Pension provision in the higher education sector: initial report

Special consultation by theHE Employers’ Pensions Forum

SectionA head

About the author

Peter Thompson is an actuary with 30 years’pensions experience. He is a director of theindependent trustee company BESTrustees plc,which he joined in April 2005 after having workedat the consultants Mercer for nearly 25 years.

Peter has been much involved with the NationalAssociation of Pension Funds (NAPF), being amember of Council from 1994 to 2005 andChairman of the NAPF from 2001 to 2003. FromAugust 2005 to March 2006 he was interim chiefexecutive of the corporate governance companyRREV. He was a member of the Railway PensionsCommission which was set up to review allaspects of pension provision in the railwayindustry and which published its final report inJanuary 2008.

Peter is a regular speaker and chairman atconferences and seminars. He has alsocontributed on pensions issues to both print andbroadcast media, including appearances on newsprogrammes on the BBC, ITV and Channel Four,as well as Money Box and the Today programme.He is a member of the Council of the PensionsPolicy Institute, and is a CEDR accreditedmediator.

Peter’s Trustee clients at BESTrustees include JSainsbury, Scottish Power and Jacob’s Bakery.He is also a member-elected pensioner trustee ofhis own scheme, the £3 billion MMC UK PensionFund.

2 Summary

13 Introduction

24 Background

35 Recent changes

46 The Hewitt report

57 Public sector or private sector?

68 Pension scheme models

710 Which design model for the higher education

sector?

812 Scheme mergers

915 Universities Superannuation Scheme issues

1016 A possible long-term solution

1117 Conclusion

18 Glossary

19 Notes

Contents

Universities UK Pensions provision 1

Pension provision in the higher educationsector: initial report

This report has been prepared for the HEEmployers’ Pensions Forum by Peter WThompson BSc FIA, Independent Actuaryand Consultant.

2

p The structure of USS, and in particular the role ofits Joint Negotiating Committee (JNC), areexamined and it is concluded that more workneeds to be done, probably outside the USSframework, to establish an agreed route forwardwith the University and College Union (UCU) thatwould enable final salary pension provision to besustained in its present form.

p Finally, a possible pensions model for the futureis outlined together with a proposal that morework needs to be done to refine such a model, andin particular how it would adapt to future changessuch as continuing improvements in mortalityrates.

p This report has been written at the request of theUniversities and Colleges Employers Association(UCEA), Universities UK and GuildHE in the lightof increasing concern among their membersabout the rising cost of pension provision. Risingcosts are partly caused by generic factors such asimproving mortality, and partly also by a numberof factors specific to the higher education sector.

p Initially the major pension schemes operating inthe sector – the Universities SuperannuationScheme (USS), the Teachers’ Pensions Scheme(TPS), the Self-Administered Trusts (SATs) andthe Local Government Pension Scheme (LGPS) –are reviewed, and recent changes made to thepublic sector schemes TPS and LGPS are noted.Key conclusions from the Hewitt report (2007)1

are listed and it is noted that some of these giveclear pointers towards possible solutions.

p The position of the higher education sector at theboundary of public sector and private sector isinteresting and has some consequences forpossible pension scheme models in the future.The major pension scheme models are reviewedand their advantages and disadvantagesexplained. Designing a pension arrangementwhich rewards or encourages particularbehaviour is described.

p In looking at possible design models for thehigher education sector, widespread adoption ofmoney purchase provision is ruled out as thereappears to be little employer appetite for such alarge-scale transfer of risk to the members.However, the option of having a “menu” of pensionoptions is worth further discussion althoughthere are substantial practical difficulties withsuch an approach.

p The widely expressed desire to achieve greatereconomies of scale is addressed from the point ofview of reducing the number of pension schemesin the higher education sector. Moving towards asingle scheme for each institution is ruled out, asis the option of adopting the TPS as the singlescheme for academic staff. However, thepossibility of USS developing in this way is worthfurther examination, which would have to includeconsulting the relevant governmentdepartments.

p The position of non-academic staff, especiallythose in pre-1992 institutions, is considered and itis suggested that merging the smaller SATs,either with each other or into a larger existingscheme such as USS or the SuperannuationArrangements of the University of London (SAUL),is a practical proposition. Some institutions havealready done this and others are understood to beconsidering this approach.

Summary

Universities UK Pension provision

1.1 This report has been prepared at the request ofUCEA, Universities UK and GuildHE in the light ofincreasing concern about the sustainability andrising cost of the pension arrangements operatedby their member institutions. Most of thesearrangements are of the “final salary” type,where pensions are based on length of serviceand salary at or near retirement.

1.2 The background to the present request stemspartly from factors which are generic to allpension arrangements – particularly cost issuescaused by improving mortality and difficultinvestment markets – and partly from factorsspecific to the sector. The sector-specific factorsinclude:

p the structure of USS;

p the extent of any influence which the sector’semployers have over USS, LGPS and TPS;

p concern over rapidly increasing costs of some ofthe smaller schemes; and

p whether some of the schemes within the sectorshould be merged.

1.3 This initial report looks at the background andconsiders possible approaches at a high level.Following consideration of this report, it is hopedthat a small number of options can be identifiedfor more detailed consideration and costing overthe coming months.

