24
This article was downloaded by: [University of Tasmania] On: 27 November 2014, At: 23:21 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Behaviour & Information Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbit20 A multi-level socio-technical systems telecommuting framework France Bélanger a , Mary Beth Watson-Manheim b & Bret R. Swan c a Virginia Tech, Blacksburg, VA 24060, USA b University of Illinois, Chicago, IL, USA c Bringham Young University, Provo, UT, USA Accepted author version posted online: 25 Jun 2012.Published online: 30 Jul 2012. To cite this article: France Bélanger, Mary Beth Watson-Manheim & Bret R. Swan (2013) A multi-level socio-technical systems telecommuting framework, Behaviour & Information Technology, 32:12, 1257-1279, DOI: 10.1080/0144929X.2012.705894 To link to this article: http://dx.doi.org/10.1080/0144929X.2012.705894 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

A multi-level socio-technical systems telecommuting framework

  • Upload
    bret-r

  • View
    228

  • Download
    3

Embed Size (px)

Citation preview

This article was downloaded by: [University of Tasmania]On: 27 November 2014, At: 23:21Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Behaviour & Information TechnologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tbit20

A multi-level socio-technical systems telecommutingframeworkFrance Bélangera, Mary Beth Watson-Manheimb & Bret R. Swanc

a Virginia Tech, Blacksburg, VA 24060, USAb University of Illinois, Chicago, IL, USAc Bringham Young University, Provo, UT, USAAccepted author version posted online: 25 Jun 2012.Published online: 30 Jul 2012.

To cite this article: France Bélanger, Mary Beth Watson-Manheim & Bret R. Swan (2013) A multi-level socio-technical systemstelecommuting framework, Behaviour & Information Technology, 32:12, 1257-1279, DOI: 10.1080/0144929X.2012.705894

To link to this article: http://dx.doi.org/10.1080/0144929X.2012.705894

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

A multi-level socio-technical systems telecommuting framework

France Belangera*, Mary Beth Watson-Manheimb and Bret R. Swanc

aVirginia Tech, Blacksburg, VA 24060, USA; bUniversity of Illinois, Chicago, IL, USA; cBringham Young University, Provo, UT,USA

(Received 8 December 2010; final version received 18 June 2012)

Telecommuting can help to create organisational efficiencies and improve competitive advantage. It has been studiedfrom a variety of perspectives, including that of transportation, management, psychology, and information systems.However, telecommuting literature, while abundant and diversified, often reports contradictory results, creatingdilemmas for practice and research. Past researchers noting such conflicting findings often identify the lack ofguiding theoretical bases as a key problem. In an attempt to explain the contradictory results found in prior researchand in practice, we review telecommuting literature and expose conceptualisation issues that need to be addressed inthe development of a telecommuting research model: telecommuting as both a context and an aspect of work, as amulti-level concept and as a time-dependent concept. The proposed multi-level model, guided by socio-technicalsystems theory, illustrates the inter-relationships of telecommuting antecedents and outcomes across levels ofanalysis and over time. The research offers a number of important implications for future research, as well as formanagers involved in or affected by telecommuting in their organisations.

Keywords: telecommuting; telework; socio-technical systems; mobile work; multilevel; time

1. Introduction

Telecommuting has become more popular than everbefore due to the ubiquitous adoption of variousinformation and communication technologies (ICT)like mobile computing devices, wireless hotspots,broadband into homes, etc. It has been used inorganisations over the years to achieve organisationaland individual benefits. Recent estimates by GartnerDataquest suggest that 25% of workers in the UnitedStates telecommuted in 2007 (Joli 2009). Otherstatistics suggest that the number of Americans whotelecommuted at least one day per month for theiremployer increased from approximately 12.4 million in2006 to 17.2 million in 2008. This represents a 74%increase since 2005 (WorldatWork 2009). Similarly,while there were 2.2 million people telecommuting inthe United Kingdom in 2003 (Smith 2004), it isestimated that more than one-third of the UnitedKingdom workforce of 29 million telecommuted in2007 (Pollster_YouGov_UK 2007).

Telecommuting, in this article, focuses on the use ofICT to replace or substitute for work environments thatrequire individuals to commute to a traditional office(Belanger 1999). In line with other reviews of research inthis area (e.g. Belanger and Collins 1998, Bailey andKurland 2002), we use the term telecommuting inter-changeably with telework, which is a term more often

used in Europe due to an initial focus on job creation ascompared to a focus on environmental concerns in theUnited States (Sørensen et al. 2008). We are specificallyinterested in work performed away from the office byemployees related to an established organisation, whichincludes mobile work when tied to an organisationalcontext but not small business owners whose place ofbusiness is only located in their home.

Telecommuting growth is a result of advances inICT, as well as organisational attempts to increaseinternal efficiencies, improve competitive advantage,and respond to calls for flexibility in work hours(Pearce 2009, Hill et al. 2010). For example, mobiletechnology enables flexibility in the timing andlocation of work activities, and makes it easier toaccommodate work and family (Towers et al. 2006). Inaddition, work is becoming increasingly global, ascorporations outsource work activities or otherwisedistribute interdependent activities across far-flunglocations (Kuldeep et al. 2009). Performing work inthis environment means that employees often work athome during non-traditional hours to communicatewith their distributed colleagues.

Telecommuting has fundamentally changed theway organisations manage and conduct business, andhow individuals and groups interrelate to accomplishwork and reach desired work outcomes (Belanger

*Corresponding author. Email: [email protected]

� 2013 Taylor & Francis

Behaviour & Information Technology, 2013Vol. 32, No. 12, 1257–1279, http://dx.doi.org/10.1080/0144929X.2012.705894

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

1999, Pinsonneault and Boisvert 2001). Not surpris-ingly, then, telecommuting impacts cannot be consid-ered solely from the point of view of the ICT usage.Rather, telecommuting involves complex inter-rela-tionships between telecommuting work environments,individual motivations, management and work prac-tices, as well as ICT use.

Telecommuting literature includes studies from avariety of fields, such as transportation, management,psychology and information systems. Because of thisvariety, the literature reveals a myriad of impacts onindividuals, groups and organisations that createunique problems for managing and facilitating tele-commuting. This literature, while abundant, oftenreports contradictory results, creating dilemmas forpractice and future research. For example, conflictingresults are particularly evident in the effects oftelecommuting on productivity. On the one hand,telecommuting is reported to have positive impacts onproductivity at the individual, group and organisationlevels through such work outcomes as increasedperformance, decreased office costs, lower absenteeismand faster response times to customers (Belanger andCollins 1998, Igbaria 1999, Pinsonneault and Boisvert2001). However, at the same time telecommutingliterature reports negative productivity impacts onindividuals and groups, such as the absence of the bestworkers when needed, decreased individual productiv-ity when working at home, problems using ICT, and aloss of synergy at all levels in the organisation(Pinsonneault and Boisvert 2001, Ward and Shabha2001, Watson-Manheim and Belanger 2002).

