined. The factors identified in the qualitative research as affecting this were seen as bureaucracy (including QAA and RAE), increased management layers, less departmental autonomy, a higher workload, pressure to conduct research (in new

154

universities), lower staff/student ratios, increased demands from students, financial stringencies, increased use of short-term contracts, erosion of pay, pressure to retire early, a greater business orientation, an increasingly ‘corporate’ culture and an increasingly commercial outlook.

6.4.4 Factors affecting satisfaction Multivariate models were used to examine the impact of a number of factors

jointly on a range of measures of satisfaction. These were: the work itself, promotion prospects, salary, total earnings, relationships with manager, job security, use of own initiative, hours, relationships with colleagues and physical work conditions.

6.4.4.1 Personal characteristics

There was no difference between women or men in satisfaction, except that women were more satisfied with salary. This may be due to women having lower expectations of salary, due to discrimination in the labour market as a whole.

Academics from ethnic minorities were less satisfied with the opportunities they have to use their own initiative and their relations with their colleagues than their white colleagues are. This is likely to indicate that academics from ethnic minorities find themselves in less satisfactory jobs.

6.4.4.2 Experience

Satisfaction with the non-pecuniary aspects of the job tends to decrease with experience for the first half of academics’ careers but then increases. Two elements where experience does not have a significant effect are satisfaction with salary and total earnings.

6.4.4.3 Grade

Professors and, to a lesser extent, senior lecturers, are generally more satisfied in their jobs than lower grades, particularly with promotions. In particular, professors are significantly more satisfied with their salary and total earnings

6.4.4.4 Contractual status

Staff on non-permanent contracts are significantly less satisfied with their promotion prospects and their job security. However, they are more satisfied with their earnings than permanent staff. This is consistent with the idea that they are willing to sacrifice earnings for other aspects of the job at this early stage of their career.

6.4.4.5 Differences between universities

Levels of satisfaction with many of the job dimensions among staff are consistently higher at one of the universities (University 2, one of the southern new universities). Whilst this could be a locational effect, a more likely explanation is that, despite many aspects of the structure of employment in higher education being set nationally, universities may affect the satisfaction of their staff. We would have expected that satisfaction with salary would show a university effect, due to

155

differences in living costs between universities. However, this does not consistently show lower satisfaction amongst those in high cost locations.

6.4.4.6 Hours

Satisfaction is lower the greater the number of hours worked. However, hours spent on research have a positive effect on satisfaction with the actual work itself, the ability to use one’s own initiative and with hours generally. Hours spent on administration appear to reduce satisfaction almost across the board.

6.4.4.7 Research ranking

Staff in five star departments tend to enjoy the work itself more than other staff (as, to a lesser extent, do staff at five-rated departments) and tend to be more satisfied with their promotion prospects. This is consistent with the idea that staff wish to work at institutions of academic excellence and the stability that the recognition of this excellence imparts is also appreciated by staff. Staff at five star departments are, however, less satisfied with their salaries (although not their total earnings). This may reflect the fact that the national pay scales in academia constrain institutions’ ability to reward the most productive workers.

These results suggest there are three separate sets of factors which determine the job satisfaction of academics:

• satisfaction with non-pecuniary elements of the current job, which explains the majority of the variation in six of the ten dimensions of satisfaction;

• satisfaction with earnings (both academic salary and other earnings); and

• satisfaction with longer-term prospects (basically promotion prospects and job security)

156

157

7 Leavers

7.1 Introduction There is little useful evidence on outflow from the university sector or where

staff go to when they leave. The main difficulty in assessing destinations is that HESA data is unreliable, with the destination of over 60% of the staff who were recorded as leaving in the sector in 2001/02 unknown. Moreover, it is liable to be biased towards recording the destinations of those who remain in the sector. However, it is the best source of data on leavers.

Using this HESA data, PREST (2000) found that, for most years between 1994-95 to 1998-99 about half of those leaving higher education institutions left the sector, although this rose to two-thirds in 1996-97 (Table 7.1). Nearly all of these left the labour market. Only a very small percentage were reported to have left for jobs outside the sector (ranging from 0.15% to 0.37% pa). However, data for 2001/02 showed greater movement to other employment, with 18% of women and 10% of men entering some other public sector employment and 9% of women and 10% of men entering the private sector (source: HESA Individualised Staff Record, 2001/02). Moreover, a fifth of women were no longer in regular employment, as opposed to an eighth of men. 13% of women and 29% of men had retired.

Table 7.1 Leavers: HEFCE-funded staff Percentage of staff Percentage of leavers

Year all

moving

moving within the

sector

moving to other

employment

retirement & other non-

employment

moving within the

sector

leaving the

sector

moving to other

employment

retirement & other non-

employment 1994-1995 2.8 1.5 0.2 1.2 52 48 5 43 1995-1996 4.3 2.2 0.3 1.8 52 48 6 42 1996-1997 5.3 1.7 0.4 3.2 33 67 7 60 1997-1998 4.5 2.0 0.3 2.2 44 56 7 49 1998-1999 4.5 2.3 0.3 1.8 52 46 6 40

• Source: PREST, 2000, from HESA data

UCEA (2002) asked universities’ personnel departments about the destinations of leavers but few were able to provide aggregate information. Therefore, in the qualitative research, heads of departments were asked about destinations of leavers. Their evidence suggested that a similar proportion of leavers left the HE sector as remained, and that business-related areas and engineering are particularly vulnerable to losing staff to other sectors. Some institutions saw a particularly high proportion of their staff retire in the previous year (2001). Based on a survey of permanent IT and computer science academic staff, the EPSRC (1999) found that most permanent staff who left went to other higher education institutions.

158

At the commencement of the study, we had intended to investigate leaving by including in the staff survey and the qualitative research, a substantial number of leavers, i.e. those who were working their notice or who intended to serve their notice. However, the participating universities could not identify these people for the sample and the numbers were lower than needed. Therefore the study has relied on academics’ stated intentions about leaving and has also based the likelihood of leaving the sector on job search actions.

The next section describes academics’ stated career plans and their stated reasons for leaving. The following section models their job search behaviour, identifying the factors which are related with searching for jobs outside the sector.

7.2 Career plans Using the staff survey, this section describes academic staff’s career plans:

both long-term, in terms of their expectations of remaining in the sector to retirement, and short-term, in their expectations of changing university or leaving the sector in the next year.

We start by examining movement between universities and academics’ expectations about moving in the next year. The following section examines first long-term plans, of remaining in the sector to retirement, and then describes expectations of leaving the sector in the next year.

7.2.1 Career plans: changing universities Five per cent of academic staff saw it as very likely they would change

universities in the next year, whilst a further nine per cent thought it quite likely (Table 7.2). These figures seem high, compared with the PREST figures121 and it may be more realistic to take those definitely expecting to change universities as the lowest estimate and this together with those who considered it very likely they would change university as the highest estimate (i.e. between two and five per cent).

Career progression and dissatisfaction at their current university were the most common reasons given for expecting to leave (Table 7.3). (The qualitative research suggested that this was not necessarily due to academic advantages of moving for one’s career, but because promotion was easier to achieve through moving.) Aspects of dissatisfaction related to moving identified in the qualitative research included racism122 and exasperation at the level of bureaucracy. Some academics were concerned at possible Government moves to concentrate research activity in a smaller number of institutions and thought it advisable to move to one of those institutions.

Other staff were driven to leave due to their contract ending and to look for a permanent job, whilst one per cent of all staff expected to leave due to redundancy. As might be expected, given the incidence of fixed-term contracts, more staff in old universities were leaving due to their contract ending. Family reasons were also important (cited by 1 in 7 expected movers) and more common for those in old universities.

121 Note that the two sets of figures are not directly comparable, because of the different sample populations. 122 This is explored in more detail in Section 8.8.

159

Table 7.2 Academic staff: Likelihood of moving to another UK university in the next year

How likely will move to another UK university in next year Total Definitely 2 Very likely or definitely 5 Quite likely, very likely or definitely 15 Neither likely nor unlikely 16 Quite unlikely, very unlikely or definitely not 68 Very unlikely or definitely not 46 Definitely not 17 No answer 2 Total 100 n 2,788


Table 7.3 Academic staff: Reasons for moving to another UK university in the next year and for not moving

Reasons for moving (moversa only), total 14 Career progression 9 Dissatisfaction at current university 6 Contract ending 5 To get a permanent job 3 Family reasons 2 Redundancy 1 None of these 0 Reasons for staying (stayersa only), total 67 No reason to move – current job is fine 25 Family reasons 25 No reason to move – other jobs/universities would be no better 18 Inappropriate stage in career to move 18 Don’t want to move house 17 Nearing retirement 10 Career progression 5 None of these 4 Those who thought it was neither likely nor unlikely they would movea and don’t know

18

Total 100 n 2785

• a ‘Mover’s are those who had said they were definitely, very or quite likely to move to another university in the following year; ‘stayers’ are those who said they were definitely not, very or quite unlikely to move to another university in the following year.


160

One of the most common reasons for expecting to stay at one’s university was that the current job was fine (25 per cent of all staff). This was matched by family considerations. (Examples of this in the qualitative research were partner’s job or children’s education.) A relatively high percentage did not want to move house. Career considerations were also very common reasons for staying put and more common in the old universities. However, 18 per cent expected to stay for a more negative reason: that they did not expect other jobs or universities to be any better. Around 10 per cent of all staff were staying put as they thought they were too close to retirement to move. This was more common in the new universities.

There were some gender differences in the reasons for staying. More women felt that jobs elsewhere would be no better and so there was no reason to move. They were also more likely to stay for family reasons. Men were more likely to stay because of nearing retirement, which reflects their older age profile.

7.2.2 Career plans: leaving the UK Higher Education sector This section examines long-term career plans (specifically whether academics

plan to remain in academia until retirement) and academics’ expectations of leaving the sector in the short term.

7.2.2.1 Remaining in academia to retirement

A high percentage of academics expected to remain in UK academia until retirement. About two-thirds of academics expected to remain in the UK until they retired (Table 7.4). However, many of these expressed some uncertainty over this. Very few thought they would definitely leave, although almost one fifth thought they were more likely to leave than not.

Table 7.4 Whether expects to stay in UK academia until retirement

Old New Male Female Total Definitely 21 22 24 18 21 Probably or definitely 65 72 72 62 67 Don't Know 15 12 13 15 14 Probably not or definitely not 21 15 15 25 19 Definitely not 5 2 3 4 4 Total 100 100 100 100 100 n 1,727 1,060 1,575 1,181 2,788


There was some difference by gender, with women being more likely to expect to leave. Those in new universities were more likely to expect to stay through to retirement than those in old, despite the old universities having an older age profile.

7.2.2.2 Expectations of leaving the sector in the following year

Turning to the likelihood of leaving the sector in the next year, six per cent of academics thought it at least very likely they would leave the sector, with a total of 14 per cent thinking it at least quite likely (Table 7.5). These figures seem high, compared with the PREST figures (Section 2.2) and it may be more realistic to take

161

those definitely expecting to leave as the lowest estimate and this, together with those who considered it very likely they would leave, as the highest estimate (i.e. between three and six per cent). Seventy per cent thought it quite unlikely (or more). Almost one quarter of those on fixed-term contracts thought they were at least quite likely to leave the sector in the following year. This compares with nine per cent on permanent contracts. There was a greater tendency for staff in old universities to expect to leave the sector, but this largely reflected the greater use of fixed-term contracts in old universities.

Table 7.5 Academic staff: Likelihood of leaving employment in UK Higher Education in the next year


Definitely 2 5 3 Very likely or definitely 3 11 6 Quite likely, very likely or definitely 9 24 15 Neither likely nor unlikely 12 19 14 Quite unlikely, very unlikely or definitely not 78 55 69 Very unlikely or definitely not 58 36 49 Definitely not 22 14 19 No answer 1 3 2 Total 100 100 100 n 1897 844 2786


Of those leaving, 43 per cent expected to move to another job and 31 per cent expected to do something else (Table 7.6). One quarter were unsure about their plans. Those on fixed-term contracts were much more likely to be planning to move to another job.

Of those expecting to move to another job, 29 per cent planned to stay in academia, but to move abroad. The qualitative research identified three types of reasons for this: the desire to return to one’s home country, career development through moving to a better university, a change in lifestyle and to escape the uncertainties and problems of the UK Higher Education sector. (It was believed that the university systems of some other countries were less bureaucratic and offered academic staff more autonomy than is now possible in UK universities.) Many planned to stay in teaching or research in the UK, but outside academia (21 per cent). However, the largest group were planning to leave academia and UK teaching and research to enter some other job in the UK (38 per cent). The pattern for those on fixed-term contracts and permanent staff differed. Permanent staff were more likely to move to an academic job abroad, whilst those on fixed-term contracts were more likely to move to a research job in the UK.

162

Table 7.6 Sector leavers: plans

Of those expecting to leave the sector Permanent Fixed-term contract Total

expect to move to another job 36 48 43 expect to do something else 43 20 31 not sure 20 29 25 no answer 1 2 2 Total 100 100 100 n 176 205 395 Of those expecting to move to a job: type of job would be likely to move to An academic job abroad 38 23 29 A non-academic job abroad 4 6 6 A research job in the UK outside academia 5 24 17 A teaching job in the UK outside academia 4 4 4 Some other job in the UK outside academia 40 38 38 no answer 8 6 7 Total 100 100 100 n 97 160 266

Together, these patterns suggest that insecurity is one of the drivers for those on fixed-term contracts to leave, as they still seem keen on research. The reasons123 given for leaving support this (Table 7.7). Over half of fixed-term contract staff who expect to leave the sector are doing so due to their contract finishing, with almost half wanting to get a permanent job. About one quarter of fixed-term contract staff who expect to leave the sector are pessimistic about their chances in UK academia and so are leaving. Retirement was the main reason identified for permanent staff leaving. However, a high percentage were leaving for unidentified reasons.

The qualitative research gave further insight into the considerations to stay. Those who had considered working outside Higher Education included Physicists who had thought about consultancy or other commercial employment, and a Social Scientist who had thought about working as a translator. However, although attractive, the alternatives were seen to lack job security and also to offer little improvement in pay. Therefore even those who considered alternative employment felt it was safer to stay where they were. For example, a lecturer in Biological Science was considering applying for research posts in industry, but only if his university contract was not renewed and he had little choice but to look for another job.

Those who had had careers outside academia seemed to be able to weigh clearly the pros and cons of leaving. The advantages to returning to these sectors were principally in job content rather than pay. For example an Engineer was tempted to return to industry because she liked seeing ‘tangible’ results from her work, which she felt industry provided more than higher education, while a senior lecturer in Education missed the school environment. On the other hand, advantages of academia 123 Respondents were asked about reasons for leaving. However, these were restricted to ‘special circumstances’ rather than factors affecting satisfaction, as satisfaction measures would be examined through modelling.

163

included job flexibility, enjoyment of the subject area and of teaching and of the relative stability of Higher Education compared, for example, to the IT and telecommunications industries.

Table 7.7 Academic staff: Sector leavers: reasons for leaving


Contract ending 0 57 32 To get a permanent job 1 47 26 Difficulty getting another UK academic job 7 28 18 Retirement 19 6 11 You want to live in another country 14 8 10 To return to home country 5 11 9 Redundancy 10 8 9 Other family reasons 10 7 8 To care for my family 6 6 7 To study for another qualification 5 7 6 Health 7 3 5 None of these 43 9 24 Total 100 100 100 n 176 205 395


Some academic staff had considered working overseas in the past, to work in Higher Education rather than another area of employment. The attraction of such a move was to have a lighter teaching load as well as a change in life-style.

Finally, those leaving academia should not all be seen as permanently lost to the sector. Twelve per cent of academics had returned to academia after a period of other employment (Table 7.8). This varied substantially by subject discipline, with breaks in academic employment particularly common in computing science (23 per cent), Art and design (19 per cent) and Social studies (15 per cent).

Such breaks tended to be short. Five per cent of all staff had left the sector for under one year in total and only two per cent had had a total period outside the sector (since entering academia) of more than five years (Table 7.9).

164

Table 7.8 Academic staff: Whether has worked continuously in academia by subject

Main subject specialism

per cent of subject specialists not worked continuously in

academia (since entering academic employment)

n

Medicine and dentistry 9 142 Subjects allied to medicine 12 370 Biological sciences 11 292 Physical sciences 12 243 Mathematical sciences (excluding computing) 9 94 Computing sciences 23 145 Engineering 9 169 Social studies, including economics 15 374 Business and administrative studies 8 160 English literature and classics 10 51 Modern languages 10 71 Other humanities 11 209 Art and design 19 53 Education 7 214 Combined studies across above subject groups 11 107 Total 12 2784


Table 7.9 Academic staff: Total time spent working outside academia

Per cent Less than 1 year 5 1 year 1 2 years 2 3-5 years 3 6-9 years 1 10 years + 1 Total 12 worked continuously 88 Total 100 n 2785


7.3 Factors affecting the likelihood of leaving UK Higher Education

7.3.1 Introduction In this section we examine the influence of reported satisfaction, wage

differentials, experience and other characteristics on the likelihood of leaving UK higher education of academics. Our survey asked how likely staff thought it was that

165

they would leave UK higher education124. Their responses are summarised in Table 7.5. In this section we investigate the influence of job satisfaction and a number of other elements on this stated likelihood of leaving. Our model of job satisfaction and intentions to leave UK higher education is set out in Appendix G.

7.3.2 Results The results of estimating our ordered probit model of the likelihood of leaving

UK higher education are presented in Table A.28 on page 284 of Appendix G. The results are discussed in more detail in Appendix G; to summarise the main results:

• Staff report that they are less likely to leave the higher their earnings.

• Staff report that they are less likely to leave the more they are satisfied with their earnings.

• The more satisfied an academic is with the non-pecuniary elements of the current job the lower their stated likelihood of leaving UK HE. ! Thus the more satisfied academics are with the elements of their jobs

such as the actual work itself, relations with managers and colleagues and hours, the less likely they are to leave the sector.

• However, there is weak evidence that the more satisfied they are with their longer-term prospects the greater the stated likelihood of their leaving, although the statistical significance of the term is not robust to specification. ! This result may be because the important influences on the likelihood

of leaving the sector are factors such as the permanency of academics’ jobs and the amount of fixed-term contract remaining for non-permanent staff.

• Academics from other EU (and EEA) countries, Australia, New Zealand and the US report that they are more likely to leave UK HE than UK (and other foreign) academics.

! Our results support the hypothesis that these staff enter academic employment in the UK after completing a higher degree in the UK, but ultimately intend to return to their home country. If this is the case, such staff will only represent a short-term solution for lower-level jobs in UK higher education unless they can be persuaded to remain in the UK.

• Staff report a lower likelihood of leaving the more years of experience in HE they have, although they do so at a decreasing rate. ! The effect of experience on stated likelihood of leaving declines for

the first thirteen to seventeen years, depending on the specification, and remains negative until experience hits the mid to late twenties.

124 Due to concern about the reliability of responses to a question on the likelihood of leaving, respondents were also asked to describe their job search activities. The relationship between responses to these questions and to the likelihood of leaving was tested and the likelihood of leaving reflected job search activity. However, analysis similar to that which follows was conducted using job search activities as the dependent variable. The results were similar (in the same direction) to those using the likelihood of leaving variable, but tended to be insignificant. Therefore, we report the analysis using the likelihood of leaving only.

166

• Individuals who have had a break in their academic career report that they are more likely to leave again. ! This may be because of the individuals’ preferences themselves or

because they work in an area where there is more flow backwards and forwards between academia and the rest of the economy.

! There is, however, little evidence that those who have previous experience of working in other countries’ HE systems are less likely to leave, ceteris paribus.

• Staff on non-permanent contracts state that they are significantly more likely to leave UK HE than their colleagues on permanent ones.

• Furthermore, as the end of a fixed time contract approaches, the more likely an academic is to leave. The greater the amount of time left a contract has to run, the less likely the individual is to leave UK HE.

• There is some evidence that academics who work longer hours are more likely to leave UK HE. ! This is less true for hours of research than hours spent on teaching or

admin. Note that this effect on leaving is over and above their influence on job satisfaction.

• Staff who feel that their workload is too high and those who feel that decisions on either individual pay, recruitment to senior posts or promotion at their current university are not at all fair are more likely to wish to leave UK higher education.

• There is evidence that the reported likelihood of leaving UK higher education is higher in the areas with strongest competition from outside academia. ! The likelihood of leaving is highest among staff working in other

(non-engineering) technology and medicine and dentistry. It is lowest in English literature and other humanities.

• There is little difference in the reported likelihood of leaving UK academia when one compares staff across institutions, ceteris paribus, although staff in two of the new universities, one northern and one southern, feel that they are more likely to leave.

7.4 Summary and conclusions

7.4.1 Career expectations A high percentage of academics (67 per cent) expected to remain in UK

academia until retirement.

Between three and six per cent appeared likely to leave the sector in the next year. Of those quite (or more) likely to leave (14 per cent), 43 per cent planned to move to another job. This was more common for those on fixed-term contracts. Those on fixed-term contracts were most likely to expect to move to a UK job entirely outside research and teaching, to a research job in the UK or to an academic job abroad. Permanent staff were most likely to expect to take up a UK job entirely outside research and teaching or an academic job abroad. Academic jobs abroad were attractive for a number of reasons: to return to one’s home country, career development through moving to a better university, a change in lifestyle and to escape

167

the uncertainties and problems of the UK Higher Education sector. (It was believed that the university systems of some other countries were less bureaucratic and offered academic staff more autonomy than is now possible in UK universities.)

The reasons for moving differed for permanent and fixed-term contracts staff125. Contract ending, the desire for a permanent job and pessimism about job opportunities in UK academia were reasons given by fixed-term contracts staff for leaving.

7.4.2 Factors affecting leaving the sector The model of expectation of leaving the sector found the following factors

important.

Dissatisfaction with the non-pecuniary elements of the job increases the reported likelihood of leaving. The effect of satisfaction with longer-term prospects is ambiguous, and possibly positive. The results of the effect of wages suggest staff appear to have already discounted the wage they could earn in academia.

Academics from other EU (and EEA) countries, Australia, New Zealand and the US are more likely to leave UK HE than UK (and other foreign) academics. Our results support the hypothesis that these staff enter academic employment in the UK after completing a higher degree in the UK, but ultimately intend to return to their home country. If this is the case, such staff will only represent a short-term solution for lower-level jobs in UK higher education unless they can be persuaded to remain in the UK.

The likelihood of leaving UK HE falls with experience, although this does so at a decreasing rate. The effect of experience on quits declines for the first thirteen to seventeen years, depending on the specification, and remains negative until experience hits the mid to late twenties.

Individuals who have had a break in their academic career are more likely to leave again, suggesting that these staff are indeed more peripatetic in nature. This may be because of the individuals’ preferences themselves or because they work in an area where there is more flow backwards and forwards between academia and the rest of the economy. It would be useful to investigate further the causes for greater movement.

Staff on non-permanent contracts report that they are significantly more likely to leave UK Higher Education than their colleagues on permanent ones. Furthermore, as the end of a fixed time contract approaches, the more likely an academic is to leave.

Academics who work longer hours state that they are more likely to leave UK Higher Education (although the statistical significance of this result is reduced if we account for satisfaction with earnings or the academic’s perceived earnings gap with alternative employment). This is less true for hours of research than hours spent on teaching or admin. Note that this effect on leaving is over and above their influence on job satisfaction. 125 Respondents were asked about reasons for leaving. However, these were restricted to ‘special circumstances’ rather than factors affecting satisfaction, as satisfaction measures would be examined through modelling. These special circumstances were not applicable to a large minority of permanent staff and so their reasons for leaving are not discussed here.

168

Those who feel that their workload is too high and those who feel that decisions on either individual pay, recruitment to senior posts or promotion at their current university are not at all fair both report that they are more likely to leave UK higher education.

There is some evidence that the likelihood of leaving UK Higher Education is higher in the areas with strongest competition from outside academia.

The reported likelihood of leaving is highest among staff working in other (non-engineering) technology and medicine and dentistry. It is lowest in English literature and other humanities.

7.4.3 Implications for policy These findings suggest a number of issues for policy:

• job content is a very important influence on retention in the sector; facilitating research is likely to increase retention; this might include increasing time and funding for research and reducing other elements of the job (such as administration) which detract from research; changes in the structure or delivery of higher education which reduce academics’ involvement in research is likely to decrease retention;

• managing hours and workload might need some consideration, as those with longer hours and who feel their workload to be too high are more likely to expect to leave;

• fixed-term contracts are a major cause of leaving the sector; although the Fixed-Term Employees Regulations 2002 will prevent the repeated hiring of staff on fixed-term contracts in the same university, it is unclear the extent to which job security will improve; as discussed in the conclusions to Chapter 5, increasing job security may need more imaginative thought to be given to the organisation of research and lecturing jobs;

• relying on recruitment from the EU, EEA, Australia, New Zealand and the US to fill gaps in British recruitment may prove problematic in the long-term, as these nationalities are more likely to expect to leave; this suggests that policy should address home recruitment and not rely on foreign recruitment to address shortages;

• academics are concerned about the fairness of human resource decisions and lack of fairness could lead to sectoral turnover; it was clear that these could be problematic (further evidence of this is given in Chapter 8); improved fairness and clarity in decisions would help increase retention;

• pay and career prospects tend to have little influence on retention.

Finally, it would be helpful if the reliability and comprehensiveness of HESA data on leavers and their destinations could be improved.

169

8 The role of human resource policies and practices

8.1 Introduction The last three chapters have identified the factors which affect recruitment and

retention into academia. In the light of these factors, this chapter considers how universities’ human resourcing policies and practices might affect recruitment and retention. The evidence on policies and practices is taken from the eight case study universities where interviews were conducted with senior representatives and with a total of 70 staff in 16 departments. In each department, interviews were carried out with the head of department, a range of staff by grade, gender and ethnicity, and postgraduate students (see Appendix A for details of the qualitative research methods). As the case study universities cover the range of types of universities and locations in England and, despite this range, their practices exhibited variations within considerable consistency, we would expect their practices and experiences to be fairly representative of those across the sector.

In each case study university, senior human resources staff and heads of department gave a description of human resource policies and standard procedures (together with any critique). Other academic staff were asked to comment on these policies and practices. The aim was to establish practice through triangulation of knowledge and experience.

The chapter looks first at employment strategy and related recruitment practice. Within this, it discusses recruitment strategy and the recruitment process used in the case study universities. This includes the pay offer and other terms and conditions. The chapter examines the issue of recruitment difficulties, and universities’ response. It then looks at promotion practice, including systems, procedures and criteria used in the case study universities. In addition to the issue of recruitment problems, the chapter looks at the extent of retention problems experienced by the case study universities. In exploring this issue, it looks at pay systems, incentives, appraisal, training and development and work allocation. The chapter also considers equal opportunities policies and practice within the case study universities.

8.1.1 A note on research staff An interesting aspect of human resource use in universities is the separation of

academic staff into researcher only and lecturing and research staff, with different terms and conditions. As described in Chapter 3, nearly all research only staff are employed on fixed-term contracts, whilst most lecturing staff are employed on permanent contracts.

In the interviews in the case study universities, respondents tended to overlook research staff in their responses, only discussing their position when prompted. The work of research-only academics appeared to be isolated, the employment of

170

researchers tended to be devolved to departments and, in the larger departments, to sub-groups within the departments. Heads of department did not appear to be much involved in the employment of research-only academics. Indeed, in the departments with more than a handful of researchers, some of the heads of departments did not know the number of researchers recruited in the previous year or in post in their department.

Previous research has shown that many contract research staff wish to remain in higher education, either as researchers or lecturers, but that many leave due to the uncertainty of fixed-term contracts and the lack of career paths within research (Hasluck et al, 2001). At the same time, as discussed above, both old and new universities are increasingly emphasising the importance of research rather than teaching.

Nevertheless, the evidence from the qualitative research suggested that, whilst development and promotion opportunities within their institutions were the norm for lecturing staff, research-only staff were not treated in this way. Only in a few case study departments were research-only staff seen as potential and likely recruits to lecturing posts, whilst none of the case study departments offered a structured career path from researcher to lecturer (whereas such paths existed for lecturer to senior lecturer and for senior lecturer to principal lecturer).

8.2 Employment strategy The human resource practices should be seen within the universities’ overall

strategy. In the research intensive old case study universities, this was to develop research excellence: recruitment stresses research excellence, as does promotion and, in some cases, so do retention practices. In some of these universities, departments which do not score highly on the Research Assessment Exercise (in some cases 5 or less) are closed. Thus, in these universities, academic staff face pressure to perform in terms of research and are encouraged to accord other activities (particularly administration, but also teaching, to some degree) with less importance.

In contrast, most of the case study new universities and the less research intensive old case study university appeared to have a less clear-cut strategy. Most stressed the importance of teaching, and, in many departments, research as well. Recruitment tended to be driven by the need to meet teaching demands; redundancies and departmental closures tended to result from low student numbers. Research excellence was not a major criterion in recruitment and, in many departments, little time was available for conducting research. However, research excellence was an important criterion in promotion to reader and professor, whilst management competence was important for promotion to principal lecturer (see Section 3.3.1 for details of the academic grade structure). Thus, in these universities, existing staff are confronted with conflicting demands, whilst recruitment may result in a pool of staff for whom progression is unlikely. In response to the funding pressures, the case study new universities appeared to be more prone to reorganisation than the case study old universities. This led to further stresses and administrative demands on academic staff.

171

8.3 Recruitment practice

8.3.1 Recruitment strategy The strategies of new and old universities towards recruitment of lecturing

staff differed in the relative emphasis on teaching and research. New universities emphasised the need for lecturing staff to deliver undergraduate courses, while the principal need of old universities was for research excellence. Where new universities’ recruitment strategies included research excellence, this was either to keep the teaching curriculum up to date, or part of a change in strategy, aimed at improving the university’s national research standing. In contrast, some heads of department in old universities talked in terms of international research standards, and of their teaching programmes as ‘research focused’. Some heads of department in new universities also talked of the need for staff with experience of ‘enterprise’, referring to experience of working in industry or commerce.

Recruitment strategy depended, to some extent, on turnover. Heads of department said that most movement was for career advancement, either to more senior posts or for jobs at more highly ranked universities. Movement in the middle grades was lower because of the need to gain research and teaching experience before applying for senior posts. However, the extent of movement was described as varying according to current career opportunities and therefore by subject and department.

In most of the case study universities, departing staff are not necessarily replaced with staff at the same level, except where they perform a particular management role. It is common practice to recruit at a more junior level, both for reasons of cost and availability. In many of the case study universities, the departure of a member of staff was followed by a review of departmental or school staffing requirements, based on income, student numbers, and research output. Decisions about staffing were made according to the school plan and with reference to the current financial position of the department or school. Therefore, vacant posts are rarely simply replaced because universities require a case to be made for this. In addition, the grade at which a vacant post is to be filled is decided following a review of resources, and in consultation with the departmental or school plan. The grade at which a vacancy is advertised will therefore be decided in negotiation between the department or school, faculty and senior management of the university. In old universities, senior Lectureships and readerships were reserved as promotable positions and so only those leaving were replaced with lecturers, whilst chairs might be replaced with chairs.

As noted above, changes in demand were identified: human resource managers and heads of department reported increasing demand for lecturers productive in research, who were able to attract research funding and achieve high research ratings in the next Research Assessment Exercise (RAE). Reflecting this, universities reported increasing ‘poaching’ of these staff by other universities and turnover was expected to increase in the run up to the next RAE. Despite bringing rewards to the university, there was a degree of ambivalence among some human resource managers about recruiting ‘stars’. One complained that,

‘They know they are in demand, especially as we approach the next RAE, and some of them have something of the Prima Donna about

172

them – like what are you going to offer to get me to come here?’ (Head of Human Resources, new university)

In very few of the case study universities did it appear that the university’s own PhD students were considered as part of the pool of potential recruits. Some of the higher status old universities, as well as those with lesser academic standing, appeared to overlook this source. Indeed, several heads of department seemed to consider it important that PhD students gained jobs at other universities to broaden their experience, although this approach was rarely taken in respect of staff once they had gained a lecturing position. As discussed in Section 3.4, some universities were giving new lecturing staff the opportunity to study for a PhD as a recruitment incentive and where they could not recruit new staff with doctorates.

8.3.2 The recruitment process As described above, the recruitment process begins with requests from

departments or schools to fill vacancies when staff leave or to meet changing teaching and management requirements. In some of the case study universities, this can take some time and it is quite usual for posts to remain vacant for some time, sometimes longer than a year, while decisions are made about use of resources and staffing needs. Indeed, some departmental heads explained that because recruitment may be delayed and take place some time after staff leave, it is sometimes difficult to ‘trace back’ and say which former post is to be replaced.

Once recruitment has been agreed, much of the responsibility devolves to departments. Departments or schools were usually responsible for drawing up a job description and person specification. The process of advertising and the organisational aspects of recruitment, for example convening an interview panel, were then carried out by human resource departments. In some old universities more of the recruitment process was devolved to departments or faculties than in new universities. However, across the eight universities, most heads of department felt that they had a significant degree of control over the recruitment process, despite the involvement of human resource departments and representation of other schools on selection boards (see below).

Little variation was found between universities in advertising practice: posts are advertised in the Guardian, the Times Higher Education Supplement and the website jobs.ac.uk. Some use is made of professional and specialist journals, but almost exclusively of scientific journals such as New Scientist and Nature. These journals were seen as particularly useful in attracting overseas applicants. Universities were unwilling to fund any further advertising on grounds of cost effectiveness. Short-listing of applicants is carried out by departments or schools and interview panels include representatives from the department, faculty, senior management and human resource department. The composition of panels depends on the grade of appointment, with departments having stronger representation where posts are junior.