1.4 This report is intended for the use of UCEA,Universities UK and GuildHE only, andresponsibility is not accepted towards any otherparties for any of the contents of this report.

1Introduction

3

4

(under the LGPS regulations on “scheduledbodies”), and employees have a right to join.

2.6 This summary describes the main features ofpension provision in the sector only. There aremany additions and exceptions, such as:

p Academic staff in post-1992 institutions inScotland are generally members of the ScottishTeachers’ Superannuation Scheme (STSS);

p Some staff in medical schools are members ofthe National Health Service Pension Scheme(NHSPS);

p The membership of USS includes someacademics in post-1992 institutions if they hadpreviously been members of USS whilst being at apre-1992 institution;

p USS is in practice open to all employees atinstitutions, including non-academic staff. Somehave taken advantage of this, including the OpenUniversity, and others are in negotiation.

2.7 These exceptions do not of themselves affect thegeneral conclusions of this report.

2.1 Most pension provision for employees in thehigher education sector is made through one offour pension schemes or groups of schemes,which are described briefly below.

2.2 The Universities Superannuation Scheme (USS)provides pensions for academic and “academic-related” (i.e. senior non-academic) staff,primarily in pre-1992 institutions (the “old”universities). USS is a funded final salary pensionscheme and indeed is one of the largest suchschemes in the UK. It has a number of unusualfeatures in its rules, one of which is the“exclusivity clause”, which requires that aninstitution participating in USS must not offeranother pension scheme to an employee who iseligible for USS membership.

2.3 The Teachers’ Pension Scheme (TPS) providespensions for academic staff in post-1992institutions (the “new” universities, many ofwhich were formerly polytechnics). It also ofcourse provides pensions for schoolteachers inEngland and Wales. TPS is “notionally funded” –that is, contributions are assessed on a quasi-actuarial basis, but in reality no funds are built upand the liability for future pensions is ultimatelyborne by the taxpayer.

2.4 Non-academic staff in pre-1992 institutions,other than academic-related staff, are generallyprovided with pensions by means of schemes setup by the individual institution. These are knownin the sector as SATs (self-administered trusts),although it should be noted that this terminologywould not be recognised outside the sector. Thelargest of these by some distance is SAUL(Superannuation Arrangements of the Universityof London), with assets of some £1.3 billion. Manyof the others are much smaller, which in somecases has given rise to concerns about theirviability. In general these are also final salaryarrangements, although there are examples ofboth career average and money purchase SATschemes.

2.5 Non-academic staff in post-1992 institutions,including academic-related staff, are mostlymembers of the Local Government PensionScheme (LGPS), operated by the local authorityresponsible for the area where the institution islocated. This reflects the origins of theseinstitutions as polytechnics when they operatedunder the aegis of local authorities. The LGPS is afinal salary pension scheme whose benefits andmember contributions are centrally determined,although each local scheme has separatefunding, administrative and employercontribution arrangements. It should also benoted that those institutions which participate inthe LGPS cannot in general withdraw from it

2Background

Universities UK Pension provision

Comments

3.8 Higher education employees only form a smallpart of the membership both of the TPS and theLGPS, and there is some concern about thelimited influence that the sector has on theoperation of these schemes. However, it isimportant in the context of pension strategy forthe higher education sector to note that both ofthese schemes have retained the final salaryconcept for existing and new members. The moveto age 65 for new entrants to the TPS, andrestrictions on the availability of early retirementin the LGPS, reflect broader government policy toencourage longer working lives in the light ofimproving mortality.

3.9 It is also interesting to note that both the TPS andthe LGPS have moved towards a structure basedon a 1/60 accrual rate, rather than the “1/80 pluscash” approach which has been traditional in thepublic sector (and which still applies in USS andSAUL, amongst others).

3.1 The LGPS and the TPS have both been the subjectof changes in recent years, which have involvednegotiation with various stakeholders includingparticipating employers, members and theirrecognised trade unions, and the relevantgovernment departments. Details can be foundon the respective websites but they can besummarised as follows.

Teachers’ Pension Scheme

3.2 The changes to the TPS were introduced witheffect from 1 January 2007. The most importantchanges are that, for new entrants (only), thepension age will be 65 rather than 60; andpensions will be based on an accrual rate of 1/60,with the option to take a tax-free cash sum, ratherthan on an accrual rate of 1/80 with an automaticlump sum of 3/80.

3.3 Benefits for those who were members before 1January 2007 are essentially unchangedalthough there are some modest improvements.

3.4 It is proposed that future increases in cost will be shared equally between employers andemployees and that there will be a maximumemployer’s contribution of 14 per cent followingthe outcome of the next actuarial valuation,although the exact mechanism by which this willhappen is still to be determined.2

Local Government Pension Scheme

3.5 The changes to the LGPS were introduced witheffect from 1 April 2008 and apply to the futureservice of existing members as well as to newentrants. Future pensions will, as in the TPS, bebased on an accrual rate of 1/60, with the option totake a tax-free cash sum, rather than on anaccrual rate of 1/80 with an automatic lump sumof 3/80. Member contributions will vary from 5.5per cent to 7.5 per cent according to pay levels.

3.6 The normal pension age will continue to be 65.Early retirement rules in the LGPS are complexand changes were made to these with effect from30 September 2006 because the existing “rule of85” was deemed to be age-discriminatory.