Potential explanations for the conflicting resultsinclude methodological limitations, in particular thelack of theoretical bases underlying much of thisresearch (Belanger and Collins 1998, McCloskey andIgbaria 1998, Bailey and Kurland 2002). While a fewattempts have been made to apply organisationaltheory to the study of telecommuting (Ward andShabha 2001, Bailey and Kurland 2002, Workmanet al. 2003, Desrochers et al. 2005, Paez and Scott2007, Mayo et al. 2009, Virick et al. 2010), littleevidence has emerged to help explain the conflictingfindings found in the telecommuting literature.

The aim of this article is to develop a conceptualmodel of telecommuting using a socio-technicalsystems (STS) theory foundation. In doing so, wereview existing literature on telecommuting andidentify three conceptualisation issues we suggestshould be addressed in the development of a strongtheoretical framework for telecommuting: telecommut-ing is often considered a context only instead of anaspect of work, telecommuting is rarely considered as amulti-level concept and the time-dependency of tele-commuting outcomes is often ignored. In developing a

framework that addresses these issues, our articlecontributes to the literature by providing a new lens toexplore telecommuting research questions.

The article is organised as follows: first, we presenta brief overview of STS. We then review existingtelecommuting literature from an STS perspective andhighlight conceptualisation issues that emerge fromthis literature. Using STS as a guide, we present amulti-level conceptual model of telecommuting thataddresses these issues. We illustrate the use of themodel with data from two organisations in the hightechnology industry before concluding with recom-mendations for future research.

2. Socio-technical systems theory

Socio-technical systems theory has its roots in thesocio-technical systems view of organisations (Tristand Bamforth 1951, Katz and Kahn 1966). The theoryhas evolved through time as it has been used and testedby researchers from various fields (e.g. Markham 1988,Hendrick and Kleiner 2001, Holden and Karsh 2009).From an STS view, organisations are open worksystems that transform inputs to desired outputs (Tristand Bamforth 1951, Pasmore 1988, Hendrick andKleiner 2002, Morrison et al. 2005). A work systemconsists of two or more persons interacting using someform of job design, hardware and/or softwaremachine(s) or tool(s), and information/knowledgewithin structure or process(es) in both internal andexternal environments. Work systems and organisa-tions are considered open in that their boundaries arepermeable, allowing interactions with their environ-ment across levels of analysis (Katz and Kahn 1966).Thus, socio-technical systems can be as simple as aperson performing a task with a simple tool or ascomplex as a large number of individuals in a multi-national enterprise working together using advancedICT. Examples of applications of STS in the ISliterature include investigations of IT investment-decision processes (Ryan and Harrison 2000) anduser acceptance of ERP systems (Lim et al. 2005).

Socio-technical systems theory incorporates factorsfrom four elements critical to transforming worksystem inputs to outputs: technology-related factorsincluded in the technical subsystem, social and people-related factors included in the personnel subsystem,organisational structures and work processes includedin the organisational structure or work/job designsubsystem, and the environment external to the worksystem. These subsystems characterise the internal andexternal contexts in which people perform their work,as shown in Figure 1.

The technical subsystem includes factors represent-ing technologies, policies and practices that describe

1258 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

the modes of production (e.g. the type and level of ICTsupport for the work), the actions individuals take onan object when performing work (e.g. the tasksthemselves), the strategy for reducing uncertainty inthe process (e.g. policies or practices whether sup-ported by ICT or not), the degree of process/workflowintegration (e.g. the degree of automation or workflowrigidity), etc. (e.g. Brown 2002). Related to telecom-muting, the technical subsystem describes such factorsas the types of ICT used when telecommuting, thefacilities available to telecommute from, the rewardand compensation plans of the organisation, task/work design when telecommuting, etc.

The personnel subsystem includes at least threetypes of factors: demographic characteristics of theworkforce, psychosocial aspects of the workforce (e.g.dimensions of personality, attitudes towards the workenvironment or the work itself, individual motivations,etc.), and the degree of professionalism required toperform the work (e.g. values, norms or expectedbehaviour patterns of the job, team and/or organisa-tion). Outcomes of the personnel subsystem primarilydescribe the way tasks are performed. Related totelecommuting, factors that are considered part of thepersonnel subsystem can include workers’ motivationsto telecommute, attitudes towards the work whiletelecommuting, beliefs or expectations for reward,compensation and/or communication when telecom-muting, personality preferences for working alone or incollaboration with others, telecommuter work/lifebalance issues and telecommuter demographics.

The organisational structure subsystem is typicallycharacterised in terms of centralisation, formalisationand complexity. Centralisation refers to the level anddegree of formal decision-making in a work system(e.g. strategic, tactical or operational). Formalisationrefers to the degree to which jobs or tasks with a worksystem are standardized. There are two types ofcomplexity assessed in relation to the STS work system:differentiation and integration. Differentiation com-plexity takes three forms – vertical, horizontal orspatial – and refers to the degree to which a worksystem or organisation is segmented into parts.Integration complexity refers to the type and numberof mechanisms that are required in the work system toensure communication, collaboration and control ofthe differentiated elements in a work system. Ingeneral, the need for integrating mechanisms goes upas the degree of differentiation increases. Related totelecommuting, the organisational structure subsystemcharacterises such aspects of the work system describ-ing the number and degree of differentiation complex-ity existing in the organisation when telecommuting,the degree to which work tasks are standardised or adhoc (formalisation) when telecommuting and thelocation and degree of formal decision-making in thework system, (centralisation) when telecommuting.

The work environment describes the relevantcharacteristics of the context within which the worksystem operates (both internal and external at what-ever level of analysis). It is critical that work systemsand organisations be able to adapt to relevant factorsin their environment. Environmental factors thatpositively or negatively affect work systems in organi-sations can be socioeconomic, educational, political,cultural or legal. For each organisation and worksystem, these factors will differ in type, quality andimportance. For example, the external environment ofthe telecommuting work system may describe thepolitical climate in relation to stakeholders externalto the telecommuting work system being analysed, theregional or national culture in relation to trust andwork, the work-group, team or organisational tradi-tions for collaborating face-to-face vs. virtually, andthe legal requirements for transacting business, pro-tecting proprietary information or providing secureICT infrastructure when telecommuting. These sub-systems continually and jointly interact with each otherand both internal and external organisational environ-ments to produce work system outcomes and organi-sational survival (Trist and Bamforth 1951, DeGreene1973, Pasmore 1988, Hendrick and Kleiner 2002). Forthe purposes of developing the multilevel telecommut-ing framework, we focus on the internal subsystemenvironments (personnel, technical and structure) andlimit our discussion of environmental conditions

Figure 1. Basic socio-technical work system model.