Due to recruitment difficulties, many of the case study universities were changing their methods of advertising posts, which traditionally involved placing advertisements in newspapers and selective journals, sitting back and waiting for applications to arrive. Many departments found advertising alone to be ineffectual, particularly for senior and specialist appointments. Human resource managers confirmed that, across their institutions, departments and schools were relying less on

173

advertising to attract applicants and more on active search methods. For the recruitment of senior staff this included forming search committees to establish a list of potential candidates who could then be approached. These bodies included staff at senior management level, with a panel involving the head of department, dean of school and a pro-vice chancellor. Once these individuals had been identified, a personal approach by a senior member of academic staff was thought to be more effective in encouraging suitable individuals to apply. Although this has implications for equal opportunities, such concerns were not expressed. An alternative approach, of using professional head-hunters was not considered effective, both on grounds of cost and the need for subject knowledge to identify suitable candidates.

Case study universities were paying increasing attention to their recruitment practices at professorial level and one university had revised the process to include an expanded ‘search’ stage. This was introduced in response to small numbers of applicants for senior posts and high drop out during the recruitment process. Where particular difficulties have been anticipated, some universities have hired consultants to head-hunt with mixed results. However, a number of heads of department said they might carry out more informal enquiries in future, particularly to find strong applicants for more senior posts.

Efforts spent in researching suitable candidates were seen as particularly worthwhile when expended in recruiting senior staff who might raise the profile of the department and attract other staff to the department or school. As a deputy director of personnel explained,

‘There is no doubt that if you have key individuals, particularly in terms of an area of research then recruitment becomes less problematic. It sends out a message that the department is committed to research in that area and gives applicants more confidence that they’ll be coming to the right place.’

Although the use of search committees was not widespread, a number of universities were encouraging departments to carry out market research on recruitment in their subject area and to use their networks to identify potential recruits. As one deputy director of personnel explained,

‘There is a shift in emphasis towards thinking how we can attract the staff that we would like rather than advertise and see who applies, and then have to re-advertise when we don’t manage to recruit anyone’.

Searches were reported increasingly to extend overseas, particularly for senior posts in old universities. The Far East, including China, was reported to be an increasingly fruitful recruiting ground both for staff and postgraduate students. Other areas of recruitment overseas included Russia and countries in mainland Europe. This is in line with the findings by HEFCE (2003) and the case studies undertaken in UCEA (2002), that staff shortages are encouraging higher education institutions to look increasingly overseas. The converse of this is the extent to which universities draw upon graduates from their own institutions rather than from other universities in

174

filling academic posts. This a not a strategy we identified, although Hoare (1994) found variations in this practice126.

Some departments in new universities had targeted firms where redundancies were being made to recruit employees with an interest in teaching or research. Universities were rarely able to recruit staff with any experience in these areas, but they were recruited on the basis of potential, and for their commercial knowledge. Approaches were also reported from firms looking for employment for staff in their redundancy programmes. These were principally engineering and computing staff, some of whom were interviewed for the research.

Some departments, for example health and law, were making increasing use of secondments and joint appointments to attract people with professional qualifications for a specified period. This was not seen as ideal because universities would prefer staff to stay for a longer period than the usual secondment of 2 or 3 years. However, secondment was seen to have the advantage of keeping a department updated with aspects of professional practice. Some departments were also offering part-time posts to some applicants where it was thought this would be an attractive option.

8.3.3 The pay offer Human resource managers saw pay as one of the main reasons for recruitment

difficulties and were looking for ways in which pay offered to new recruits could be increased. The grade at which staff were recruited was not usually negotiable, since this is usually agreed with the faculty, with reference to the responsibilities of the post, before recruitment. However, flexibility within the grade was very frequently exercised, so that a recruit would be placed nearer the top of the scale than at the entry point. As is common elsewhere, the point on the scale would take into account current salary. However, matching existing salaries was most difficult when applicants were from outside the Higher Education sector (for example in the case studies in business, education and health).

The main factor affecting the pay offer was current salary, although, where recruits were coming from higher paid sectors outside higher education, this could not always be matched. Some departmental heads also mentioned that comparability inside their department was taken into account, but this was not universal. Others mentioned experience and age. However, a senior lecturer said that the university had raised his offer so that ‘I would earn more than my wife’.

Universities were also using a number of formal ways of increasing salaries in the harder to recruit areas. These included ‘Golden Hello’ payments (prior to the Government scheme), market supplements and retention payments. Golden Hellos were one-off payments in the region of £4,000. Some universities offered ‘relocation payments’ of roughly £8,000 to cover the moving costs of key recruits. Old universities were more likely to offer introductory payments than new universities because a greater proportion of their recruitment is from the international academic market. Recruitment payments, including Golden Hellos, were usually very closely targeted at certain subject areas and posts where recruitment difficulties had been experienced. Human resource managers were not in favour of the Government’s

126 Hoare’s (1994) analysis is based on details of individuals’ degrees listed in the Commonwealth University Yearbook.

175

Golden Hello initiative because they saw it as prescriptive. One director of personnel argued,

‘It’s a ‘one size fits all mentality. It is unsubtle to what we are trying to achieve and corrupts what we are doing here. We might not want staff in the areas that HEFCE have defined as key. What we’d like is to decide the areas where recruitment bonuses would work best.’

Some heads of department were also critical of Golden Hello payments because they would result in new staff being paid more than existing staff.

Market supplements were paid by a number of old and new universities. These are regular payments aimed at redressing some of the imbalance in salaries in Higher Education and in the external market, usually in the private sector. Therefore they were most commonly awarded in such areas as computer science. They were used as much for retention as for recruitment, since they were generally awarded to all staff in a particular subject area or group rather than new arrivals. However, there were indications that universities were targeting these payments more closely (see below). Some human resource managers were uncomfortable with the idea of market supplements for recruitment and, to a lesser extent, retention payments because they were seen to create discrepancies in conditions between staff in different subject areas and lead to resentment. Moreover, it was argued that few areas suffer prolonged and intractable difficulties, so that the need for such measures is often temporary. The deputy head of a computing department in a new university explained that a market supplement of £3,700 had been introduced during a period of recruitment difficulties, but that this was soon found to be unnecessary. As he explained,

‘In IT there is a business cycle and with unerring accuracy the market supplement came in at exactly the wrong time. Post 2000 there was a big shake out, following the build up to the millennium and the IT industry has been in recession ever since. At the time market supplements were being introduced we were getting 28 applicants for a PL post. They came in bang on time when you could have recruited easily, but that’s the public sector for you’.

Departments competing with sectors outside higher education tended to believe pay differentials caused problems. However, often, competition appeared to be confined to early career stages127) when pay differentials between academia and other careers were often small. Therefore, in these cases, either departments were wrong in their belief that pay differentials were causing problems or the problem was that, in making their career choice, potential entrants compared long-term earnings in academia and elsewhere. However, for some subjects and posts, competition does continue into higher levels of academia. For example, in education and health staff were recruited from outside higher education at all stages of their professional careers. In these two subjects, recruitment difficulties had arisen due to pay increases in teaching and in the health service, resulting in major recruitment problems. The Head of one of these departments thought that targeted pay increases (for staff in these

127 This was because as careers progressed, the skills of those in academia and many other careers increasingly diverged, restricting transferability.

176

subjects) would be problematic, reducing morale across the university, and that the general slippage in academic pay needed rectifying.

8.3.4 Other terms and conditions Pay was not seen as the only factor discouraging potential applicants.

Universities referred to other aspects of academic work and the environment which could be improved. These included office space, facilities and equipment which had been improved where recruitment problems were severe. Universities, both old and new, were also offering incentives to new staff to gain research experience through induction awards and start-up funding. Others offered reduced teaching and administrative loads to allow more time for research, although in some cases the reductions did not materialise (or were not maintained) leading to discontent and, perhaps, turnover. A number of staff from different departments in one of the case study new universities said that at recruitment, the university had emphasised allowing relatively high time for research, but once employed this diminished rapidly. Not surprisingly, this was seen to lead to dissatisfaction and turnover. All universities offered teacher training to new lecturing staff, which some saw as a recruitment incentive. Where universities and departments could not recruit staff with higher degrees, they were offering applicants the opportunity to study for a masters degree or doctorate, through reductions in teaching time. In some cases these were formally offered as ‘teaching fellowships’, a sub-lecturer post. This has the dual benefit of developing research skills and allowing promotion and salary increases. These are awarded as short-term contracts, usually of around 3 years. Universities reported mixed success with this measure: one new university was disappointed when all of its new psychology recruits left after completing their PhDs.

8.3.5 Response to recruitment difficulties We have described the difficulties experienced by some case study universities

in attracting either sufficient applicants or those of the preferred quality. Universities had two options when faced with a poor field of applicants: to re-advertise in the hope of attracting a better field; or to lower standards. Re-advertising was not generally thought to be worthwhile, so the second option was often taken and compromises made. A few heads of department referred to this as ‘dumbing down’. There were indications that this was more common in new universities than in old. One human resource manager explained that,

‘Heads of school will typically say that the candidate they recruited met the requirements. They perhaps weren’t ideal, they didn’t have much of a research track record, perhaps didn’t come across as that keen, but they were fine. So when I say that recruitment hasn’t been a problem it’s because we aren’t looking for the stars. We want people of high quality but we aren’t the Waitrose of the Higher Education world’. (new university)

Some heads of school felt this was not an option for them because of the risk of losing current high performing staff. As the head of Engineering at an old university explained,

177

‘We would defer recruitment rather than appoint someone we didn’t think was right. We have to ensure that they are as good as the staff we already have. We put 70 staff in the last RAE and we have to make a judgement according to what we have here as well as what there is outside. We can’t lower our standards, it wouldn’t be good for the school and it might even lead to good staff leaving.’

The consequences of deferring recruitment were seen as equally serious for some departments. Failure to recruit resulted in increased teaching loads for existing staff and it was feared that these might move elsewhere. Some departments, particularly in new universities, made increased use of temporary, part-time staff to cover for gaps in permanent staff. This had the disadvantage of increasing the administrative load on the core workforce, since temporary staff only teach, rather than carry out research or administration. Some universities increased the long-established practice of paying postgraduate students to carry out some teaching, partly to make up for shortfalls in lecturing staff. A further option for universities was to cut course modules. This carried the risk of disappointing students and affecting enrolments, and was seen as a ‘last resort’. It was more usual for research activity and income to be affected by recruitment difficulties than for teaching to be reduced. This included the supervision of research students. However, the consequences for teaching were sometimes serious and included lack of continuity when course tutors changed and delivery by non-experts in the field.

The longer-term responses to recruitment difficulties in the case study universities included changing traditional advertising methods and making academic jobs more attractive, principally through enhancing pay. These are discussed further below.

One factor leading to discontent and perhaps higher turnover was differences between the job and expectations of the job at recruitment. In some cases this could be due to inexperience and lack of knowledge. For example, a head of department in Law at a new university said,

‘The job can seem attractive to people in professional practice who want a life outside work. Beginners often think they are going to be old-fashioned university lecturers. They don’t know about the demands of the QAA and so on. They want to teach and do research and don’t realise how unrewarding a lot of teaching is and that they have to get their own research – the university doesn’t hand it to you on a plate’.

In some cases, it seemed to stem from universities’ attempts to lure applicants, through promises of low teaching loads and time for research, which either did not materialise or which were withdrawn within a few years. In the case study universities, this problem was only found in the new universities.

8.4 Promotion practice Promotion systems and procedures can play a major role in retention both

within the sector and within a university, affecting the need to move for career progression and satisfaction. The need to move depends on the relative extent of

178

internal and external promotion opportunities, whilst perceived fairness and the appropriateness of procedures appear to affect satisfaction and retention in the sector.

Both internal and external promotion systems are important in universities, with the approach varying not only between, but also within, universities. Universities also vary in their use of individual-based promotion and job-based promotion, that is whether promotion merely recognises that a person is performing at a higher level (individual-based) or whether it moves a person into a higher level vacancy (job-based). Procedures appeared to vary little between case study universities: annual promotion rounds were used in all the case study universities, although for some staff some other procedures were used. However, the criteria and transparency of the process did vary, affecting both access to promotion and satisfaction. These aspects are described below.

8.4.1 Promotion systems In the case study universities, three main systems for promotion were

identified.

The first, used in the old universities for all lecturing posts and in new universities for promotion to reader or professor, entailed an annual promotion round, with individuals applying for promotion. The number of promotions was usually capped and varied in the narrowness of ring-fencing (to departments, schools or faculties). Thus those seeking promotion competed for the available promotions with others (sometimes in very different subjects) and therefore meeting promotion criteria did not guarantee promotion.

The other two standard approaches were found in new universities, for promotion to senior lecturer and to principal lecturer. Promotion from lecturer to senior lecturer was often automatic, either after a few years as a lecturer, on achievement of certain qualifications or on reaching the top of the lecturer scale. Promotion to principal lecturer tended to be job-based, with posts carrying specific management responsibilities and the post filled through advertisement (including external). Thus, with the major exception of promotion to principal lecturer, promotion tended to be individual-based and mainly conducted through an annual promotion round.

One case study university did not have annual promotion rounds. The director of human resources referred to annual rounds as ‘dreadful things’. In this institution the route to promotion was by vacancy only, advertised externally and internally. However, job-based promotion could also be problematic, if vacancies rarely arose, particularly if staff took on responsibilities at a higher grade but saw no promotion. Indeed, a new university was introducing an individual-based system for promotion to principal lecturer, in parallel with the job-based system, due to lack of vacancies at this level.

Slight variations in these systems in a few case study universities increased flexibility somewhat. In one old university, biannual promotion rounds were held, although only for promotion to readerships and chairs. This and other old universities had no cap on promotion to these levels. Self-nomination was not always used. In two case study universities individuals could not apply and, instead, heads of department or faculties nominated individuals. In both, consideration had been given recently to switching to self-nomination, but been rejected. For example in one, it has been

179

decided not to change it in this way because of concern that it would only benefit staff with more confidence. At the same time, human resource staff were concerned that the present system encouraged the ‘blue-eyed boy’ syndrome in promotion.

In several of the case study universities, promotions were also agreed outside the annual promotion round to prevent staff from leaving. In some cases, this could result in immediate promotion, whilst in others the formal procedures of the promotion round would be followed. Most heads of department could only speculate about how they might encourage a member of staff to stay, because they had never done so, although human resource managers said it was practised in their university. Therefore, a number talked of how they would explain the career opportunities within the university, if they felt the member of staff had not appreciated these. As a head of history in a new university explained,

‘I would talk through their perceptions about their job and if I felt they misunderstood what their prospects were at the university I would try to dissuade them, but if it was a good move then I wouldn’t stand in their way’.

Other heads of department also expressed a reluctance to persuade members of staff against leaving for better jobs.

In some cases, this ‘backdoor’ approach was formalised to the extent that the university only allowed accelerated promotion if a formal job offer had been received. In other cases, only belief that the person was likely to leave was required.

8.4.2 Procedures and criteria Annual promotions rounds involve a process of application and assessment,

with the degree of formality and paperwork varying within universities and according to the seniority of the post. For example, in one old university, promotion to senior lecturer involves a written application endorsed by the head of department approved by the faculty and then ‘rubber stamped’ by the Vice Chancellor; in another, external referees are required. Promotion to reader or chair involves interview and external assessment. As with recruitment, the more senior the post, the more senior are the external assessor and representatives on the interview panel. Promotion and appraisal were not linked in any of the case study universities. The procedure took many months, for example nine months from application to being promoted in one university.

Promotions were made against grade criteria, but these were seen as insufficiently clear by some human resource managers and heads of department. Therefore some universities were in the process of clarifying promotion criteria (to improve ‘transparency’) and making criteria known to staff. Having clear criteria for promotion was seen by some to dispense with the need for an interview with the promotion applicant, since evidence could be set out against pre-determined criteria and an objective assessment made. However, even where clear promotion criteria were in place, some heads of department felt that these were not always interpreted fairly. At the same time, some human resource managers felt it was necessary to apply different standards, given the different labour markets for academic subjects and recruitment and retention difficulties. For example one director of personnel remarked,

180

‘Nationally it’s easier to get a chair in economics than in other disciplines and we are having to create more chairs and increase pay to take this into account. If we don’t, they’ll go somewhere else for their chair, so we just can’t use the same criteria for staff regardless of their academic area.’

Promotion criteria, in old and new universities, included activity in the four areas of research, teaching, administration and enterprise. Universities were concerned that the criteria used should take account of the full range of academic activity and not disadvantage staff making a valuable contribution in one area, but not another. Traditionally research activity, including income generation and output have been valued most highly at old universities, particularly in appointments to reader and professor. Although some old universities were aiming to increase the weighting of other activities, particularly contributions to teaching and administrative responsibilities, research continued to be the pre-eminent criterion. A head of department considered this problematic as it was a disincentive to taking on administrative tasks (a view which was backed up by junior academics). Moreover, he commented that ‘research excellence’ was measured in relation to the RAE, distorting research towards that and providing short-term payback.

New universities, in contrast, have traditionally used teaching contribution, and associated administration, as a key criteria for promotion and promotion to principal lecturer placed emphasis on administration. Pressure to raise the research standing of their institutions was leading many new universities to place more emphasis on research than teaching in their promotions criteria. This was seen as problematic by some heads of department who felt the change in emphasis could negatively affect the promotion prospects of good teaching staff.

Length of service sometimes entered the criteria. For example, in one old university, lecturers had to have reached the top of the scale before they could apply for senior lectureships (although this restriction did not apply to senior lecturers applying for readerships or for any grades applying for chairs). In a new university, a human resource interviewee commented,

‘There has been assumption that promotion to principal lecturer should be by time-serving, whereas we would do better to emphasise quality of output. There has been a lid on promotion because schools want to save cash, so we’re reviewing how best to measure quality of research and teaching to see how this could be more open and we can get more through.’

One university required a masters degree or higher for promotion to senior lecturer.

The transparency of the procedures varied. It was clear that stated procedures and criteria were not universally followed and that staff had varying views on what these procedures and criteria were. A head of department in a new university said,

‘There is a tension between the criteria as laid down and how it is defined and interpreted.’

181

Many staff who had not yet considered promotion were unaware of the procedures and criteria. However, some other staff also appeared not to know the actual procedure or criteria. A head of department in a new university thought this was, in part, the fault of the university, saying ‘The criteria are not always clear’.

Also problematic was the extent to which procedures and criteria were not followed. As reported above, this occurred to reduce turnover sometimes leading to differences in treatment between subjects. However, it also occurred in other circumstances, with criteria not being properly applied. It was clear that this was seen as unfair and was a major cause of dissatisfaction for academics.

A human resource specialist in a new university said,

‘still sometimes there is a tap on the shoulder and people are put into positions that were never openly advertised, even within the college’.

8.4.3 Response to retention difficulties/changes: promotion systems Case study universities were looking more generally at progression and

promotion, from a concern to make criteria more transparent and to make the process smoother and faster. The principal need was seen to identify and reward strong performers before they start job-searching. Some universities were awarding retention payments to targeted groups or individuals, usually ‘star’ researchers in key areas who might be poached by other institutions. For example, one old university identified grade 5 research staff in economics for additional payment. These payments were funded from the HEFCE rewarding and developing staff initiative (KPMG, 2005).

It was less usual for universities to look for ways to recognise the achievements of junior staff and to ease movement up the academic career ladder. Some department heads spoke of the need to assist the career development of junior lecturers, so that they might achieve promotion within the institution rather than by leaving, but this was rarely identified as a way of improving staff retention. A human resource specialist in a new university said,

‘There is no formal system for identifying potential among existing staff but we are starting to formalise this. We have no fast-track system and it is all very ad-hoc and not managed’.

The career development needs of contract research staff appeared to be even less of a concern, with only one university actively seeking career ladders for this group of academic staff, and few departments seeing them as a source of recruits to permanent teaching posts.

Some universities were moving away from systems of nomination for promotion by senior staff to self-nomination. Universities were also improving management training, partly with the aim of assisting managers with staff development, in order to prevent turnover in their schools or departments (see below). Staff management was also identified as an area of change, which might improve staff performance and assist in meeting criteria for promotion. Most universities had introduced annual appraisal only in the past few years. The following section explores the issues of appraisal, development, management and promotion.

182

One case study university was changing to a new grading system for academic staff. Four job families were planned, each with 7 levels with ‘inscriptions’ for each. The idea behind this change was to make the promotion route clearer and to value the contribution of academics who are making a contribution other than to teaching and research, for example enterprise and intellectual property. A performance-related stage will be added to the top of each scale so that good performance can be rewarded. This move was part of the university’s response to HEFCE’s call for greater transparency across the pay structure.

8.5 Pay system Below professorial level, most university academic staff are employed on pay

scales set nationally. Reflecting their different grade structures new and old universities have separate payscales for academic staff (see Chapter 3, Table 3.2). Old universities have ‘discretionary’ points at the top of each lecturing scale (above Lecturer A) and for the Grade II research scale. Professors are paid on an individualised basis in all universities, with a minimum level set nationally. Additional responsibilities payments (e.g. to heads of department) may be made. Progression up the scale is normally by one increment per annum until the top of the scale (or the first discretionary point) is reached. One of the new universities differed in their approach, with performance-related pay for all senior lecturing staff, enabling salaries to increase by up to 10 per cent. The system was based on people with similar experience and output receiving similar pay. This was being subverted through the use of pay incentives for recruitment (see above) and for retention (see below).

A similar criticism to that of promotion was raised: that staff took on duties such as programme leader with no additional pay. A human resources specialist said,

‘We’re still living with a payment scheme that relates to a different era of employment. In the past achieving programme leader was high stakes and carried a lot of status, which could then lead to promotion, but now the sector is driven by research and having management or administrative duties is viewed negatively. Not only does it not have status in itself, it means you have less time to do research.’

8.5.1 Pay level This section concentrates on the pay system. However, the level of pay is also

of interest. Much of the discussion about recruitment and retention problems focused on pay relative to other sectors. Human resource specialists tended to consider the level of pay to be low, resulting in recruitment and retention problems.

8.5.2 Retention incentives Retention incentives, together with inflation of pay on appointment for hard to

fill jobs (whether through a market supplement or through appointment to a higher than normal point for the recruit’s skills and experience) were leading to a departure from a long-standing approach of setting pay based on experience. As has been described in Chapter 6, this caused some discontent amongst academics, although others recognised the need for such payments.

183

The single increment progression up each scale was seen as a problem for retention. A human resource manager said,

‘The deans say they are not able to keep pace with staff expectations of salary. Junior staff begin on a low salary and they can get experienced and very good in quite a short period of time. The only way they can get a higher salary is to leave the university for other institutions. These are typically younger and enthusiastic staff with strong academic ability’. Human resources manager, new university.

They then illustrated the problem with the following example,

‘A member of staff left for a more highly paid post and then re-applied for a university job 6 months later. Because the job he left for was more highly paid, he was able to re-enter at a salary of at least £5K more. This was thought to be very bad for staff morale.’

This human resource manager attributed the problem to the lack of accelerated progression up the scales, lack of merit pay and no regrading of jobs if responsibilities increased. This university had proposed a scheme which would have allowed acceleration at the discretion of the director of the academic school. The criteria and process were believed to be insufficiently transparent and there were concerns about ‘blue-eyed boy syndrome’.

Case study universities were increasingly using pay incentives to keep valued staff, particularly in the hard to recruit areas and with senior research productive staff. A number of universities, new and old, allowed departments and schools to increase the pay of staff planning to leave, through a range of measures: movement up the scale, fast-track promotion or retention payments. Precisely what would be offered to potential leavers would depend on their reasons for leaving and what was being offered by the competitor institution or employer.

This had been formalised into the structure in some of the case study universities. For example, in a new university, the human resources department meets with schools to discuss staffing issues. Schools flag up potential leavers and incentives may be offered to these staff to stay if they are working in a key area for the university (in terms of attracting funding, other staff or students). In an old university, senior Management identified departments where retention was particularly important. Within these departments, star researchers are identified and may receive a cash bonus. However, in most of the case studies, retention payments were ad hoc in response to an individual saying they were leaving.

Despite the growth in use of retention incentives, many of the human resource staff interviewed did not see them as particularly effective and also saw them as having undesirable side effects on morale.

A number of heads of department said they had offered promotion or pay increases to encourage staff to stay. In most cases these had been unsuccessful, sometimes assisting staff to gain a higher pay offer for their new post but not encouraging them to stay. It was found that these were almost always offered too late, and that the member of staff had already decided to leave. Where they were successful, they had involved large salary increases for staff already on relatively high

184

salaries. The head of engineering sciences at an old university reported that two professors in his school had been offered salary increases of £25,000, taking their salaries to £80,000. One of these professors had been offered a post at a UK university with a salary of £90,000. Although effective in these cases, salary increases of this scale are clearly beyond the budgets of most university departments.

Most heads of department were opposed to offering financial incentives to stay, on the grounds that they are likely to be ineffective and that they are divisive. Therefore as a head of Sociology in an old university stated,

‘I wouldn’t want highly valued staff to go without having the chance to talk to them about how their career could develop here but we don’t cut deals with people’.

In some old universities, retention payments had been given on an ad-hoc basis for some years and, although they were often not effective in encouraging staff to stay, universities were planning to use them more strategically. As stated above, part of the reason for their ineffectiveness was that incentives were offered too late, once staff had made the decision to go. Therefore a number of universities were concerned to identify what one called ‘significant contributors’ to the university’s output, who were likely to be poached by other institutions. Heads of department were being asked to identify these individuals, whose pay and conditions might then be improved. Some universities were formalising their practice into a ‘retention framework’, which might include facilities and staffing support as well as salaries. Whether these will work in practice will depend to some extent on the support of heads of department. Many interviewed were opposed to differential pay systems for staff within their department, seeing these as divisive.

Finally, the use of pay in recruitment and retention had been leading to an increasing split between higher education related skills and experience and pay, with similar people being paid differently within and across departments. In some case study universities, reorganisation had had similar effects, whilst other problems arose from pay not reflecting additional responsibilities. For example, in a new university, re-structuring had resulted in some principal lecturers no longer having the additional responsibilities which originally went with their post, but retaining their pay. Conversely, some lecturers and senior lecturers were carrying out responsibilities of a higher grade on no additional pay. This university, along with others, was working on job evaluation to try to address the issue.

8.6 Appraisal, training and development Human resource managers saw induction and appraisal as having an important

role to play in staff integration and career development. However, these and other management practices were not identified as key factors in retention of academic staff, either by human resource managers or by heads of department, although some aspects appeared to affect staff satisfaction and would affect career development.

8.6.1 Recruits All case study universities had an induction process. In some this was

delivered at a number of levels: institution, faculty and department. Induction was

185

generally brief, typically consisting of half a day in the case of a faculty or departmental induction. New universities’ induction processes were of longer standing than in old universities and were often longer and more in-depth. The regularity with which induction training was held varied, ranging from termly to annual. Induction was generally supported in principle, although not universally; one head of department in an old university thought it was a waste of time, stating that ‘Academic staff don’t need an induction process’. However, in practice, the picture was more mixed, with some respondents considering it good and others criticising it as too brief and not providing the appropriate information.

Some case study universities were also introducing mentoring schemes for new staff in their first year at the institution, and one new university operated a peer observation scheme in which an experienced member of staff observed and commented on a new lecturer’s performance. In another, the mentor observed teaching, reporting this to the head of department. Although the principle of mentoring was generally supported, effectiveness varied, and some universities did not have adequate systems to ensure poor mentoring was addressed.

Teaching staff with little teaching experience would take the Postgraduate Certificate in Teaching and Learning in Higher Education (PGCTLHE). However, one university was developing its own qualification for new staff. This was mainly designed to improve teaching quality, but was also aimed at improving retention. In some universities, teaching loads were reduced in recognition of the additional time demands for new lecturers, but this was not the practice in all of the case study universities, which could cause problems as junior lecturers had to acquire teaching and research skills whilst having a full teaching load.

8.6.2 Appraisal In most universities and departments appraisal schemes were either relatively

new or had been recently changed. However, most were developing along similar lines.

Appraisals tended to be development focused (and not linked to pay or promotion). Therefore, it was intended that appraisals should result in personal plans. These are then used to draw up a staff development plan and to provide specific training. However, some heads of department said that this policy had not yet been put into practice, or that courses met only generic skill needs, for example in information technology, or that funding was inadequate to meet training needs.

Views on the usefulness of appraisal varied. One human resource manager commented that they thought ‘younger staff expect to be managed more than in the past, in terms of getting feedback on their work, being developed and having appraisals’ and that unmet expectations might be a factor in staff turnover. For example, the heads of both departments in the case study at one new university described the process as being very useful, in that it enabled development needs to be identified and that these then fed into the university staff development policy. However, one did consider that more training could be made available and that too little was provided outside teaching. In another university, where the head of department saw it as useful, one of the members of the department said ‘I’m not sure it’s taken terribly seriously’. Some heads of department were not convinced of the need for induction and appraisal. A head of physics at an old university expressed the view that,

186

‘Staff are appraised every time they submit a paper or make a presentation at a conference. Yet more appraisal is a waste of time.’

Universities were looking for ways to link staff appraisal more directly with promotion. This included better recording of achievements and progress to feed into the promotions exercise. For this reason a number of universities had reduced the interval between appraisals to a year. Some universities were also using appraisals to assess eligibility for performance related pay, for which they have been given Government funding since 2001. Appraisals were also seen as important in identifying poor performance, although some human resource managers were of the view that this problem is not properly addressed in universities.

8.6.3 Training and development Human resource managers stated that heads of department do not always have

the skills needed to develop staff. It was apparent that many heads of department did not see staff development as one of the responsibilities of their post. Many expressed the view that training courses are available for those who are interested, but that career development is largely an academic’s own affair, and a result of their own efforts in research and publications. Lack of interest was acute in respect of research staff, who tended to be overlooked in all respects by heads of department. Universities tended to devolve budgets for training to departments and, in some cases, such budgets were very small. However, others were considered to cover needs.

Human resource managers and heads of department said their universities offered training in a range of areas, including teaching, research, admin. and community links. A number gave examples of short courses provided in diverse areas including staff recruitment, web-page design, distance learning and time management. A number of universities produced a training directory listing courses available. Some heads of department also referred to opportunities to attend conferences and events.

Although the link between appraisal and training was acknowledged to be weak, human resource managers and nearly all heads of department saw provision of training as good. However, it cannot be assumed that heads of department, who usually hold the purse strings, support training. Lack of departmental support potentially leads to reduced access to training in those departments. While many heads of department were in favour of improving training opportunities, a number commented that further training makes academic staff more attractive to other institutions, suggesting some ambivalence. Some also commented on the cost of training, and at their institution’s reluctance to send staff on external courses.

A number of universities were developing management training for heads of department to assist them in identifying staff development needs. The intention in some universities was to smooth the process of promotion and to reduce levels of turnover in the junior grades. Although this is clearly necessary, such developments were not yet in place or were in the very early stages. Therefore, despite having an active training department, one human resource manager considered that management development at the university was ‘appalling’ compared to the non-university sector.

Development can be affected through other means than formal training. In some departments, the head tried to ensure staff lectured in their specialist area. Attendance at conferences was recognised as important and the possibilities for this varied. For example, in one department in a new university, the budget was negligible

187

for this, other than for new lecturers who had an allocation of £500 for three years. In a department of an old university the budget could usually fund junior lecturers to go to conferences that were important for them. Although it rarely ran to funding others, the head of department considered that this was less important as established staff could usually build up their own funding. Library and laboratory resources are also important and severe lack of funding was mentioned as problematic in a new university. One university organised staff within departments into subject strands and various forms of funding were only available to subject strands. Reorganisation (due to financial constraints) had reduced the number of strands from three to two, cutting off some funding sources (and the possibility of submission for the RAE) to those continuing to specialise in the third strand.

In some of the high ranking old universities, interviewees commented on the need for academic staff to gain experience in more than one institution (including, in some cases, industry). This could lead to staff turnover, but one university mentioned seconding staff and also that staff returned having left. (In this case, this could be due to the high reputation of the department concerned.)

Some staff in vocational subjects (recruited from outside higher education) faced the challenge not only of learning to teach, but also to research. Some staff had little research training in their previous jobs. To address this shortcoming, one of the new universities had a scheme which enabled staff to be released to do a Masters or a PhD, with their post covered in their absence. The university also provided research development training and organised support groups for research. A further problem was identified in the time available for conducting research. This appeared to be a constraint in the case study new universities and the low research intensive old university. Similarly, some heads of department complained of the lack of time for research (and sometimes teaching as well). This was in all types of universities. For example the head of department in one of the research-intensive old universities said,

‘I’ve done no academic work for 18 months, nothing but personnel work, finance, involving mountains and mountains of bureaucracy. I’m only doing it because no other bugger would’.

The issue of heavy and competing demands within the academic job is discussed in more detail below.

8.7 Work demands and work allocation

For lecturing staff, the relative demands between teaching, research and administration is important, not only to take into account individual preferences, but also because research output tends to be the main route towards promotion. This has traditionally been the case in old universities but is now increasingly a criteria for promotion in new universities.

Academic staff time was formally allocated to teaching, research and administration, with the amount of administration in part dependent on taking on additional responsibilities. Systems for the allocation of time varied, with some case study universities having time allocation rules common across the university (for example providing maximums for each type of work). In others, time allocation was largely determined at departmental level. Other aspects which varied were whether

188

grade or additional responsibilities affected time allocated to specific tasks. For example, in some, all academic staff had similar teaching loads and other responsibilities were allocated on an ad hoc basis. In some, for example a new university, staff were classified research active or not: the former did little administration, but the latter, while doing no research, had greater teaching loads, as well as administration. This caused problems as it was recognised that research was valued and was also the route to career progression (despite principal lectureships largely being about administration, see above). The system did not seem to affect how well work allocation was judged to work by either managers or academic staff. Formal university-wide workload allocation systems tended to be complex. Not all staff understood the computation for allocation and, in most of these cases, they tended to see the system as unfair.