3.7 As in the TPS it is proposed that employers andemployees will share future increases in the costof pension provision. The LGPS is currentlyconducting a public consultation as to how thisproposal should be implemented, and whetherthere should be a maximum employer rate ofcontribution in the future.3

3Recent changes

5

6

4.6 Finally, the responses to questions 16 and 17indicated a general belief that the presentarrangements are only partially understood andthat members did not fully appreciate the value ofa pension. It is clear that there is work whichcould be done to improve both these aspects ofpension provision within the higher educationsector.

4.1 In the second half of 2007, Hewitt, a pensionsadviser, was commissioned by Universities UK,GuildHE and UCEA to produce a report onpensions strategy for the higher education sector.Its report was essentially an attitude surveywhich asked over 160 higher educationinstitutions about their attitudes towards pensionprovision for their employees, putting forwardtwenty questions each with a “point” and a“counterpoint” view. There was an overallresponse rate of 53 per cent. Each set ofresponses was rated, both for the averageanswers within the range 1 to 10 and for thedispersion of those answers.

4.2 As one would expect, some of the responses weremore helpful than others. Some showed evidenceof a wide variety of views within the sector with noconsensus of opinion. Some, on the other hand,provided evidence of strongly held views acrossthe sector, which are helpful in indicatingpossible ways forward and in ruling out someother possibilities.

4.3 Those responses which indicated the strongestviews, starting with the strongest (using a fairlybasic methodology to combine both the absolutevalues of the answers and their dispersion) are asfollows:

p The sector should work together foradministrative economies of scale (question 6);

p Risk should be shared between the institutionand the employees (question 14);

p Pension benefits should be comparable to thoseprovided in the public sector (Question 20);

p The cost of early retirement should be met byeach institution (question 18);

p Members’ contributions should vary to reflect thetrue cost of their benefits (question 13);

p The higher education sector should negotiate onpension issues as a single group (question 5).

4.4 Some of these responses give clear pointers as tothe type of pension benefits that might be foundacceptable (or not) in the higher education sector,whereas others are more indicative of the processthat the sector might use to reach a generallyacceptable conclusion.

4.5 A few other sets of responses are notable, notbecause of a particular strength of view butbecause of a lack of support for a particular view.For example, although the question on the degreeof change (question 15) indicates that somechange is widely expected, it shows little appetitefor radical change. This view is supported by theresponse to other questions, for examplequestion 19 on employment models.

4The Hewitt report

Universities UK Pension provision

seems to indicate that the sector as a whole doesnot have the appetite for risk transfer whichwould be inherent in a large-scale move to moneypurchase; the answer to question 14 in the surveybears this out in that there was very clearly noappetite whatever amongst the respondents for asolution in which the employees would bear allthe risk.

5.1 The higher education sector is in some respectscuriously positioned at the boundaries of thepublic and private sectors. It is still dependent formuch of its activities on government funding, butreceives an increasing proportion of its incomefrom tuition fees, research grants and othersources (such as alumni and vacation courses).The post-1992 institutions were public sector inorigin as polytechnics run by local authorities.Nevertheless, it has become increasingly clear inrecent years that financial disciplines importedfrom the private sector are an essential part ofgovernance in the higher education sector.

5.2 Pension arrangements in the sector tend toreflect this background. USS is, and has alwaysbeen, very clear that it is a private sector schemeand not a public sector scheme; it carries nogovernment guarantee and pays a levy to thePension Protection Fund like other private sectorschemes. The same applies to SAUL and theSATs.

5.3 LGPS and TPS, on the other hand, are clearlypublic sector pension schemes and the latter isunfunded. The LGPS is funded although it is notsubject to the normal funding rules (under thePensions Act 2004) and political as well asactuarial considerations have sometimes playeda part in setting employer contribution rates.Most employers in LGPS have a tax-raising powerand there is no realistic prospect that a localgovernment pension scheme would be allowed tobecome insolvent.

5.4 In recent years, as is well known, most privatesector employers have moved away from finalsalary pension provision, at least for newemployees and sometimes also for existingmembers. Final salary schemes have beenreplaced by money purchase or, in some cases,career average schemes. These have differentrisk characteristics which are explored later inthis paper. It is fair to say that smaller employersin particular have moved towards moneypurchase provision.

5.5 Public sector pension schemes have themselvesbeen the subject of revision in recent years,including but not limited to the LGPS and the TPSas mentioned above. However, they have ingeneral continued with a defined benefit model,rather than the more radical shift to moneypurchase which has characterised much of theprivate sector.

5.6 Against this background it is interesting to notethe strength of feeling expressed by respondentsto the Hewitt report that pension benefits in thehigher education sector should be comparablewith the public rather than the private sector. This

5Public sector or private sector?

7

8

accrual rate and the rate at which a member’spre-retirement earnings are revalued. In general,for a given overall cost:

p members with long service are likely to receiveless from a career average scheme than from afinal salary scheme, while members with shorterservice are likely to do somewhat better;

p the higher the rate of pre-retirement revaluation,the more weight will be given to earnings earlierin the working lifetime, rather than later, and viceversa. In other words, the lower the rate of pre-retirement revaluation, the less will be thechange from benefits calculated on a final paybasis.