Behaviour & Information Technology 1259

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

external to the organisation. Socio-technical systemssuggest that there is interaction between the subsys-tems. For example, the technical subsystem andorganisational structure conjointly affect the processof transforming antecedent inputs to outputs. Inreality, the three internal subsystems operate underthe principle of joint causation, meaning they areaffected by other causal events in the telecommutingenvironment (Katz and Kahn 1966, Hendrick 2002).This principle of joint causation leads to the STSconcept of the joint optimisation of a work system.

Joint optimisation refers to the potential or abilityof work systems to accomplish their intended trans-formative process(es). Typically, once designed, thetechnical subsystem is relatively stable, and therefore,it is often left to the personnel subsystem to adapt tothe organisational structure and external environment.For example, when telecommuting, workers may haveto continuously interact with the ICTs and with theorganisational structure (e.g. team coordination),when performing tasks and producing outcomes.Consequently, the way telecommuters perform theirtasks may be significantly altered over time whencompared to workers who perform the same tasks butare not telecommuting. Thus, we argue that interac-tions between the telecommuting personnel, technicaland organisational structure subsystems result inoutcomes for telecommuters at the multiple levels ofanalysis. Outcomes at the group and organisationallevels of analysis either emerge or are realised from theaggregation of individual level telecommutingoutcomes.

The concepts of joint causation and joint optimisa-tion highlight feedback that exists over time and acrosslevels of analysis as the work system seeks to reachstability or steady state. Thus, an STS perspectivesuggests that studies telecommuting should be con-ceptualised as multi-level and time-dependent. More-over, while this perspective enforces the need toexamine telecommuting as a context in which work istaking place, it also highlights the need to considertelecommuting as an aspect of the work itself. As wehave discussed, the way tasks are performed byemployees may be different when they are telecommut-ing than when they are not. We next conduct a briefsurvey of telecommuting literature focusing our atten-tion on these conceptualisation issues.

3. Telecommuting literature

There has been substantial research on telecommutingover the past decades as telecommuting started to gainacceptance in organisations. With few exceptions,empirical research on telecommuting started to appearin publications in the early 1980s. Rather than review

the early literature, we turn to several in-depth reviewspublished between 1998 and 2002 to summarise priortelecommuting research before examining more recenttelecommuting literature. Table 1 shows the mainfindings of these literature reviews. The reviewsconclude that telecommuting research largely hasbeen non-theoretical, highlighting the need for theore-tical frameworks. This is not surprising as much of theearly telecommuting research has taken the form ofinvestigating its advantages and disadvantages. Tworeviews note serious methodological weaknesses in theempirical studies conducted, including small samplepilot studies, only bivariate relationships studied, and afocus on cross-sectional surveys of users.

In Table 2, we present a samples of studiesconducted from 2002 onward. Our literature searchwas conducted using the ABI/Inform database and thesearch terms telecommuting and telework, as well asvariations such as telecommuter and teleworker. Thetable reveals a number of studies that explicitly addresstelecommuting or mobile work for employees relatedto an organisation at least one day per week.1 We alsoinclude a sample of recent papers addressing mobilework, where some time was spent working at home.2 Insection 3.1, we briefly discuss the findings from thesestudies. Then, in section 3.2., we analyse this literaturefrom an STS perspective.

3.1. Review of recent literature

In an effort to resolve some of the inconsistencies inprevious studies, Gajendran and Harrison (2007)conducted a meta-analysis of 46 studies in whichdata were collected in natural settings involving a totalof 12,883 employees. They find that telecommuting haslargely beneficial consequences, both proximal, e.g.increased autonomy and lower work-family conflict,and distal, e.g. job satisfaction and performance. Somenegative effects were found with high-intensity tele-commuting (more than 2.5 day/week). In particular,relationships with co-workers appeared to be harmed.Hunton and Norman (2010) found that teleworkprogrammes with the option of working at home orat a satellite office were associated with increasedorganisational commitment for employees while simi-lar programmes with only a full-time work at homeoption did not result in an increase. An ethnographicstudy by Whittle and Mueller (2009) found significantdissatisfaction with telework expressed by a team ofconsultants at a high-tech firm. However, the con-sultants espoused more positive feeling to their clients.

Other recent studies continue to investigate factorsinfluencing telecommuting adoption. Perez et al.(2004) investigate why individual adoption of tele-commuting has remained far below predicted levels.

1260 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

They argue that a stronger theoretical basis is neededto investigate problems in this domain and propose amodel of telecommuting adoption based the technol-ogy acceptance model (TAM). The model is multi-level, focusing on ease of use and usefulness oftelecommuting in the organisation (supply side)instead of the usual measures of frequency oftelecommuting by individuals (demand side). Severalrecent studies, i.e. Peters and Heusinkveld (2010) andHornung et al. (2009) have examined the role of themanager in the telecommuting adoption process.Peters and Heusinkveld (2010) found that the percep-tion of improvements in work outcomes and socialcosts/benefits vary among managers from different

‘occupational communities’, e.g. CEOs’ beliefs differfrom HR managers’. Their findings suggest thattelecommuting initiatives should pay much moreattention to the institutional environment and man-agers’ subcultures.

Khalifa and Davison (2008) investigate the useful-ness of the Theory of Planned Behaviour (TPB) toexplain continuance of telecommuting by individualsafter initial adoption. They find that habit cannegatively moderate the relationship between attitudetowards telecommuting and intention to telecommute.They also find that subjective norms, or the beliefs ofthe individual that important others think he shouldtelecommute, have a moderate effect on intention.

Table 1. Telecommuting literature reviews.

Authors Years covered Focus of review Main findings Concerns raised

Bailey andKurland (2002)

1984–2001 Who participates intelework?Why they do?What happenswhen they do?