Whilst the problems for many staff in the less research-intensive ranking old university and the new universities may have resulted from higher demands overall, a number of specific issues stood out. First, the management of the system of work allocation was poor in some university departments. It appeared sometimes to be either ad hoc or allocation based on ‘the path of least resistance’, with administrative work allocated to the most willing or least resistant staff. A senior lecturer in engineering at a new university complained that, ‘It’s a bit ad hoc…people do get dumped on’. A senior lecturer in another new university said, ‘I just avoid them and people get tired of asking. I don’t see the purpose of doing it’.

Being a programme leader was seen as increasing administrative tasks at the expense of research activity (and hence career prospects), without any form of compensation. One senior academic in a new university explained,

‘There is a tendency for younger staff to have work dumped on them by the wily old lags. They are bright and enthusiastic and get encouraged to take on tasks. They can become programme leaders relatively quickly, but there is no extra pay for the work they take on.’

Finally, the treatment of new academic staff varied. For example, in a new university, new lecturers were given the full teaching load. In addition they had to take their lecturing qualification, placing substantial demands on new lecturers. Other universities gave lighter teaching or administrative loads to new lecturers, in recognition of the additional time teaching would take. This was generally seen as fair. However, as a recruitment incentive for the research active, some universities offered lighter teaching loads to recruits who were not new to lecturing. This caused discontent amongst existing staff. It could also cause discontent amongst the recruits if the lighter teaching loads did not materialise:

‘I was promised a low teaching load and then they load more and more modules on you’.

8.8 Equal opportunities This section concentrates on three aspects of equality arising from the case

studies. Firstly, the beliefs about equality of opportunity of human resource managers and heads of department. Secondly, the actions of the universities to secure equal

189

opportunities. Thirdly, the examples of lack of equality given by members of staff, together with other evidence identified in the interviews. It also provides evidence, from the survey of academic staff, of the extent to which academic staff believe that academia offers equality of opportunity.

Much of the evidence is based on perceptions. Within the study it was not possible to examine how well founded these perceptions were. However, other research has suggested discrimination in academic employment in respect of gender (Dearden et al., 2003) and ethnicity (Carter, Fenton and Modood, 1999). Even in respect of staff numbers, respondents would tend to give a perception, rather than actual headcount. The danger of this, even with very small numbers, was exemplified by a human resources manager’s description of the senior management team (comprising ten people) who initially said this was 50:50 male female, but on actual counting found it to 25:75 female to male.

A number of human resources managers emphasised the existence of a policy on equal opportunities. One manager in a new university described theirs as ‘rigorous’, adding that it produces regular updates on recruitment and retention in relation to diversity. A principal lecturer in another new university also believed that equal opportunities policy was strong and firmly embedded in practice. One principal lecturer in a new university believed that positive discrimination is often practised,

‘Well, realistically, there is a lot of positive discrimination. I think this is typical throughout academia in the UK in how this affects both recruitment and promotion and workload. It doesn’t/hasn’t affect[ed] me apart from going for grants and fellowships. They have special criteria. There’s lots of things you can’t apply for because you don’t come under the remit of their positive discrimination.’

Heads of department, most of whom were male, were asked about the relative opportunities for male and female academics. They frequently expressed the view that these are the same, either within their department or school, in the university or both. A number referred to the presence of women in senior positions, for example Dean or on the Senior Management Team in support of this, while acknowledging that this was a relatively recent development. Some heads of department felt that women faced disadvantages, including time out of an academic career for childcare, but few referred to institutional barriers, such as sexist attitudes and behaviour. One exception was the (female) head of English at an old university who argued,

‘The culture of the laboratory is laddish and a lot of senior male figures in the university are uncomfortable with assertive and successful women’.

However, previous research has shown a lack of faith in universities’ commitment to Equal Opportunities (with 32 per cent considering their institution not very or not at all committed to Equal Opportunities) (Carter, Fenton and Modood, 1999) and the staff survey showed that many staff believed that disability, age, religious affiliation, ethnicity and gender affected progression within academia128. 128 Note this does not mean that they believe there are problems at their own university. This was not asked.

190

Nearly two-thirds of staff thought that age affected progression (Table 8.1). This broke down into 25 per cent seeing age as detrimental to younger academics and 55 per cent believing it detrimental to older academics (with 19 per cent seeing it as detrimental to older and younger). Almost half of academics believed that gender affected progression in academia, almost all seeing it as detrimental to women. Disability was believed to be detrimental to progression by 38 per cent of academics and ethnicity believed by about one quarter to affect progression (with nearly all seeing it as detrimental to ethnic minorities). Only 10 per cent believed religious affiliation or belief affected progression.

Table 8.1 Equal opportunities, staff beliefs Whether believes….. affects people's chances of progression in academia Disability Age

Religious affiliation or belief

Ethnicity Gender

A great deal 4 11 1 3 6 Quite a lot or a great deal 15 35 2 10 20 A little, quite a lot or a great deal 38 64 10 26 44 Not at all 32 20 66 52 40 Don't Know/Refused 30 17 24 21 16 Total 100 100 100 100 100 n 2,785 2,785 2,785 2,785 2,785 • Source: NIESR/DfES Staff survey, 2004

Staff discussed some of the issues at their own institution in the qualitative research. A senior lecturer in an old university said:.

‘Universities, as employers, are not very good at promotion of women, ethnic minorities and people with disabilities. [My university] is like others. Senior positions are still male dominated and they don’t have large numbers of staff from ethnic minorities.’

The qualitative research found that some staff had not considered equality issues. If they considered that the academic workforce did not reflect the population, they saw this was due to extraneous factors – although some of these, such as childcare difficulties, might be seen by others as an equality responsibility of the university. In the qualitative research nearly all academic staff tended to concentrate on the issue of equal opportunities for women and few considered other potentially under-represented groups, including ethnic minority groups and disabled people. Moreover, whether they identified inequality or not, few seemed concerned about the issue, unless it affected their own position. There were a few exceptions to this, including some whose views were influenced by the disparity between the diversity of their students and the uniformity of the staff.

8.8.1 Gender Most heads of department in the qualitative research said they believed that

opportunities for women in academic life were improving. Although many referred to the poor representation of women in senior posts, the number of senior academic staff, and particularly those in management roles, was seen to have increased significantly in recent years. Some heads of department felt nonetheless that some

191

features of an academic career work to women’s disadvantage. Particular reference was made to the damage that career breaks and part-time working can have on an academic career. It was also thought that women with children participate less in international conferences than their male colleagues, and that this has a detrimental effect on their careers. A small number of managers referred to the culture of universities, particularly the older more prestigious institutions. This was seen as an enduring source of inequality. For example,

‘It is a bit male and clubby and it would be nice to be more diverse. …In British universities there are comparatively few women in senior positions [compared to US]’.

Similar sentiments were expressed by a female Head of department describing a ‘laddish’ culture (see Section 8.8).

Interestingly, amongst those who thought there was no gender discrimination, where departments had more women than men staff, this was often taken as proof of equality of opportunity, whereas if there were few female staff this was not seen to be a possible indicator of inequality. A number of academic staff commented on the under-representation of females amongst staff in comparison with their student body. One lecturer at an old university compared the equal gender split amongst those doing doctorates in the department to the very low number of female staff (which he thought typical of UK universities). He attributed this to women not returning after maternity and career breaks. At the same time, he thought that gender affected promotion prospects (in his university and in UK universities more generally) and that it was ‘much tougher for women’. A law professor, in an old university, commented on the low number of female professors in his department,

‘this ‘seems not uncommon’ in law schools generally. I’m not sure whether this is direct or indirect discrimination, or just ‘how it is’. It seems to be the case that women don’t do well in law schools in general in the UK.’

Some examples of gender discrimination were recounted: a senior lecturer at a new university said that he had been appointed at a higher level than his experience should have allowed so that he could earn more than his wife. A female colleague in the same department believed there was substantial gender discrimination at the university. She was frustrated by colleagues’ ‘old fashioned and male attitudes.’ She elaborated on this view as follows:

‘There’s general intolerance and discrimination. They don’t treat female colleagues as evenly and as well as they should…. The culture here is very male dominated – typical of this region and many staff are local’.

In the view of this senior lecturer gender discrimination was evident in a number of areas, including selection of staff for submission for the RAE and workload, with women having heavier teaching loads and extra work associated with teaching and course administration. These practices were seen to lead to higher turnover among female staff. A confidential HEFCE report into submission rates to

192

the RAE found that women in the 31 to 49 age range were less likely than men of the same age to be submitted to the RAE.

The issue of workloads was taken up by a female senior lecturer in a new university, who said,

‘Women seem to get given more work because they can multi-task better. Also, women have a harder job saying no and want to please people’

Another, in an education department, commented on the fact that gender affected expectations about the nature of work and sexist peer pressure,

‘There are cultural issues. The department has a primary school ethos and some female colleagues are expected to be doing work like tidying up files, putting notices on walls and such-like. Women who didn’t do those kinds of things leading up to our Ofsted inspection had some negative comments. This is typical of a primary school, but not the university itself.’

When asked about the effect of having and raising children on academic careers, staff views were highly varied and did not appear to be related to whether the respondent was a parent themselves. In the main, compared with most other jobs, academic jobs were seen as easier for coping with bringing up children, largely due to the high degree of control academics have over their hours of work. Childcare provision varied, with some universities providing childcare, although this could be oversubscribed. Views on whether caring for children affected promotion prospects varied. For example, the following statements were from staff in the same department,

‘Yes, it does affect promotion. You have less time and less flexibility’

‘Yes, it can affect people’s promotion prospects – obviously.’

‘As far as I’m aware it doesn’t have an impact on promotion prospects.’

Where staff felt it did affect promotion prospects, the type of reason given was time demands and maternity/career breaks. These breaks were seen to have a particularly serious effect on research activity, a key criterion in promotion. It was also reported that inadequate arrangements were made for maternity cover, leading to the stressful situation of staff having to arrange it themselves. The issue of part-time working by mothers was also raised. There was evidence that some universities would not allow principal lectureships to be conducted part-time, whilst one university was investigating whether the promotion criteria were fair for part-time staff. Of particular concern was the publications requirement which part-time staff may have more difficulty meeting. Some staff believed this is particularly likely to affect mothers’ career progression. Apart from the issue of part-time working, the difficulties of balancing an academic job and busy home life were recognised sometimes to take their toll on female academics. As a male lecturer commented,

193

‘If you’re tired it’s difficult to keep your thinking at the cutting edge.’

However, this was not always seen as unfair by academics, for example one female senior lecturer, in a new university, thought that women were at least partly to blame for their poorer success rate because ‘…they are not prepared to commit the energy and hours to get promotion.’ She explained she had experienced few problems herself because her partner had always taken a share of the childcare and household workload. The importance of household division of labour was also brought out by one male senior lecturer who said, ‘Childcare isn’t an issue. My wife doesn’t work.’

Twenty-eight per cent of academic staff had children (31 per cent of men and 25 per cent of women). The survey of academic staff found that amongst staff with children, 72 per cent were not dissatisfied with the extent that the job allowed them to meet their childcare responsibilities (Table 8.2). The percentage not dissatisfied was greater amongst women than men.

Table 8.2 Childcare provision, staff with children

How satisfied with extent job allows to meet childcare responsibilities Male Female Total

Somewhat, mostly or completely satisfied 56 74 63 Completely satisfied 7 10 8 Mostly or completely satisfied 41 50 44 Neither satisfied nor dissatisfied 12 5 9 Somewhat, mostly or completely dissatisfied 31 21 28 Mostly or completely dissatisfied 11 7 10 Completely dissatisfied 5 2 4 Refused 1 1 1 Total 100 100 100 n 488 294 788


In respect of caring for dependant adults, a higher percentage of those with adult care responsibilities were satisfied with the extent that their job allowed them to meet those responsibilities.

194

Table 8.3 Adult care provision

Whether satisfied with extent job allows to meet adult dependant care responsibilities Percent of

staff

Percent of staff with eldercare responsibilities

Somewhat, mostly or completely satisfied 7 57 Completely satisfied 2 15 Mostly or completely satisfied 5 44 Neither satisfied nor dissatisfied 3 23 Somewhat, mostly or completely dissatisfied 2 19 Mostly or completely dissatisfied 0 5 Completely dissatisfied 0 2 no dependent adults 88 Total 100 100 n 2,785 328 • Source: NIESR/DfES Staff survey, 2004

8.8.2 Ethnicity The picture with regard to ethnic minority representation is a more complex

one. As discussed earlier, many of the departments in the qualitative research recruit extensively from overseas and this practice is increasing. With the exception of the Far East, the main recruiting ground is Europe. Therefore, overseas recruitment has not led to greater ethnic diversity in these departments, although it has led to greater diversity of nationality. Most of the case study universities’ human resource managers and heads of department acknowledged that their academic workforce is largely white. Moreover, even where ethnic minorities are represented in academic grades, numbers in senior positions (and especially in management roles) are small. It would appear that British born ethnic minorities are particularly poorly represented. The under-representation of ethnic minorities was recognised by some of the academic staff interviewed. As a senior lecturer stated,

‘In terms of diversity, there’s not so many women as you’d expect, not many Asians [either men or women]. I’m not sure why this is but University-wise, the minorities are not over-represented by any means.’ (White, male, new university)

A confidential HEFCE report into submission rates to the RAE, found no significant difference by ethnicity. However, there were differences approaching significance and it cautioned that the situation should be monitored.

In one university, a human resources manager reported that the university was more active on issues of race than gender, but this seemed to involve basic measures: establishing a policy, setting targets, producing a report and having a statement in job advertisements. Certainly one of the departmental heads thought that the main action was with respect to students and that ethnic minorities were not well represented amongst staff.

One respondent in the qualitative research reported racial and religious discrimination against themselves. The person, from an ethnic minority, described a range of incidents, relating to himself, which suggested differential treatment and

195

harassment by senior staff. However, he had too little confidence in the university to report the incidence, as he felt this would affect his career. His main aim was to move universities.

8.8.3 Age Mixed messages were given about ageism in the case study universities.

Amongst respondents, age was often seen as synonymous with grade and status. Therefore, ‘new lecturers’ and ‘young lecturers’ were often interchangeable terms. One respondent considered that both younger and older were at a disadvantage in recruitment and retention and a reader gave an example of a colleague who failed to gain promotion, ostensibly, because ‘he was too young’.

In several of the universities, early retirement was still encouraged and in the survey six per cent of staff aged 40 and over reported having been offered early retirement in the previous five years (Table 8.4).

Table 8.4 Early retirement

Per cent of

all staff Per cent of those aged 40 and over

Offered early retirement: in the last year 2 3 in the last two to five years 3 5 in the last five years 4 6 Total aged 40 and over 60 100 aged under 40 40

Total 100 n 2,785 1,681


One head of department in a new university said early retirement was practised in a ‘fairly nasty’ way and a professor in another new university described it as ‘forced’ for all senior academics over the age of 50, as part of a cost cutting exercise. A reader, in an old university, said that

‘I expect to retire ‘around 60’. At this point, I’ll ‘become too expensive’, and it’s likely that the university will consider replacing me with two young cheaper staff. There may be pressures from the university to take early retirement.’

The pressure for early retirement was believed to have led one university to target young people as recruits and in another university an example was given where older applicants, who had been in post-doctoral posts for some time, were looked upon less favourably. Another respondent at an old university said that ‘Age is always considered but it’s not a major issue.’

At the same time, some departments tended to recruit a combination of young recruits and people in their 50s moving from industry or target the latter group. However, this was because of long standing recruitment problems in the subject areas

196

of computing and engineering. Therefore, where necessary, considerations of age were put to one side. As a senior lecturer, said ‘if you’re good enough, you’re young enough’

8.8.4 Other Very few respondents in the qualitative research commented on equal

opportunities in relation to disability. The only two who did pointed out how few, if any, disabled staff there were.

Respondents were not asked about class discrimination specifically, in either the qualitative or quantitative research, to the surprise of one respondent. This respondent, a professor at an old university, said that he felt that working class applicants and academics were substantially disadvantaged. This was due to a snobbishness, certainly in many of the old universities in which he had worked. He referred to his experience of recruitment exercises where he was convinced that selectors dismissed, somewhat sneeringly, working class applicants, merely due to their background. A respondent at a new university, from a wealthy background, felt that he experienced the obverse of this discrimination, that he was harassed because of his wealthy background. It is possible that both behaviours exist in UK universities.

The Head of a Psychology department reported a problem in recruiting non-British staff. It is important for psychology departments to have British Psychological Society recognition (so that degree students gain exemptions for BPS qualifications). This requires 75% of staff to have BPS recognised qualifications, which can be a problem for those qualified abroad, and results in foreigners having less chance of recruitment. A similar problem occurred in a Health department, where knowledge of the Health system in the UK was important for doing the job, disadvantaging applicants from abroad.

8.8.5 Equality initiatives Human resource departments were engaged in a number of initiatives aimed at

increasing the representation of women, and to a lesser extent ethnic minorities, in the academic workforce. These were focused on the representation of women in management positions, including senior management teams. One old university had organised a series of sessions for junior and middle grade academics, with an external facilitator, to discuss staff development issues and to improve skills such as CV writing and IT. This university was looking into aspects of the academic career which might disadvantage women, for example promotion practices in relation to part-time working (see above). Among other initiatives, another old university was exploring ways in which the ‘masculine’ culture of their institution, which included college dinners, could be changed to be more welcoming to women. However, not all human resource managers gave such consideration to issues of gender equality. One director of human resources expressed his irritation at the HEFCE requirement for targets in relation to equal opportunities:

‘I just said OK, we’ll have a target of 20% female managers in the academic area by 2005, but this will be against a backdrop of reducing people in these posts. I just said 20% to keep them quiet. They’re quite happy with it, don’t know how we’re going to do it though (laughs).’ (new university)

197

However, the importance of the HEFCE requirement was illustrated by a new university, which expected the targets to encourage internal applicants for senior posts.

Work on race equality was largely at the level of policy formation, for example the development of a diversity strategy or policy statement. Some universities were holding diversity awareness training courses for staff. One was including race equality issues in all training courses, where relevant. A number were wording advertisements in order to encourage more applications from ethnic minorities and were planning to advertise posts in ethnic minority newspapers. In discussions about race equality, human resource managers did not distinguish between academic and non-academic posts, British born and overseas employees. Some gave the impression that the issue is about numbers or percentages employed, and made comparisons with the local population, for example ‘We are close to the [county] average’ rather than representation in senior posts. There was also a tendency among heads of department to point to the activities of their human resource departments as evidence that the university was, in practice, an equal opportunities employer and that promotion prospects were therefore the same for ethnic minorities as for their white colleagues.

8.9 Summary and conclusions Given the variation in practice in the case study universities and improvements

in practice some reported, it is clear that universities can influence recruitment and retention. Most of the case study universities had mixed practices (good and bad) and seemed to be trying to address some of the problem areas. However, their opportunity for manoeuvre is obviously constrained by finance, affecting, for example, pay levels, promotion opportunities, facilities and training and work demands. Moreover, the national pay scale sets the main parameters for pay (below professorial level). Below, we summarise the main ways in which practices were likely to affect recruitment and retention.

8.9.1 Ethos and human resource structure The ethos and human resource structure (particularly in the old universities)

was very much of individualism. This does not sit entirely well with an increasing use of current managerial techniques (monitoring, appraisal, emphasis on development etc.). The problem was not just of staff resistance to these, but also that it was not clear that the structure enables such systems to be delivered adequately. In particular, the system dominant in old universities of rotating heads of department needs to be considered. On average, this system will reduce management expertise of the heads and so both the system, role and training for heads should be looked at. One of the case study universities provides an example of trying to tackle the lack of management expertise. This university was developing management training for Heads of department, in order to improve competence and commitment to appraisals.

Tension between the departments and both the university, in general, and the Human Resources department, in particular, did not seem uncommon. The general impression was of academic staff strongly focused on the research, teaching and external activities of their jobs with wider university matters seen as a distraction. This extended to aspects of human resource practices (e.g. appraisal). In the case

198

studies, not only were administrative demands seen as burdensome, but also often as unnecessary and even counterproductive. Some of these demands stemmed from organisational change. If such changes and other administrative demands could be reduced, that should be helpful. Where they cannot, perhaps greater communication between what appeared as two sides would lessen the resistance to administrative work.

8.9.2 Employment strategy Universities are in a difficult position. They have financial pressures which

lead to careful examination of the need to recruit and, at times, difficulties recruiting appropriate staff. These pressures lead to recruitment practices which have implications both for successful recruitment and for retention of academic staff in the university sector (as well as within the university).

Financial pressures lead to careful consideration of the need to fill vacancies and to create posts. These can result in lengthy delays in recruitment, resulting in higher workloads for existing staff. Failure to recruit can have the same effect. Whilst temporary measures may be taken to minimise workload, in the case study universities, these covered replacement teaching (or the post being left vacant). As temporary staff tend not to conduct administration, existing staff are liable to have seen an increased workload, particularly in administration. Where temporary cover is not provided, staff would also see additional teaching. We have seen that both higher numbers of hours (and to a degree, higher numbers of hours on administration) and perceptions that workloads are too high reduces retention in the sector. Alternatively, if academics maintain their workload, this is liable to be through reducing research time. Lower research time is also associated with increased likelihood of leaving the sector as well as resulting in lower research ratings and reduced income for the university and department.

The case study universities described proactively searching for recruits. This should be helpful in reducing recruitment difficulties. However, where this is directed at overseas recruitment, depending on the nationality of recruits, this could reduce retention in the sector, as EU, US, Australian and New Zealand academics are more likely to leave. The approach has also to be carefully scrutinised to avoid infringing equality of opportunities, since it may favour academics who are known to senior staff and disadvantage those without such connections to UK universities. The practice may therefore perpetuate current patterns of gender and ethnicity in UK universities.

The targeting of staff outside the sector may also help fill vacancies. However, without support and training (not only in teaching but also in research), there may be problems in enabling these recruits to progress. This may lead to turnover. One of the case study new universities described a Catch-22 situation, where they provided such support (with reduced teaching time whilst doing a Masters degree or PhD), only to see, on completion, staff moving to more research intensive universities. Whilst such turnover would continue to be problematic, its effects might be somewhat mitigated by such training being centrally funded, rather than covered by individual universities. It might also be enforced by a ‘bond’ agreement requiring staff to stay for a fixed period. However, this would not meet universities’ needs for a stable pool of academics to develop to senior positions. Another problem with recruiting from outside the sector is the lesser knowledge of the job. It was mentioned that there can be a lack of awareness of all the roles and pressures of an academic job. This is likely

199

to decrease retention in the sector. Obviously, in part, the solution lies in providing more information prior to recruitment.

The level of pay was correctly identified as a problem for recruitment and a number of approaches taken to enable either recruits or selected groups to be paid more (market supplements, Golden Hellos and appointing at a higher point on the scale). Whilst these approaches may result in easier recruitment, they are a major cause of dissatisfaction to many existing staff and are seen as unfair. Perceptions of unfairness stimulate leaving the sector. Moreover, some of these practices run the risk of infringing equality of opportunity (notably, where existing pay is taken into account or where mobility enhances pay). Appointing at a higher point also encourages movement within the sector in order to raise individual pay.

Finally, the approach of offering recruits greater time for research may reduce recruitment difficulties. However, this is liable to be perceived as unfair by existing staff and, as already said, perceptions of unfairness increase turnover from the sector. (Note that reduced teaching contact time for a period, whilst a new lecturer develops their courses is unlikely to be seen in this light.) Moreover, where the additional research time does not materialise or is quickly eroded, the recruits themselves are more liable to leave.

8.9.3 Promotion practice The majority of academic staff considered promotion decisions to be unfair

(see Chapter 6 Table 6.10) and perceived unfairness reduces retention in the sector (see Chapter 7). A number of promotion practices found in the case study universities would increase perceptions of unfairness and so may be detrimental to the sector. These were:

• offering promotion to staff to prevent them leaving (this practice excited very strong feelings amongst respondents in the qualitative research);

• lack of promotion for those already performing at the level of a higher grade; this is difficult to tackle where finance and hence promotions are restricted, but was being addressed by one case study university;

• opacity of and staff ignorance of procedures and criteria (some case study universities were seeking to clarify these);

• differential application of criteria across subjects at institutional level, in response to different external labour markets, for example in professional areas such as law and accountancy;

• the relative weight given to the various job roles; in particular, lack of recognition for administrative work (except for Principal lectureships) and to teaching and the increasing stress (including at new universities) on research output; there was some shift in this at some of the case study universities, although in at least one case this appeared to be paying lip service to recognising non-research roles, an approach which is liable to increase feelings of unfairness.

The practice of offering promotion to staff to prevent them leaving may sometimes be effective in retaining selected staff, but at the expense of morale and retention of other staff. However, many of the respondents could only point to cases

200

where this approach had failed (i.e. the person offered promotion left). Indeed, this approach may be counterproductive in retaining valued staff, as it encourages them to look for other jobs.

There are a number of other problems with the system which may increase turnover. Progression can be slow and is inhibited by requirements to reach the top of a scale prior to applying for promotion. This will encourage staff to seek promotion in other universities or to leave the sector. The number of senior positions in new universities is particularly low, slowing progression further. Given the pay levels for lower level positions, this may encourage staff to leave the sector. The problem for researchers is even greater. The combination of pay levels and slow promotion may also reduce recruitment into the sector.

Annual promotion rounds are inflexible. The process of applying for promotion can be laborious and rejection may be discouraging. Again, this is particularly a problem where the number of promotions are capped, irrespective of the quality (and current tasks) of applicants.

The drivers of many of these practices which reduce retention in the sector tend to be financial (curtailing promotion opportunities) and shortages (leading to competition within the sector). However, some are under universities’ own control, as could be seen by changes made in some of the case study universities.

8.9.4 Pay system The salary level, being comparatively poor within the UK, is liable to increase

difficulties attracting entrants to the sector, and to reduce retention.

The ways in which universities have been seeking to retain valued staff will have the same implications for sectoral retention as flexibility over pay at recruitment and offering promotion to people because they may leave. The lack of additional payments for additional tasks is similar to the lack of promotion despite performing at a higher grade level and is likely to lead to feelings of unfairness.

The introduction of performance pay is interesting. If conducted fairly and transparently, it might be expected to increase retention. However, the case study universities said that it was already being subverted. On top of this, a well-known problem of performance pay systems is that the majority of people believe their performance is above average and so they should receive above average pay rises. As this is impossible, morale is affected and systems may be seen as unfair.

Appraisal, training and development practices have a number of implications for retention and it is interesting that the case study universities appeared to be at an early stage of implementing many practices which are common in other large organisations.

Induction training, where well-conducted, should ease recruitment for new entrants especially, as should mentoring. Introducing good appraisal schemes to identify development needs and, as in one case study university, linking them to promotion, should encourage staff not only to improve performance, but, where linked to promotion and well-implemented, to improve the process of promotion and its perceived fairness. However, it is essential that implementation is good and that development needs are not only identified, but also addressed. Otherwise, appraisal raises false expectations. This appeared to be a problem in some of the universities, either due to financial constraints restricting training opportunities or due to lack of

201

commitment by Heads of department. The introduction of management training in one case study university for Heads of department to assist them in identifying development needs, with the aim of reducing turnover of junior staff and to ease promotion was very interesting.

Other actions which should improve satisfaction at work included ensuring staff lectured in their own specialism, adequate provision of library and laboratory facilities and ensuring adequate time for research.

Finally, in at least some, if not all, case study universities and departments, research staff appeared to be excluded from these practices and to be overlooked by Heads of department. Greater emphasis on development and training might, first, serve to increase the retention of such staff and also increase their suitability for permanent lectureships, thus both increasing retention in the sector and reducing recruitment problems.

8.9.5 Work demands and work allocation Our research has shown that the higher hours of work increase turnover from

the sector, as does higher administrative time, and that more hours spent on research reduce turnover. This makes allocation of tasks within the universities important. Task allocation was conducted at the departmental level in the case study universities and the departments varied in how this was done. Thus some aimed to equalise work demands and to spread administrative tasks evenly; others did not appear to do this and it should increase retention in the sector if allocation was made more equal.

The classification of staff as research active or not (for the RAE) may cause problems, unless the aim is to encourage the non-research active to leave the sector.

Finally, the practice, reported in one case study university, of promising low teaching loads (to enable concentration on research), but not delivering this, is liable to ease recruitment, but increases turnover.

8.9.6 Equal opportunities There was widespread belief amongst academic staff that progression in

academia was affected by a number of personal characteristics, particularly age (especially detrimental to older academics) and gender. The ways in which certain groups might be disadvantaged varied from the subtle (laddish culture) to the formal and direct (pressure for early retirement).

The extent and quality of human resource practices designed to ensure equality of opportunity varied greatly between the case study universities, which exhibited a range of good and not so good practices.

Given the widespread belief amongst academic staff that there were various forms of inequality and the relationship between perceived fairness and leaving the sector, the lack of high quality action of equal opportunities in some universities is likely to increase turnover from the sector. However, this might not occur if academia is seen to be no different from other sectors of the economy. Nevertheless, the use of early retirement obviously increases the likelihood of academics leaving the sector.

202

203

9 Conclusions

Universities have undergone major changes in the past 20 years. Growth in student numbers, the conferment of university status on polytechnics, changes in the funding regime and greater emphasis on accountability have led to substantial change in the nature of academics’ jobs. These have been combined with greater funding pressures and changing funding regimes. There have been major changes in the inspection, monitoring and quality review approaches (including the introduction of new teaching quality assessment and audit approaches and the RAE). Academics are under greater pressure to produce research outputs. These changes have altered not only the amount of work demanded of academic staff, but also its content. In any sector, change alone tends to increase turnover. In this case, a large majority of staff consider that academic jobs have worsened since they entered the sector. Wages in academia are both perceived to be and are relatively low (compared with people with higher degrees in the rest of the economy), which should increase recruitment difficulties. Job satisfaction is lower than amongst employees in general. Thus, it would not be surprising if shortages had been appearing.

However, for individual universities, turnover is, at most, only slightly higher than that for highly qualified staff in the whole economy. Some recruitment and retention problems have been identified in previous studies, but the case study universities did not seem to consider that they had real problems, although problems in some subjects and for some posts at certain times were recognised. However, there was evidence from the qualitative research that vacancies were not always filled and that standards of recruits were not always as high as desired. Thus, it appears that there are shortages, reducing the quality of staffing in the sector and causing problems of delivery or overwork.

This concluding chapter draws out policy implications of the findings and identifies gaps in our knowledge. Our findings suggest there are a number of ways in which individual universities and the Government could raise recruitment and retention of academics in the sector. Some university–level actions may merely improve the position of one university over another (i.e. a zero-sum game). What is needed is approaches to expand recruitment and retention for the sector. Many, but not all of these would have financial implications.

Since the start of the study, there have been a number of changes in higher education which affect academics’ employment; others are imminent (see Appendix H). Other major changes include the framework agreement for pay (see www.UCEA.ac.uk) and changes to the Research Assessment Exercise for 2008 (see www.rae.ac.uk). It is too early to say whether these changes will address some of the issues raised in this chapter.

204

9.1 The labour pool Academic employees come from two main sources: students (both UK-

domiciled and foreign) (34 per cent) and employees from other sectors in the UK (42 per cent). Both are major sources. Most of the remainder are from employment abroad (21 per cent). A high percentage of entrants are foreign (40 per cent), comprising 10 per cent foreign students studying in the UK, five per cent of foreign students studying abroad and 24 per cent foreign employees (16 per cent employed abroad and eight per cent employed in the UK).

Around one third of current academics made a career change to enter academia, generally from higher level occupations (managerial, professional, associate professional and technical), rather than from lower level jobs (via being a mature student). Career changers are more common in new universities and in Business and administrative studies, Computing, Subjects Allied to Medicine, Education, Social studies and Engineering. Other important, but much smaller, sources of entrants are employees from foreign higher education institutions and foreign research institutes.

9.2 Pay University academics’ pay is relatively low compared with similarly qualified

employees in the UK. Starting pay is low and remains comparatively low for most academics for most of their career. It is only when academics reach their mid-50s that their pay slightly overtakes that of similarly qualified employees. Although high pay was not an important consideration in career choice for the majority of research students, it was for a substantial minority and was a disincentive to entering academia. This was exacerbated by misconceptions about academic pay, as research students tended to underestimate it. Many were also unaware of the additional financial benefits through pension provisions. We were unable to examine the role of pay on career decisions of non-entrants from other sources, but it seems unlikely that relatively low pay would not also reduce this supply. In respect of retention in the sector, the level of pay did not appear to have any effect.

At the same time, UK academic pay was higher than academic pay in a number of European countries and in Australia and New Zealand. This should ease recruitment from these countries and may have contributed to the expansion of recruitment from abroad. Forty per cent of recruits are foreign and it seems likely that this percentage will increase (as universities target foreign recruits, as advertising methods are increasingly international and with the fall in PhDs awarded to UK-domiciled students). Amongst research students, there was a greater propensity for non-British EU nationals to want to go into UK academia. However, pay in academia was higher in the US (for comparable academic staff), thus enhancing the attraction of US academic employment over UK.

Possible policy responses to this situation include raising pay, in order to attract more British (as well as foreign) nationals and increasing recruitment effort (probably most effectively with targeted foreign nationals). However, the increasing reliance on foreign nationals is itself potentially problematic, as academic staff from certain countries (specifically, those from the USA, Australia, New Zealand and the European Union) are more likely to leave the UK academic sector. Many research students were unaware of the pensions offered in academia. These add substantially to

205

the overall pay and benefits package. Raising awareness amongst research students of the actual levels of pay (throughout an academic career) and of the pension benefits ought to improve the supply to the sector129.