6.8 The exact outcome for any individual or group ofindividuals does of course depend largely on theiractual pay progression, both in absolute termsand in relation to the rate of pre-retirementrevaluation assumed, and models can beconstructed to demonstrate a wide range ofpossible outcomes. Nevertheless, the generalpoints made above still apply; they show that acareer average pension scheme can be used, ifdesired, as a means to distribute pension benefitsin a way different from that which would beachieved under a final salary structure.

Money purchase (defined contribution)

6.9 In a money purchase pension scheme, employerand members pay a pre-determined rate ofcontribution into the scheme; this is invested andthe outcome for the member is whatever theaccumulated “pot” will purchase at the point ofretirement. The rate of contribution may be fixedas a percentage of salary; it may be related to themember’s seniority; or the member may be ableto choose a rate, within limits, with the employerpaying a corresponding rate which typically variesaccording to the member’s chosen rate to give a“thrift plan” effect.

6.10 Whichever pattern is used, the employer’s liabilityis restricted to the agreed rate of contribution andthe employer has no liability beyond this. Themember takes the investment risk as well as thelongevity risk. It is widely believed that a largeproportion of the money purchase schemes set upin the last decade will produce pensions atretirement which will be inadequate for themember to live on, although many such schemesare in their relative infancy. It is also worth notingthat members cannot accurately assess theirlikely pension from a money purchase schemeuntil shortly before retirement.

6.11 Money purchase schemes are best suited tothose cases where the employer has no appetite

6.1 In this section the key characteristics of the mainpension scheme models are considered. It isimportant to realise that these are not the onlypossible designs of pension scheme that could bemade available; they do however cover the mainoptions, and most others can be characterised asbeing similar to, or a combination of, the modelsdescribed below.

Final salary (final pay)

6.2 In a final salary pension scheme, the member’spension at retirement is determined by referenceto length of service and pay at or shortly beforeretirement (or leaving service). Membercontributions are normally a fixed proportion ofpay with the employer paying the balance of cost.The concept underlying a final salary scheme isthat the member’s standard of living inretirement should be related to that which theyenjoyed shortly before retirement, as measuredby their regular earnings in the last few years oftheir working life.

6.3 In this type of arrangement, the employer takesnearly all the risk: investment, mortality, paygrowth and inflation risks may all affect the costof benefits and the member simply pays a fixedproportion of pay. The member’s only materialrisk is that of insolvency of the employer,although the introduction of the PensionProtection Fund has largely mitigated this risk.

6.4 Final salary schemes are most suitable for:

p long-serving employees;

p employees whose pay rises in real terms,especially where such rises continue throughouttheir career; and

p employers who are able to take financial risk.

Career average

6.5 In a career average pension scheme, each year’spay is revalued to retirement using a pre-determined index (for example, the Retail PricesIndex (RPI) or national average earnings), and thetotal is then multiplied by the accrual rate. Again,member contributions are normally a fixedproportion of pay with the employer paying thebalance of cost.

6.6 The employer’s exposure to risk is less, in thatpay growth (especially late in career) does notaffect past service liabilities. However, theemployer is still exposed to the financialconsequences of mortality and investment risks.

6.7 The distribution of member outcomes within acareer average scheme depends crucially on the

6Pension scheme models

Universities UK Pension provision 9

for, or ability to take on, financial risk; in somesuch cases, members may actually be better offin a money purchase arrangement since theirpension should be unaffected in the event that theemployer fails.

Other designs

6.12 Other designs, such as cash balance plans(where the scheme funds an amount of cash atretirement rather than a pension) andcombinations of final salary and money purchase,have met with limited success both in the UK andin the United States. These generally work bysharing risk, for example, investment andmortality risk, between the parties in variousways. They can be complex to communicate andadminister, although this should not be aninsurmountable obstacle to such a solution if thisis thought to be the most appropriate in thecircumstances.

General

6.13 It is important to realise and accept that there isno one pension scheme model which is always“good” and one which is always “bad”, eventhough various commentators have tried topresent this as being the case. There are good andbad final salary schemes, just as there are goodand bad money purchase schemes. The keydeterminant of the generosity of a pensionscheme is the level of contributions (employerand members combined) made to the scheme.For the same total contributions, any pensionscheme will produce broadly the same level ofoverall benefits. What the choice of design does isto influence how this overall outcome isdistributed between the beneficiaries, forexample:

p long-stayers versus early leavers;

p high-flyers versus plodders;

p those who live to a ripe old age versus those whodie early;

p members versus their dependants;

p those who are successful, or lucky, in theirinvestment decisions versus those who are not.

6.14 It follows that, if an employer or a group ofemployers can determine which behaviours theywish to encourage and reward amongst themembers of their workforces, they can design apension scheme to reflect the desiredbehaviours. The same is, of course, true of theremuneration package more generally, but that isbeyond the scope of this report.

10

consideration of a solution based on moneypurchase provision is therefore ruled out.

7.6 After discussion of the behaviours to be rewarded,the next consideration is the cost parameterswithin which the sector feels it can afford tooperate. As described above, whether a particularpension scheme can overall be described as“good” or “bad” depends less on its particulardesign than on the level of contributions whichcan be afforded by the participants. In this areathe input of those operating within the highereducation sector is indispensable. However, it isworth noting that the question in the Hewittsurvey on the cost of pension arrangements(question 12) elicited a wide spread of responseswith an average close to the mid-point of therange, which does not indicate a general view thatcurrent costs are too high.