Male professionals andfemale clericalworkers predominateNo clear evidencetelework increases jobsatisfaction andproductivity

Research needs to gobeyond individualteleworkersResearch shouldrevisit why peopletelecommuteResearch shouldemphasise theory-building using existingtheories

Belanger andCollins (1998)

1983–1996 Distributed workarrangements ingeneral; mostliterature is ontelecommuting,most commonform of sucharrangements

Organisations usetelecommuting to cutcosts and as incentiveto best employeesIndividual, work,organisational, andtechnology variablesmust fit to lead tosuccessful outcomes

Previous researchfocuses on bivariaterelationships onlyNo theoreticalunderpinningsNeed more empiricalresearchLittle hypothesis-driven research

McCloskey andIgbaria (1998)

1982–1995 Review of theempirical researchfor futuredirections

Telecommuting researchfocused on five areas:small pilot studies,extent oftelecommutingparticipation, beliefsand perceptions oftelecommuting(perceived impactsonly), advantages anddisadvantages oftelecommuting, andwork and familyissues

Lack formal definitionsof telecommutingMethodologicalweaknesses, i.e. smallor poorly constructedsamples, cross-sectional surveys onlyLack control ofextraneous variables,(i.e. employmentstatus, job type,telecommutingparticipation level)

Pinsonneault andBoisvert (2001)

1983–1999 Impacts oftelecommuting onindividuals andorganisations, andmanagerialimplications

Schedule flexibility, nointerruptions, andcommute savings asbenefits; professionaland social isolation asdrawbacks

No particular concernsraised

Siha and Monroe(2006)

1979–2002 Explore state oftelecommuting todate, identifyfailure and successfactors

Developed theoreticalmodel of antecedentsand outcomes ofsuccessfultelecommutinginitiatives

Need to betterunderstandorganisational,managerial andworker motivations

Behaviour & Information Technology 1261

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Others identify technology support as an importantfactor for employees to consider telecommuting (Bakeret al. 2006), highlighting the importance of training fornon-telecommuters.

Studies have also investigated work-related con-sequences of telecommuting for the individual, withfindings that teleworking makes it much more difficultfor employees to access tacit knowledge (Lee et al.2007), that communications among teleworkers isreduced with better collaborative technologies butincreasingly centralised around the manager (Belangerand Allport 2008), and that professional isolationamong teleworkers is negatively associated with jobperformance and increases with time spent on tele-working (Golden et al. 2008). Golden and Veiga (2005)investigate inconsistent findings in research on tele-commuter job satisfaction. They found a curvilinearU-shaped relationship between the extent of telecom-muting per week and job satisfaction, suggesting thatcertain jobs may become more difficult to performeffectively when telecommuting frequently. Golden(2007) found that prevalence of telecommuting isnegatively associated with non-telecommuting co-worker satisfaction; this relationship is influenced bythe amount of time co-workers telecommute, theextent of face-to-face interactions and job autonomy.O’Neill et al. (2009) found that there are differences incertain personality and motivational traits related toteleworker and non-teleworker effectiveness.

Researchers have continued to examine the inter-play of consequences of telecommuting in work andnon-work contexts. Golden (2006) found that therelationship between the extent of telecommuting andjob satisfaction was mediated by quality of interactionsin work-oriented and family-oriented relationships.Hartig et al. (2007) study the relationship betweentelecommuting and stress mitigation, and the potentialeffect on the restorative role of the home. Theirfindings indicate that teleworking women experienceless effective restoration than their non-teleworkingcounterparts while men experience more effectiverestoration. Similarly, Araujo (2008) found that theuse of ICT for to constructing space-time workpatterns has different implications for each gender.Results suggest that ICT is more useful to men thanwomen in organising their space-time work patterns.In the context of nomadic computing, Cousins andRobey (2005) also found that professionals skilled attechnology use were able to control the boundariesbetween their personal and business social contexts.

Other findings in this area include the fact that non-telecommuters have higher levels of work-familyconflict, time-based conflict, strain-based conflict(strain from work activities limits participation infamily activities) and higher perceptions of work

interference with family than telecommuters; andmale telecommuters have higher levels of behaviour-based conflict (behaviour effective in home-environ-ment is counterproductive in work environment) thanfemales (Madsen 2006). However, the use of mobiletechnology enables flexibility in the timing andlocation of the performance of work activities, andmakes it easier to accommodate work and family(Towers et al. 2006, Sørensen and Al-Taitoon 2008). Italso increases expectations of colleagues in the workplace and family members, leading to greaterworkload.

3.2. STS conceptualisation issues

The review reveals several limitations in the concep-tualisation of telecommuting: telecommuting as both acontext and an aspect of work; telecommuting as amulti-level concept and telecommuting as a time-dependent concept. These limitations for each studyare identified in the final column of Table 2. We nextdiscuss the findings from these studies in the context ofthese conceptualisations issues.

3.2.1. Telecommuting as an embedded aspect of work

Telecommuting is often considered the context orenvironment in which an individual’s work is carriedout, and not an aspect of the work itself. Theassumption seems to be that the same work will bedone in the same way at home facilitated by ICT as inthe office, but with less interruption (Westfall 2004).During telecommuting, the type and characteristics ofthe ICT artefacts are crucial to defining the context ofwork and its impacts on telecommuting work out-comes. In the context of telecommuting, use of ICTscan be both an enabler and a constraint to facilitatingdesirable work outcomes at several levels of analysis inorganisations. So, even though the setting where thework takes place is different, as well as methods ofaccessing information and interacting with colleagues,little research has examined how work is actuallytaking place (Orlikowski and Barley 2001). Fortu-nately, recent research is moving in this direction.Golden (2007) collects data from non-telecommutersto understand the implications of telecommuting onwork performance throughout the workplace, not juston individuals. In one of the few studies that begins toaddress the integration of a particular technology intothe work process, findings from Towers et al. (2006)suggest that the use of mobile technology enablesflexibility of timing and location of work, and makes iteasier to accommodate work and family but alsoincreases expectations of both spheres, leading togreater workload. Others identify both enablers and

1262 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

2.

Recentresearchontelecommuting.

Authors

Researchquestion(s)

Methodologyandsample

Main

findings

Conceptual

Lim

itations1

Kakihara

andSørensen

(2004)

Discuss

theem

ergence

ofthe

mobileprofessional

Adhocobservationand

interviews:62mobile

professionalsin

Tokyo;3

casespresentedin

thepaper

Needto

broaden

ourconceptionofmobility;

mobileprofessionalinvolves

locational,

operational,andinteractionalmobility

2,3

Perez

etal.(2004)

Developtheoretically-based

model

oftelework

adoption

Conceptual

Model

ofteleworkingadoptionbasedon

TechnologyAcceptance

Model

(TAM)

1

CousinsandRobey

(2005)

Exploredpractices

of

individualnomadic

computingusers

after

implementationofa

nomadic

computing

environment

1case

(financialinstitutionin

USA);observations,

documents,andinterviews:4

loanofficer;2interviewed

again

3monthsafter

change

Althoughusers

experiencedcontradictory

outcomes

asthey

soughtresolutionsto

the

dilem

masposedbywork

andnonwork

dem

ands,allusers

reported

effectivenessin

theircomputingpractices

1,3

Golden

andVeiga(2005)

Reconcile

inconsistentfindings

inresearchontelecommuter

jobsatisfaction

Survey:onefirm

,data

from

321telecommuters

Curvilinearrelationship

betweenextentof

telecommutingandjobsatisfaction.