Various targeted recruitment and retention approaches have been used (e.g. Golden Hellos, market supplements, performance points, retention points) as, seemingly, cost-effective ways to use pay to deal with specific recruitment and retention problems. However, perceived unfairness in these approaches (as well as in promotion and recruitment) was found to reduce retention in the sector. Certainly, enhancing pay of recruits (above comparable levels for existing staff) and of valued staff expected to leave are regarded as ‘unfair’ and excite strong feelings. Market supplements and Golden Hellos are also seen as unfair. Performance pay must be implemented well to avoid perceptions of unfairness. However, even so, the problem of the average employee tending to believe their performance is above average and so deserving a higher than average pay rise, means that more employees will be dissatisfied by performance pay than satisfied. Thus, at worst, such approaches may be counter-productive (or cost-ineffective). Universities would be able to improve retention in the sector through ensuring that the implementation of discretionary pay is conducted fairly, justifiably and transparently.

9.3 Job content Job content is a very important influence on recruitment and retention of

academics. Throughout the study a repeated message was that academics were driven by the desire to do research and neither enjoyed administrative tasks nor always could see the benefits of these. Research students were most attracted to academia by the opportunity to do their own research. For art and design lecturers an important factor was to fund artistic practice. Retention in academia appeared to be affected by the number of hours spent doing research and by the number of hours spent in administration (with higher research hours associated with lower expectations of leaving academia and the converse for administrative hours). Therefore, the balance of the three main aspects of the job (research, teaching and administration) and the time to conduct each of these properly is of major importance for recruitment and retention in the sector.

Research is a key attraction of academic jobs: having the time (and funding) to conduct research to one’s own agenda. Research is also one of the major influences on satisfaction. Those who spent more time on research were less likely to expect to leave the sector. This did not vary between old and new universities. Teaching is a much lesser draw (than research) to an academic career for research students (although, possibly, of greater attraction for career changers), although it is an important enhancer of satisfaction to staff once in the job. However, increases in the amount of teaching and changes in the nature of teaching (e.g. increasingly demanding students, a ‘production-line approach’) reduce satisfaction. Just as academics were overwhelmingly keen on research, they were overwhelmingly critical of the amount and nature of administration they had to do. Those who spent more time on administration were more likely to expect to leave the sector. Criticism of 129 The Teachers Pension Scheme (the scheme for academics predominant in new universities) is currently under review, including the pension age (raising the pension age to 65) (see http://www.teachernet.gov.uk/educationoverview/briefing/currentstrategy/pensions/).

206

administration seemed to be related to three types of administration: quality control procedures (notably the TQA and the RAE); organisational changes; and changes in teaching demands (e.g. new courses). The impression was of staff being stretched and stressed to meet administrative demands, many of which they did not see as useful for improving research and teaching and some of which were seen as stemming from administrators making unnecessary changes.

For recruitment and retention, there are a number of implications. Increases in the research component of the job is likely to reduce recruitment and retention difficulties (and, conversely, decreases in the research component are likely to increase them). Thus one way to tackle recruitment and retention difficulties is to ensure greater research time. Conversely, the introduction of ‘teaching only’ and ‘research only’ universities (or lecturing jobs) would be liable to exacerbate recruitment and retention problems.

As well as increasing research time, reducing administrative time would improve retention. This might be tackled in a number of ways: expansion of staff with a reduction in the staff/student ratio so that the administrative load is spread more thinly; providing greater administrative support; and reducing administrative demands. To reduce administrative demands, perhaps greater attention should be paid not only to the gains which might be made (from, for example, change, new reporting systems and from assessment) but also to their negative impacts (i.e. balancing gains and losses of any change or demand). This would be important at both Governmental and university levels e.g. funding changes have led to universities responding to changes in student numbers through departmental closures, amalgamations, changes in courses, all of which affect staff administration. At the same time, if staff could be convinced more often of the usefulness of the administrative demands (and the changes requiring them), administration might not cause so much dissatisfaction and so its effect on retention might be reduced.

Satisfaction with the non-pecuniary aspects of the academic job tends to decrease over the early part of careers and then increases towards the end of the career. Professors, and to a lesser extent senior lecturers, are on the whole happier in their jobs than those at lower grades. This suggests that it may be particularly important to focus on improving the job content of those in mid-career and of those failing to progress up the hierarchy.

9.4 Workload Connected with job content is workload, as part of the squeeze on research is

due to higher teaching and administrative demands. Staff who work longer hours are more likely to expect to leave the sector with the effect primarily emanating from higher teaching and administration hours. Those who feel their workload is too high are also more likely to expect to leave.

Thus the evidence argues for reducing workload in order to improve retention, with the reduction mainly falling on teaching and administration (and taking into account the findings on job content, primarily the latter).

207

9.5 Other aspects of the job Academics have long been seen as benefiting from autonomy, flexibility, long

holidays and so forth in their jobs and these were found to be important in affecting recruitment and retention.

As well as the ability to do research, other job aspects important in attracting research students include a good working environment, variety, freedom to use initiative and seeing tangible outcomes from their job. Close behind these factors are autonomy in the job, control of their research, career prospects, collaboration and flexibility of working hours. Also important are good physical work conditions, helping people and job security. For career changers other motivations include a better lifestyle. Examples of drivers of this include stress in previous job, shift work, routine working and wanting variety at work, academia being ‘more relaxed and informal’, career progression requiring movement into management (and wanting to stay with one’s profession). Few were driven through lack of further progression in their previous careers outside Higher Education, through being made redundant or through retirement.

Academic staff also value more subtle elements of their job, such as the support of their peers and the ability to participate in the wider academic community. Dissatisfaction with non-pecuniary elements (including relations with managers, being able to use one’s own initiative, relations with colleagues and physical work conditions) increased the likelihood of leaving the sector.

Increasingly managerialist approaches to higher education (e.g. quality control measures, including the RAE and teaching quality assessment) tend to reduce autonomy, whilst financial constraints affect physical work conditions and the ability to participate in the wider academic community. These types of changes will exacerbate recruitment and retention problems.

9.6 Job security The majority of entrants to academia are recruited to fixed-term research

contracts. It appeared that the tendency to recruit to research jobs and to fixed-term contracts had grown over time.

The use of fixed-term contracts is a disincentive to potential entrants and a major cause of academics leaving the sector. Academic jobs tend not to meet research students’ (and probably potential career changers’) aspirations for job security: 81 per cent of research students saw it as at least very important to have a permanent job within two years of gaining their degree and were aware of the lack of security in academic jobs. Staff on fixed-term contracts were much more likely to leave the sector and could be seen as being driven out of the sector due to insecurity: they tended to leave due to their contract ending, the desire for a permanent job or pessimism about job opportunities in UK academia. On leaving the sector, many plan to enter research jobs elsewhere and academic jobs abroad, suggesting that many remain interested in research.

Whilst the employment of young staff on fixed-term contracts may seem a rational strategy from the point of view of the university or department in the light of the fact that it is difficult to determine their research and teaching abilities, this may be rather short-sighted. Not only do new and potential staff desire a permanent post,

208

but the uncertainty created by such posts increases the likelihood of their leaving not only the university but UK higher education as a whole.

The necessity for the use of fixed-term contracts is unclear. The two explanations for the use of fixed-term contracts are a) that it is necessary as a sorting system for selection into lectureships: prior selection is not possible, so a larger pool than required as lecturers are trained and then compete for jobs; and b) that research jobs require highly specialised skills, research demands change (for example, as research grants end) and so researchers with the wrong skills leave (i.e. transferability of researchers between projects is not possible). This latter also assumes that research is not largely determined on the basis of researchers’ skills but that of others (i.e. lecturers). Obviously, the system of funding is an important factor in this. Moreover, the finding from the qualitative research that Heads of department were not always aware of the number, let alone skills, of the research staff in their department, suggests that, in some departments at least, the issue of transferability could be given more consideration.

The Fixed Term Regulations, introduced in October 2002, give employees comparable employment protection and rights as permanent employees and seek to address the problem of repeated fixed-term contracts. However, it is not clear the extent to which the regulations will enhance job security.

To increase the potential supply of academics and to reduce loss from the sector, the insecurity facing researchers needs to be reduced. This may need imaginative approaches as to how job security (with career development) might be provided. These might include: combining research and teaching; systems to facilitate transfers between research posts; systems to facilitate combining research posts; the development of a formal career route from research to lecturing. We would suggest that decisions about how this could be done should consider the structure of the academic sector, including the appropriateness of the lecturing/research division and the funding system.

9.7 Career prospects

Academic jobs tend not to meet research students’ (and probably potential career changers’) aspirations for career progression and this was one of the main disincentives to an academic career. Amongst academic staff, dissatisfaction with career prospects was high, although this did not appear to affect retention in the sector.

Academic careers offer a clearly defined career hierarchy, progressing from temporary research posts through Lecturer, Senior Lecturer to Professor. However, for those who wish to remain in research-only posts (i.e. and not enter the lecturing grades), the career structure is not good, with few senior research-only posts. The potential for career progression (in all academic jobs) is relatively low in new universities. It is likely that dissatisfaction with career progression is, in part, due to relatively low pay levels, as promotion is essential to achieving more adequate levels of pay.

There was substantial concern amongst academics about the fairness of decisions on recruitment to senior posts and promotion at their university. Perceptions of unfairness did lead to staff leaving the sector. Universities tend to have detailed

209

(and sometimes very onerous) procedures for promotion, which are designed to ensure fairness. However, it was clear that, whilst the process might be followed, the substance might not be. Moreover, the difficulties of balancing need for certain types of staff (highly marketable groups) against equal (‘fair’) criteria across all staff was problematic.

Perceptions of the fairness of promotion decisions are likely to be influenced by the extent of opportunities for promotion: the more opportunity, the less important fairness becomes (and the more likely that each individual will have experienced promotion). Thus, expanding the opportunities for promotion may increase retention. This is likely to require additional funding. Again, it is also possible that the emphasis on promotion is, in part, a result of relatively low pay levels. If so, increasing the pay of academics might also increase retention via reducing the importance of perceived unfairness of promotion decisions.

The criteria for promotion, except to Principal Lecturer (new universities) whilst varying amongst universities, tended to stress research achievement and to minimise the importance of administrative achievement. This seemed problematic for universities, in that it reduced the willingness of staff to take on essential administrative duties. It also tended to penalise ‘helpful’ staff. It seems likely that this also had a detrimental effect on the promotion of women. It would be useful to both universities and to staff if all aspects of the job were rewarded through promotion. Although universities were moving towards this, it appeared that the changes for some, despite changes in promotion criteria, may be more formal than real.

A further problem in promotion was identified in respect of Principal Lectureships. These did tend to take into account administrative achievements. However, further promotion, to Professor, tended to be based on research achievement. This meant that promotion to Principal Lecturership reduced the likelihood of further career progression. Again, the main problem with this is the relatively low maximum salary for Principal Lecturers, whilst achieving a professorship enabled substantial further pay progression.

9.8 Careers advice The evidence from both research students and from current academics

suggests that a substantial minority of entrants to academia have not seriously considered alternatives and may lack information on their options. For research students entering academia, it appears often to be a form of drift, to a job which is known, although this is usually driven by strong research interest. This suggests that the provision of more-effective careers advice may reduce, rather than increase, entry to academia.

However, although research students generally appeared to be well-informed about the nature of academic employment, they tended to underestimate academic salaries and to be unaware of the substantial boost to income provided by pension provisions. Better information on salaries and pensions may therefore increase the supply of research students into academic jobs. It seems likely that potential career changers are less well informed about the nature of academic jobs and ways to inform these about opportunities in academia would be useful.

210

9.9 Discrimination130 Discrimination (particularly if perceived to be worse than elsewhere) is liable

to act as a disincentive to the recruitment and retention of discriminated against groups. The evidence to date suggests discrimination in pay and/or promotion by gender and pay by ethnicity. However, substantially more research is needed in this area. Certainly, a majority of staff believed age affected opportunities in academia, nearly half believed gender did, over a third believed disability did and more than a quarter ethnicity.

However, there is no reason to believe that academia has overcome the problems of discrimination present in other sectors. Indeed, the extent of devolvement of recruitment, promotion and task allocation to departments, without extensive appropriate training, suggests that discrimination is less likely to be addressed than in certain other sectors. The rotation of the post of Head of department, common in old universities, is likely to add to this, reducing the ability to develop relevant expertise.

In addition to the gender and ethnic pay discrimination suggested by other research, specific issues this study raised in respect of discrimination were:

• the progression route and requirements for progression (for example, work time pressure, the need to attend conferences, and the onus to achieve quickly) may indirectly discriminate against carers and, most commonly, mothers; whilst flexibility over working time is helpful for carers, according the same access to flexibility to non-carers is likely to enhance, relatively, the latter’s job prospects, as non-carers are better able to take advantage of flexibility (e.g. able to work the hours which maximise their work output).

• few academic jobs are done part-time and the introduction of more flexibility in this respect might encourage the recruitment and retention of women

• the lower recognition given to administrative achievement may discriminate against women (and, possibly, ethnic minorities); this hinges on how administrative tasks are allocated and whether allocation takes advantage of a possible greater willingness of women to accept administrative tasks

• the use of early retirement and the expectation that staff at certain levels should be a certain age tends to perpetuate age discrimination

• it may be that the degree of devolvement of power to the departments has a further detrimental effect, making it difficult for those encountering problems to have them addressed.

Whilst addressing the above issues would be helpful, our main conclusion in respect of discrimination and, potentially disadvantaged groups, is that the area is substantially under-researched.

130 By discrimination we include both direct and indirect discrimination, including institutional discrimination. We do not mean to suggest that all is deliberate.

211

9.10 Other issues Increasing the supply of British research students should increase the supply

of academics, without reducing retention. This might be achieved through more funding for PhD students or addressing the debt built up during earlier degrees.

A number of aspects of academic employment that staff feel are important for their satisfaction did not affect their likelihood of leaving. Those who said that the RAE, the QAA and the general direction of higher education policy lowers their satisfaction by a lot are no more likely to expect to leave UK Higher Education than those who do not.

9.11 Management and human resource practices The evidence from the case study universities showed differences in approach

to recruitment and human resource practices. Most of the case study universities had mixed practices (good and bad) and seemed to be trying to address some of the problem areas. Tension between the department and both the university in general and human resourcing did not seem uncommon.

The general impression was of academic staff strongly focused on the research, teaching and external activities of their jobs with wider university matters seen as a distraction. This also extended to aspects of human resource practices (e.g. appraisal). In the case studies, not only were administrative demands seen as burdensome, but also often as unnecessary and even counterproductive. Some of these demands stemmed from organisational change. If such changes and other administrative demands could be reduced, that should be helpful. Where they cannot, perhaps greater communication between what appeared as two sides would lessen the resistance to administrative work.

The ethos and human resource structure (particularly in the old universities) was very much of individualism. This does not sit entirely well with an increasing use of current managerial techniques (monitoring, appraisal, emphasis on development etc.). The problem was not just of staff resistance to these, but also that it was not clear that the structure enables such systems to be adequately delivered. In particular, the system dominant in old universities of rotating heads of department needs to be considered. On average, this system will reduce management expertise of the heads and so both the system, role and training for heads should be looked at. For example one of the case study universities was developing management training for Heads of department, in order to improve competence and commitment to appraisals.

9.12 Further research A number of important gaps in our knowledge, which are relevant to policy

development in higher education, have been identified in the report. The main areas where further research is needed are as follows.

1. Career-decisions amongst undergraduates and Masters students. It was recognised in the initial stages of the study that the supply of academics was affected by decisions taken at undergraduate level onwards. However,

212

resources limited the study to research students. It would be useful to identify the factors that affected progression from undergraduate to Masters student to research student and the role that career decisions (and, particularly, consideration of an academic career) played. Identifying the role of tutors in progression and career decisions would also be useful.

2. Hourly paid staff. There is very little information on hourly paid staff or their role and the study was unable to explore this. There are a number of important issues about hourly paid staff, including:

• the extent and pattern of use (e.g. as lecturers, researchers) • the reasons for using hourly paid (e.g. uncertainty, cost, flexibility,

fixed-term need) • the role of hourly paid academic employment in an academic career: to

what extent is hourly paid an entry point (e.g. research students tutoring; for career changers), a survival strategy (for academics who cannot get longer-term contracts or full-time jobs), separate from an academic career (interesting/remunerative work for those with other jobs or activities) and an exit route towards retirement; how much does it encourage career changers into academia;

3. Career-changers. The evidence on recruitment to the sector was largely based on a survey of existing research students and of academic staff. The former enables an analysis of an important part of the labour pool for academic recruits. However, the latter can only describe what influenced those career-changers who are currently academics. Thus we have not been able to examine fully the factors dissuading potential recruits from entering an academic career from employment in other sectors. Further research into this is needed.

4. Discrimination. As we have said, there is little evidence on discrimination in academia, although a small number of studies point to problems of discrimination. It would be useful to examine the evidence on under-representation of women and ethnic minorities and pay and promotion differences by gender and ethnicity across the higher education sector. This has only been done for selected subjects or using simple descriptive techniques.

It would also be useful to examine the incidence and nature of equal opportunity policies and practices in universities in order to identify how equality opportunity issues might be better tackled. Previous research in this area is now relatively dated.

5. Further research on leavers. The survey of academic staff based its research into leaving the sector on respondents’ intentions. It would be useful to follow up respondents to find out which did leave. This would enable the validity of our assumption that leavers and those with the intention to leave are similar and, if not, then to identify the factors affecting leaving.

6. International flows of academics. The study identified a range of flows between UK and foreign academia (British academic staff working abroad for a period, foreigners employed in English academia and the greater propensity for certain nationalities to expect to leave UK academia). It is unclear the extent to which movement to academia abroad (particularly of British staff) represents a loss to UK academia (notably, the extent to which it is temporary and that it provides valuable development). Moreover, the reasons for international movement are not entirely clear. It would be useful to examine

213

the nature of their reasons. This might also include flows from developing countries, which were not considered in detail in the current study.

7. Human resource practices and management. The research suggested separation between management and human resourcing on the one hand and academic staff on the other. Ethos, organisation and job structure seemed to encourage this separation. This is liable to reduce effectiveness and, probably, leads to some of the aspects of dissatisfaction (with hours, administration and career progression) which encourage staff to leave the sector. Research to investigate the organisation of jobs and the roles of different sets of employees and their interaction might identify how this could be improved.

8. Fixed-term contracts. It is apparent that the use of fixed-term contracts increases recruitment and retention problems. However, the question remains whether their elimination, or reduction (particularly through providing greater job continuity, rather than just different contractual terms) is desirable and, if so, how this could be achieved. Research to examine the effectiveness of the structure of academic jobs and into fixed-term contracts and lecturing on permanent contracts and into whether and how this structure might be improved would be useful.

214

215

Bibliography

Academy of Medical Sciences, (2000), The tenure-track clinician scientist. London Akerlof, G.A., Rose, A.K., and Yellen, J.L., (1988), ‘Job switching and job

satisfaction in the U.S. labor market’, Brookings Papers on Economic Activity, 2, pp. 495-582.

Anderson, M.S., (ed.) (1998), The experience of being in graduate school: An exploration, New Directions for Higher Education, No. 101, San Francisco: Jossey-Bass.

Antel, (1985), Antel, J.J. (1985), ‘Costly Employment Contract Renegotiation and the Labor Mobility of Young Men’, American Economic Review, 75, pp. 976-991.

Austin, A.E., (2002), ‘Preparing the Next Generation of Faculty: Graduate School as Socialization to the Academic Career’, Journal of Higher Education, 73.1, pp. 94-122.

AUT (1995), Promoting Professionalism, Association of University Teachers. AUT (2003) ‘New family friendly rights from 6 April 2003’, circular to LA

representatives, AUT, London. Banerjee, D. S., and Gaston, N., (2004), ‘Labour market signalling and job turnover

revisited’, Labour Economics, 11, pp. 599-622. Bartel, A., Borjas, G., (1981). In: Rosen, S. (Ed.), Wage Growth and Job Turnover:

An Empirical Analysis. Studies in Labor Markets. University of Chicago Press for the NBER, Chicago, IL.

Bagihole, B. (2002) ‘Challenging Equal Opportunities: changing and adapting male hegemony in academia’, British Journal of Sociology of Education, 23(1) pp19-33.

Bailey, T., (1997), ‘Review of Assessing educational practices: The contribution of economics’, Journal of Economic Literature, 35(1), pp. 146-148.

Baimbridge, M., (1995), ‘Academic Staff Salaries: Equity and Exploitation’, Applied Economics Letters, December 1995, 2(12), pp. 469-72.

Balzer, W. K., Kihm, J. A., Smith, P. C., Irwin, J. L., Bachiochi, P. D., Robie, C., Sinar, E. F., and Parra, L. F., (1997), Users’ manual for the Job Descriptive Index (JDI; 1997 Revision) and the Job In General scales. Bowling Green, OH: Bowling Green State University.

Bellamy, S., Morley, C., and Watty, K., (2003) ‘Why Business Academics Remain in Australian Universities Despite Deteriorating Working Conditions and Reduced Job Satisfaction: an intellectual puzzle’, Journal of Higher Education Policy and Management, 25.1, pp. 13 – 28.

Bett, M., (1999), Independent Review of Higher Education Pay and Conditions. Norwich, Stationery Office.

Bisset, I.M., Fisher, S.G., Macrosson, W.D.K., and Phillips, C.W., (2000), ‘The Selection of Academic Staff’, Higher Education Management, 12.3, pp. 129-43.

Blackaby, D., and Frank, J., (2000), ‘Ethnic and other minority representation in UK academic economics’, Economic Journal, 110, pp. F293-311.

216

Blackaby, D., Leslie, D., Murphy, P., and O’Leary, N., (2002), ‘White/Ethnic Minority Earnings and Employment Differentials in Britain: Evidence from the LFS’, Oxford Economic Papers, 54, pp. 270-97.

Blake, M., and LaVille, I., (2000), Who Applies for Research Funding? Key factors shaping funding application behaviour among women and men in British higher education institutions, London: Welcome Trust.

Booth, A., Burton, J., and Mumford, K., (2000), ‘The position of women in UK academic economics’, Economic Journal, pp. F312-33.

Booth, A., Frank, J., and Blackaby, D., (2003), ‘Outside Offers and the Gender Pay Gap: Empirical Evidence from the UK Academic Labour Market’, presentation to the Royal Economic Society Annual Conference, Warwick.

Bryson, C., (1999), ‘Contract Research: the failure to address the real issues’, Higher Education Review, 31.2, pp. 74-94.

Bryson, C., and Barnes, N., (2000a), ‘Working in Higher Education in the United Kingdom’, in Tight (ed.) (2000), pp 147-85.

Bryson, C., and Barnes, N., (2000b), ‘The Casualisation of Employment in Higher Education in the United Kingdom’, in Tight (ed.) (2000), pp 186-241.

Caplow, T., and McGee, R.J., (2001), The academic marketplace, New Brunswick, N.J. and London: Transaction, 26.3, pp. 247-54.

Card., D., (1999), ‘The Causal Effect of Education on Earnings’, in O. Ashenfelter and D. Card (eds.) Handbook of Labor Economics, Amsterdam and New York: North Holland.

Cartter, A.M., (1976), Ph.D.’s and the Academic Labor Market. New York: McGraw Hill.

Carter, J., Fenton, S., Fenton, S., and Modood, T., (1999), Ethnicity and employment in Higher Education, PSI report number 865, London, Policy Studies Institute.

CHEMS, (2000) Academic Staff Salaries and Benefits in Seven Commonwealth Countries, 1998-99, Association of Commonwealth Universities.

CIPD, (2002), Labour turnover 2002, Chartered Institute of Personnel and Development.

Chevalier, A., Dolton, P., and McIntosh, S., (2002), ‘Recruiting and Retaining Teachers in the UK: An Analysis of Graduate Occupation Choice From the 1960s to the 1990s’, Discussion Paper 21, Centre for the Economics of Education, London School of Economics.

Chartered Institute of Personnel and Development (2002) Labour Turnover 2002: Survey Report. London: CIPD.

Clark, A.E., (1997), ‘Job satisfaction and gender: Why are women so happy at work?’, Labour Economics, 4, pp. 341-72.

Clark, A.E., (2001), ‘What really matters in a job? Hedonic measurement using quit data’, Labour Economics, 8.2, pp. 223-42.

Clark, A.E., Georgellis, Y., and Sanfey, P., (1998), ‘Job Satisfaction, wage changes and quits: evidence from Germany’, Research in Labour Economics, 17, pp. 95-121.

Clark, A.E., and Oswald, A.J., (1996), ‘Satisfaction and comparison income’, Journal of Public Economics, 61, pp. 359-81.

Coaldrake, P., and Stedman, L., (1999), Academic Work in the Twenty-first Century: Changing roles and policies, Occasional Paper 99H, Higher Education Division, department of Education, Training and Youth Affairs.

Coe, T., and Boddington, A., (1999), Study of the Factors Affecting the Career Choices of Chemistry Graduates, report to the Royal Society of Chemistry.

217

Coe, T., and Boddington, A., (2003), Recruitment and Retention of Women in Academic Chemistry, report to the Royal Society of Chemistry.

Cothal, J., Callan, V., and Härtel, C.E.J., (1997), ‘Recruitment and Selection of Academic Staff: perceptions of department heads and job applicants’, Journal of Higher Education Policy and Management, 19.2, pp. 99-110.

Court, S., (1998), Opportunity Blocks: A survey of appointment and promotion in UK higher education, Association of University Teachers, London.

Court, S., (1999), ‘Negotiating the research imperative: the views of academics on their career opportunities’, Higher Education Quarterly, 53.1, pp. 65-87.

CVCP (1997) Clinical academic careers. (The Richards Report). London. Dearden, L., Goodman, A., Griffith, R., and Tesser, D., (2003), ‘Career Progression in

Economics: A Scoping Report’, Royal Economics Society. Deaton, A. and Muellbauer, J. (1980) Economics and Consumer Behaviour, Cambridge

University Press, Cambridge. Dollard, M.E., Winefield, A.H., and McGirk, E.T., (1992), ‘Study into stress among

Correctional Services Officers, South Australia: Techsearch. Dolton, P., van-der-Klaauw, W., (1994), ‘The Turnover of UK Teachers: A

Competing Risks Analysis’, New York University, C.V. Starr Center Economic Research Report No. 94-21.

Dolton, P., van-der-Klaauw, W., (1999), ‘The Turnover of UK Teachers: A Competing Risks Analysis’, Review of Economics and Statistics, 81.3, pp. 543-50.

Dutta, J., Sefton, J. and Weale, M., (1999), ‘Education and Public Policy’, Fiscal Studies, 20.4, pp. 351-386

EC, (2000), Promoting Excellence Through Mainstreaming Gender Equality, Report of the ETAN Expert Working Group on Women and Science.

Ehrenberg, R.G., (1991), ‘Academic Labor Supply’, in Economic Challenges in Higher Education, Clotfelter, Ehrenberg, Getz and Siegfried (eds.) Chicago: University of Chicago Press.

Ehrenberg, R.G., (1992), ‘The Flow of New Doctorates’, Journal of Economic Literature, 30, pp. 830-75.

Ehrenberg, R.G., (2003), ‘Studying Ourselves: The Academic Labor Market’, Journal of Labor Economics, 21.2, pp. 267-87.

Ehrenberg, D.G., Kasper, H., and Rees, D.I., (1991), Faculty Turnover in American Colleges and Universities’, Economics of Education Review, 10.2, pp. 9-110.

Enders, J., (2000), ‘Academic Staff in Europe: Changing Employment and Working Conditions’, in Tight (ed.) (2000), pp. 7-32.

EPSRC (1999), Attraction & Retention of Staff and Students in Research in IT & Computer Science, Swindon: EPSRC.

Farber (1999) ‘Mobility and Stability’, Chapter 37 of Ashenfelter, O., and Card, D., (eds.), Handbook of Labor Economics, Amsterdam, Elsevier.

Freeman, R.B., (1978), ‘Job Satisfaction as an Economic Variable’, American Economic Review, 68, pp. 135-41.

Freeman, R.J., (1971), The Market for College-Trained Manpower, Cambridge, Mass.: Harvard University Press.

Freeman, R.J., (1975), ‘Supply and Salary Adjustments to the Changing Science Manpower Market: Physics 1948-1973’, American Economic Review, 65, pp. 287-311.

Freeman, R.B., (2000), ‘The Changing State of Economics in the United Kingdom and United States’, Economic Journal, 110, F355-7.

218

Gibbons, R., and Waldman, M., (1999), ‘Careers in Organizations: Theory and Evidence’, in Orley C. Ashenfelter and David Card (eds.), Handbook of Labor Economics, Volume 3b, Chapter 36, Amsterdam: Elsevier.

Golde, C.M., and Dore, T.M., (2001), ‘At Cross Purposes: What the experiences of doctoral students reveal about doctoral education.’ A report prepared for The Pew Charitable Trusts, Philadelphia, PA. www.phd-survey.org

Graves, P.E., Marchand, J.R., and Sexton, R.L., (2002), ‘Hedonic wage equations for higher education faculty’, Economics of Education Review, 21, pp. 491-6.

Guttman, L., (1954), ‘Some Necessary Conditions for Common Factor Analysis’, Psychometrika, 19.2, pp. 149-61.

Hagedorn, L.S., and Sax, L. J. (1999), ‘Marriage, children, and aging parents: The role of family-related factors in faculty job satisfaction’, mimeo, University of Southern California, submitted to American Behavioral Scientist

Hagle, T.M., and Mitchell, G.E., (1992), ‘Goodness-of-fit Measures for Probit and Logit’, American Journal of Political Science, 36.3, pp. 762-784.

Halsey, A., (1992), The Decline of Donnish Dominion, Oxford: Clarendon Press Harman, H.H., (1976), Modern Factor Analysis, Chicago: University of Chicago Press. Harris., G., (1997), ‘Is job sharing worthwhile? A cost-benefit analysis in UK

universities’, Higher Education, Jan, 33.1, pp. 29-38. Hasluck, C., J. Pitcher and C. Simm (2001). Academic Research Careers in Scotland:

A Longitudinal Study of Academic Contract Research Staff, Their Jobs and Their Career Patterns. Edinburgh/Coventry: Scottish Higher Education Funding Council and Institute for Employment Research.

Heap, H. P. S., 1999, ‘Making British universities accountable: in the public interest?’, Paper prepared for the Conference towards a New Economics of Science, University of Notre Dame, 13-16 March 1997, Chicago University Press, US.

HEFCE (2000), Review of research, Publication 2000/37. HEFCE (2002), Academic staff: trends and predictions, Publication 2002/43. HEFCE (2003), Appointment, retention and promotion of academic staff in higher

education institutions, Report to the HEFCE by the Scottish Council for Research in Education, University of Glasgow and Nottingham Trent University.

HEFCE (2005a) Higher Education Workforce in England – Trends, profiles and projections.

HEFCE (2005b) Staff employed at HEFCE funded HEIs: trends, profiles and projections, April.

Hinsz, V.B., Nelson, L.C., (1990), ‘Testing models of turnover intentions with university faculty’, Journal of applied social psychology, 20.1, pp.68-84.

HM Treasury, 2002), SET for success: The supply of people with science, technology, engineering and mathematical skills. (Report of Sir Gareth Roberts' Review), April 2002.

House of Commons Science and Technology Committee (2002a) ‘Science and Technology - Second Report’ 10 April 2002. London: House of Commons. http://www.parliament.the-stationery-office.co.uk/pa/cm200102/cmselect/cmsctech/507/50702.htm

House of Commons Science and Technology Committee (2002b) ‘Science and Technology – Eighth Report’ 6 November 2002. London: House of Commons. http://www.parliament.the-stationery-office.co.uk/pa/cm200102/cmselect/cmsctech/1046/104602.htm

219

Husbands, C. and A. Davis (2000). The teaching roles, institutional locations, and terms and conditions of employment of part-time teachers in UK Higher Education. Journal of Further and Higher Education 24.3: 337-362.

IRS (2002), Recruitment and retention of staff in UK higher education: A survey and case studies, report commissioned by HEFCE, SCOP, UCEA and UUK.

Jackson, M.P., (1997), ‘The impact of discretionary pay in UK universities’, Higher Education Management, 9, pp. 99-113.

Jenkins, A., Vlackman, T., Lindsay, R., and Paton-Salzburg, R., (1998), ‘Teaching and research: student perspectives and policy implications’, Studies in Higher Education, 23.2, 127-141.

Jones, J.R., Huxtable, C.S., Hodgson, J.T., Price, M.J., (2003), ‘Self Reported Work Related Illness in 2001/2002 Results from a Household Survey.’ HSE

Jovanovic, B., (1979), ‘Job matching and the theory of turnover’, Journal of Political Economy, 87, pp. 972-990.

Kaiser, H.F., (1970), ‘A Second Generation Little Jiffy’ Psychometrika, 35, pp. 401-17. Keep, E and Mayhew, K., (1995), ‘The economic demand for higher education and

investing in people’. In Coffield, F. (ed) Higher Education in a Learning Society, Durham, Durham University.

Keep, E., and Sisson, K., (1992), ‘Owning the Problem: Personnel Issues in Higher Education Policy Making in the 1990s’, Oxford Review of Economic Policy, 8.2, pp. 67-78.

Keep, E., Storey, J., and Sisson, K., (1996), ‘Managing the employment relationship in higher education: Quo vadis?’ in Cuthbert, R. (ed) Working in Higher Education, SRHE/OUP.

Kinman, G., and Jones, F., (2004), ‘Working to the Limit: Stress and work-life balance in academic and academic-related employees in the UK’, AUT.

KPMG (2005) Evaluation of Rewarding and Developing Staff in HE initiative 2001-02 to 2003-04. A report to HEFCE by KPMG LLP, May. http://www.hefce.ac.uk/pubs/rdreports/2005/rd14_05/

Laafia, I., and Larsson, A., (2001), Women in public research and higher education in Europe, Eurostat, ‘Statistics in Focus’ – Theme 9.