A “menu” of pension options

7.7 One option canvassed in the Hewitt report, whichreceived a generally positive response (althoughnot as positive as the six responses listed insection 4.3), was the possibility of employershaving a “menu” of possible pension options tooffer their staff. The counter proposal, underwhich each employer would individually decidethe pension arrangements to offer its employees,received relatively little support.

7.8 Under the “menu” approach, employers would beable to select, from a limited pre-determined list,what pension arrangements to offer theiremployees. There is clearly a large number ofoptions that could be included, but those listed inthe Hewitt report were:

p 1/60 final salary from age 60

p 1/50 career average scheme from age 65

p 6 per cent defined contribution (money purchase)plan.

7.9 Each option would have different rates ofemployee and employer contributions associatedwith it.

7.10 On the face of it, this seems like an attractiveapproach, allowing institutions a degree offlexibility but restricting the choice sufficiently sothat the sector could still benefit from thoseeconomies of scale which were so clearly desiredby survey respondents (question 6). The samewould certainly not apply if each institution wereto be able to choose its own pension provision.

7.11 There are, however, many practical issues interms of the implementation and administrationof a “menu” approach, and some of these will bematerial. Examples are:

7.1 As described above, pension provision within agiven cost framework can be used to rewardbehaviours which are determined to be beneficial. Employers in the higher educationsector clearly feel that they should negotiate onpension issues as a single group. This willtherefore require a consensus on what thesebehaviours should be; whether they are the samefor all institutions; and whether they are the samefor all staff within each institution. Employmentmodels have changed in many sectors in recentdecades and there is some evidence from theHewitt report that the same is happening in thehigher education sector. So, for example,institutions may, or may not, wish to reward andencourage:

p long service;

p flexibility and innovation;

p loyalty to a particular institution, or to the sectoras a whole;

p staff moving within the higher education sector,or outside it;

p staff staying on beyond age 60, or even 65,provided they are fit and healthy.

7.2 There may be other behaviours which the sectorwishes to encourage or discourage as well asthose listed. Any of these can give useful pointersas to possible pension scheme designs for thefuture. There is also the question of risk, and theextent to which institutions in the sector are ableand willing to accommodate pension risk withintheir overall operations (especially given the veryhigh proportion which payroll costs representwithin the total costs of operating a highereducation institution).

7.3 There is also the output from the Hewitt surveymentioned above (on page 6). The following twoanswers give the strongest pointers for thisdiscussion:

p Risk should be shared between the institutionand the employees (question14)

p Pension benefits should be comparable to thoseprovided in the public sector (question 20)

7.4 The view on risk being shared (as opposed to allthe risk being borne by one party or the other) wasvery clear, and as already pointed out there wasno appetite at all for a solution which involvedemployees bearing all the risk.

7.5 The combination of these two answers leadsinexorably to a conclusion that there is littleappetite in the higher education sector for awidespread adoption of money purchase pensionprovision as the way forward. Further

7Which design model for the higher education sector?

p who chooses which pension arrangement to offer,and whether it should be the same for allemployees of an institution;

p whether the overall pay package would beadjusted in some way if a cheaper pensionarrangement were offered;

p whether the chosen arrangement would beoffered to existing employees (in place of theircurrent pension scheme), or just to newemployees;

p if existing employees were to be included, howpast service would be handled;

p what would happen if an employee were to moveto another institution which participated in thesame scheme but offered a different menuoption;

p whether the “menu” would be offered under theaegis of one of the existing schemes (USS, forexample) or as a new arrangement;

p if the latter, how the USS “exclusivity clause” andthe right of non-academics at post-1992institutions to join the LGPS would be addressed;

p how funding, sectionalisation and cross-subsidyissues would be handled;

p the consequences for existing arrangements,whether USS, TPS, LGPS or SATs.

7.12 These issues can all be expanded upon if desired,and they are not necessarily insuperable, but itmust be clearly understood that theimplementation of a “menu” approach would befar from straightforward. If this approach is to beexamined further, the issues of scheme designdiscussed in section 6 must be considered so thatparticipants in the sector can be satisfied that thisapproach would achieve the desired outcome.

Universities UK Pension provision 11

12

8.7 It is worth exploring what these perceived gainsmight be. As shown in the table above,contributions to USS and TPS are virtuallyidentical, although there is of course noguarantee that this will continue in the future.New entrants to TPS are already joining with apension age of 65 and, other things being equal,this would suggest that the contribution rates toTPS will, over time, be lower than they wouldotherwise have been. The TPS is of course anunfunded public sector scheme, so it carries aGovernment guarantee of its benefits, whichmembers may perceive as an advantage; on theother hand, the higher education sector has littleinfluence over the TPS, which is primarilyintended for teachers in state schools.

8.8 Economies of scale are not likely to be a majorissue here: both USS and TPS are very largeschemes and already benefit as far as possiblefrom the scale of their operations. It appears thatthe main benefit of having a single scheme for allacademic staff is one of perception rather thancontrol or economies of scale.