Relationship

moderatedbytask

characteristics

1,3

JarvenpaaandLang

(2005)

Examineexperiencesofmobile

technologyusers

inHong

Kong,Japan,Finland,and

theUnited

States

33internationalfocusgroups

with222urbanmobile

devices

in4countries

Identifies

eightcentralmobiletechnology

paradoxes

thatshapeuserexperience

and

behavior;suggestpossible

designfeatures

thatrelate

totheparadoxes,anddiscusses

how

thesefeaturescould

bebettermanaged

2

Mazm

anianet

al.(2005)

Examineuse

ofwirelessem

ail

devices

amongmobile

inform

ationprofessionals

andtheirspouses,including

effects

ofsocialpresence,

physicalandvirtual

interactions,andpublic

environments

69semi-structuredinterviews

in3organisationswith

mobileprofessionals

People

differentiate

wirelesse-maildevices

from

other

communicationtechnologiesin

term

sofpatterns,norm

s,andexperience

of

use.Participants

report

constantand

sustained

interactionwiththeirdevices,at

allhours

andlocationsoftheirday

1,2,3

Baker

etal.(2006)

Istherearelationship

between

form

softechnologysupport

andem

ployee

reactionsto

telecommuting?

Survey:20firm

sin

Australia,

50respondents

who

telecommute

atleastoneday

per

week

Supportmore

closely

relatedto

technologyhas

more

impact

onreactionsto

telecommuting

thansupport

less

closely

relatedto

technology.

3

Golden

(2006)

Examines

role

relationships

playin

mediatinglink

betweenextentof

telecommutingandjob

satisfaction.

Survey:onefirm

,data

from

294telecommuters

U-shaped

relationship

betweenextentof

telecommutingandjobsatisfaction,

mediatedbyquality

ofinteractionsin

work-

orientedandfamily-orientedrelationships.

2,3

Madsen(2006)

Investigatesdifferencesin

work

familyconflictbetween

telecommuters

andnon-

telecommuters

Survey

oftelecommuters

and

non-telecommuters

atseven

firm

sin

MidwestUSA

Non-telecommuters

havehigher

levelsofwork-

familyconflict,time-basedconflict,strain-

basedconflict,andhigher

perceptionsof

work

interference

withfamilythan

telecommuters

do.Male

havehigher

levels

ofbehaviour-basedconflictthanfemales

telecommuters

3

(continued)

Behaviour & Information Technology 1263

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

2.

(Continued)

Authors

Researchquestion(s)

Methodologyandsample

Main

findings

Conceptual

Lim

itations1

Schweitzer

andDuxbury

(2006)

Assessstate

oftelecommuting

inCanadain

1999andin

2001

Survey

data

collectedin

1999

and2001from

over

20,000

employeesand6,300

employersin

Canada

Increase

innumber

offirm

soffering

telecommuting-from

3%

in1999to

7%

in2001.Participants:highly

educated,well-

paid,male

andfemale

knowledgeworkers

withdependentcare

responsibilities

1

Towerset

al.(2006)

Investigate

theshifting

boundaries

betweenhome

andwork

fornon-

telecommuters

usingmobile

technology

Survey

andinterviewdata

from

aCanadiangovernment

agency.845surveysand61

interviews

Mobiletechnologyenablesflexibilityoftiming

andlocationofwork.Easier

toaccommodate

work

andfamilybutincreases

expectationsofboth

spheres

andleadsto

greaterworkload.

3

BrodtandVerburg

(2007)

Identify

enablers

andbarriers

forsuccessful

implementationofmobile

work

inpractice

5cases;15in-depth

interviews

Adequate

skills,sufficientcommitmentanda

system

aticpreparationare

key

enablers

tothesuccessofmobilework

environments

Potentialbarriers

formobilework

environments

arise

mainly

from

thechanges

ofwork

processes

andwork

stylesofmobile

workers

2

GajendranandHarrison

(2007)

Studyconsequencesof

telecommutingfor

individuals;throughwhat

psychologicalmechanisms

doeff

ects

occur;under

what

conditionsdostrongest

effects

occur

Meta-analysisof46studiesin

naturalsettingsinvolving

12,883em

ployees

Telecommutinghaslargelybeneficial

consequences.Somenegativeeff

ects

found

withhigh-intensity

telecommuting(m

ore

than2.5

day/w

eek)

1,3

Golden

(2007)

Investigateswhether

prevalence

oftelecommuters

inanoffice

impactswork

outcomes

ofnon-

telecommuters

Survey

of240educatednon-

telecommutingprofessionals

inahigh-technology

companyin

theUS

Prevalence

oftelecommutingisnegatively

associatedwithco-w

orker

satisfaction;

relationship

influencedbyamountoftime

co-w

orkerstelecommute,level

ofFTF

interactionsandjobautonomy

3

Hartig

etal.(2007)

Investigate

tradeoffsbetween

stress

mitigationof

teleworkingandpossible

effectonrestorativefunction

ofhome

Survey:58teleworkersand43

non-telew

orkersin

one

Swedishcompany

Telew

orkersandnon-telew

orkersexperience

homeas

place

ofrestoration;telew

orking

women

reportless,andmen

more,effective

restorationthannon-telew

orkingcounterparts

2,3

Lee

etal.(2007)

Examinegapbetweentelework

andcentralwork

inaccessingtacitknowledge

anduse

ofcommunication

media

forknowledge

exchange

Studyin

seven

largeJapanese

companies:58survey

responsesand35

interviewees:telecommuters,

andmobileandsatellite

office

workers

Significantdiscrepancies

betweentelework

and

central-work

inknowledgeaccessibilityand

inpatternsofmedia

usageforknowledge

access.Telew

orkingmadeitmore

diffi

cultto

accesstacitknowledge

1,3

Sanchez

etal.(2007)

Explorestherelationship

betweenteleworking

adoption,workplace

flexibility,andfirm

perform

ance

Structuredinterviews:

managersat479sm

alland

medium-sized

firm

sin

northwestSpain

Firm

perform

ance

ispositivelyrelatedto

teleworkingprograms.Telew

orkingfirm

suse

more

flextime,

havemore

employees

involved

injobdesign,are

more

intensively

managed

byresultsanduse

more

variable

compensation

2,3

(continued)

1264 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

2.

(Continued)

Authors

Researchquestion(s)

Methodologyandsample

Main

findings

Conceptual

Lim

itations1

Araujo

(2008)

Investigateshow

university

lecturers

dealwithspace

and

timein

professional

activities.

Interviews:38university

lecturers

intw

oPortuguese

universities

Use

ofIC

Tto

construct

space-tim

ework

patternshasdifferentim

plicationsforeach

gender.

1,3

Axtellet

al.(2008)

Looksatdifferencesbetween

train-basedmobilework

and

office

environments.Explore

themain

impactsoftrain

mobilework

ontasks

Surveys:350in

UK

plus19

interviews

Majority

oftasksconducted

ontrain

are

sociallyindependentin

nature

(withoutthe

needforcommunicationwithothers).