Lancaster, K. (1971) Consumer Demand: A New Approach, Columbia University Press, New York.

Lindsay, R., Breen, R., and Jenkins, A., (2002), ‘Academic Research and Teaching Quality: the views of undergraduate and postgraduate students’, Studies in Higher Education, 27.3, pp. 309-27.

Long, J.S., (1997), Categorical Limited Dependent Variable Models, Sage. Machin, S., and Oswald, A., (1999), Signs of Disintegration: A Report on UK

Economics PhDs and ESRC Studentship Demand, Report to ESRC (original report submitted November 1998, final revision June 1999).

Machin, S., and Oswald, A., (2000), ‘UK economics and the future supply of academic economists’, Economic Journal, 440, pp. F334-49.

McKelvey, R.D., and Zaviona, W., (1975), ‘A statistical model for the analysis of ordinal level dependent variables’, Journal of Mathematical Sociology, 4, pp. 103-120.

McNabb, R., and Wass, V., (1997), ‘Male-Female Salary Differentials in British Universities’, Oxford Economic Papers, July 1997, 49(3), pp. 324-43.

Mallam, U., (1994), ‘A national research study on factors influencing faculty turnover at selected Nigerian colleges of technology/polytechnics’, Higher education, Mar 1994, Vol.27, No.2, pp.229-238.

220

Maxwell J., and Murphy, D., (2003), Academic Staff Salaries and Benefits in Seven Commonwealth Countries, 2001/02002, Association of Commonwealth Universities.

McInnes, C., (1999), ‘The work roles of academics in Australian Universities’, Canberra: department of Education, Training and Youth Affairs.

Millward, N., Stevens, M., Smart, D., and Hawes, W.R., (1992), Workplace Industrial Relations in Transition: the ED/ESRC/PSI/ACAS Surveys, Aldershot: Dartmouth Publishing.

Mincer, J., (1986), ‘Wage changes and job changes’, Research in labor economics 8. Greenwich, CT: JAI Press, pp. 171-191.

Mobley, W.H., (1977), ‘Intermediate linkages in the relationship between job satisfaction and employee turnover, Journal of Applied Psychology, 62, pp. 237-40.

Moore, W.J., Newman, R.J., and Turnbull, G.K., (1998), ‘Do Academic Salaries Decline with Seniority?’ Journal of Labor Economics, 16.2, pp. 352-66.

Murphy, K. M., and Welch, F., (1990), ‘Empirical Age-Earnings Profiles’, Journal of Labour Economics, 8.2, pp. 202-29.

NATFHE, (2001), ‘Know your rights as a lecturer’, Guardian, Monday October 15, 2001.

Nicholson, E.A., and Miljus, R.C., (1972), ‘Job Satisfaction and Turnover among Liberal Arts College Professors’, Personnel Journal, 51, pp. 840-5.

Ong, L.L., and Mitchell, D., (2000), ‘Professors and hamburgers: an international comparison of real academic salaries’, Applied Economics, 32, pp. 869-76.

Oshagbemi, T., (1996), ‘Job Satisfaction of UK Academics’, Education Management & Administration, 24.4, pp. 389-400.

Oshagbemi, T., (1999), ‘The impact of age on the job satisfaction of university teachers’, Research in Education, 59, pp. 95-108.

Pearson, R., Buchan, J., Bevan, S., Jackson, C., and Stock, J., (1990), The Recruitment and Retention of University Academic and Academic Related Staff, IMS Paper 157, Brighton.

Pissarides, C.A., (1992b), ‘Search Theory at Twenty-One’, London School of Economics: Centre for Economic Performance Discussion Paper No. 90.

PREST (Policy Research in Engineering, Science and Technology University of Manchester) (2000) Impact of the Research Assessment Exercise and the Future of Quality Assurance in the Light of Changes in the Research Landscape, Report to the Higher Education Funding Council for England.

Provan, D., (2001) Academic Staff Salaries and Benefits in Six Commonwealth Countries, 2000-2001, Association of Commonwealth Universities.

Rees, D., (1994), ‘Does Unionisation Increase Faculty Retention?’, Industrial Relations, July 1994, 33(3), pp. 297-321.

Rice, R.E, Sorcinelli, M.D., and Austin, A.E., (2000), Heeding new voices: Academic careers for a new generation, New Pathways Inquiry #7, Washington, DC: American Association for Higher Education.

Rich, J., (1999), ‘Gender Segregation in the Academic Staff of Universities in Great Britain, 1980/81 – 1993/94’, Economica, November 1999, 66(264), pp. 509-31.

Roback, J., (1982) ‘Wages, Rents, and the Quality of Life’, Journal of Political Economy, 90, pp. 1257-78.

Rosen, S., (1979), ‘Wage-Based Indexes of Urban Quality of Life’, in P. Mieszkowski and M. Strazheim (eds.), Current Issues in Urban Economics, Baltimore: Johns Hopkins University Press.

221

Smith, P. C., Kendall, L. M., and Hulin, C. L., (1969), The measurement of satisfaction in work and retirement. Chicago: Rand McNally.

Stapleton, D.C., (1989), ‘Cohort Size and the Academic Labor Market’, Journal of Human Resources, 24, pp. 221-52.

Stevens, P.A, (2004), ‘Academic Salaries in the UK and US’, National Institute Economic Review.

Tight, M., (ed.), (2000), Academic Work and Life: What it is to be an Academic, and how this is Changing, London: JAI.

Tolbert, P.S., Simons, T., Andrews, A., and Rhee, J., (1995), ‘The effects of gender composition in academic departments on faculty turnover’, Industrial and labor relations review, April 1995, Vol.48, No.3, pp.562-579.

UCEA (2002), Recruitment and retention of staff in UK higher education: A survey and case studies, report commissioned by the HEFCE, SCOP, UCEA and UUK.

University Grants Committee, 1984, A Strategy for Higher Education into the 1990s, UGC

University of Plymouth (2003) Occupational stress in higher education institutions, Phase I results for all HEIs, April. University of Plymouth, Institute of Health Studies. http://www.ihs.plymouth.ac.uk/~stresshe/worddoc/51077.doc

Van-der-Berg, B., Siegers, J., and Winter-Ebmer, R., (1998), ‘Gender and Promotion in the Academic Labour Market’, Labour, 12(4), pp. 701-13.

Ward, M., (2001), ‘The Gender Salary Gap in British Academia’, Applied Economics, October 2001, 33(13), pp. 1669-81.

Ward, M.E., and Sloane, P.J., (2000), ‘Non-Pecuniary Advantages Versus Pecuniary Disadvantages; Job Satisfaction among Male and Female Academics in Scottish Universities’, Scottish Journal of Political Economy, 47.3, pp. 273-303.

Winefield, A.H., Gillespie, N., Stough, C., Dua, J., and Hapuararchchi, J., (2002), ‘Occupational Stress in Australian Universities: A National Survey’, National ertiary Education Union.

Windmeijer, F.A.G., (1995), ‘Goodness-of-fit Measures in Binary Choice Models’, Econometric Reviews, 14.1, pp. 101-116.

Worrall, L. and Cooper, C. L. (1998) The Quality of Working Life: 1998 survey of managers' changing experiences, London: Institute of Management.

Zoghi, C., (2003), ‘Why have public university professors done so badly?’, Economics of Education Review, 22, pp. 45-57.

222

223

Appendix A: The qualitative research

A.1 The scope of the qualitative research Recruitment and retention of staff encompass a wide and complex range of

issues and the study included a strong qualitative component to identify the full range of experience and perspectives (and not only the most pressing concerns). The purpose of the qualitative research was to capture this range, in order to assist in the design of the quantitative survey and to assist in the explanation of the quantitative survey findings. The scope of the qualitative research for the study was, therefore, wide. It covered issues of strategy, policy and practice, at central and departmental level, and the experiences and perspectives of individuals working or studying at postgraduate level in Higher Education.

Interviews were conducted with staff at three levels and with postgraduate students. At a senior level, interviews were held with human resource managers and heads of department (or in some cases, school) to identify the main issues and concerns for universities and departments in relation to recruitment and retention. This included identifying problems, strategies and practices at the institutional level (channelled through human resource/personnel departments) aimed at dealing with recruitment and retention difficulties and identifying problems and practice at a departmental level. Interviews with staff and postgraduates were designed with two purposes: to assist in the design of surveys of staff and students by identifying the main issues affecting the decision to enter, remain in and leave employment in Higher Education; and to explore decision-making processes in a more detailed way than is possible in a structured survey. The qualitative research therefore explored intrinsic and extrinsic features of the job and how other factors (such as personal circumstances) interact with these and enhance or diminish their importance. Interviews also explored alternatives to working in Higher Education, whether these had been considered or whether future job changes were anticipated.

The fieldwork was conducted between July 2003 and July 2004. Most interviews with human resource managers were conducted between July and October 2003 and most interviews with departmental staff and research students were conducted between January and July 2004.

A.2 The sample At the start of the study, face-to-face interviews were carried out with senior

representatives of 13131 universities examining recruitment and retention issues, including university strategies and practices affecting recruitment and retention. In most cases the Director of Human Resources/Personnel was interviewed (often

131 Including one university which subsequently decided not to participate in the study.

224

together with other human resource/personnel staff). In some cases, other senior staff were interviewed, for example, the pro-Vice Chancellor responsible for academic staffing.

From the 12 participating universities, eight were selected to take part in case studies. These were selected to include universities with differing status and a range of average RAE score and types of location. Four were old universities, three were new universities and one was a College of Higher Education132. Two universities were in London (one new and one old university), the others were from large cities and towns of varying sizes.

In each university, the case study focused on two departments. The selection of subject areas, and therefore schools or departments, was guided by the following principles:

• to reflect, broadly, the distribution of academic staff in higher education by discipline, e.g. about half should be science and engineering, half arts and social sciences;

• to exclude minor subjects

• to cover subjects with differing experience of recruitment and retention problems, i.e. to include departments which had been identified in the initial interviews as having recruitment or retention problems (e.g. Education) and some where there was a lack of problems (e.g. History);

• to cover a range of subjects with differing potential career paths, i.e. to include subjects:

- for which movement between academic and non-academic employment is thought to exist throughout the career (e.g. law);

- which offer early choice of career in the specialist subject outside academia as well as inside (e.g. psychology);

- where pursuit of a subject interest tends to restrict one to academic employment (e.g. history);

• to cover some subjects in both old and new universities, in order to explore differences in experience between the two;

• to include departments with both higher and lower RAE performance;

• medicine was excluded, as those on clinical rates were excluded from the study and those in medicine but not on clinical rates are likely to be either atypical or to move between clinical and non-clinical rates

The subject areas and university schools/departments selected are listed in

Table A.1.

132 For ease of exposition and to protect anonymity, from this point the College of Higher Education will be referred to as a university (and, specifically, a new university).

225

Table A.1 Case studies: subject areas

new universities old universities Engineering Engineering

Computer Science Computer Science Law Law

Art and Design Biological Science Education English

Health Management History Physics

Psychology Sociology

Four groups were interviewed in these selected departments/schools. These were:

1. heads of department/school, or assistant heads;

2. first-time entrants to academic posts in Higher Education recruited in the past two years;

3. current staff with more than two year’s service; and

4. post-graduate students (as potential academic staff).

It had also been intended to interview staff who had resigned and were currently serving out their notice. However, it was not possible to identify these because heads of department were not aware of any staff in such circumstances.

In most cases sampling was carried out using lists provided by the head of department, with standard information about grade and additional information, for example length of service. In a small number of cases, staff lists were obtained from university web-pages. Postgraduates were selected where departments had these and where heads or assistant heads of department provided lists. Sampling of staff and postgraduate students was carried out to secure a broadly representative mix by seniority, age, sex and ethnic group. This was to obtain the experiences and perspectives of staff in different circumstances. These were then approached by letter, with information about the study, and interviewed by telephone using a semi-structured discussion guide.

In total, telephone interviews were conducted with 70 academic staff and postgraduate students across the 16 departments, as follows:

• 15 heads (or assistant heads) of department (or school) to provide an overview of recruitment and retention issues and practice at this level;

• 47 members of academic staff, from junior researchers and lecturers to professors, to explore reasons for working in Higher Education, experiences and future intentions.

• eight postgraduate students to explore reasons for pursuing a higher degree and career intentions.

226

227

Appendix B: The staff and student surveys

B.1 Aim The aim of the surveys was to provide representative data on academic staff

(lecturing and research grades) and of research students in England, whilst, for resource reasons, conducting the surveys in 12 universities only. This limitation guided our selection of universities and the sample sizes in each university. However, only ten universities actually took part in the staff survey and nine universities in the research student survey (see page 229).

B.2 The university sample The purpose of the sampling process was to provide a sample of academics

and of students from which whole-sector estimates of the factors affecting recruitment and retention could be made, from which differences by job factors (e.g. subject and seniority) and by personal factors could be estimated and from which differences by organisational differences (for example, type of institution) could be estimated. The first purpose required the sample to be random, the second required the sample to cover a wide range of subjects and types of respondent, whilst the third required a spread of types of institution.

B.2.1 University sample population The population comprised all English higher education institutions excluding:

• institutions with substantial concentration of subjects; these were excluded in order to ensure that the survey sample adequately covered a range of subjects and to facilitate generalisation from the sample; those excluded comprised all Specialist Institutions133 (Table A.2) (about half of which were arts, design and drama colleges), with the exception of:

- College of St Mark & St John - London School of Economics & Political Sciences - School of Oriental and African Studies - St Martin's College - Trinity & All Saints.

133 A concern was expressed that the exclusion of most Specialist Institutions would, without structuring, result in the study being unlikely to be able to investigate the main subjects of the Specialist Institutions (i.e. education, nursing and health studies, drama, design, art, agriculture). With the exception of agriculture, this was seen as very unlikely, as for each of these subjects (cost centres) at least 70 per cent of academic staff were in the proposed sample (i.e. 30 per cent or fewer were in the excluded Specialist Institutions and institutions with fewer than 200 academic staff) and, with the exception of agriculture, each cost centre appears in at least 84% of included institutions.

228

• institutions with fewer than 200 academic employees; these were excluded to ensure the analysis could be conducted by institution. Those excluded (in addition to small specialist institutions excluded due to limited subject spread) are listed in Table A.3.

Table A.2 Specialist institutions excluded from the sample

Art, music, drama colleges The Arts Institute at Bournemoutha Central School of Speech and Drama Conservatoire for Dance and Dramaa Cumbria College of Art & Design (Cumbria

Institute of the Arts) Dartington College of Arts Falmouth College of Arts Kent Institute of Art & Design Northern School of Contemporary Dance Norwich School of Art & Design Ravensbourne College (Ravensbourne

College of Design and Communication) Rose Bruford College Royal Academy of Music Royal College of Art Royal College of Music Royal Northern College of Music Surrey Institute of Art & Design (The Surrey

Institute of Art & Design, University College)

The London Institute (combines St Martin’s and others)

Trinity College of Music. Welsh College of Music and Drama Wimbledon School of Art

Other specialist institutions (area of specialism)

Bishop Grosseteste College (education) Cranfield University (management and

engineering) Institute of Cancer Researcha (biosceonce,

clinical medicine, nursing) London Business School (management) London School of Hygiene & Tropical

Medicine (clinical medicine) Newman College (Education) RCN Institute (also known as the Institute of

Advanced Nursing Education) (nursing) Royal Veterinary College (Veterinary) School of Pharmacy (Pharmacy) St George's Hospital Medical School (clinical

medicine and related) University of London Institute of Education

(education)

Agricultural colleges

Harper Adams Agricultural College Royal Agricultural Collegea Writtle College

• Source: HEFCE Report 98/10 Funding of specialist institutions: Report of the advisory panel chaired by Sir Stewart Sutherland March 1998 http://www.hefce.ac.uk/pubs/hefce/1998/98%5F10.htm and a ‘Public resources for teaching and student numbers in HEFCE-funded institutions: 2001-02’ Tables 1-5 http://www.hefce.ac.uk/pubs/hefce/2002/02%5F42.htm

• (The following, listed in the HEFCE Report 98/10, were excluded due to mergers with other institutions: Bretton Hall (now University of Leeds); Homerton College (now Cambridge University), Loughborough College of Art & Design (now Loughborough University); Wye College, University of London (now Imperial College); The College of Guidance Studies (dissolved and now Canterbury Christchurch); Westhill College (now Birmingham University); Westminster College, Oxford (now Oxford Brookes).

229

Table A.3 Other institutions with fewer than 200 staff excluded from the sample

Institution staff n Trinity and All Saints College 130 College of St Mark and St John 135 University of London (Institutes and activities) 141 St Mary's College 143 University College Chichester 181 York St John College 193 • source: HESA ISR 2001/02

B.2.2 University sample structuring The university sample was structured by college/other higher education

institution, average RAE (2001) score134 and location (London and elsewhere) (Table A.4). Colleges of Higher Education are listed in Table A.5; London Universities are listed in Table A.6.

Table A.4 University sample structuring

Population Sample

HEIs Staff Intended HEIs

Actual HEIs, staff

Actual HEIs,

students Colleges 11 3,919 1 1 1 London 21 23,658 2 2 2 RAE <5 13 10,376 1 1 1 RAE 5+ 8 13,282 1 1 1 Other locations RAE < 4 30 30,235 5 4 3 RAE 4 - <5.5 17 22,641 2 1 1 RAE>5.5 13 25,084 2 2 2 Total 92 105,537 12 10 9 • source: HESA ISR 2001/02. Staff on clinical grades are excluded (as are selected specialist

institutions and institutions with fewer than 200 academic staff, see Section B.2.1).

The actual number of universities participating in the surveys differed from that intended. One participant university failed to progress to providing a sample (and so only participated in the qualitative research). Another informed the research team two weeks before the start of the fieldwork that data protection considerations meant they would need to use an opt in. In the time scale it was not possible to replace these two universities. A third university was not willing to supply a research student sample (for confidentiality reasons) and we had decided that the omission of this single university would be acceptable. 134 The university RAE score is the THES RAE assessment, which calculates an average RAE score per member of staff based on the 2001 exercise and averages over all staff: all staff were counted with non-submitted staff allocated the departmental score minus two and scores calculated on a seven-point scale.

230

Table A.5 Colleges of Higher Education

Bath Spa University College Bolton Institute of Higher Education Buckinghamshire Chilterns University College Canterbury Christ Church University College Chester College of Higher Education Edge Hill College of Higher Education King Alfred's College, Winchester Liverpool Hope University College Southampton Institute St Mary's College University College Chichester University College Northampton University College Worcester York St John College Table A.6 Universities in London

Birkbeck College, University of London Brunel University City University University of East London Goldsmiths College, University of London University of Greenwich Imperial College of Science, Technology and Medicine, University of London King's College London, University of London Kingston University London Metropolitan University London School of Economics and Political Science, University of London Middlesex University Queen Mary University, University of London Royal Holloway, University of London School of Oriental and African Studies, University of London South Bank University Thames Valley University University of Surrey, Roehampton University College London, University of London University of Westminster • source: http://www.hefce.ac.uk/pubs/hefce/2002/02_42/02_42anna.xls

The sample of universities was drawn at random within the sampling structure.

The Vice-chancellor (or equivalent) was contacted to seek agreement for participation. Discussions were then held with a senior member of staff responsible for human resources (normally the head of human resources, but in some cases a senior academic) to ensure the university could supply the appropriate samples of staff and students.

231

In addition to the 12 universities already described, six other universities were contacted, but declined to participate. In two of these, a decision against participation was taken by the Vice Chancellor. In a further two, a decision was made by the Head of Human resources, on the grounds of the additional work entailed in sampling and selecting email addresses. In a further two cases, an initial meeting took place with the Head of Human Resources and other senior staff, but was followed by a decision against participation, on the grounds of the additional work required and the demands on academic staff.

B.3 The staff sample Academic staff on clinical rates were excluded from the study. This was a

policy decision, due to clinical rates being separately determined and due to the somewhat different patterns of employment for this group. Academic staff on clinical rates comprise five per cent of all HEI academic staff in England (5983 out of 118546) (Table A.7). The main subject areas affected were medicine and dentistry (77 percent of staff excluded), subjects allied to medicine (8 per cent excluded) and biological sciences (3 per cent excluded).

Table A.7 Academic staff on clinical rates

% by subject

on clinical rates

n

all staff

% of all staff on clinical

rates Subject of academic discipline group no yes (1) Medicine & dentistry 23 77 5,460 70 (2) Subjects allied to medicine 92 8 9,493 12 (3) Biological sciences 97 3 15,490 7 (7) Mathematical sciences 99 1 3,738 0 (J) Combined 98 2 2,030 1 (Z) Invalid subject of academic discipline supplied

95 5 9,043 8

Total 95 5 118,546 100 n 118,546 • Fewer than 0.5 per cent of academic staff were on clinical rates for other subjects • Source: HESA ISR 2001/02

Through discussions with the participating universities at the start of the study, it became clear that hourly paid staff could not be included in the survey. This was because few of the case study universities were able to identify hourly paid academic staff (i.e. lecturers and researchers) (they were not distinguishable from other hourly paid staff) and because email addresses were not held.

Within each university, all eligible academic staff were included in the survey. This was to ease sample selection for the universities. However, one university required an opt out. As the nature of this university (a high ranking, research intensive university) meant that it would be an important supplier of future academics, we did not wish to exclude it from the study. For reasons of confidentiality, the non-opt out staff numbers are not given for this university. The total staff sample from all the

232

universities supplied to NIESR/BMRB was 8663. The sample was stratified by the groups listed in Table A.8. (Groups have been aggregated to maintain confidentiality.)

Table A.8 Staff sample

number of universities staff sample

London RAE <5 and RAE 5+ 2 1779 Other locations RAE < 4 and college 5 3183 RAE 4 - <5.5 and RAE>5.5 3 3701 Total 10 8663

Data has been aggregated to maintain confidentiality

B.4 The student sample Within each university, all students were included in the survey. As with staff,

this was to ease sample selection for the universities. The sample was stratified by the groups listed in Table A.8. (Groups have been aggregated to maintain confidentiality.)

Table A.9 Student sample

number of universities

student sample

London RAE <5 and RAE 5+ 2 531 Other locations RAE < 4 and college 4 833 RAE 4 - <5.5 and RAE>5.5 3 3153 Total 9 4,517

Data has been aggregated to maintain confidentiality

B.5 Fieldwork The surveys were web-based. Staff and students were contacted by the

university to seek permission to include them in the sample. The initial contact explained the purpose of the study. For all but one university, permission was assumed to be granted unless the recipient said otherwise (i.e. an opt out). Those included in the sample were then sent an email informing them about the survey, providing a unique personal identifier and inviting them to log on to complete the questionnaire. Non-respondents were sent two reminders at approximately three weekly intervals.

Staff were first contacted by the survey team on the 13th and 14th May 2004. Students were first contacted between 17th and 27th May 2004. The surveys closed 7th July 2004.

233

B.6 Response rates The total number of respondents was 2805 for the staff survey and 1330 for

the student survey. Excluding the opt in university, the staff survey had a response rate of 32%. The student survey response rate was 29%.

B.7 Weighting For staff, weighting factors were applied to each element of the sampling

structure to make each reflect its relative size in the population. Differential weights were applied to science and non-science within cells for the same purpose. The weights applied are given in Table A.10.

Table A.10 Weights

Staff weight

Students weight

science 0.8092435 1.228593431 Colleges non-science 0.8927722 1.57862256 science 1.562129 1.228593431 London RAE <5 non-science 0.7025301 1.57862256 science 1.1201705 0.793499719 London RAE 5+ non-science 1.30158 1.314286462

Other locations science 0.7583255 1.228593431 RAE < 4 non-science 1.17128 1.57862256 science 0.8733812 0.793499719 RAE 4 - <5.5 non-science 1.62916 1.314286462 science 0.8632424 0.793499719 RAE>5.5 non-science 1.25769 1.314286462

For students, a similar approach would have led to some relatively large weights. Instead, average weights for those in institutions with an RAE rating of under 4 and 4 and higher were constructed, again with separate weights for science and non-science (Table A.10).

B.8 Bias Key characteristics of survey respondents were checked against key

characteristics of staff and of students as provided by HESA data.

B.8.1 Staff The characteristics of staff in the survey were similar to that reported in the

HESA data in terms of gender, part-time working and subject. Permanent staff were somewhat over-represented in the survey and thus findings based on the full dataset are likely to be biased towards the interests of permanent staff.

234

Table A.11 Staff: comparison of characteristics: survey and HESA data

Survey HESA % Female 42 38 % PT 11 13 % Permanent 68 55 Subject Medicine and dentistry 5 1 Subjects allied to medicine 13 8 Biological sciences 11 15 Agricultural and related subjects 1 1 Physical sciences 9 12 Mathematical sciences 3 3 Computing science 5 5 Engineering 6 7 Other technology 0 2 Architecture, building and planning 1 1 Social studies, including economics & social/economic geography 14 15

Business and administrative studies 6 5 Librarian and Information sciences 0 1 English literature and classics 2 4 Modern languages 3 2 Other humanities 7 5 Art and Design 2 5 Education 8 5 Combined studies across above subject groups 4 2

• Source: NIESR/DfES Staff survey and HESA Individualised Staff Record 2001/02 (excluding clinical staff)

• Survey data are weighted

B.8.2 Research students The survey reflected research students well in terms of subject. The survey

slightly over-represented women.

235

Table A.12 Research students: comparison of characteristics: survey and HESA data

Survey HESA % Female 49 43 Subject Medicine and dentistry 7 9 Subjects allied to medicine 8 5 Biological sciences 16 11 Agricultural and related subjects 1 2 Physical sciences 14 12 Mathematical sciences 1 2 Computing science 4 4 Engineering 11 11 Other technology 1 2 Architecture, building and planning 1 2 Social studies, including economics & social/economic geography 11 12

Business and administrative studies 2 4 Librarian and Information sciences 1 English literature and classics 4 4 Modern languages 2 2 Other humanities 10 7 Art and Design 1 3 Education 1 7 Combined studies across above subject groups 4 1

• Source: NIESR/DfES Student survey and HESA Student Record 2001/02 • Survey data are weighted • HESA student data relate to data for postgraduate students whose degree is mainly by research in

England in the academic year 2001/02.

236

237

Appendix C: HESA staff and student data

C.1 HESA staff data set. Our analysis of numbers of academic staff is based on data supplied by the

Higher Education Statistical Agency (HESA), extracted from the Staff Individualised Record 2001/02 (ref 21349). The Staff Individualised Record provides data in respect of the characteristics of members of academic staff employed under a contract of employment by a higher education institution in the UK. Thus, the record does not contain information on staff providing a higher education in further education colleges (FECs) that are not funded directly by the Higher Education Funding Council England (HEFCE). Academic staff are defined as those whose primary employment function is teaching only, teaching and research, or research only. Staff data relate to individual academic appointments of at least 25% of a full-time equivalent number of staff active during the academic year, 1 August 2001 to 31st July 2002. Part-time staff working half-time for the whole year and full-time staff working for six months of the year only would both be shown as 50% FTE. A member of staff working half-time for half of the year would be shown as 25% FTE. In other words, FTE includes part-time and part-year proportions for the individual, including if necessary the product of part-time and part-year work.

There are a number of exclusions from the data, the first by HESA, the others by ourselves in order to make the population similar to that of the universe from which our case study HEIs are drawn. The University of North London was not included in the data supplied by HESA because the university has asked that its individual level data is not released at this time. Therefore all analysis, based on the HESA staff individualised record excludes University of North London. In addition to this, after consultation with the DfES, we have restricted our sample in two ways: we have limited our analysis to English higher education institutions because of differences in the funding structures in other UK countries and we have removed staff who are paid on clinical rates, e.g. doctors and dentists (these staff may or may not hold honorary contracts with the NHS). This was done using field 23 ‘Clinical Status’.

Note that there are staff in the Individualised Staff Record with their FTE recorded as less than 25%. The majority of these records relate to staff who arrived or left during the year. The rest are staff who hold multiple contracts.

Note that the data for the staff individualised record are essentially based on contracts rather than members of staff. Staff who are employed on multiple contracts/appointments are included in the record in the individual contracts/appointments HESA coverage conditions. Because of the different internal systems and practices within institutions, HESA accepts either a single record for these staff or multiple records. Institutions returning multiple contractual appointments for a single member of staff must use the same staff identifier, so that they relate to a single appointment, on which the return is based. Because of the

238

variation in recording practice, one cannot be certain how many returns relate to multiple post holders. However, internal work by HEFCE has used counts of persons at institutions, i.e. if an individual held a more than one post at a single HEI, they were counted as a single observation, but if they held posts at more than one HEI they were counted as more than one. This work shows that the use of contracts rather than persons has little effect on the conclusions to be drawn from the data. As we can see from Table A.13 and Table A.14 below, the percentages are identical whether one uses the contract count or person at institution count, with the exception of the ‘percentage of staff record reported as leaving the institution or unknown’ for the destinations ‘At another UK HEI’ and ‘unknown’, which are one per cent different depending on which count is used. Even in this case the differences are not large enough to make an impact on the interpretation one puts on the pattern of the data.

Table A.13 Comparison of HESA Individualised Staff Record for 2001/2 using contract counts and persons at institutions – Employment in previous year

Number of staff records % of staff records % of staff records reported as new to institution (or unknown)

Previous employment Contracts Persons at

HEI Contracts Persons at HEI Contracts Persons at

HEI At same institution 92,511 90,756 84% 84% n/a n/a

At another UK HEI 4,191 4,164 4% 4% 24% 24%

Other 10,044 9,978 9% 9% 57% 57% Unknown 3,348 3,253 3% 3% 19% 19% Total 110,094 108,151 100% 100% n/a n/a

Source: Unpublished Internal work by HEFCE Columns for ‘contracts’ relate to counts of staff contracts as used in this report – final column relates to recorded employment in previous year. Columns for ‘persons at HEI’ related to individuals working at HEIs. That is, the same person working at more than one institution within the reporting year will be counted more than once. Where the same person at the same institution is reported with different previous employments or destinations, the category nearest ‘at the same institution’ as ordered in the tables is taken.

239

Table A.14 Comparison of HESA Individualised Staff Record for 2001/2 using contract counts and persons at institutions – Destinations

Number of staff records % of staff records % of staff records reported as leaving the institution (or unknown)

Previous employment Contracts Persons at

HEI Contracts Persons at HEI Contracts Persons at

HEI At same institution 96,031 94,418 87% 87% n/a n/a

At another UK HEI 1,606 1,594 1% 1% 11% 12%

Other 4,921 4,816 4% 4% 35% 35% Unknown 7,536 7,323 7% 7% 54% 53% Total 110,094 108,151 100% 100% n/a n/a

Source: Unpublished Internal work by HEFCE Columns for ‘contracts’ relate to counts of staff contracts as used in this report – final column relates to recorded employment in previous year. Columns for ‘persons at HEI’ related to individuals working at HEIs. That is, the same person working at more than one institution within the reporting year will be counted more than once. Where the same person at the same institution is reported with different previous employments or destinations, the category nearest ‘at the same institution’ as ordered in the tables is taken.

The primary employment function of a member of staff relates to the contract of employment and not the actual work undertaken. Therefore, ‘teaching only’ staff are those whose contract of employment states that they are employed only to undertake teaching, ‘teaching and research’ staff are those whose contract of employment states that they are employed to undertake teaching and research; and ‘research only’ staff those whose contract of employment states they are employed only to undertake research.

The terms of employment describe the type of appointment a member of staff has with the institution at the date the data are returned to HESA or date of leaving if earlier. Permanent staff are those who are employed on a contract of employment that states the member of staff as permanent or on an open-ended contract. This includes term-time only staff who are employed on an open-ended contract. Fixed-term contract staff are those employed for a fixed period of time or have an end date on their contract of employment. This includes staff on rolling fixed-term contracts. Hourly paid/casual staff are those employed casually by the institution, not on a permanent contract of employment. This includes temporary staff, staff employed on irregular or occasional contracts and those paid on an hourly basis.

The grade structure indicates a staff member’s grade for their present employment. Groups of grades have been devised with regard to the different grading scales used within different institutions (Earl, 2001). Grades have not, however, been linked to salary information. ‘Professors’ includes heads of departments, professors, former UAP scale researchers (grade IV), clinical professors and those appointed professors on a locally determined scale. ‘Senior lecturers & researchers’ includes principal lecturers, senior lecturers (former UAP/CSCFC scales), former UAP scale researchers (grade III), clinical senior lecturers and those appointed senior or principal lecturers on a locally determined scale. ‘Lecturers’ includes lecturers, senior lecturers

240

(former PCEF scale), clinical lecturers and those appointed lecturers on a locally determined scale. ‘Researchers’ includes all research grades not listed above and those researchers appointed on a locally determined scale.

The Nationality field defines the country of legal nationality. Where a member of staff has dual nationality and one of the nationalities is UK, they are coded as UK. Where neither of the nationalities is UK, for coding purposes preference is given to those nationalities within the European Union.

Because of data protection issues, the information is provided in the following bands: £5000 & under, £5001 to £10000, £10001 to £15000, £15001 to £20000, £20001 to £25000, £25001 to £30000, £30001 to £35000, £35001 to £40000, £40001 to £45000, £45001 to £50000, and over £50000.

C.2 HESA Student Data The data for our analysis of research postgraduates come from the Combined

Student/Module for the 2001/2 academic year. ‘Research postgraduates’ is defined as those students undertaking study for a doctorate or Masters degree by research. This was combined with the First Destination Supplement (FDS), relating to those students who left in 2001/2. The analysis was not limited to students from English higher education institutions because the appropriate pool of domestic entrants into Higher Education academic post is the whole of the UK.