8.9 Moving USS members to TPS would have majorand serious consequences both for USS, and forthe institutions which currently have members inUSS. These arise from the fact that a move to TPSwould in effect close USS, at least to newmembers and possibly also to future accrual (ifexisting members were also to join TPS). Thiswould probably cause USS to re-examine itsinvestment strategy with a move away from thecurrent high level of equity investment. Also, ifinstitutions were to cease membership of USSaltogether, it is probable that a “Section 75” debtwould arise which would be likely to impose anintolerable financial burden on such institutions.It seems unlikely that any actual or perceivedgains would be worth this kind of financial risk.

8.10 From the members’ point of view, new entrants toTPS do of course join with a normal pension age of65, which may itself prove an insuperableobstacle to a transfer of existing members fromUSS to TPS. The position of academic-relatedstaff would have to be considered; currently theseare in USS if in pre-1992 institutions, or in LGPS ifin post-1992 institutions. It is not clear whetherthey could join TPS and the alternative would befor them to seek admission to the relevant LGPS.

8.11 Ultimately it seems that the obstacles topositioning TPS as the single scheme foracademics are probably insuperable and thisoption is not therefore pursued further.

8.12 The alternative would be for USS to become thesingle scheme for academic staff, either for newentrants to the sector only, or including all

8.1 One way of achieving the economies of scale soclearly desired by the participants in the sectorwould be to reduce the number of pensionarrangements on offer. There are severalpossible approaches, and the following areconsidered further in this section:

p One scheme for each institution

p One scheme for all academic staff

p One scheme for all non-academic staff

p One scheme for the sector

One scheme for each institution

8.2 Under this approach, each institution would haveone pension arrangement for all of its own staff,academic and non-academic. The concept couldbe extended so that smaller institutions couldperhaps join in a grouped scheme with otherinstitutions in their area or region. In effect,academic staff would join the SAT operated by theparticular institution where they were working.

8.3 It is not clear that this approach offers anyadvantages, and indeed there would beconsiderable disadvantages in splitting (forexample) USS into many component parts.Further, academic and non-academic staff do notnecessarily have the same needs or the samecareer paths. In addition, the history andexperience of the higher education sectorsuggests that “larger is better” whereas splittingUSS and the other large schemes would have thereverse effect.

8.4 Consequently, this option is not consideredfurther.

One scheme for all academic staff

8.5 At present, academic staff are generallymembers either of USS (in pre-1992 institutions)or TPS (in post-1992 institutions). Contributionrates are as follows:

USS TPS

Member 6.35% 6.4%

Institution 14% 14.1%

8.6 This approach suggests that all academic staffwould become members either of USS or of TPS,and therefore that the other one would cease tohave members in the higher education sector.Either approach (USS members into TPS, or TPSmembers into USS) has disadvantages and theissue therefore is whether these are too great tooutweigh any perceived gains from having asingle scheme for all academic staff.

8Scheme mergers

post-1992 institutions, including academic-related staff, are in the LGPS covering the areawhere the institution is located.

8.18 Because academic and academic-related staffare excluded from SATs, the average salary ofmembers of these schemes is low. The averagesalary of SAUL members at 31 March 2005 (thelast actuarial valuation) was £20,1495 whereasthat for USS at the same date was £33,281.6 Lessthan 0.5 per cent of SAUL’s active members earnin excess of £50,000 a year.

8.19 Contributions to these schemes vary widely. Theemployer contributions to SAUL are 13 per cent ofpayroll, whereas contributions to SATs tend to behigher with half reporting employer contributionsin excess of 17.5 per cent.7 LGPS contributionsare lower than SATs, with over half being below15 per cent.

8.20 The average size of SATs indicates that somemust be small, of the order of £20 or £30 million,which leads to the conclusion that someeconomies of scale must be obtainable if thesecould somehow be merged together or absorbedwithin a larger scheme. Both SAUL and USS havemade it clear that they are happy to absorb otherschemes by merger; some such mergers havealready taken place (the Open University, forexample, has undergone a full merger of itsscheme with USS) and others are being activelyconsidered.

8.21 So one option for many SATs would be for them tojoin USS or SAUL, subject to the legal and fundingrequirements set out by those schemes and bypensions legislation more generally. Joiningcould be for future new entrants only; for allactive members for their future service; for allactive members including their past service; or afull merger including a transfer of deferredmembers and pensioners. Any funding deficittransferred would have to be paid off over anagreed period, typically not exceeding ten years,calculated on funding assumptions set by thereceiving scheme. This may appear onerous, andcertainly some institutions have commented onthe perceived “cost” of joining USS. However, anyinstitution not transferring its past serviceliabilities would need to consider carefully itsfunding plan for such liabilities, and whether thelong-term intention is to secure the liabilities insome way, either by the traditional means ofannuity purchase or through one of the newervehicles operating in this market. The institutionwould have to be careful to avoid a large residualliability falling on the institution a decade or morehence.

existing TPS members. At present, this wouldhave little immediate financial impact oninstitutions because the contribution rates to bothschemes are very similar.

8.13 However, the result if new recruits only were to beadmitted to USS would be an influx of youngermembers to USS, which ought to reduce theoverall rates payable to USS. The remaininghigher education population in TPS wouldgradually age, and the contribution rate to TPSought to increase gradually; however, as TPS is anunfunded scheme and the proportion of itsmembers in higher education institutions issmall, the extent to which this would actuallyhappen in practice is not clear.