People

maketechnologicaltask

and

contextualadaptationsto

allow

them

towork

toconduct

sociallyinterdependent

work

(needforcommunicationwithothers)

1,2,3

Belanger

andAllport

(2008)

Explore

theeff

ects

of

improvem

ents

intechnology

forteleworkerson

communicationspatterns

Case

study:6monthspre

and

post

technologychange

surveysandinterviewsof13

teleworkersandtheir

managem

ent

Communicationfrequency

amonggroup

mem

berswasreducedafter

thetechnology

improvem

entbutbecamemore

centralised

aroundthedistrictmanager

1

Golden

etal.(2008)

Examineprofessionalisolation

ofteleworkersandthe

relationship

withjob

perform

ance

andturnover

intention

Survey:matched

sample

of261

professional-level

teleworkersandtheir

managersin

onehigh-tech

firm

Telew

orker

professionalisolationisnegatively

associatedwithjobperform

ance;im

pact

isgreaterwithmore

timespentteleworking.

Unexpectedly,professionalisolationis

relatedto

lower

turnover

intention

1,3

Khalifa

andDavison

(2008)

Investigate

theapplicabilityof

TPBto

explain

intended

continuance

level

of

telecommuting

Survey:101telecommuters

incompaniesin

North

America

Attitudehassignificanteff

ectonintentionbut

isnegativelymoderatedbyhabit;subjective

norm

sexertmoderate

effectonintentionto

telecommute

2,3

Sanchez

etal.(2008)

Analyse

contributionofHR

throughadoptionof

telework

orworkplace

flexibilitypractices

tofirm

perform

ance

Survey:HR

managersand

CEOsin

Spanishfirm

s,156

responses

Accessto

HR

commitmentpractices

leadto

intensity

oftelework

adoption,whichis

positivelyassociatedwithfirm

perform

ance

2

Wajcmanet

al.(2008)

Investigate

propositionthat

mobilephonedissolves

boundaries

separatingwork

andhome,

thusextending

thereach

ofwork

Data

collectedfrom

individuals

aged

15years

andolder

inhouseholdsin

Australia,

from

Marchto

May2007,

usingaquestionnaire,

aphonelogandatime-diary

Resultsindicate

thatmobilephoneisnot

primarily

awork

extensiondevice.

The

volumeofwork-relatedtrafficoutsideof

hours

ofem

ploymentwasfoundto

below

andthemain

usesofthemobilewerefor

contactingfamilyandfriends

1,2,3

HislopandAxtell(2009)

Explore

multi-locationwork

of

teleworkers

Interviews:18consultantsfrom

2sm

allUK

consultancies

Spacesusedbymulti-locationworkersvary

significantlyfrom

theworkplacesofworkers

whowork

predominantlyfrom

asingle

location;multi-locationworkerstypically

haveto

investmore

effort

increatingand

producingaworkplace

inthelocationsthat

theirwork

takes

them

to

2,3

(continued)

Behaviour & Information Technology 1265

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

2.

(Continued)

Authors

Researchquestion(s)

Methodologyandsample

Main

findings

Conceptual

Lim

itations1

Hornunget

al.(2009)

Studyofidiosyncraticdeals(i-

deals)whereworkersshape

theirem

ployment

arrangem

ents

through

negotiatingindividualised

employmentconditions

facilitatedbytelecommuting

Survey:263supervisors

managingtelecommuting

employeesin

Germanpublic

administration

Resultsindicate

thatsupervisors

canbe

motivatedto

authorise

effectivei-dealsin

the

contextofbroader

organisationalprograms

thatlegitim

ate

andsupporttheirnegotiation

2,3

Mayoet

al.(2009)

Usingacontingency

perspectiveanddata

from

122CEOsofSpanishfirm

s,thispaper

examines

what

makes

afirm

likelyto

adopt

telecommuting

Survey:102Spanishand

PortugueseCEO’scompared

topublicbusinessdata

Empiricalevidence

showed

thattelecommuting

iscorrelatedwithsm

allorganisationalsize,a

highproportionofinternationalem

ployees,

andvariable

compensation

2,3

O’N

eillet

al.(2009)

Explore

personality

and

motivationaltraitsrelatedto

teleworker

perform

ance

and

satisfaction

Survey:156em

ployeesfrom

8organisationsin

Canada:78

teleworkersand78non-

teleworkers

Certain

personality

andmotivationaltraitsare

relatedto

teleworker

andnon-telew

orker

effectiveness,butsometraitsshow

differentialvalidity.Thereare

also

situationaldifferences(children,tenure,etc.)

1,2,3

WhittleandMueller

(2009)

Attem

ptto

further

understand

how

work

isconducted

when

usingflexible

working

practices

Ethnographic

studyover

three

years

usingobservationand

semi-structuredinterviewsof

10managem

entconsultants,

allmem

bersofateam

Identify

a‘gap’betweenthevisionsofseamless

knowledgesharing,synergisticvirtual

teamwork,andstrongsocialnetworks

producedbytheconsultants

forthebenefit

ofclients

andthetalesofisolation,

disconnection,disaffection,andcynicism

observed

when

clients

werenotpresent.

More

specifically,theauthors

highlightthe

politics

involved

intherepresentationof

flexibility

1

Hillet

al.(2010)

Explore

theinfluence

of

workplace

flexibilityon

work-lifeconflict;

specificallyexamine

flexibilityin

where(w

ork-at-

home)

andwhen

(perceived

schedule

flexibility)workers

engagein

work-relatedtasks

Survey:globalsample

of

workersin

75countries

(n¼

24,436)

Across

allfourgroupsofcountries,thebenefit

ofwork-at-homeisincreasedwhen

combined

withschedule

flexibility.

Workplace

flexibilityisbeneficialboth

toindividuals(intheform

ofreducedwork-life

conflict)

andto

businesses(intheform

of

capacity

forlonger

work

hours).However,

work-at-homemaybeless

beneficialin

countrieswithcollectivistcultures

1,2

HuntonandNorm

an

(2010)

Investigate

impact

oftelework

arrangem

ents

on

organisationalcommitment

andrelationship

between

telework

arrangem

ents,

organisationalcommitment,

andtask

perform

ance

Longitudinalstudy(2

years)of

160participants

(medical

coders)

in5hospitals:

telework

arrangem

ents

across

hospitalsdiffered

with

onecontrolgroup;

unbalancedrandomised

26

26

1design

Participants

inthreeofthetelework

arrangem

ents

exhibited

significantincreases

inallaspects

ofcommitment.In

groups

workingexclusivelyathome,

organisational

commitmentwasequivalentto

thecontrol

groups.Foundorganisationalcommitment

mediatesrelationship

betweenthetelework

arrangem

ents

andtask

perform

ance

1

(continued)

1266 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

barriers to mobile work. Importantly, mobile work canbe hampered by existing business processes (Brodt andVerburg 2007) and mobile workers need to be creativein developing their work spaces (Hislop and Axtell2009). For certain jobs, however, mobility may allowboth individual and organisational needs to be met(Kakihara and Sørensen 2004, Sørensen and Al-Taitoon 2008).