Note that the problem from which the FDS suffers with respect to non-response affects students reading for postgraduate research more than undergraduates. This is because the FDS has only included qualifiers from full-time study, and the majority of PhDs are completed whilst in ‘writing-up’ mode. The FDS only accounts for about 38% of the qualifiers.

241

243

Appendix D: National and international comparisons

D.1 Data The data for this study come from a mixture of surveys and censuses. The

primary data issue is how to identify higher education academics. This was done using information on occupation and industry where available. In countries where it was possible to classify data by both these dimensions, individuals were considered to be higher education academics if they were in an appropriate occupation (e.g. lecturer or researcher) and working in the higher education sector. Otherwise, individuals were identified by their occupation only. For some the occupational classification could only be obtained at a level that precluded the separation of higher education academics from individuals who worked more broadly in the tertiary sector (e.g. Denmark). These are noted below.

Another issue to consider is whether to examine earnings net or gross of taxation. One would prefer all of the data to be net of taxation so we could compare take-home pay. Unfortunately, the data for the majority of countries was only available on a gross basis and so in what follows we compare gross annual salaries. We do provide the UK figures in gross and net form to allow comparison to the French data, which is only available net of direct taxation.

Unfortunately, we could not study some of the European countries that we would wish to because of data considerations. The EU legislation instructing member states to run labour force surveys (Council Regulation (EC) No 577/98) does not stipulate income as information that is required. Therefore, although all EU member states have labour force surveys, their coverage of income is patchy. In Germany, there are two potential surveys, the Mikrozensus and the EVS. The Mikrozensus does collect information on income and occupation. However, the income information is not considered reliable, as one third of all interviews are completed by proxy. The EVS is a survey of households’ incomes and expenditures and so income is total household income and not individual earnings. Moreover, the EVS does not contain occupational information with which to identify academics. The Italian and Spanish labour force surveys do not collect information on earnings and so are of no use to our study.

D.1.1 UK The UK data come from the four quarters of the 2001 Labour Force Survey

(LFS). Academics are defined using the three-digit 1992 Standard Industrial Classification (SIC) and the three-digit 2000 Standard Occupational Classification (SOC). Academics are defined as being in SOC codes 2311, 2319, 2321, 2322 and 2329. The main SOC unit group for academic staff is 2311, which refers to ‘higher education teaching professionals’. Other teaching staff from SOC unit group 2319 are defined as ‘teaching professionals not elsewhere specified’ if they were also coded as working in SIC class 80.3 ‘higher education’. Also included as academics were

244

individuals who were in SOC minor group 232 ‘research professionals, i.e. unit groups 2321 ‘scientific researchers’, 2322 ‘social science researchers’ and 2329 ‘researchers not elsewhere specified’ if they were also coded as working in SIC class 80.3 ‘higher education’. The two measures of salaries used are gross weekly wages (variable grsswk) and net weekly wages (netwk).

In order to determine the accuracy of our data we compare the mean wages in the LFS and HESA staff data. Note that we do not have the actual salaries of staff in the HESA Staff Record, but a series of bands: a lower band of £5,000 and under, and £5000 bands between £5,001 and £50,000 and an upper band of over £50,000. Therefore, we take the mean point of each band and replace those in the upper band with £50,000. It is clear from this comparison that the two means are very similar both in total and for men and women individually.

Table A.15 Comparison of mean wages in LFS and HESA data

Mean wages LFS HESA Men £31,919 £32,526 Women £26,112 £28,552 Total £29,409 £31,150

D.1.2 US The US data come from the Merged Outgoing Rotation Group of the 2001

Current Population Survey. The industry an individual worked in was determined using the three-digit 1980 Census of Population Industry Classification (CPIC). The appropriate CPIC code was 822 ‘Colleges and universities’. The individual’s occupation was determined using the four-digit 1980 Census of Population Occupation Classification (CPOC). The majority of academics were identified as individuals whose occupation was CPOC code 221 ‘teachers, postsecondary’ and whose CPIC industry code was 822 ‘Colleges and universities’. Unfortunately the CPOC does not have an occupational group for researchers per se. Therefore, these individuals were identified using a number of scientific occupations (both in the social and natural sciences) in combination with the CPIC code. Unfortunately it is impossible to differentiate between teachers and researchers teaching in the HE sector and the state colleges. The wage data used are ‘earnings per week’ (earnwke).

D.1.3 Australia The Australian data come from the 2001 Census of Population and Housing,

provided by the Australian Bureau of Statistics. Academics are those appearing in occupational group 2421 ‘University lecturers and tutors’. We also tried an alternative classification including other professionals who worked in the academic sector (industry group 8431), where other professionals defined as individuals in 2-digit occupation groups 21 ‘Science, building and engineering professionals’, 22 ‘Business and Information Professionals’, 23 ‘Health Professionals’, 24a ‘Other Education Professional’, and 25 ‘Social, art and miscellaneous professionals’. However, even though this group excluded managerial and associate professional occupations, the relative size of this group was two-thirds of the size of the university lecturers and tutors group and so was unlikely to reflect academics who primarily undertake

245

research. The earnings data refers to gross income (including pensions and allowances). This data is recorded in 14 bands, from AUS$1-39 to AUS$1,500 or more. When calculating statistics based on this data, the Australian Bureau of Statistics uses the mean value of each band, except the upper band which they set to AUS$1,500. This effectively sets an upper limit on earnings that does affect some of our figures (in particular the upper quartile and decile points of men’s earnings). It also means that the mean of Australian academic earnings may be understated. This does not of course affect the median or any percentile points below this limit. Note also that once we break the data down by occupation and gender we cannot also break the data down into quartiles or deciles.

D.1.4 New Zealand The New Zealand data come from the 2001 Census of Population and

Dwellings. Higher education academics in New Zealand were identified by occupation and industry. The relevant occupational group in the NZSCO99 V1.0 classification system is 23 Teaching Professionals. The relevant industry in the NZSIC96 V4.0 classification system is N8431 ‘Higher Education’. As with Australia, we also tried an alternative classification including other professionals who worked in higher education, but this was unsuccessful at identifying non-teaching research staff.

D.1.5 Canada The Canadian data come from a 20% sample of the 2001 Census, supplied by

Statistics Canada. The relevant occupational group in the 1991 Standard Occupational Classification was E111 ‘University professors’ and E112 ‘Post-Secondary Teaching and Research Assistants’. The former are exclusively in the HE sector, although the latter, which includes post-docs and teaching assistants, does include some teaching and research assistants in the FE sector. However, it does not include the primary teaching staff in the Canadian HE sector. These are included in group E12 ‘Colleges and other vocational instructors’. The salary data refer to gross employment income.

D.1.6 Denmark The Danish data come from Statistics Denmark’s StatBank data archive. The

occupational group used was 231 ‘College, university and higher education teaching professionals’. This occupational group will include college teaching professionals outside the HE sector. The earnings data refer to gross income and come ultimately from public sector’s computerised pay transfer systems. Because this is a census rather than a survey, the data refer to all 13,818 academics in the Danish HE sector.

D.1.7 France The French data come from the Enquête Emploi 2001. Academics in higher

education were identified by a combination of industry and occupation. The relevant industry group in the NAF classification was 803Z ’enseignement supérieur’ (Higher education). The relevant occupational groups were 3411 ‘Professeurs arégés et certifiés’ (aggregate and certified Professors), 3415 ‘Enseignants de l’enseignement supérieur’ (teachers of higher education) and 3421 ‘Chercheurs de la recherché publique’ (Researchers of public research). Teachers in 3411 ‘Enseignants de l’enseignement supérieur’ generally teach at secondary schools and colleges, but

246

those with agrégés and certifiés certification may also teach in HE. Around one in five of such teachers (in 3411) teach in HE. They make up around 30% of our sample (around 60% are 3415). The data do not include people working in FE, which come under NAF industry classification 804C ‘formation des adultes et formation continue’ (lit. ‘training of the adults and continuous training’). Income information is only available net of taxation. Therefore whenever information on French earnings is presented we always provide net as well as gross income for the UK.

D.1.8 Sweden The Swedish data come from Statistics Sweden’s StatBank. Higher education

academics were identified as occupation 231 ‘College, university and higher education teaching professionals’. Unfortunately, we were unable to narrow down this classification any further.

D.1.9 Japan We also have a figure for mean academic wages in Japan, which we include

for comparison. This comes from the website of the Statistics Bureau of the Japanese Ministry of Public Management, Home Affairs, Posts and Telecommunications.

D.1.10 Additional tables Table A.16 and Table A.17 present the data underlying the percentile and

decile charts in the text.

247

Table A.16 Academic real earnings quartiles

1st quartile 2nd quartile 3rd quartile UK (gross) Total 26,000 30,992 37,180 Male 26,104 32,370 39,403 Female 25,012 29,016 32,500 US Total 24,074 35,054 49,097 Male 26,628 38,555 52,599 Female 21,010 28,050 41,364 Australia Total 26,669 32,754 38,354 Male 27,973 34,339 38,354 Female - - - NZ Total 20,018 25,726 31,023 Male 21,998 27,583 35,033 Female 18,044 23,478 28,659

Total 21,274 28,449 35,456 Men 22,942 29,196 37,542

Canada (professors)

Women 18,773 25,031 32,536 Total 8,915 13,882 18,774 Men 10,712 14,598 20,854

Canada (assistants)

Women 8,541 13,344 17,081 Sweden Total 16,962 21,329 26,283 Male 16,895 22,840 28,550 Female 16,374 20,153 23,176 UK (net) Total 17,446 21,580 25,220 Male 17,992 22,308 26,936 Female 16,523 20,280 23,400 France Total 15,868 19,835 24,596 Men 17,190 21,158 26,447 Women 14,943 18,513 19,990

Real salaries converted using PPP rates from OECD Main Economic Indicators * Because the wages for Australia are based on a banded variable, they have an upper limit of

£38,354. Figures for female Australian academics are unavailable once we break them down by gender and

quartile’

248

Table A.17 Academic real earnings deciles

1st decile

2nd decile

3rd decile

4th decile

5th decile

6th decile

7th decile

8th decile

9th decile

UK Total 21,788 25,012 26,988 29,016 30,992 32,978 35,984 39,000 44,980 Male 22,350 25,012 27,581 30,524 32,370 35,256 38,376 43,004 50,123 Female 21,611 24,024 26,109 27,976 29,016 30,992 31,980 34,788 38,111

US Total 15,251 21,886 25,862 30,385 35,054 39,978 45,595 52,599 68,648 Male 17,509 24,549 29,436 33,668 38,555 42,925 49,097 56,101 70,910 Female 14,007 19,624 22,433 25,533 28,050 33,668 37,169 44,866 54,714

Australia Total 20,992 25,313 27,870 30,325 32,754 35,209 37,638 38,354 38,354 Male 22,399 26,694 29,251 31,782 34,339 36,896 38,354 38,354 38,354 Female 0 0 0 0 0 0 0 0 0

Canada professors Total 14,012 19,606 23,356 25,031 28,449 30,868 33,373 37,538 41,715 Men 14,022 20,858 24,476 26,281 29,196 31,289 35,876 38,792 41,718 Women 13,665 16,854 20,858 23,773 25,031 29,196 31,285 33,430 36,223

Canada assistants Total 6,672 8,343 10,715 12,511 13,882 15,183 17,461 20,854 25,030 Men 6,673 8,911 12,511 12,598 14,598 15,574 18,773 21,025 25,031 Women 6,672 7,592 8,947 11,677 13,344 14,602 15,853 19,192 24,129

New Zealand Total 14,039 18,776 21,261 23,607 25,726 27,844 29,963 32,604 41,516 Male 16,786 20,580 23,269 25,426 27,583 29,741 31,898 38,210 44,563 Female 11,988 16,198 19,202 21,353 23,478 25,551 27,623 29,695 31,767

France Total 11,240 15,247 17,190 18,513 19,835 21,158 23,802 25,733 30,414 Male 11,240 15,868 18,182 19,835 21,158 23,802 24,992 27,505 33,059 Female 10,579 13,885 15,604 17,164 18,513 19,174 19,835 21,158 25,786

UK (net) Total 14,352 16,796 17,992 19,760 21,580 22,776 24,128 26,416 30,004 Male 14,404 17,295 18,616 21,029 22,308 24,024 25,688 28,610 33,020 Female 14,144 15,756 17,420 17,992 20,280 21,580 22,256 24,128 26,416

Real salaries converted using PPP rates from OECD Main Economic Indicators * Because the wages for Australia are based on a banded variable, they have an upper limit of

£38,354. Figures for female Australian academics are unavailable once we break them down by gender and

decile

249

Appendix E: Intra-UK earnings and UK/US academic earnings

In this appendix we set out our econometric analysis of intra-UK earnings and UK/US academic earnings.

E.1 Background In section 4.3 we saw how the earnings of academics in higher education

across the world vary. However, the analysis has taken little account of differences in the composition of academic staff between countries, which may affect earnings. It is important to take into account many of the other factors that may vary in order truly to compare like with like. We have already seen, for example, how the proportions of part-time staff vary between the UK and the US. If we rely solely on the figure for the whole of academia we would obtain an altogether too sanguine view of the wages of UK academics relative to their US counterparts, particularly for males.

In this appendix we employ an econometric method to test whether there is a statistical difference in earnings once differences in education, gender and ethnicity have been controlled. The analysis addresses two questions. It pursues the international theme, to explore the extent of the earnings differential between the US and the UK, once differences in characteristics have been taken into account. It also turns to the domestic labour market to explore academic pay relative to that in the rest of the UK economy.

E.2 Empirical Model The data once more come from the UK Labour Force Survey Spring 2001 to

Winter 2001 and the US Current Population Survey 20001. In order to remove any potential bias to our results, we focus our analysis on the graduate sector of the economy (i.e. on those individuals with at least a degree) since this sector of the labour market reflects the choices and opportunities open to potential and actual academics.

In order to account for the differences in academic and non-academic earnings, we model the log of weekly wages (w) as dependent on a polynomial in age135, whether the individual only has a first degree (i.e. the baseline is a higher

135 Many studies use a quadratic age/potential experience term. However, given the results of Murphy and Welch, it would seem appropriate to include higher-order age terms in our specification to avoid the bias they report. According to Murphy and Welch (1990) a quadratic earnings function will underestimate the rise in earnings in the early years and overestimate it in the middle years of an individual’s working life, with a slight overestimate of the decline in earnings over the final years. After some experimentation we settled on the cubic as our preferred specification for the UK and quadratic for the US.

250

degree) (degree) and a vector of control variables (C) to account for differences between men and women, different degree subjects136, ethnic groups etc137, that is:

( ) ik

kkCdegreeageageagesw εβαααααα +++++++= ∑33

232

222110ln (1)

It may be the case that there are important differences between the structure of lifetime earnings of men and women, or academics and non-academics, those with degrees compared to those with higher degrees. Therefore, we include interaction terms in our initial specification to account for this. In the results presented below we have excluded those additional interaction terms that were not found to be statistically significant. The full set of variables used in the analysis are set out in Table A.18

136 Only around half of the UK respondents report their degree subject, so those who did not report were chosen as the baseline. 137 For the UK estimation we also include a dummy variable for each of the first three quarters because the data are made up of four appended quarters of data to ensure a large enough sample size. Note that there are no repeat observations because income questions are only asked to individuals in waves 1 and 5.

251

Table A.18 Variables used in empirical analysis

Variable name academic = 1 if HE academic, = 0 otherwise female = 1 if female , = 0 otherwise degree = 1 if only has degree , = 0 otherwise pt = 1 if job is part-time , = 0 otherwise nonwhite = 1 if non-white ethnicity, = 0 otherwise married = 1 if married, = 0 otherwise children number of children female×variable female interacted with variable academic×variable academic interacted with variable degree×variable degree interacted with variable Subject of degree (UK) or teaching (US) sg_medic =1 if science sg_sci =1 if medical or medical related sg_eng =1 if engineering sg_socb =1 if social science or business sg_lang =1 if languages sg_human =1 if humanities sg_arts =1 if arts sg_educ =1 if education UK-only variables healthpr =1 if has health problem q1 =1 if sampled in Spring 2001 survey q2 =1 if sampled in Summer 2001 survey q3 =1 if sampled in Autumn 2001 survey

E.3 Results

E.3.1 UK The results of the estimation of our model in the UK are presented in Table

A.19. We can see that wages increase over an individual’s lifetime, although the effect is non-linear – they do so at a decreasing rate and tail off at the end of working life. This can be seen more clearly when we consider the predicted lifetime earnings profiles in Chapter 4 (Figure 4.8 on page 79). People remaining in academia have a different lifetime earnings profile to graduates working outside academia. Academic earnings start from a lower point, but increase faster over the span of an academic career.

252

Table A.19 UK Results

Coef. s.e. academic -0.485*** 0.109 female×academic 0.083 0.054 age 0.262*** 0.017 age2 -0.005*** 0.000 age3/1000 0.027*** 0.003 female×age -0.008*** 0.001 academic×age 0.010*** 0.002 pt -1.021*** 0.017 academic×pt -0.102 0.066 female 0.121 0.101 nonwhite -0.079*** 0.021 degree -0.089*** 0.012 sg_medic 0.246*** 0.061 sg_sci 0.230*** 0.059 sg_eng 0.225*** 0.059 sg_socb 0.265*** 0.059 sg_lang 0.105 0.068 sg_human 0.008 0.065 sg_arts 0.031 0.066 sg_educ 0.064 0.065 female×medic 0.029 0.098 female×sci -0.027 0.098 female×eng -0.077 0.105 female×socb -0.061 0.096 female×lang 0.042 0.104 female×human 0.117 0.103 female×arts -0.037 0.104 female×educ 0.100 0.101 q1 -0.042*** 0.015 q2 -0.024* 0.015 q3 0.004 0.014 Constant 1.845*** 0.226 Observations 10,269 F32 257.29*** Adjusted R2 0.436

• * significant at 10%; ** significant at 5%; *** significant at 1%

Once we account for age, qualifications and ethnicity, UK academics earn less than non-academics, ceteris paribus. However, one has to be careful to interpret the significant negative coefficient on academic in the wage equation as evidence of a ceteris paribus earnings gap on its own because of the non-linear nature of the

253

equation (i.e. because of the inclusion of the academic×age term)138. In fact, the wage an individual can earn as an academic will be below that which they could earn elsewhere until they reach their mid-fifties when the position is reversed. The experiences of men and women working in higher education are different. Female academics do not fare well relatively to their male colleagues within academia. However, women’s wages generally fall relative to men’s over their working lives. This may be due to the effect of periods out of the labour force on experience (i.e. on the number of years service they have) or because there is inequality in promotion behaviour.

E.3.2 US The results of our estimation for the US are presented in Table A.20. The

earnings profile of US academics is similar to those in the UK, although it is flatter. The earnings of US academics increase over their working lives (see Figure 4.9 in the main text). However, unlike in the UK, US academic wages reach a plateau at the end of their career. The average US academic continues to earn more over their entire working life, although the amount their earnings increase slows over their career.

138 Note also that we are saying that academics earn less than non-academics not that the return to academia is negative. It may be that there are other unobserved differences between academics and non-academics.

254

Table A.20 US Results

Coef. s.e. academic -0.818*** 0.078 female×academic 0.132*** 0.037 age 0.067*** 0.003 age2 -0.001*** 0.000 female×age -0.004*** 0.001 academic×age 0.014*** 0.002 pt -0.322*** 0.014 female×pt -0.242*** 0.017 academic×pt -0.283*** 0.041 female -0.079*** 0.029 nonwhite -0.139*** 0.012 female×nonwhite 0.112*** 0.017 degree -0.180*** 0.009 female×degree -0.058*** 0.013 sg_medic 0.029 0.047 sg_sci 0.147*** 0.010 sg_eng 0.260*** 0.011 sg_socb 0.139** 0.063 sg_lang -0.021 0.064 sg_human 0.045 0.085 sg_arts -0.155* 0.085 sg_educ 0.030 0.132 Constant 5.551*** 0.056 Observations 39,187 F22 501.16 Adjusted R2 0.219

• * significant at 10%; ** significant at 5%; *** significant at 1%

255

Appendix F: The likelihood of students entering academia

In this section we will investigate the likelihood of our research postgraduates seeking academic employment. We do this by using their responses to the question ‘What best describes your career intentions?’ From the responses, set out in Table 5.25, we create a categorical variable (y) with the career intentions ordered to reflect the strength of the preference for an academic career. Because it is extremely difficult to account for people who do not know or have not considered the issue in decision models, in our primary model we concern ourselves only with the individuals who have some idea of their career intentions (i.e. the first five rows in Table 5.25)139. The categories in our dependent variable are as follows:

Category Career intentions % of sub sample*

% of full sample*

y = 0 ‘You'd prefer a career outside academia (and an academic job first is not important)’ 21.4 19.2

y = 1

‘You are aiming for a career outside academia but an academic job would help in this career’ and ‘You may go in for academia because you cannot think of other careers you want to have’

15.7 14.1

y = 2 ‘A number of careers are of equal interest with academia’ 22.6 20.2

y = 3 ‘You are keen to have an academic career’ 40.4 36.2 • *Unweighted • Career intentions come from question tabulated in Table 5.25

As a check on our results, we will estimate a second model, which does include a further category for those who do not have any idea of their career intentions or who have not considered the issue. In this case the categories of our dependent variable are as follows:

139 These individuals represent 92 per cent of the weighted sample, and 90 per cent of the unweighted sample.

256

Category Career intentions % of sample*

y = 0 ‘You don't know what you want to do’ and ‘You are not thinking about careers’ 10.4

y = 1 ‘You'd prefer a career outside academia (and an academic job first is not important)’ 19.2

y = 2

‘You are aiming for a career outside academia but an academic job would help in this career’ and ‘You may go in for academia because you cannot think of other careers you want to have

14.1

y = 3 ‘A number of careers are of equal interest with academia’ 20.2

y = 5 ‘You are keen to have an academic career’ 36.2 • * Unweighted • Career intentions come from question tabulated in Table 5.25

F.1 Model Our analysis utilises an ordered probit regression model (McKelvey and

Zaviona, 1975) to estimate the relationship between a categorical ordered (or ordinal) dependent variable describing the respondent’s career intentions and a set of independent variables. The explanatory variables include a set of variables relating to personal characteristics. A number of these are interacted to account for differences in behaviour between men and women, since most respondents are entering the period of their lives when labour market choices are most affected by family considerations and these are likely to have differential impacts on men and women. It is possible that these will not be as important as they are for less-qualified individuals since the latter tend to have children at an earlier age. We also include variables to account for whether the respondent’s parents have worked in teaching or research jobs either in or out of higher education. However, it may be the case that these factors will have a stronger influence on the decision to undertake higher education and to progress to a higher degree than the decision to continue into an academic career. Nevertheless, since the respondents have already chosen to undertake postgraduate education, if parental occupation has an additional influence on whether an individual becomes an academic this will have important policy implications.

Perhaps more important than parental experience of academia at this stage of the decision process is the individual’s own prior experience of academic employment. We therefore include variables to account for whether the individual worked as a lecturer or researcher prior to undertaking their current degree and whether they have done so during it. One would expect a priori the former group to be more likely to continue into academic employment as they may be undertaking a research degree in order to top up their academic credentials. For those currently teaching or undertaking paid research, its effect on their entering into academia may go either way, it may give them a taste of the academic life or put them off it.

One important dimension of the recruitment difficulties facing institutions of higher education is the variation across subject areas. Therefore, we include subject

257

dummies in order to determine whether there are differences that could lead to recruitment difficulties in the future (the baseline subject area is social studies). Finally we include university dummies in order to pick up any unobserved heterogeneity at the institution level.

The variables included in our analysis are set out in Table A.21.

Table A.21 Variables used in empirical analysis of academic career intentions

Variable Description Female Female Children Has children aged 16 or under Num children Number of children aged 16 or under Over 30 Is aged 30 or over Non-white Non-white ethnic group EU Pass Holds an EU or EEA passport Non-UK EU Holds neither a UK nor EU/EEA passport Disabled Has disability that affects work Partner Lives with spouse or partner Caring Is providing care for any adults Debt Expects to be in debt on completion of course Female×Children ‘Female’ interacted with ‘Children’ Female× Num children ‘Female’ interacted with ‘Num children’ Female×Partner ‘Female’ interacted with ‘Partner’ Prior lecturer Has worked as lecturer in HE prior to current degree Prior researcher Has worked as researcher in HE prior to current degree Cur lecturer Has done paid work as lecturer during current degree

Cur researcher Has done paid work as researcher during current degree

Parent academic Parent worked as a academic (lecturing or research) in HE

Parent other HE Parent worked in another job in HE Parent research Parent worked as researcher outside HE Parent teacher Parent worked as teacher/lecturer outside HE

Non-PhD Is not registered for PhD (i.e. registered for Masters by research or ‘other’ degree)

First class Has first class undergraduate degree Lower Has a lower second class degree or lower DK degree class Do not know degree class

University dummies (baseline = old southern university 2) Old Sth Uni 1 Old southern University 1 New Sth Uni 1 New southern University 1 New Lon Uni New London University Old Lon Uni Old London University New Nth Uni New northern University Old Nth Uni Old northern University New Nth Uni 2 New northern University 2

Subject area dummies (baseline = Subjects allied to medicine)

258

The model is estimated by log likelihood with population weights applied. In order to assess the fit of our model to the data we include the pseudo-R2 measure due to McKelvey and Zavioma (1975). Measuring the goodness-of-fit for models with qualitative limited dependent variables is more difficult than is the case in the linear regression model. The familiar goodness-of-fit measure in the linear model, the coefficient of determination (R2) is not applicable since the estimated probabilities cannot be compared with the ‘true’ probabilities, which are unknown140. Windmeijer (1995) conducted a simulation study that compared a number of pseudo-R2 measures for a binary outcome variable with the OLS R2 measure of the underlying latent variable model. He came to the conclusion that it was the measure due to McKelvey and Zaviona (1975) which most closely approximated this ‘true’ value. Hagle and Mitchell (1992) undertook a similar study and came to the same conclusion141.

F.2 Results The results are presented in Table A.22. We estimate both models for the

whole sample of research postgraduates sampled (including a dummy variable for those who are not currently studying for a PhD) and then on a restricted sample of PhDs only. The results are fairly consistent across specifications. The goodness-of-fit is higher for the two models that exclude the people who do not know what they will be doing next. The fit is also slightly better for the specifications that concentrate on the students working for a PhD only.

Most of the personal characteristics included in the model do not appear to have a significant effect on the desire for entering academia. There are no significant differences between males and females, ceteris paribus. Any differences between men and women therefore are likely to be the result of differential rates of entry to higher degrees by research. Since there are similar numbers of men and women in our sample overall, this is likely to be due to the differential rates across subject area and so the direct effect on the likelihood of research postgraduates pursuing an academic career will be picked up in the subject dummies. The effect of having children appears to be negligible. Students who are over 30 years of age have a slightly stronger intention to enter academia, although this effect is only significant at the 10% level in the first specification in the table. Again students from non-white ethnic backgrounds are slightly more likely to enter academia, but this effect is only significant at the 10% level in the second specification.

Students with an EU passport from outside the UK express a stronger desire to progress from their course into academic employment. This result is consistent across specifications.

Turning to prior experience of working in higher education, those who have worked as a lecturer in HE prior to current degree appear more likely to enter academia. This result is statistically significant at the 1% level and is consistent across specifications. Since the survey is of research postgraduates and not entrants to academic employment, it is likely that this effect represents academics returning to the sector and those already working in the sector who feel that they need to top-up their qualifications in order to get an academic job, or a better one. Doing paid work

140 Windmeijer (1995) 141 For more on this topic see Long (1997)

259

as lecturer or researcher during their current degree appears to have a positive effect on students’ desire for an academic career. Indeed those who do so are more likely to enter academia. This result is also statistically significant at the 1% level and is consistent across specifications. Having one or more parents who have worked in higher education appears not to have any effect on the likelihood of entering academia. Any such influence may have its impact earlier on in the process, affecting the likelihood of attending university or choosing to undertake postgraduate study. The only significant effect of parental occupation appears to be that of having a parent that has worked as a researcher outside higher education. The cause of this is uncertain, but it may reflect an increased probability of becoming a researcher outside HE.

Once more, the effect of previous qualifications may operate through the decision to undertake a postgraduate research degree rather than the likelihood of entering academia. Certainly, the 40 per cent of research postgraduates who have first class degrees is much higher than the rest of the graduate population. Research postgraduates with first class undergraduate degrees have a stronger desire to enter academia than those with upper second class degrees (the baseline in the estimated model), as do those with lower second class degrees or lower. The reasons for these two effects are most likely different. Only ten per cent of our respondents have a lower second class degree or lower, whereas forty per cent have firsts142. It is likely that the former had to overcome considerable difficulty in order to study for a higher degree.

Students whose research is in physics and engineering and technology are less likely to wish to enter into academia (the baseline is social studies). Students in languages and other humanities are more so, although the latter effect is not significant if we include the ‘don’t knows’.

142 Note that for four per cent of the sample, we do not know the degree classification of their first degree (i.e. dkdeg=1)

260

Table A.22 Ordered probit results: Academic career intentions

Main specification Including don't knows Full sample PhDs only Full sample PhDs only

-0.358 -0.399 -0.034 -0.231 Female (0.426) (0.453) (0.410) (0.413) 0.051 0.101 0.241 0.458* Children (0.246) (0.264) (0.248) (0.249) 0.012 0.023 0.036 -0.000 Num children (0.100) (0.112) (0.097) (0.102) 0.167* 0.139 -0.029 -0.032 Over 30 (0.089) (0.093) (0.085) (0.089) 0.144 0.167* 0.113 0.105 Non-white (0.089) (0.095) (0.084) (0.092) 0.197* 0.215** 0.237** 0.244** EU Pass (0.104) (0.109) (0.095) (0.101) 0.150 0.081 0.204* 0.123 non-UK EU (0.122) (0.128) (0.107) (0.115) 0.267 0.433* 0.010 0.271 Disabled (0.226) (0.254) (0.211) (0.238) 0.115 0.081 0.067 0.015 Partner (0.109) (0.114) (0.101) (0.105) 0.009 0.013 0.000 -0.034 Caring (0.141) (0.152) (0.143) (0.155)

-0.156** -0.101 -0.088 -0.070 Debt (0.070) (0.074) (0.064) (0.068) 0.197 0.176 0.034 0.104 Female ×

Num children (0.201) (0.219) (0.202) (0.206) 0.199 0.194 -0.053 0.121 Female ×

Children (0.422) (0.451) (0.406) (0.410) -0.039 -0.010 0.002 0.010 Female ×

Partner (0.145) (0.153) (0.135) (0.143) 0.381*** 0.333** 0.314** 0.293** Prior lecturer (0.128) (0.132) (0.122) (0.126) -0.049 -0.001 -0.032 -0.021 Prior researcher (0.105) (0.111) (0.094) (0.100) 0.150** 0.137* 0.202*** 0.238*** Cur lecturer (0.072) (0.076) (0.066) (0.070)

0.251*** 0.183** 0.218*** 0.199*** Cur researcher (0.082) (0.086) (0.074) (0.077) 0.003 0.021 0.031 0.029 Parent academic (0.111) (0.117) (0.096) (0.104) 0.134 0.212 0.104 0.169 Parent other HE (0.159) (0.176) (0.145) (0.160) -0.289* -0.357** -0.136 -0.227 Parent research (0.163) (0.175) (0.132) (0.144) 0.045 0.042 0.017 0.004 Parent teacher (0.083) (0.089) (0.074) (0.080)

• Robust standard errors in parentheses • * significant at 10%; ** significant at 5%; *** significant at 1%

• Weighted using weight

(Table continues overleaf)

261

Table A.22 Ordered probit results: Academic career intentions(continued)


0.037 -0.061 Non-PhD (0.108) (0.102)

0.210*** 0.180** 0.251*** 0.215*** First class (0.074) (0.079) (0.068) (0.073) 0.212* 0.216* 0.251** 0.247** Lower (0.118) (0.126) (0.110) (0.117) 0.272* 0.187 0.133 -0.002 DK degree class (0.158) (0.171) (0.169) (0.187) 0.099 0.057 0.054 0.148 Medicine and

dentistry (0.164) (0.183) (0.153) (0.168) -0.113 -0.119 -0.071 -0.041 S SA Med (0.153) (0.169) (0.138) (0.152) -0.155 -0.234 -0.113 -0.001 Biological sci +

Agriculture (0.310) (0.398) (0.293) (0.344) -0.355** -0.395*** -0.355*** -0.349*** Physical sciences (0.141) (0.150) (0.123) (0.132) 0.152 0.169 0.032 -0.096 Mathematical

sciences (0.302) (0.390) (0.287) (0.337) -0.066 -0.065 -0.017 -0.031 Computing

sciences (0.212) (0.219) (0.190) (0.201) -0.433*** -0.514*** -0.359*** -0.405*** Engineering & oth

technlonogy (0.137) (0.150) (0.121) (0.135) 0.402 0.237 0.427 0.274 Business and

admin. studies (0.307) (0.313) (0.296) (0.306) 0.464** 0.595*** 0.269 0.415* Languages (0.191) (0.219) (0.186) (0.216) 0.410** 0.391** 0.187 0.190 Other humanities (0.172) (0.184) (0.158) (0.170) -0.017 -0.138 -0.109 -0.177 Art & design +

Education (0.272) (0.292) (0.253) (0.268) -0.051 -0.034 0.048 0.063 Combined studies (0.184) (0.197) (0.162) (0.180) -0.025 -0.066 0.093 0.052 Old Sth Uni 1 (0.123) (0.126) (0.108) (0.111) 0.156 0.273 0.210 0.275 New Sth Uni 1 (0.252) (0.266) (0.275) (0.301) 0.959** 1.177*** 0.489 0.569 New Lon Uni (0.380) (0.445) (0.427) (0.487) 0.135 0.031 0.147 0.119 Old Lon Uni (0.145) (0.152) (0.133) (0.139) 0.196 0.287 0.339 0.403* New Nth Uni (0.235) (0.247) (0.210) (0.232) 0.067 0.047 0.053 0.043 Old Nth Uni (0.102) (0.103) (0.092) (0.095) -0.024 -0.024 -0.276 -0.252 New Nth Uni 2 (0.190) (0.195) (0.199) (0.209)

262

Table A.22 Ordered probit results: Academic career intentions (continued)


-0.632** -0.764** -1.279*** -1.535*** τ1 (y=1) (0.302) (0.322) (0.289) (0.300) -0.143 -0.282 -0.528* -0.761** τ2 (y=2) (0.301) (0.322) (0.289) (0.300) 0.486 0.343 -0.137 -0.369 τ3 (y=3) (0.302) (0.322) (0.289) (0.300)

0.418 0.186 τ4 (y=4) (0.289) (0.301) Observations 1,189 1,058 1,326 1,175 LPL -1476.9 -1309.4 -1925.2 -1691.1 M&Z R2 0.212 0.220 0.166 0.186

• Robust standard errors in parentheses • * significant at 10%; ** significant at 5%; *** significant at 1%

• LPL = Log pseudo-likelihood • Weighted using weight

263

Appendix G: Job satisfaction and intentions to leave academia

In order to understand and analyse the issues surrounding the retention of academic staff, we use a structural model of job satisfaction and intentions to quit based on models in the job satisfaction (e.g. Clark and Oswald, 1996; Clark, 1997, 2001; Ward and Sloane, 2000; Lydon and Chevalier, 2002) and labour turnover/quit behaviour (e.g. Jovanovic, 1979; Farber, 1999; Gibbons and Waldman, 1999) in the literature. Conventional models of intentions to quit and turnover behaviour (e.g. Brewer, 1996) concentrate on the pecuniary aspects of the current job and potential alternatives, e.g. current salary. Whilst they acknowledge the importance of non-pecuniary factors, the datasets used for analysis seldom contain information on these. If an attempt is made to model the non-pecuniary aspects of a job, it is done so via a limited set of job characteristics, such as total hours of work. Most studies of job satisfaction do not assess the implication of their results for actual worker behaviour, i.e. the propensity to quit. In this study we are able to combine the two in order to examine the influences on the various elements of job satisfaction and the effect of job satisfaction on intentions to quit.