8.14 The consequences for TPS would also have to bethought through, not least because such a movewould have implications for the financial positionof the TPS as a whole. As the TPS is an unfundedscheme, any change which materially affects itsflow of new entrants could disturb its cashflowpattern with consequent implications for othercontributors. This would be an area where theinput of HM Treasury and DIUS would have to besought and a convincing business case for such amove would have to be constructed. At this stageit is not clear that such a business case exists.

8.15 A final consideration is whether, even if abusiness case could be constructed, post-1992higher education sector institutions do actuallywish to withdraw from TPS. From the viewpoint ofmembers (and presumably their trade unionrepresentatives), it is difficult to see why theywould wish to withdraw voluntarily from ascheme whose benefits are backed by theGovernment.

8.16 The case for USS becoming the single scheme foracademics is not therefore straightforwardalthough it has fewer obvious drawbacks than theTPS option. This may therefore be worthy offurther examination, including at least initialdiscussions with relevant governmentdepartments, to see if the option deservesdetailed analysis.

One scheme for all non-academic staff

8.17 At present, non-academic staff in pre-1992institutions, other than academic-related staff,are generally in a pension scheme operated bythe institution (a SAT), except in the London area,where the centralised scheme SAUL covers thesestaff. SAUL is a large scheme with assets ofaround £1.3 billion, whereas other such schemesare much smaller, with average assets of £68million as at 31 July 2005.4 Non-academic staff in

Universities UK Pension provision 13

8.22 An alternative to a full merger with SAUL or USSmight be for institutions operating SATs to cometogether in regional groupings in order to takeadvantage of some at least of the economies ofscale that are available to larger schemes. Thisfeels rather like a “halfway house”, and inaddition there would be other practicaldifficulties, one of which would be determiningthe receiving scheme for this purpose. This couldbe a new vehicle set up specifically as a receivingscheme (which has been done in other sectors,such as the Co-Op), although this would tend toreduce the cost savings which might otherwise beachievable; or one of the existing SATs, dependingon legal advice, could be chosen as the recipientscheme.

8.23 None of this addresses the issues facing post-1992 institutions which have members in theLGPS. As mentioned earlier, these institutionsgenerally cannot withdraw from the LGPSwithout changes to regulations, and consultationwith members or their trade unions would alsoprobably be necessary. The regulatory issues,and the lower employer contribution rates to theLGPS, mean that further consideration of theposition of non-academic staff in post-1992institutions should be deferred.

One scheme for the higher education sector

8.24 This is really just an extension of the variouspossibilities outlined in the precedingparagraphs. However, it is not obvious that theeconomies of scale to be obtained from such alarge pension scheme would be significantenough to justify some of the practicalchallenges, including the sheer scale of themerged operation (which would probably be thelargest funded pension scheme in the UK),governance issues, the workload on trustees, andso on. Given the issues around the public sectorschemes discussed earlier, a more realisticoption may be to move gradually towards onescheme for the pre-1992 higher education sector.

8.25 If some of the suggestions proposed in thissection were to be adopted, the number ofschemes in the pre-1992 sector would beexpected to reduce materially. If in due course itwere to appear that further economies of scalecould realistically be achieved by a second roundof mergers, then this is an issue which could beaddressed at that time.

14

Universities UK Pension provision

9.1 Although USS is a very large (around £30 billion),and generally well governed, funded pensionscheme, some difficulties and anomalies havearisen, partly from its legal structure, and partlyfrom the relationships built up within the sectorover the years.

9.2 USS was set up in 1975 as a successor to theFederated Superannuation System forUniversities (FSSU), and its legal structurereflects the political and industrial relationsenvironment prevailing in the 1970s. In particular,USS has a “Joint Negotiating Committee” (JNC),whose purpose (inter alia) is “to approveamendments to the rules proposed by the trusteecompany”.8 This in effect gives the JNC a vetoover any changes to the rules of USS.

9.3 The membership of the JNC comprises fivemembers appointed by Universities UK, fiveappointed by UCU, and an independent chairman.In default of agreement, the chairman has acasting vote, and if the chairman is not preparedto use that vote then a “built-in deadlock” ensues.This has led recently to a situation whereproposed changes to USS rules, in particularbringing in a later normal pension age, have notbeen able to proceed.

9.4 This is a difficult situation which will requirecareful and delicate handling, probably over along period, to resolve. Changing the structure ofUSS to reduce or remove the “blocking power” ofthe JNC would itself require the consent of theJNC, which in the circumstances is unlikely to beforthcoming. There seems little doubt thatchanges will be required to USS if the presentfinal salary pension model is to be sustained inthe higher education sector over the comingyears, and all parties will need to understand thatthese changes are essential rather than merelydesirable.

9.5 The second set of issues involving USS is therelationship between USS and the employers inthe higher education sector. This seems to bemore an issue of communication and style ratherthan substance. Both parties seem to wish toimprove the relationship and this will be bestdone by more frequent meetings and genuineattempts to understand the other’s point of viewand constraints.