In reality, the act of telecommuting may enable newand evolving conditions on the nature of the workitself that must be taken into account. As others havepointed out, ‘Under a new conceptualisation, telework[telecommuting] might come to be seen as one of manymechanisms individuals enact to cope with thedemands of the modern workplace’ (Bailey and Kur-land 2002).

3.2.2. Telecommuting as a time-dependent concept

An important issue in telecommuting research is theamount of time spent on telecommuting (Ramsower1985, McCloskey and Igbaria 1998, Belanger et al.2001) as telecommuters’ experiences with telecommut-ing will impact their attitudes towards this workarrangement and subsequent outcomes, although it isnot often explicitly identified in most studies. Thepassing of time can affect productivity and adoption oftelecommuting by individuals. For example, initially aworker may experience the flexibility enabled bytelecommuting as a positive experience. Yet, challengesmay surface over time, e.g. the inability to get timelyfeedback from people, or difficulties working withtechnology, that may reduce the worker’s propensity totelecommute and lower her satisfaction withtelecommuting.

The frequency of telecommuting also appears toinfluence satisfaction with telecommuting. While manystudies acknowledge variation in frequency, e.g.number of days per week, few have investigated theinfluence on outcomes. In one of the few studies toexamine this question, Golden and Viega (2005) founda curvilinear U-shaped relationship between thefrequency of telecommuting per week and job satisfac-tion. The relationship was moderated negatively byincreased task interdependence and positively by lowlevels of job discretion. The researchers suggest certainjobs may become more difficult to perform whenfrequency of telecommuting is increased.

Many theories offer ‘snapshots’, focusing on a one-time experience with no accounting for the compound-ing effects over time. Few researchers have taken thisissue into account, although some, e.g. Cousins andRobey (2005), Scheitzer and Duxbury (2006), Belangerand Allport (2008), did collect data from telecommu-ters at two points in time. Time can be investigatedT

able

2.

(Continued)

Authors

Researchquestion(s)

Methodologyandsample

Main

findings

Conceptual

Lim

itations1

PetersandHeusinkveld

(2010)

Examinetherole

ofmanagers

inthetelework

adoption

process;especiallyinfluence

ofinstitutionalcontexton

attitudetoward

telecommuting

Survey:96CEOsand380HR

managersin

Dutch

organisations

Perceived

improvem

entsofwork

outcomes

and

socialcosts/benefits

vary

amongmanagers

from

differentoccupationalcommunities,

e.g.,CEOs’beliefs

differ

from

HR

managers’.Telecommutinginitiatives

should

paymuch

more

attentionto

the

institutionalenvironmentandmanagers’

subcultures

2,3

Notes:

1Weidentified

twolimitationssuggestedbytheSTStheory

when

studiesdid

nottakeinto

accountthemulti-level

nature

oftelecommuting(3)andwhen

studiesdid

nottaketheeff

ectof

timeinto

consideration

(2).

Inaddition,weidentified

aconceptuallimitation

through

theliterature

review

forstudiesthatconsidered

telecommutingonly

asacontextasopposed

toconceptualisingtelecommutingasanem

bedded

aspectofwork

(1).

Behaviour & Information Technology 1267

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

from a diachronic perspective, i.e. across differentperiods, or from a synchronic perspective, i.e. withinone period. Both are important in telecommuting.Recent studies have begun to investigate changingtemporal patterns of work when telecommuting(Towers et al. 2006, Araujo 2008).

Even though there are few studies acknowledgingthis, the effect of time on telecommuting is central tounderstanding the implications of such work arrange-ments since time is required for most effects to occurwithin or across levels of analysis. For example, theimpacts of telecommuting on a team may not beobserved until individual team members have alreadyexperienced telecommuting over time. Alternatively,different individuals on the team may have differenttemporal patterns of work that may have to bereconciled. For that reason, a telecommuting concep-tual model should take into account compoundingeffects of telecommuting over time.

3.2.3. Telecommuting as a multi-level concept

Telecommuting research often fails to recognise multi-ple levels of analysis. Yet, outcomes for telecommutershave effect on and are affected by their co-workers’,managers’, teams’ and subordinates’ outcomes (Pearl-son and Saunders 2010). Consequently, researchersshould investigate the multiple levels of effects oftelecommuting (Perez et al. 2004). For example, tele-commuting arrangements are commonly expected tomotivate employees and reduce organisational costswhile improving individual and organisational produc-tivity (Leonard 2000, Bailey and Kurland 2002). Yet,most prior research has been conducted at either theindividual or organisational level of analysis (Bailey andKurland 2002). We found few examples of studiesconsidering these multiple effects concurrently althoughsome studies implicitly recognise the multi-level natureof telecommuting in general (Jarvenpaa and Lang 2005)or its multi-level outcomes (e.g. Sanchez et al. 2008,Whittle and Mueller 2009). Most organisational andgroup level outcomes studied in telecommuting researchtend to be simple aggregations of individual-leveltelecommuting experiences (Pinsonneault and Boisvert2001). For example, some studies measure factors at theindividual level but make inferences at the organisa-tional level (Cousins and Robey 2005). The emphasis onthe individual in previous research is perhaps notsurprising as the phenomenon of telecommuting chal-lenges our understanding of an organisation. Since theindustrial age, a core tenant of the business organisationhas been the separation of the individual’s home fromthe workplace. Thus, much research has been concernedwith the impact of this major change on the individual(Orlikowski and Barley 2001).

Another interesting aspect that is rarely studied inrelation to telecommuting is the group as a level ofanalysis, even though research is needed at this level(Bailey and Kurland 2002). Some studies examinetelecommuters’ relationships with co-workers and findchanges in communication patterns among groupmembers (Belanger and Allport 2008) or negativeimpacts on co-worker relationships with increasedlevels of telecommuting (Gajendran and Harrison2007, Golden 2007). No telecommuting studies, how-ever, seem to focus specifically on the group level ofanalysis. It is not clear why there has been so little suchresearch. It may be that initial research on telecom-muting in the USA was in the transportation field,where emphasis was on individual travel patterns andenvironmental impacts of telecommuting. Yet, virtualteams are increasingly important in organisations, andemployees are almost always members of at least one,and often, multiple teams (Belanger and Watson-Manheim 2006). Some research areas do focus ongroups in distributed settings: Computer SupportedCooperative Work (CSCW) and virtual teams. Yet,both CSCW and virtual team literatures do notspecifically address telecommuting. Additionally,CSCW research often focuses on the technology itself,such as interfaces for distributed work. Virtual team’sliterature has not focused on the location of individualin general or telecommuting in particular; researchershave primarily investigated antecedents and emergentprocesses influencing team effectiveness (e.g. review byMortensen et al. 2009).