An important but often-ignored fact about self-reported satisfaction is that it is conditional on the current job having been chosen previously on the basis of expectations about unknown elements of the job and then experienced. Whilst the pay and other financial aspects of employment (e.g. pensions) are often known with certainty before a contract is made, many of the non-pecuniary aspects (such as how well one will get on with one’s colleagues) are not. Once one understands this, the fact that many studies report high levels of satisfaction is no longer a mystery. People choose jobs that they wish to undertake, given their expectation of pecuniary and non-pecuniary aspects of the job. High levels of reported job-satisfaction do not necessarily mean that most jobs are inherently satisfactory, but rather that most people’s expectations are fulfilled; and that those for whom they are not are likely to shift jobs in order to find a better match. Also, if the nature of the job changes, such as pressure to increase the number of hours undertaken or how they are spent (i.e. the relative balance of the more and less pleasurable aspects of the job) job satisfaction will fall and the likelihood of an individual leaving will increase in just the same way as it would if wages were cut or grew at a slower than expected rate. Another type of change that might affect job satisfaction and/or the likelihood of an individual leaving are changes in an individual’s circumstances. The arrival of children may mean that staff reassess the balance between the pecuniary and non-pecuniary aspects of the job. Pleasure in the job itself may be less important and salary more so, when there are more mouths to feed.

We outline our general model in section G.1. We investigate the determinants of the job satisfaction of academics in section G.2 and the factors influencing the reported likelihood of leaving UK higher education in section G.3. Further discussion of the results is included in the text of chapters 6, 7 and 9.

264

G.1 The General Model Our model has essentially two stages: in the first stage we consider the

determinants of academics’ job satisfaction; the less satisfied an individual is with their job, the more likely they are to consider leaving academia.

In economic terms, the utility a job provides is a product of a number of factors, the most obvious of which is the wage it pays, others include the hours of work the job-holder must undertake, the environment he or she must work in (both in terms of the physical and social environment) and the longer-term prospects it offers in terms of job security and possibilities for promotion.

Consider the following static utility function

(1) ( )zxhwuUv ,,,+=

where U is utility relating to non-work aspects of life, w represents earnings (including other non-wage financial benefits, together these make up what we call the pecuniary benefits of a job), h represents hours worked, x represents non-pecuniary aspects of work, z represents personal characteristics likely to affect utility derived from work (such as whether the individual has a family etc). Like many studies, we assume that overall utility is separable in U and u(.)143 and concentrate our analysis on job-related utility144.

If an individual is offered a job j, they will accept the job offer if

(2) ( )[ ] ( )iijij zxhwuzxhwuE ,,,,,, 000>

where w0 and u0 represent the pecuniary and non-pecuniary benefits of the individual’s current state respectively145 and E[.] is an expectations operator. It is possible that u can vary across individuals, e.g. long hours may be less attractive for people with children. From this we can see that the worker will quit if they think there is an alternative job k such that

(3) ( )[ ] ( )ijjjikkk zxhwuqzxhwuE ,,,,,, >−

where q represents the costs associated with quitting for alternative employment.

This model is closely related to the worker-side of the classic turnover model of Jovanovic (1979), where worker-firm matches are experience goods146. In Jovanovic (1979), dissatisfaction – that is, reality being worse than one’s expectation – reveals itself in a positive relationship between the probability of separation and firm-specific tenure at low levels of tenure. As the individual learns more about the 143 This assumption is generally only implicit, with studies taking u(.) as the starting point (e.g. Clark and Oswald, 1996). 144 Note that we will allow the impact of certain aspects of the job on other utility generating activities through u(.). 145 Note that in addition to the benefits of a current job, w0 could also be the benefits received in unemployment. 146 In Jovanovic (1979), the wage offered is based on the firms’ expectation of the individual’s marginal product.

265

job, they are more likely to find aspects of it fail to meet their expectations. After this learning period, where expectations are confronted with outcomes, the separation probability will decrease with tenure as those with a higher probability of leaving do so and those with a lower probability remain (this is also because little learning takes place and there is only a small probability that the expected marginal product will decline sufficiently to cause the worker to move to a new firm). Moreover, if there is heterogeneity in the rate at which individuals find out whether their expectations are true or not, the relationship between tenure and separations will be negative. This heterogeneity can be due to heterogeneity of workers themselves or due to some random influence on the hazard rate at which the true values of u(.) for the job are revealed.

Because of the uncertainty surrounding which other jobs are available and the probability of actually obtaining them, we can generalise (3) to obtain

(4) ( ) ( )[ ]{ }qzxhwuEzxhwufP ikkkijjj −+−= ,,,,,,

That is, the probability of leaving UK HE, P, is a negative function of utility in the current job, a positive function of the expectation of utility outside UK HE, and a negative function of the costs of quitting147.

As is common with studies of satisfaction, we evaluate u(wj, hj, xj, zi) using measured job satisfaction. When it comes to the elements of E[u(wk, hk, xk, zk)], this is rather more difficult. We do have a measure of what respondents expect that they could earn if they worked outside academia, E[u(wk)]. Other elements and the effects of the costs of quitting are more difficult, and we aim to pick these up with other variables, such as subject dummies and variables to account for whether individuals have experience of work outside academia. Moreover, it is likely that individuals’ expression of satisfaction may not only be influenced by their current job, but also by their expectations of alternatives.

G.2 The Job Satisfaction of Academics Our model of job satisfaction is a generalisation of models such as Clark

(1997), Ward and Sloane (2000) and Lydon and Chevalier (2002). Our model differs because it explicitly models the links between the influence of the characteristics of a job (i.e. the elements of job satisfaction) and intentions to the UK higher education sector.

In our model, job satisfaction depends on a number of elements of the job, as outlined in Table 6.3 on page 136. Before entering employment, individuals will form an expectation of what academic and alternative jobs will yield in terms of these elements and choose the one which offers the best set of features. Following Clark and Oswald (1996), Clarke, (1997, 2001), Ward and Sloane (2000) and Lydon and Chevalier (2002), we call the job which offers the best set of features, the one that offers the highest utility. Once an individual takes up a job and experiences the true value of the elements of the job, they will remain in the job unless they think that

147 Note that in this report we are interested in the probability of leaving UK higher education rather than merely the probability of leaving the job.

266

there is an alternative job which offers a higher level of utility, once one accounts for the costs associated with searching for alternative employment and changing jobs.

We have noted that our model links models of satisfaction with Jovanovic’s (1979) classic model of job-matching and turnover, where worker-firm matches are experience goods148. In common with this model and its descendants, dissatisfaction caused by reality being worse than one’s expectation leads to a positive relationship between the probability of leaving and experience at the early stages of workers’ time in a job because they are likely to find aspects of it fail to meet their expectations149. After this learning period, where expectations are confronted with outcomes, the likelihood of leaving will decrease, since dissatisfaction is caused by a dissonance between expectations and reality. The fact that average reports of satisfaction tend to be high can be due to the fact that either: (a) individuals do have some useful knowledge of what the elements of the job are and/or (b) those who are dissatisfied with their job tend to leave. The implication is that reported satisfaction will on average tend to increase with experience, with the exception of low levels of experience, where it will increase.

G.2.1.1 The satisfaction of temporary staff

We have so far assumed that individuals will be offered a job that they find satisfactory. However, universities are not entirely certain of the productivity of potential staff. This problem is particularly acute at the beginning of an academic’s career. Indeed, academia is one of the few sections of the labour market where there is a readily accessible measure of workers’ productivity – research output. Publications in peer-reviewed journals represent an instrument whereby potential employers can form expectations of future productivity. The outcome of the research assessment exercise (RAE) on the academic labour market has been such that university departments will tend to be populated by individuals of a similar academic standard (particularly in terms of the RAE assessment criteria). The RAE has created an incentive for departments to employ staff who are as good or better than the current average in the department, in terms of research output, and to dispense with staff who do not ‘make the grade’. The outcome of this state of affairs is likely to be university departments with similar levels of productivity150.

The existence of such a metric of productivity creates a problem for individuals with low levels of academic experience, particularly new academics. This means that these may have to endure jobs with lower levels of job satisfaction than more experienced academics with similar levels of productivity, in order to obtain academic credentials. This may create a dissonance between the levels of satisfaction that they feel they ought to be experiencing and those which the market is willing to offer. This is a potential explanation for the high proportion of new academics in

148 In Jovanovic (1979), the wage offered is based on the firms’ expectation of the individual’s marginal product. 149 If an individual’s expectations turn out to be correct, they will of course not quit unless a new job comes into being that offers higher utility (after accounting for job search and moving costs). 150 Unfortunately, we do not have information on publications at the individual level, only departmental RAE scores.

267

fixed-term research posts, a group which suffers particular retention difficulties (Bett, 1999)151.

G.2.1.2 Satisfaction in a dynamic framework

Thus far we have concentrated on individuals’ short-term expectations about a job. They will also have longer-term expectations about their career paths. There are a number of reasons why they may have to reassess their situation: the nature and terms of the job may change152, they may not receive promotion or their own circumstances may change. For example, young, single academics may be willing to accept research or teaching assistant posts with a fixed term, because they place a lower value on their own time and require a lower level of consumption than those who have families. Indeed, early research posts may be seen as investments. These changes will affect the utility an individual gains from a job and thus their reported satisfaction with it and their propensity to search for alternative employment or quit for another post.

Thus it can be seen that what is important for reported satisfaction is not merely the level of the factors influencing job satisfaction – the number of hours, the amount of time spent on certain aspects (research, administration etc) – but rather changes in them. These changes also include the change that comes about when a job is sampled and an expectation is revealed.

G.2.2 Analysis We investigate the determinants of satisfaction using ordered probits. In order

to account for potential correlations between the equations, we estimate them as a system of seemingly-unrelated ordered probits (Weesie, 1999). The variables used in the analysis are outlined in Table A.23. The dependent variables are ordered categorical variable denoting satisfaction with each aspect of the job where 1 = Completely dissatisfied, 2 = Mostly dissatisfied, 3 = Somewhat dissatisfied, 4 = Neither satisfied nor dissatisfied, 5 = Somewhat satisfied, 6 = Mostly satisfied, 7 = Completely satisfied. Explanatory variables include terms for experience, staff grade, permanence of contract and hours worked. Because not all staff undertake teaching, research and administration, we include total hours and separate variables for hours of research and administration. In common with other studies of this nature (e.g. Lydon and Chevalier, 2002) we include these terms as logarithms (taking the value of zero where staff report that they do not spend any hours on them). We also include dummy variables to account for differences by subject area and university. Because only a little over half of our respondents report what they expect to earn if they worked outside higher education, we exclude this from our initial specification. We do however report results including this variable in our salary and total earnings equations in Table A.25.

151 Another explanation for universities suffering recruitment difficulties among fixed-term research staff is caused by the uncertainty of re-employment. If staff value the security that tenured positions offer, and the possibility of a permanent post following a temporary one is uncertain, it is entirely rational for individuals to increase their job search intensity as they approach the end of their contracts. 152 By job here we include other roles within the university that make up the expected career path that begins with the initial post the individual is hired to fill.

268

Table A.23 Variable used in the analysis of job satisfaction

Variable Explanation e Experience. Years employed in UK higher education e2 Experience squared GProf Grade = Professor or head of department GSLect Grade = lecturer Non-perm Not on permanent contract htotal Log of total hours of work hresearch Log of hours spent on research hadmin Log of hours spent on administration RAE5* department rated 5* in last RAE exercise RAE5 department rated 5 in last RAE exercise RAE4 department rated 4 in last RAE exercise w log of annual earnings

w* log of annual earnings would expect to earn if worked outside academia

Non-white Non-white ethnic group Female female Children Has children Female×Children Female interacted with Children Num children Number of children Female×Num children Female interacted with Num children Married Married Female×Married Female interacted with Married

University dummies (baseline = old southern university 2) Old Sth uni 1 Old southern university 1 New Sth uni 1 New southern university 1 New Lon uni New London university New Sth uni 2 New southern university 2 Old Lon uni Old London university New Nth uni New northern university New Sth uni 3 New southern university 3 Old Nth uni Old northern university New Nth uni 2 New northern university 2 Subject area dummies (baseline = Subjects allied to medicine)

The results of our estimation on the full sample are set out in Table A.24 (those for the reduced sample are presented in Table A.25). At the bottom of the table are the Log-pseudo likelihood, the likelihood ratio χ2 test of joint significance of the coefficients and an R2 goodness-of-fit measure due to McKelvey and Zaviona (1975) for the independent equations to give us an indication of the fit of the model to the data153. These indicate that there is some variation in how our equations fit the data. 153 Note that the methodology used to estimate the satisfaction ordered probits as a system involves first estimating the individual models separately and then using the variance covariance matrices to calculate robust standard errors. The goodness of fit measures are calculated based on the independent estimates and so are likely to under report the fit of the more efficient final system of equations.

269

All of them are in the range one would expect for this type of ordered categorical data. The likelihood ratio test that the regressions are informative, the joint test of significance of the coefficients of each equation is accepted at the 1% level for all equations.

First of all, we can see that our results support the earlier result that there are no significant differences between the satisfaction of men and women with many of the elements of the academic employment. It is only with respect to their salary and total earnings that women report significantly different levels of satisfaction. In both cases, women report higher levels of satisfaction (this is not the case when we reduce the sample to those who report what they feel they could earn outside academia (Table A.25) although married women are significantly more satisfied with their salary and total earnings in this case). Members of non-white ethnic minorities are less satisfied with the opportunity they have to use their own initiative, the hours they have to work and their relations with their colleagues than their white colleagues are. This may indicate one of two things – either non-white staff have higher expectations than white staff over these dimensions of the job or they find themselves in jobs where these dimensions are less satisfactory. It seems unlikely that the former is the case, and therefore suggests that they are finding themselves in less satisfactory jobs154. There is some evidence also that they are less satisfied with their salary and total earnings, although only the latter is statistically significant, and then only at the 10% level. The dissatisfaction with total earnings becomes stronger (and more significant from a statistical standpoint) when we confine our sample to those who report what they feel they could earn outside the higher education sector. The fact that non-whites feel less dissatisfied with their salary than they are with their total earnings suggests that there may be greater wage equality within higher education than there is in the areas where academics work to supplement their salaries.

As we would expect, given our discussion above, the relationship between experience and job-satisfaction (where it is statistically significant) is non-linear. Satisfaction with the non-pecuniary aspects of the job tends to decrease with experience for the first half of academics’ careers but then increases. Whether this can be seen as a pure learning effect as noted in the introduction is unclear, as the results imply that it can take ten or twenty years to learn whether one likes the job or not. In a cross-sectional study such as ours, it is difficult to distinguish between true experience and other age-related effects, which may have a part to play in explaining job-satisfaction among English academic staff.

Two elements where experience does not have a significant effect are satisfaction with salary and total earnings. This is as we would expect, as the wage structures in UK HEIs are highly structured and pay scales can be seen by job applicants before they apply for and accept job offers. The one influence on earnings that the individual can predict less well is promotions and the coefficients are significant and of the expected sign; the non-linear relationship of satisfaction decreasing over the early career and then increasing is supported by the data.

Professors and, to a lesser extent, senior lecturers are generally happier in their jobs than lower grades – particularly, and unsurprisingly, with promotions. Professors tend on the whole to be more satisfied with their jobs than senior lecturers. In

154 Our supposition is based on the lack of evidence that ethnic minority employees have higher expectations in these areas and that disadvantage in employment is common for ethnic minorities.

270

particular, professors are significantly more satisfied with their salaries and total earnings whereas senior lecturers are not significantly different from lower grade staff. One explanation for this result is that if all staff have similar levels of ability, but jobs are rationed so that only a lucky few get promoted, then those that do not get promoted will exhibit higher levels of dissatisfaction. Another extreme case is where the ability of staff varies and those that are promoted are the ones of the highest ability. In this case, if lower ability staff feel they are of the same ability as those who are promoted (or higher), they are will exhibit higher levels of dissatisfaction. The truth may lie somewhere between these two extremes. One area where senior lecturers are significantly more satisfied than professors and lower grades is with their physical work conditions. It may be the case that senior lecturers receive better offices etc than lower grades and are thus more satisfied than lower grades, but that professors do not receive an additional one and compare themselves with senior lecturers when considering their satisfaction.

We suggested in the introduction that staff on non-permanent contracts are likely to be less satisfied with elements of their job than other staff because temporary jobs reflect an investment at the beginning of an academic career, rather than a career job in itself. Our results show that staff on non-permanent contracts are significantly less satisfied with their promotion prospects and their job security. They appear to be more satisfied with the actual work itself than permanent staff, although this result is only statistically significant at the 10% level. It is interesting to note that they are more satisfied with their earnings than permanent staff. This is consistent with the idea that they are willing to sacrifice earnings for other aspects of the job at this early stage of their career.155

There are few clear patterns in the variation in satisfaction across subject areas. Academic staff working in medicine and dentistry, biological and physical sciences are more dissatisfied with their job security than those working in other areas, although those in physical sciences appear to be more satisfied with their salaries and other earnings.

155 Note that staff on non-permanent contracts are, on average, ten years younger than permanent staff.

271

Table A.24 Results – satisfaction

Actual work itself

Promotion prospects Salary Total

earnings Rel. with manager

Job security

Use own initiative Hours Rel. with

colleagues Physical work conditions

-0.112 -0.128 0.240*** 0.205** -0.067 0.008 0.087 0.021 0.090 0.058 Female (0.086) (0.083) (0.088) (0.088) (0.083) (0.083) (0.084) (0.083) (0.085) (0.082) -0.103 0.019 -0.056 -0.063 0.070 -0.028 0.038 0.007 0.106 0.049 Children (0.103) (0.099) (0.105) (0.106) (0.099) (0.100) (0.100) (0.099) (0.101) (0.098) 0.226 -0.178 -0.258 -0.239 -0.196 -0.060 0.105 0.060 -0.133 -0.028 Female× Children (0.161) (0.154) (0.164) (0.164) (0.154) (0.156) (0.156) (0.154) (0.158) (0.154) 0.058 -0.044 0.006 0.028 0.003 0.018 0.014 0.027 0.022 -0.002 Num children (0.038) (0.037) (0.039) (0.040) (0.037) (0.037) (0.037) (0.037) (0.038) (0.036) -0.035 0.083 0.049 0.039 0.042 0.002 -0.076 -0.015 0.002 0.008 Female×

Num children (0.065) (0.062) (0.066) (0.066) (0.062) (0.063) (0.063) (0.062) (0.064) (0.062) -0.049 -0.019 0.007 -0.031 -0.096 -0.078 -0.010 0.138* 0.006 0.084 Married (0.073) (0.071) (0.075) (0.075) (0.071) (0.071) (0.072) (0.071) (0.072) (0.070) 0.074 0.065 0.110 0.147 0.105 0.193* 0.011 -0.176* -0.014 -0.060 Female× Married (0.105) (0.101) (0.107) (0.108) (0.101) (0.102) (0.103) (0.102) (0.103) (0.101) -0.024 -0.038 -0.102 -0.161* -0.014 0.034 -0.267*** -0.154* -0.270*** -0.069 Non-white (0.083) (0.081) (0.085) (0.085) (0.081) (0.080) (0.081) (0.081) (0.082) (0.080) -0.019** -0.058*** -0.007 -0.002 -0.045*** -0.024*** -0.012 -0.039*** -0.026*** -0.020*** e (0.008) (0.008) (0.008) (0.008) (0.008) (0.008) (0.008) (0.008) (0.008) (0.008) 0.001** 0.001*** 0.000 -0.000 0.001*** 0.001*** 0.000 0.001*** 0.001*** 0.001*** e2 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) 0.183** 0.991*** 0.319*** 0.370*** 0.368*** 0.645*** 0.326*** 0.086 0.071 -0.029 GProf (0.080) (0.078) (0.090) (0.090) (0.077) (0.079) (0.079) (0.076) (0.078) (0.076) -0.069 0.272*** 0.084 0.069 0.076 0.169*** 0.036 -0.107* -0.092 -0.186*** GSLect (0.063) (0.061) (0.066) (0.066) (0.061) (0.061) (0.062) (0.061) (0.062) (0.061) 0.116* -0.475*** 0.356*** 0.316*** 0.022 -1.447*** -0.068 0.055 -0.079 -0.096 Non-perm (0.063) (0.061) (0.068) (0.068) (0.061) (0.065) (0.062) (0.061) (0.062) (0.061)

272

Actual work itself



Job security



0.050 -0.250*** -0.521*** -0.454*** -0.217*** -0.148** -0.149** -0.853*** -0.108* -0.107* htotal (0.060) (0.060) (0.080) (0.078) (0.061) (0.060) (0.059) (0.079) (0.059) (0.058) 0.069*** 0.035 -0.054** -0.058** 0.018 -0.015 0.067*** 0.136*** -0.001 0.018 hresearch (0.023) (0.022) (0.024) (0.024) (0.022) (0.023) (0.023) (0.023) (0.023) (0.022) -0.135*** -0.027 -0.103*** -0.109*** -0.042 -0.054** -0.072*** -0.205*** -0.111*** -0.078*** hadmin (0.028) (0.027) (0.029) (0.029) (0.027) (0.027) (0.027) (0.028) (0.027) (0.027) 0.253*** 0.191** -0.166** -0.101 0.135* 0.109 0.148* 0.088 0.011 -0.002 RAE5* (0.081) (0.077) (0.081) (0.081) (0.078) (0.078) (0.079) (0.078) (0.079) (0.077) 0.168** 0.073 -0.177** -0.141** 0.047 0.091 0.039 -0.061 -0.071 -0.073 RAE5 (0.070) (0.067) (0.071) (0.071) (0.067) (0.068) (0.068) (0.067) (0.069) (0.067) 0.062 0.031 -0.053 -0.015 -0.083 -0.144** -0.008 0.007 -0.169** -0.256*** RAE4 (0.072) (0.069) (0.073) (0.073) (0.069) (0.070) (0.070) (0.069) (0.070) (0.069) -0.001 0.003 0.911*** 0.795*** 0.005 0.010* 0.000 0.001 0.004 0.002 w (0.006) (0.006) (0.108) (0.108) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) -0.033 -0.099 -0.132 -0.149 -0.128 -0.269** 0.010 0.075 -0.095 0.050 Medicine and

dentistry (0.115) (0.110) (0.118) (0.120) (0.110) (0.111) (0.112) (0.111) (0.113) (0.109) -0.132 -0.151* 0.008 -0.029 -0.167* -0.348*** -0.050 -0.091 -0.093 -0.093 Biological

sciences (0.093) (0.089) (0.094) (0.094) (0.089) (0.090) (0.090) (0.089) (0.091) (0.088) 0.273 -0.100 -0.238 -0.314 0.085 0.096 0.208 -0.230 -0.048 -0.423** Agriculture and

related subjects (0.212) (0.205) (0.210) (0.211) (0.204) (0.209) (0.210) (0.203) (0.210) (0.200) 0.136 -0.076 0.245** 0.242** 0.001 -0.288*** 0.130 0.083 0.065 -0.057 Physical sciences (0.101) (0.096) (0.102) (0.103) (0.097) (0.097) (0.098) (0.097) (0.099) (0.096) 0.125 0.233* 0.134 0.114 0.111 0.057 0.253* 0.201 0.005 0.212 Mathematical

sciences (0.135) (0.129) (0.136) (0.136) (0.131) (0.131) (0.133) (0.131) (0.133) (0.129) 0.064 -0.055 0.191* 0.215* 0.117 0.030 0.059 -0.062 -0.200* 0.388*** Computing

sciences (0.114) (0.109) (0.115) (0.115) (0.110) (0.110) (0.111) (0.111) (0.112) (0.110) -0.131 -0.062 0.079 0.033 -0.050 -0.068 -0.045 -0.179* -0.073 0.048 Engineering (0.111) (0.108) (0.114) (0.114) (0.108) (0.108) (0.109) (0.108) (0.110) (0.107)

273

Actual work itself



Job security



0.505 -0.055 0.192 0.234 0.343 0.091 0.276 0.104 0.043 0.045 Other technology (0.362) (0.333) (0.349) (0.350) (0.341) (0.339) (0.342) (0.334) (0.344) (0.328) -0.453** -0.551*** -0.179 -0.112 -0.220 -0.249 -0.213 -0.340* -0.395** -0.267 Architecture and

planning (0.192) (0.190) (0.196) (0.201) (0.186) (0.189) (0.188) (0.188) (0.190) (0.185) -0.047 0.104 0.118 0.120 -0.110 -0.066 -0.091 -0.102 -0.221*** -0.091 Social studies (0.084) (0.081) (0.087) (0.088) (0.081) (0.082) (0.082) (0.081) (0.083) (0.081) -0.039 -0.213** -0.140 -0.093 -0.035 0.139 -0.077 -0.074 -0.243** 0.167 Business and

admin. studies (0.108) (0.104) (0.111) (0.111) (0.104) (0.104) (0.105) (0.104) (0.106) (0.104) 0.159 0.234 0.223 0.072 0.660* 0.391 0.152 0.144 -0.387 -0.073 Librarianship &

info. science (0.372) (0.347) (0.383) (0.383) (0.363) (0.362) (0.358) (0.347) (0.359) (0.349) 0.076 -0.038 0.187 0.108 -0.208 -0.271* -0.175 -0.293* -0.005 -0.210 English lit. And

classics (0.168) (0.160) (0.164) (0.166) (0.160) (0.163) (0.163) (0.160) (0.165) (0.159) 0.288** 0.183 0.021 -0.057 0.002 0.036 0.101 0.224 0.243* 0.521*** Modern

languages (0.147) (0.142) (0.156) (0.156) (0.143) (0.145) (0.144) (0.144) (0.146) (0.142) 0.046 0.139 0.141 0.132 0.011 -0.142 -0.123 -0.160* -0.196** -0.134 Other humanities (0.101) (0.096) (0.104) (0.104) (0.097) (0.098) (0.098) (0.097) (0.099) (0.096) -0.284* -0.290* -0.071 -0.152 -0.272* -0.422*** -0.383** -0.260* -0.236 -0.302* Art and design (0.159) (0.157) (0.174) (0.174) (0.155) (0.156) (0.156) (0.157) (0.159) (0.156) 0.265*** 0.064 0.029 0.048 0.149 0.164* 0.170* -0.048 0.033 0.105 Education (0.099) (0.094) (0.101) (0.102) (0.095) (0.095) (0.096) (0.094) (0.097) (0.094) 0.037 -0.078 0.055 0.105 -0.056 -0.148 0.199 0.125 0.181 0.039 Combined studies (0.124) (0.118) (0.124) (0.124) (0.119) (0.120) (0.122) (0.119) (0.123) (0.119) 0.097 -0.291*** -0.248** -0.186* 0.101 0.191* 0.284*** 0.146 -0.088 0.116 Old Sth uni 1 (0.101) (0.096) (0.099) (0.099) (0.097) (0.099) (0.099) (0.097) (0.098) (0.096) 0.322*** 0.322*** 0.244* 0.243* 0.259** 0.292** 0.266** 0.147 0.109 -0.174 New Sth uni 1 (0.124) (0.119) (0.127) (0.127) (0.121) (0.121) (0.121) (0.119) (0.122) (0.118) 0.065 -0.162* -0.061 -0.034 0.005 -0.345*** -0.088 0.095 0.008 -0.888*** New Lon uni (0.101) (0.098) (0.105) (0.105) (0.098) (0.098) (0.099) (0.098) (0.100) (0.098)

274

Actual work itself



Job security



0.231** 0.091 -0.010 0.016 0.062 0.031 0.025 -0.052 0.157* -0.163* New Sth uni 2 (0.094) (0.091) (0.097) (0.097) (0.090) (0.091) (0.092) (0.091) (0.093) (0.090) 0.120 0.045 -0.230*** -0.211*** 0.283*** -0.002 0.258*** 0.169** 0.101 -0.060 Old Lon uni (0.079) (0.075) (0.080) (0.080) (0.076) (0.076) (0.077) (0.075) (0.077) (0.075) 0.198 -0.083 0.089 0.026 0.213* -0.589*** 0.004 0.031 0.104 -0.607*** New Nth uni (0.128) (0.122) (0.136) (0.136) (0.122) (0.122) (0.123) (0.123) (0.125) (0.122) 0.157* -0.051 -0.162* -0.169* 0.230** -0.000 0.089 0.039 0.308*** -0.133 New Sth uni 3 (0.094) (0.090) (0.098) (0.098) (0.091) (0.091) (0.092) (0.091) (0.093) (0.090) -0.057 -0.046 -0.032 -0.070 0.215*** 0.239*** 0.077 0.067 0.059 -0.062 Old Nth uni (0.066) (0.064) (0.067) (0.068) (0.064) (0.065) (0.065) (0.064) (0.065) (0.063) 0.044 -0.096 0.077 0.111 -0.186 -0.519*** -0.147 0.245** -0.103 -0.421*** New Nth uni 2 (0.128) (0.123) (0.135) (0.135) (0.124) (0.124) (0.125) (0.124) (0.126) (0.123) -2.346*** -2.468*** 5.946*** 4.893*** -2.615*** -2.703*** -2.685*** -5.221*** -3.202*** -2.534*** τ1 (0.242) (0.232) (1.033) (1.033) (0.234) (0.233) (0.235) (0.295) (0.241) (0.227) -1.733*** -1.993*** 6.545*** 5.513*** -2.156*** -2.292*** -2.224*** -4.669*** -2.668*** -1.947*** τ2 (0.232) (0.231) (1.033) (1.033) (0.232) (0.231) (0.230) (0.292) (0.232) (0.225) -1.142*** -1.453*** 7.179*** 6.147*** -1.750*** -1.852*** -1.832*** -4.001*** -2.174*** -1.364*** τ3 (0.230) (0.230) (1.034) (1.034) (0.231) (0.230) (0.228) (0.290) (0.230) (0.224) -0.974*** -0.871*** 7.519*** 6.544*** -1.364*** -1.413*** -1.496*** -3.641*** -1.763*** -1.057*** τ4 (0.229) (0.229) (1.035) (1.035) (0.231) (0.230) (0.227) (0.289) (0.228) (0.224) -0.224 -0.374 8.030*** 7.032*** -0.983*** -0.909*** -0.794*** -3.187*** -1.139*** -0.539** τ5 (0.229) (0.229) (1.036) (1.035) (0.230) (0.229) (0.226) (0.288) (0.227) (0.223) 1.509*** 0.628*** 9.142*** 8.105*** 0.097 0.162 0.375* -2.104*** 0.267 0.637*** τ6 (0.230) (0.231) (1.039) (1.038) (0.230) (0.229) (0.226) (0.286) (0.227) (0.224)

Observations 2,706 2,662 2,361 2,339 2,692 2,695 2,705 2,690 2,706 2,698 LPL -3,464.2 -4,763.7 -4,257.9 -4,235.1 -4,630.7 -4,432.4 -3,997.9 -4,567.0 -3,774.3 -4,668.2 χ2 215.6*** 418.4*** 245.8*** 225.9*** 204.2*** 1082.8*** 196.3*** 708.3*** 156.3*** 94.1*** M&Z R2 0.087 0.160 0.106 0.099 0.073 0.354 0.073 0.236 0.060 0.102

• Robust standard errors in parentheses, * significant at 10%; ** significant at 5%; *** significant at 1%

275

Levels of satisfaction with many of the job dimensions among staff at university 2 (one of the southern new universities) appear to be higher than those at the other universities (recall that the baseline is one of the highly-rated southern old universities). There seems no reason to believe this was merely a locational effect, as, whilst the location might be thought to be good, other case study universities were in good locations. It, therefore, was likely to be a function of the university. In the qualitative research, the main difference between this university and the other case study universities was that many respondents praised the ethos of the university and its treatment of staff, describing it as caring, respectful of individuals and flexible over individual needs. It is this, perhaps, that led to greater satisfaction amongst its staff. The universities where satisfaction with salary is lower are located in the south of England where living costs are higher. However, this dissatisfaction is not present at all of the southern and London universities. Indeed the satisfaction at the first of the highly-rated southern old universities (and the London one) is significantly higher than the other.