9Universities Superannuation Scheme issues

15

16

10.1 Pension provision in the higher education sectoris, at present, mostly on a final salary basis, forgood reasons connected with pay andemployment patterns in the sector. Most of thesereasons remain valid, and it is clear that many ofthose operating in the sector regard final salaryprovision as an important part of theremuneration package, particularly for academicand academic-related staff.

10.2 The key issue for the sector is how final salarypension provision can remain affordable andsustainable at a time when life expectancy iscontinuing to increase, to an extent that manypeople who retire at age 60 can expect to spendnearly as long in retirement as at work. Manypension schemes in both the public and privatesectors have moved away from a retirement age of60 to one of 65, with retirement before 65 being ona cost-neutral basis. Some have made provisionfor retirement ages to continue to increase in thefuture if life expectancy continues to improve.

10.3 A sustainable pension scheme, in any sector,must be affordable to both the employer and themembers; must be fair as between differentcategories of member; and must provide areasonable pension for the long-serving retiree. Ifany of these criteria are not satisfied, then eitherthe employers or the members (or both) will losefaith in the scheme and it will eventually losecredibility.

10.4 A possible approach for the future could includesome of the following features:

p Normal pension age 65

p Early retirement (in normal health) on a cost-neutral basis

p Future improvements in life expectancy dealtwith by:

u automatic raising of normal pension age; or

u sharing of future cost increases; or

u adjusting pension benefits to reflect the longerperiod of payment; or

u some combination of these things.

10.5 There are numerous possible combinations of theabove factors – raising normal pension age,sharing future cost increases, and adjustingpension benefits, in the light of improving lifeexpectancy (generally described as “calibration”)– which could be explored with a view toimproving the sustainability of defined benefitprovision in the higher education sector. This is anarea where more work should be done,incorporating proper costings and considerationof alternatives.

10A possible long-term solution

Universities UK Pension provision

11.1 This report is a high-level overview whichattempts to define the main pension problemsfacing employers in the higher education sectorand to offer some options to be explored further inthe coming months.

11.2 The following areas are specificallyrecommended for further investigation:

p Discussion of the employee behaviours whichshould be rewarded and encouraged throughpension scheme design

p The advantages and disadvantages of a “menu”approach

p The possibility of one scheme for all academicstaff, based on USS

p Scheme mergers for non-academic staff

p Governance issues within USS

p Options for a long-term solution based on finalsalary benefits with calibration.

11Conclusion

17

18

SectionA head

TPS: http://www.teachernet.gov.uk andhttp://www.teacherspensions.co.uk

USS: http://www.usshq.co.uk

Career average A type of defined benefit pensionscheme in which pension is based on earnings ineach year of membership, revalued in line with anindex

Defined benefit A pension scheme in whichpension is based on pay, service, or other valuesfixed in advance

Defined contribution A pension scheme in whichpension is based on the contributions made andthe investment return which these have produced

DIUS The Department for Innovation,Universities and Skills

Final salary/final pay A type of defined benefitpension scheme in which pension is based onlength of service and pay at, or shortly before,retirement or leaving service

FSSU The Federated Superannuation System forUniversities, set up in 1913 and the forerunner ofUSS

Guild HE Representative organisation for highereducation colleges, specialist institutions andsome universities.

HE higher education

JNC The Joint Negotiating Committee of USS

LGPS The Local Government Pension Scheme

Money purchase Same as Defined Contribution

NHSPS National Health Service Pension Scheme

RPI Retail Prices Index

SAT Self-Administered Trust: used in thiscontext to mean a pension scheme operated by ahigher education institution for its own non-academic staff

SAUL Superannuation Arrangements of theUniversity of London

STSS Scottish Teachers’ SuperannuationScheme

TPS Teachers’ Pension Scheme

UCEA Universities and Colleges Employers’Association

UCU University and College Union

USS Universities Superannuation Scheme

UUK Universities UK

Websites

LGPS: http://www.lgps.org.uk

SAUL: http://www.saul.org.uk

Glossary

Universities UK Pension provision

1 Hewitt (2007) Strategic enquiry into the pension

arrangements for the higher education sector

2 Details of the new TPS contribution structure can be

found at www.teachernet.gov.uk

3 Communities and Local Government (2008) Sustaining

the local government pension scheme in England and

Wales. www.xoq83.dialpipex.com

4 Derived from data in BUFDG (2006) Twenty questions

about pension provision in higher education

5 Actuarial valuation as at 31 March 2005, SAUL website

6 Actuarial Valuation Report as at 31 March 2005, USS

website

7 BUFDG (2007) SAT and LGPS pensions schemes in UK

higher education: 2005–06 data

8 USS (2007) Report and Accounts for the year ended 31

March 2007

Notes

19

About Universities UK

This publication has beenproduced by Universities UK,which is the representative bodyfor the executive heads of UKuniversities and is recognised asthe umbrella group for theuniversity sector. It works toadvance the interests ofuniversities and to spread goodpractice throughout the highereducation sector.

Universities UKWoburn House20 Tavistock SquareLondonWC1H 9HQ

telephone+44 (0)20 7419 4111

fax+44 (0)20 7388 8649

emailinfo@UniversitiesUK.ac.uk

webwww.UniversitiesUK.ac.uk

© Universities UKISBN 1 84036 178 6May 2008