There is a negative relationship between total hours worked and satisfaction with all elements except the actual work itself. This reflects the general premise that people generally prefer fewer hours of work than more. However, for those who do it, hours spent on research have a positive effect on satisfaction with the actual work itself, the ability to use one’s own initiative and with hours generally. Note that this is the effect of an hour of research, leaving the total number of hours worked unchanged. The net affect of an additional hour of research on satisfaction (i.e. accounting for both the coefficient on htotal and hresearch) is only positive with respect to the actual work itself. Hours of work spent on administration have a negative effect on satisfaction with all dimensions of academics’ job satisfaction, with the exception of their promotion prospects and relations with their managers. It may be the case that the positive benefits of administration with respect to these two factors offset the negative effects on general satisfaction. Alternatively, it may just be the case that there is no significant relationship between the amount of time spent on admin and satisfaction with promotion prospects and relations with one’s manager.

The fact that the negative effect of hours spent on research and administration is greater than extra hours generally, suggests that staff who put in extra effort in these tasks do not feel that they are adequately rewarded. The fact that the more time staff spend on research the more satisfied they are with the actual work itself is consistent with the idea that research is a non-pecuniary benefit of academic work. That is, research is one of the factors that offsets the low salaries in academic relative to alternative employment. Conversely, hours spent on administration appear to reduce satisfaction almost across the board.

Staff in five star departments tend to enjoy the work itself more than other staff (as, to a lesser extent, do staff at five-rated departments) and tend to be more satisfied with their promotion prospects, their relations with their manager and their ability to use their own initiative. This is consistent with the idea that staff wish to work at institutions of academic excellence, and the stability that the recognition of this excellence imparts is also appreciated by staff. Staff at five star departments are, however, less satisfied with their salaries (although not their total earnings). This may reflect the fact that the national pay scales in academia constrain institutions’ ability to reward the most productive workers. This conclusion is supported by the results reported in Table A.25, where the inclusion of a measure of individual staff’s own assessment of their abilities – the wage they would expect to command outside

276

academia – is included. In this case, the negative effect of RAE5* on satisfaction with salary and total earnings disappears.

Earnings are not significantly correlated with satisfaction with any of the non-pecuniary aspects of the job. This suggests that staff are able to consider the pecuniary and non-pecuniary aspects of their job independently. This cannot, however, tell us whether lower wages are traded off with other aspects of the job, because there is not enough within-academia variation in these factors. In order to answer this question, one would have to have a sample that included responses for people in non-academic jobs.

Apart from the effect of being in a five-star department noted above, the inclusion of the expected outside wage has little effect on the coefficients in the results for satisfaction with salary and total earnings except for that on current wages and experience. The constraint that the coefficients on w and w* are equal and opposite is rejected at the 1% level for both equations. We can, therefore, reject the hypothesis that it is only relative earnings that count for satisfaction with earnings. The lack of a significant effect of experience in these specifications may reflect the fact that individuals’ perception of their own worth outside UK academia diverge from their actual salaries over their career.

Table A.25 Including alternative salary

Salary Total earnings 0.085 0.054 Female (0.105) (0.105)

0.018 -0.009 Children (0.129) (0.129)

-0.286 -0.253 Female×Children (0.214) (0.214)

0.022 0.040 Num children (0.050) (0.050)

0.045 0.028 Female× Num children (0.092) (0.092)

-0.106 -0.132 Married (0.088) (0.088)

0.328** 0.352*** Female×Married (0.128) (0.128)

-0.138 -0.210** Non-white (0.102) (0.102)

-0.014 -0.001 e (0.011) (0.011)

-0.000 -0.000 e2 (0.000) (0.000)

0.509*** 0.536*** GProf (0.117) (0.117)

0.115 0.084 GSLect (0.081) (0.081)

0.294*** 0.263*** Non-perm (0.082) (0.082)

-0.587*** -0.504*** htotal (0.100) (0.095)

-0.043 -0.044 hresearch (0.030) (0.030)

277

Salary Total earnings -0.085** -0.091** hadmin (0.036) (0.036)

-0.012 0.038 RAE5* (0.097) (0.097)

-0.025 -0.003 RAE5 (0.088) (0.089)

0.083 0.120 RAE4 (0.089) (0.089)

1.758*** 1.474*** w (0.152) (0.149)

-0.992*** -0.868*** w* (0.089) (0.089)

-0.192 -0.190 Medicine and dentistry (0.145) (0.147)

-0.069 -0.101 Biological sciences (0.113) (0.113)

-0.113 -0.239 Agriculture and related subjects (0.236) (0.237)

0.325*** 0.316*** Physical sciences (0.122) (0.122)

0.216 0.241 Mathematical sciences (0.168) (0.168)

0.244* 0.293** Computing sciences (0.134) (0.134)

0.058 0.023 Engineering (0.131) (0.131)

0.611 0.708* Other technology (0.413) (0.414)

-0.416* -0.341 Architecture and planning (0.239) (0.247)

0.204* 0.168 Social studies (0.107) (0.107)

-0.026 0.015 Business and admin. studies (0.132) (0.132)

0.229 0.022 Librarianship & info. science (0.511) (0.510)

0.006 -0.066 English lit. And classics (0.210) (0.210)

-0.108 -0.154 Modern languages (0.211) (0.211)

0.145 0.164 Other humanities (0.134) (0.134)

-0.053 -0.161 Art and design (0.215) (0.216)

-0.171 -0.113 Education (0.129) (0.129)

-0.088 0.004 Combined studies (0.155) (0.154)

-0.166 -0.114 Old Sth uni 1 (0.117) (0.117)

0.338** 0.284* New Sth uni 1 (0.161) (0.161)

0.028 0.034 New Lon uni (0.132) (0.132)

0.116 0.144 New Sth uni 2 (0.118) (0.118)

278

Salary Total earnings -0.119 -0.080 Old Lon uni (0.095) (0.095)

0.061 0.017 New Nth uni (0.175) (0.175)

-0.126 -0.127 New Sth uni 3 (0.121) (0.122)

0.015 -0.017 Old Nth uni (0.080) (0.080)

0.300* 0.355** New Nth uni 2 (0.163) (0.163)

3.964*** 2.601* τ1 (1.450) (1.445)

4.616*** 3.269** τ2 (1.451) (1.445)

5.296*** 3.956*** τ3 (1.452) (1.446)

5.654*** 4.363*** τ4 (1.452) (1.446)

6.208*** 4.899*** τ5 (1.453) (1.447)

7.309*** 5.952*** τ6 (1.455) (1.448)

Observations 1,621 1,610 LPL -2,839.295 -2,847.043 χ2 326.64*** 279.23*** M&Z R2 0.204 0.175

• Standard errors in parentheses • * significant at 10%; ** significant at 5%; *** significant at 1%

G.2.3 Is there a single overall measure of satisfaction? Many studies utilise a single measure for overall job satisfaction, with the

assumption (either implicit or explicit) that the factors mentioned above are intermediate determinants. This is sometimes done by extracting the first principal component of reported satisfaction with these elements. Beyond the statistical objection to performing factor analysis on categorical variables, our discussion above suggests that this may be an overly simplistic view of job satisfaction and that there may be a number of separate and possibly orthogonal elements that need to be considered. A more appropriate methodology is to perform a factor analysis on the predicted values of latent variables assumed to underlie these categorical reported measures of satisfaction and examine how much of the total variation in satisfaction with these ten elements can be explained by the extracted factors156. In this section we perform principal component analysis of the latent variable, calculated on the covariance matrix.

Because of the reduction in sample size imposed by the inclusion of w* in our analysis, we restrict our analysis here to the predicted values obtained from the analysis outlined in Table A.24. The results of performing principal-components factor analysis on our predicted values of the ten elements of job satisfaction are

156 For more on factor analysis, see for example Harman (1976).

279

reported in Table A.26. This table reports the factor loadings, along with the eigenvalues and the proportion explained by the factors extracted with eigenvalues of more than one157.

Table A.26 Factor loadings for overall job satisfaction

Element of satisfaction Factor 1 Factor 2 Factor 3 The actual work itself 0.815 -0.044 -0.201 Promotion prospects 0.431 0.120 0.800 Salary 0.011 0.986 -0.167 Total earnings 0.015 0.986 -0.161 Relations with manager 0.874 0.051 0.139 Job security -0.045 0.181 0.917 Being able to use own initiative 0.892 0.037 0.107 The hours you work 0.801 -0.124 -0.336 Relations with colleagues 0.736 -0.014 -0.066 Physical work conditions 0.751 0.000 -0.045 Eigenvalue 4.158 2.013 1.724 Proportion explained 0.416 0.201 0.172 Cumulative 0.416 0.617 0.790

There are three factors with eigenvalues of more than one and together these explain 79 per cent of the variation in the ten latent satisfaction variables158. The first factor explains just over half of the common variance of the predicted satisfaction variables. The factor loadings suggest that this factor explains much of the variation of six of the factors: respondents’ relations with their manager, being able to use their own initiative, the hours they work, relations with colleagues and physical work conditions159. We label this factor ‘satisfaction with non-pecuniary elements of the current job’. The second factor merely reiterates what we saw in the analysis above, namely that the explanation for both salary and total earnings is highly correlated. Moreover, the analysis on the whole sample suggests that the pecuniary factors are quite separate from the non-pecuniary ones. The final factor in the analysis of the whole sample includes satisfaction with promotion prospects and job security. We call this factor ‘longer-term prospects’.160

These results suggest that there are in fact three separate sets of factors which determine the job satisfaction of academics. The most important from the viewpoint of most economists, namely earnings, is distinct from the other dimensions of the job. Satisfaction with their longer-term prospects explains almost as much of the total variation. The majority of the variation in six of the job dimensions is explained by a 157 The so called ‘Kaiser-Guttman rule’ after Guttman (1954) and Kaiser (1970). 158 In what follows, we will refer to the ten predicted latent satisfaction variables as ‘predicted satisfaction variables’ for brevity. 159 It also explains some of the satisfaction with promotion prospects. 160 Note that promotion prospects also enter into the principal ‘non-pecuniary’ factor, although with a lower factor loading than the other non-pecuniary factors.

280

single common factor, which we call ‘satisfaction with non-pecuniary elements of the current job’. It is important to note that these proportions of variance explained are not the same as weights – they do not rank the relative importance of these factors. However, the influence of these factors on the likelihood of leaving can be assessed and this is explored in the next chapter (see Section G.3.2).

G.3 Factors affecting the likelihood of leaving Higher Education

G.3.1 Introduction In this section we examine the influence of reported satisfaction, wage

differentials, experience and other characteristics on the likelihood of leaving UK higher education of academics. Our survey asked how likely staff thought it was that they would leave UK higher education161. Their responses are summarised in Table 7.5. We will investigate the influence of job satisfaction and a number of other elements on this likelihood of leaving by means of ordered probits, as used in the analysis above. The variables included in the analysis are set out in Table A.27.

Table A.27 Variables used in analysis of likelihood of leaving

Variable Explanation w log of annual earnings

w* log of annual earnings would expect to earn if worked outside academia

No w* Individual does not report w* S1 First factor of satisfaction (non-pecuniary elements) S2 Second factor of satisfaction (pecuniary elements) S3 Third factor of satisfaction (longer-term prospects) Age 55+ Aged 55 years or over Non-white Non-white ethnic group Female female Married Married Female×married female interacted with married Children Has children aged 16 or under Female× Children Female interacted with Children Num children Number of children aged 16 or under Female× Num children female interacted with children

EU Pass Holds non-UK EU or EEA passport OZNZ Holds Australian or New Zealand passport USA Holds US passport

161 Due to concern about the reliability of responses to a question on the likelihood of leaving, respondents were also asked to describe their job search activities. The relationship between responses to these questions and to the likelihood of leaving was tested and the likelihood of leaving reflected job search activity. However, analysis similar to that which follows was conducted using job search activities as the dependent variable. The results were similar (in the same direction) to those using the likelihood of leaving variable, but tended to be insignificant. Therefore, we report the analysis using the likelihood of leaving only.

281

Variable Explanation Foreign Holds other foreign passport HQ UG Highest qualification = UG degree HQ Masters T Highest qualification = Taught masters HQ Masters R Highest qualification = Masters by research Study PhD Is studying part time for PhD e Experience. Years employed in higher education e2 Experience squared Break from academia Has taken career break from academia

Non-UK HE Has worked in non-UK HE for more than one year Career change Has changed career to enter academia Prof: manager Occupation prior to HE: Manager or senior official Prof: prof Occupation prior to HE: Professional Prof: assoc Occupation prior to HE: Associate professional Prof: admin Occupation prior to HE: Administrative or secretarial Prof: other Occupation prior to HE: Some other job PT Working part-time Non-perm Not on permanent contract End of contract Months to end of contract (fixed-term contracts only) Staff-type (baseline = teaching staff) Researcher-only Staff type = researcher only Lecturer/researcher Staff type = lecturer and researcher GProf Grade = Professor or head of department GSLect Grade = lecturer htotal Total hours worked hresearch Hours spent on research hadmin Hours spent on administration RAE lowers sat RAE lowers satisfaction a lot QAA lowers sat QAA lowers satisfaction a lot

Policy lowers sat The general direction of higher education policy lowers satisfaction a lot

Teach no Would prefer to spend no time teaching Research all Would prefer to spend all time on research Admin no Would prefer to spend no time on administration Workload to high Consider total workload to be very much too high

Not fair Find decisions on either individual pay, recruitment to senior posts or promotion at current university not at all fair

RAE 5* department rated 5* in last RAE exercise RAE 5 department rated 5 in last RAE exercise RAE 4 department rated 4 in last RAE exercise Subject dummies University dummies

G.3.2 Results The results of estimating our ordered probit model of the likelihood of leaving

UK higher education are presented in Table A.28. We estimate four specifications of

282

the model. Because of the large numbers who do not report w*, the first specification includes a dummy variable for these non-reporters (No w*) and the second excludes these. The goodness-of-fit statistic suggests that the latter is marginally a better fit to the data. The third specification replaces the w and w* with the second principal component factor from the satisfaction analysis – satisfaction with pecuniary factors (salary and total earnings). The fit of this equation is slightly lower than the first two specifications. Finally, because of the co-linearity between w and w* discussed below, we include a further specification with the difference between current and expected alternative salary, w-w*.

The effect of both current earnings (w) and expected non-HE earnings (w*) on the likelihood of leaving UK HE is negative, with the latter being slightly smaller than the former. One reason for this result is that it is due to the high degree of co-linearity between the two terms – individuals who earn more in UK academia expect to earn more outside UK academia162. Because of this, in the final column of Table A.28 we report the results of a specification with the difference between current and expected earnings w–w*. This term is almost always negative163, implying that most academic staff feel they can earn more outside UK academia. The sign on the coefficient of the wage difference variable is positive, implying that the likelihood of leaving UK academia falls the greater current earnings are relative to expected earnings, the coefficient is, however, not significant. When w and w* are replaced with the satisfaction with earnings factor (factor 2), this is found to be negatively related to the likelihood of leaving – i.e. the more satisfied an academic is with their earnings, the less likely they are to leave.

Turning to the other satisfaction factors, the first factor – satisfaction with non-pecuniary elements of the current job – has a negative and statistically significant effect on the likelihood of leaving UK HE. Thus the more satisfied academics are with the elements of their job such as the actual work itself, relations with managers and colleagues and hours, the less likely they are to leave the sector. This is not true for the third factor – longer-term prospects. The effect of this on academics’ propensity to leave UK HE appears to be positive – that is, the likelihood of their leaving actually increases with their satisfaction with their longer-term prospects – although the statistical significance of the term is not robust to specification. This result may be because what is important is factors such as the permanency of academics’ jobs and the amount of fixed-term contract remaining for non-permanent staff that are the important influences on the likelihood of leaving the sector; both terms are statistically significant and of the expected sign (see below).

In periods of staff shortages, it is common to look abroad for solutions. What implication does this have for the future supply of academics in the UK? Do foreign academics remain? Our results are unambiguous: Academics from other EU (and EEA) countries, Australia, New Zealand and the US are more likely to leave UK HE than UK (and other foreign) academics. Our results support the hypothesis that these staff enter academic employment in the UK after completing a higher degree in the UK, but ultimately intend to return to their home country. If this is the case, such staff will only represent a short-term solution for lower-level jobs in UK higher education unless they can be persuaded to remain in the UK.

162 A regression of w* on w yields a coefficient on w of 1.03 and an R2 of 0.999. 163 It is only positive for ten per cent of those for whom it can be calculated.

283

Previous work on employee turnover suggests that the likelihood to leave jobs generally and possibly UK HE in particular would decline with tenure (see Farber, 1999; Gibbons and Waldman, 1999). Our results support this hypothesis, with the likelihood of leaving UK HE falling with experience, although doing so at a decreasing rate. The effect of experience on quits declines for the first thirteen to seventeen years, depending on the specification, and remains negative until experience hits the mid to late twenties.

There is evidence that individuals who have had a break in their academic career are more likely to leave again, suggesting that these staff are indeed more peripatetic in nature. This may be because of the individuals’ preferences themselves or because they work in an area where there is more flow backwards and forwards between academia and the rest of the economy. There is little evidence that those who have previous experience of working in other countries’ HE systems are less likely to leave, ceteris paribus, although note that it relates to UK academics; the combined effect of being from another EU or EEA country, Australia, New Zealand or the US and having worked in a foreign HE institution on the likelihood of leaving UK HE is still positive.

Although the effect of the ‘longer-term prospects’ factor on the likelihood of leaving UK academia appeared counter-intuitive, the permanency of academics contracts and the contract time remaining were important. Staff on non-permanent contracts are significantly more likely to leave UK HE than their colleagues on permanent ones. Furthermore, as the end of a fixed time contract approaches, the more likely an academic is to leave. The negative significant coefficient on End of contract implies that the greater the amount of time left a contract has to run, the less likely the individual is to leave UK HE.

Academics who work longer hours are more likely to leave UK HE, although the statistical significance of this varies across specifications. This is less true for hours of research than hours spent on teaching or admin. Note that this effect on leaving is over and above their influence on job satisfaction.

When we consider the aspects of academic employment that staff feel is important for their satisfaction, we find that few of them appear to affect their likelihood of leaving. Those who say that the RAE, the QAA and the general direction of higher education policy lowers their satisfaction by a lot are no more likely to leave UK HE than those who do not. The exceptions to this are those who feel that their workload is too high and those who feel that decisions on either individual pay, recruitment to senior posts or promotion at their current university are not at all fair, who are both more likely to leave UK higher education.

There is evidence that the likelihood of leaving UK higher education is higher in the areas with strongest competition from outside academia. The likelihood of leaving is highest among staff working in other (non-engineering) technology and medicine and dentistry. It is lowest in English literature and other humanities.

There is little difference in the likelihood of leaving UK academia when one compares staff across institutions, ceteris paribus, although staff in two of the new universities, one northern and one southern feel that they are more likely to leave.

284

Table A.28 Results – Likelihood of leaving UK higher education

full reduced wage dif wage sat -0.297* -0.307** -1.446 -0.148*** w (0.155) (0.155) (0.882) (0.054)

-0.221** -0.226*** w* (0.086) (0.087)

0.123 w – w* (0.083)

-2.497*** No w* (0.922)

-0.776** -0.696** -0.836** -0.838*** S1 (0.326) (0.323) (0.325) (0.325)

-1.218** S2 (0.505)

0.711 0.591 0.843* 1.029** S3 (0.450) (0.446) (0.447) (0.449)

-0.185*** -0.176*** -0.202*** -0.190*** Age 55+ (0.065) (0.065) (0.065) (0.065)

-0.242 -0.164 -0.269 -0.344* Non-white (0.203) (0.203) (0.203) (0.204)

0.356 0.270 0.431 0.413 Female (0.272) (0.270) (0.270) (0.271)

0.126 0.119 0.140 0.127 Married (0.099) (0.098) (0.098) (0.099)

-0.257 -0.224 -0.289* -0.222 Female×married (0.159) (0.158) (0.158) (0.162)

-0.090 0.116 -0.101 0.112* Children (0.064) (0.144) (0.063) (0.064)

-0.107 -0.155 -0.125 -0.109 Female× Children (0.089) (0.257) (0.089) (0.089)

-0.078 0.069 -0.061 -0.058 Num children (0.116) (0.063) (0.116) (0.116)

-0.253 -0.077 -0.307 -0.188 Female× Num children (0.220) (0.089) (0.219) (0.224)

0.293*** 0.314*** 0.287*** 0.296*** EU Pass (0.092) (0.092) (0.092) (0.091)

0.721*** 0.715*** 0.697*** 0.706*** OZNZ (0.246) (0.245) (0.242) (0.245)

0.401* 0.363 0.408* 0.382* USA (0.231) (0.232) (0.231) (0.231)

-0.119 -0.087 -0.117 -0.119 Foreign (0.118) (0.120) (0.117) (0.117)

0.111 0.136 0.132 0.104 HQ UG (0.091) (0.091) (0.091) (0.091)

0.173 0.197 0.191 0.165 HQ Masters T (0.144) (0.143) (0.144) (0.144)

0.162** 0.186** 0.168** 0.166** HQ Masters R (0.074) (0.075) (0.074) (0.074)

-0.142 -0.163* -0.124 -0.134 Study PhD (0.087) (0.088) (0.086) (0.087)

-0.026** -0.025** -0.033*** -0.026** e (0.013) (0.013) (0.013) (0.013)

0.001** 0.001** 0.001*** 0.001** e 2 (0.000) (0.000) (0.000) (0.000)

0.269*** 0.280*** 0.259*** 0.265*** Break from academia (0.090) (0.090) (0.089) (0.089)

285

full reduced wage dif wage sat -0.172 -0.220 -0.142 -0.162 Non-UK HE (0.275) (0.271) (0.272) (0.272)

-0.029 -0.025 -0.036 -0.035 Career change (0.066) (0.066) (0.066) (0.066)

0.001 -0.000 -0.004 -0.008 Prof: manager (0.117) (0.116) (0.117) (0.117)

0.143* 0.158** 0.139* 0.146** Prof: prof (0.074) (0.074) (0.074) (0.074)

0.034 0.044 0.046 0.042 Prof: assoc (0.105) (0.105) (0.104) (0.104)

-0.134 -0.143 -0.105 -0.105 Prof: admin (0.150) (0.152) (0.150) (0.151)

0.105 0.118 0.117 0.117 Prof: other (0.098) (0.097) (0.097) (0.097)

-0.169 -0.193 0.034 -0.190 PT (0.125) (0.125) (0.108) (0.124)

1.769** 1.607** 2.043*** 2.136*** Non-perm (0.694) (0.687) (0.685) (0.684)

-0.020*** -0.020*** -0.020*** -0.020*** End of contract (0.003) (0.003) (0.003) (0.003)

0.134 0.167* 0.159 0.155 Researcher-only (0.097) (0.098) (0.097) (0.097)

0.187** 0.194** 0.174** 0.175** Lecturer/researcher (0.084) (0.083) (0.084) (0.083)

-0.454 -0.348 -0.769* -0.495 GProf (0.469) (0.464) (0.452) (0.470)

-0.384 -0.323 -0.508* -0.440 GSLect (0.275) (0.273) (0.271) (0.273)

0.541*** 0.450** 0.318 0.304 htotal (0.200) (0.220) (0.200) (0.213)

-0.185*** -0.094 -0.210*** -0.053 hresearch (0.070) (0.078) (0.071) (0.076)

0.277* 0.098 0.195 -0.017 hadmin (0.144) (0.155) (0.141) (0.150)

0.063 -0.060 0.033 0.056 RAE lowers sat (0.166) (0.081) (0.168) (0.166)

-0.157 -0.128 -0.142 -0.148 QAA lowers sat (0.324) (0.110) (0.326) (0.328)

0.000 -0.128* 0.007 -0.006 Policy lowers sat (0.151) (0.068) (0.152) (0.151)

0.086 0.121 0.076 0.098 Teach no (0.160) (0.077) (0.158) (0.159)

-0.225* 0.084 -0.222* -0.226* Research all (0.135) (0.055) (0.135) (0.135)

0.152 0.034 0.155 0.148 Admin no (0.097) (0.054) (0.097) (0.096)

0.248*** 0.159*** 0.249*** 0.239*** Workload to high (0.070) (0.052) (0.070) (0.070)

0.268*** 0.249*** 0.276*** 0.261*** Not fair (0.055) (0.056) (0.055) (0.055)

0.049 0.043 0.043 -0.023 RAE 5* (0.115) (0.115) (0.115) (0.117)

-0.005 -0.008 -0.016 -0.096 RAE 5 (0.096) (0.096) (0.096) (0.100)

-0.212** -0.213** -0.220** -0.263*** RAE 4 (0.100) (0.101) (0.100) (0.101)

286

full reduced wage dif wage sat -0.006 -0.011 0.007 -0.009 norae (0.071) (0.072) (0.071) (0.071)

0.351** 0.315* 0.407** 0.311* Medicine and dentistry (0.162) (0.162) (0.162) (0.166)

0.004 -0.028 0.049 0.018 Biological sciences (0.118) (0.118) (0.117) (0.118)

-0.084 -0.089 -0.098 -0.215 Agriculture and related subjects (0.205) (0.208) (0.200) (0.208)

0.474* 0.408 0.559** 0.640** Physical sciences (0.253) (0.252) (0.252) (0.251)

0.229 0.195 0.269* 0.292* Mathematical sciences (0.164) (0.163) (0.163) (0.164)

0.126 0.098 0.150 0.237* Computing sciences (0.129) (0.130) (0.129) (0.134)

0.055 0.067 0.058 0.068 Engineering (0.138) (0.138) (0.138) (0.139)

1.090*** 1.007** 1.188*** 1.282*** Other technology (0.403) (0.399) (0.398) (0.405)

-0.480 -0.446 -0.463 -0.541* Architecture and planning (0.305) (0.301) (0.307) (0.307)

-0.431** -0.411** -0.445** -0.439** Social studies (0.192) (0.191) (0.192) (0.192)

-0.118 -0.116 -0.129 -0.181 Business and admin. studies (0.143) (0.141) (0.142) (0.144)

-0.812 -0.851 -0.851 -0.783 Librarianship & info. science (0.531) (0.528) (0.529) (0.532)

-0.516*** -0.556*** -0.454** -0.432** English lit. And classics (0.193) (0.189) (0.192) (0.192)

0.086 0.020 0.199 0.135 Modern languages (0.327) (0.328) (0.327) (0.325)

-0.496*** -0.484*** -0.482*** -0.469** Other humanities (0.185) (0.183) (0.185) (0.186)

-0.392 -0.405 -0.351 -0.461* Art and design (0.256) (0.258) (0.253) (0.256)

0.308** 0.291** 0.307** 0.356*** Education (0.135) (0.134) (0.134) (0.136)

0.340 0.288 0.405* 0.425* Combined studies (0.246) (0.243) (0.244) (0.244)

0.352* 0.308 0.380* 0.304 Old Sth uni 1 (0.194) (0.193) (0.194) (0.196)

0.038 0.026 0.020 0.122 New Sth uni 1 (0.158) (0.158) (0.156) (0.164)

-0.117 -0.140 -0.132 -0.147 New Lon uni (0.126) (0.127) (0.127) (0.126)

0.068 0.033 0.075 0.077 New Sth uni 2 (0.120) (0.119) (0.120) (0.120)

0.262 0.228 0.255 0.206 Old Lon uni (0.173) (0.172) (0.172) (0.175)

0.536* 0.461* 0.658** 0.641** New Nth uni (0.279) (0.277) (0.275) (0.276)

0.398* 0.343* 0.436** 0.384* New Sth uni 3 (0.209) (0.208) (0.208) (0.210)

-0.088 -0.099 -0.091 -0.107 Old Nth uni (0.084) (0.084) (0.083) (0.084)

-0.088 -0.124 -0.029 -0.041 New Nth uni 2 (0.211) (0.209) (0.209) (0.209)

287

full reduced wage dif wage sat -7.538*** -7.601*** -2.301*** -3.362*** τ1 (1.789) (1.794) (0.701) (0.825)

-6.544*** -6.602*** -1.310* -2.369*** τ2 (1.787) (1.793) (0.700) (0.824)

-5.951*** -6.007*** -0.718 -1.777** τ3 (1.787) (1.793) (0.700) (0.824)

-5.380*** -5.437*** -0.149 -1.208 τ4 (1.788) (1.793) (0.701) (0.825)

-4.837*** -4.893*** 0.392 -0.667 τ5 (1.787) (1.793) (0.700) (0.825)

-4.539** -4.594** 0.688 -0.370 τ6 (1.787) (1.793) (0.703) (0.827)

Observations 2,312 1,600 1,600 2,312 LL -3699.2 -3692.9 -3705.0 -3703.0 M&Z R2 0.195 0.200 0.190 0.192

Standard errors in parentheses

* significant at 10%; ** significant at 5%; *** significant at 1%

288

289

Appendix H: Recent and proposed policy developments

The following lists Higher Education initiatives since the survey fieldwork and relevant to recruitment and retention. The list was drawn up by the DfES.

RAE: following the review by Sir Gareth Roberts on research assessment in 2002 and wide ranging consultation, the UK published the arrangements for the RAE 2008 in February 2004. Information on the 2008 RAE <http://www.rae.ac.uk/default.htm>

The Science and Technology Select Committee published its report on the RAE in September following two Inquiries. The link is:

http://www.parliament.the-stationery-office.co.uk/pa/cm/cmsctech.htm#reports

The OST administered New Academic Fellowships scheme to help improve the career prospects of PhD researchers in the UK was launched in May 2004. <http://www.ost.gov.uk/research/fellowship.htm>

Relationship between Teaching and Research: In November 2003 Alan Johnson (then Minister of Lifelong Learning, Further and Higher Education) and Lord Sainsbury (Minister for Science and Innovation) set up the Higher Education Research Forum to consider, among other things, the relationship between teaching and research. The Forum produced its Advice to Ministers in July. Link to the site is:

<http://www.dfes.gov.uk/hegateway/hereform/heresearchforum/index.cfm>

On 11 March the European Commission adopted a European Charter for Researchers and a Code of Conduct for the Recruitment of Researchers. The Link is:

http://europa.eu.int/eracareers/index_en.cfm?l1=15&CFID=8202636&CFTOKEN=15187d4-00069ad2-165f-127a-94f7-8307fc480000 The Science and Technology Committee published its report on its Inquiry into ‘Strategic Science Provision in English Universities’ in April 2005. <www.parliament.uk/s&tcom>

Bureaucracy and administrative burdens: i) The Regulatory Impact Assessment process led by the Cabinet Office requires a regulatory impact assessment to be conducted on the impact of any new regulations (including HEIs). More information on:

http://www.brtf.gov.uk/docs/pdf/localdeliveryres.pdf

http://www.brtf.gov.uk/pressreleases/2005/lessismore.asp

290

ii) The HERRG is an independent 'gatekeeper' group for higher education established in summer 2004 by the Minister for Lifelong Learning, Further and Higher Education. Its remit includes scrutiny of new initiatives and existing burdens on higher education with a view to minimising bureaucracy: For more information please visit: http://www.dfes.gov.uk/hegateway/hereform/improvingregulation/index.cfm"

Teaching and Quality Developments

Centres for Excellence in Teaching and Learning (CETLs) There are amongst 74 Centres to be created across England, with a total funding allocation of £315 million over five years, from 2005/06 to 2009/10. This is HEFCE’s largest ever single funding initiative in learning and teaching. The CETLs are designed to promote excellence across all subjects and aspects of teaching and learning in HE. Funds awarded are to be used to recognise and reward excellent teachers, and to enable investment in staff, building, and equipment to support successful learning.

www.hefce.ac.uk/learning/tinits/cetl/final

HE Academy The Higher Education Academy was launched in October 2004, to work with the higher education community to enhance all aspects of the student experience. It aims to provide, a UK-wide focus for enhancing the student experience, and to help institutions, subject communities, individuals, the Government and funding bodies to provide a helpful environment for student learning.

www.heacademy.ac.uk

HE Leadership Foundation The Leadership Foundation was established in spring 2004 and is a joint initiative of Universities UK and HEFCE. The Foundation aims to improve the supply - and demand - of leadership development opportunities in HE, identify and disseminate good practice and raise the profile of HE leadership, governance and management.

http://www.lfhe.ac.uk/

HEFCE workforce report From Autumn 2005 the Higher Education Funding Council for England (HEFCE) will produce an annual report providing a broad strategic review of workforce issues in HE. Analysis in the report will be based on the Higher Education Statistics Agency’s new Individualised Staff Record. The intention is that the first report will be published on HEFCE’s website by Autumn 2005.

Copies of this publication can be obtained from:

DfES PublicationsP.O. Box 5050Sherwood ParkAnnesleyNottinghamNG15 0DJ

Tel: 0845 60 222 60Fax: 0845 60 333 60Minicom: 0845 60 555 60Oneline: www.dfespublications.gov.uk

© National Institute of Economic and Social Research 2005

Produced by the Department for Education and Skills

ISBN 1 84478 523 8Ref No: RR658www.dfes.go.uk/research

Documents

1.DfES Recruitment and Retention Report Jul 2005