The lack of emphasis on groups in telecommutingliterature may have obscured some important tele-commuting relationships. For example, what might bethe consequences for group performance if onemember is at home? Alternatively, what might be theconsequences to companies if individuals in severalgroups are simultaneously working remotely? In highlydistributed companies, the consequences may bedifferent than in more traditional firms where mostemployees are at the same site. Often, employees havedifferent levels of telecommunication access fromhome, potentially affecting the performance of theoverall group. For example, some groupware degradesto the lowest network characteristic or speed. In otherwords, in a group where one employee is on a poorconnection, it will have negative effects on the entiregroup’s access speed. While this represents only oneexample, it is clear that the lack of emphasis on grouplevel effects of telecommuting may obscure theexistence of compounding effects of individual factorson work outcomes across levels of analysis in anorganisation.

A few studies have examined firm-level conse-quences. Sanchez et al. (2008) found that access to

1268 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Human Resource commitment practices leads toincreased intensity of telework adoption, which ispositively related to firm performance. Sanchez et al.(2007) found that firm performance is positivelyrelated to the existence of teleworking programmes.In a study with respondents spread across 75 countries,Hill et al. (2010) found that workplace flexibility isbeneficial to individuals (due to decreased work-lifeconflict) and to firms (increased capacity for longerwork hours).

In summary, there are potential consequences todrawing telecommuting research conclusions withoutconsidering multi-level factors. For example, in con-sidering multi-level issues in telecommuting, there isthe potential of misaligned incentives across levels.Employees may be rewarded for individual perfor-mance but expected to be available as a ‘team player’by their manager, which also may be an importantfactor affecting promotion.

3.2.4. Summary

The brief review of recent literature likely does notshow the full extent of research that has been publishedon telecommuting. Yet, in looking at both pre-2002and recent telecommuting work, it is clear that issuesare slowly emerging in the literature while the socialand organisational environments are changing rapidly.These changes might be the result of individualsworking longer and longer hours, or our transition tothe digital economy where it is easier to transport workwherever one is, as long as ICTs are available. Indeed,the increased use of mobile technologies to commu-nicate and access information has accelerated theblurring of lines between work and non-work activities(Kakihara and Sørensen 2004, Cousins and Robey2005, Towers et al. 2006, Axtell et al. 2008).

Orlikowski and Barley (2001) point out that ITpractitioner-focused literature has been consistentlyoptimistic that the pervasive use of computers, andespecially access to the Internet, has created conditionsthat should lead to substantial increases in the numberof telecommuters. On the other hand, organisationaltheorists are pessimistic about the spread of telecom-muting; pointing out that the work practice is contraryto well-established social and cultural understandingsof how work takes place. Orlikowski and Barley (2001,p. 157) conclude: ‘Yet despite these insights, neitherview comes to grip with the social dynamics oftelecommuting because neither has investigated howpeople integrate telecommuting into their daily lives’.This is particularly important as social and organisa-tional changes are taking place that further inserttelecommuting into people’s daily lives. For example,as more corporations have activities that span the

globe, people increasingly work non-traditional hours,and much of this work takes place at home.

Despite these significant changes in the way work isconducted, findings from our literature review indicatethat researchers continue to extensively focus on jobsatisfaction of telecommuters, adoption of telecommut-ing by individuals and organisations, impact on jobsatisfaction after adoption and the relationship to work-family conflict. This may not be an unexpected trend; asOrlikowski and Scott (2008) point out, managementresearch generally does not consider materiality of anykind and IT in particular.3 Moreover, there is usually atime lag between the surfacing of changes in theorganisational work environment and the developmentof academic research addressing these new realities(Belanger et al. 2002). However, we believe a corelimitation of this research continues to be the lack oftheoretical grounding due in part to the inherentcomplexity of the phenomenon. Telecommuting sym-bolises a major shift in understanding of work andprofessional activities for individuals, organisations andsociety, in addition to representing the deep penetrationof ICT usage into the fabric of life at each level.Managers need to be better equipped to deal with anduse telecommuting strategically in their organisations.Moreover, enterprise-wide use of wireless technologyand mobile devices will challenge management practicesand stimulate changes in organisational design toaccommodate the increasing ability to work anywhere,including remote locations (Sørensen et al. 2008).

This review of the telecommuting literature from anSTS perspective reveals that important conceptualisa-tion issues are implied but not typically considered byresearchers, and highlights the lack of theoreticalfoundations in this domain. The next section thereforeaddresses the call for increased use of establishedorganisational theory to guide future research andexplain past paradoxes (Bailey and Kurland 2002) byproposing a multi-level conceptual model oftelecommuting.

4. A multi-level socio-technical systems telecommuting

framework

Socio-technical systems theory is used to guide thedevelopment of the multilevel framework. Socio-technical systems is first utilised to define and integratedifferent aspects of telecommuting specific to thetelecommuters’ personnel subsystem, technical subsys-tem and organisational structure. It is also used toexplicitly identify and account for the effects of bothtime and multi-level factors that impact telecommutingwork systems.

Socio-technical systems theory can be used totheorise and analyse how multiple factors jointly

Behaviour & Information Technology 1269

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

influence telecommuter work outcomes and workermotivations to telecommute over time. Additionally, itcan be used to conceptualise how telecommutingresults in multi-level outcomes by jointly impactingindividual, group and organisational level socio-technical factors – such as individual-level satisfac-tion/dissatisfaction, team-level productivity and orga-nisational-level absenteeism.

As discussed earlier, consequences of telecommut-ing at the group and organisational level have beenunder researched as compared to the individual-level.As interdependent tasks have become more spatiallydistributed, the need to understand the group-levelimplications becomes more important. Recent researchindicates that the configuration of the work group mayinfluence communication patterns and ultimatelyperformance (O’Leary and Mortensen 2010). Theimplication of different configurations, e.g. one iso-lated person as compared to multiple isolated people,is just one area of group level analysis that could beexplored. In addition, many firms approach telecom-muting from a strategic perspective, but little researchhas addressed organisational-level implications. Wepropose to address these limitations with the Multi-Level Telecommuting Framework presented in Figure2, which is further described below.

In our framework, we use the concepts of jointcausation and joint optimisation to highlight feedbackthat exists over time and across levels of analysis as thework system seeks to reach stability or steady state.The principles of joint causation and joint optimisationcan be represented and operationalised with theconcept of fit (Venkatraman 1989). For example, ininformation systems, fit has been used to study the

impact of task-technology fit on outcomes (Vessey andGalletta 1991, Goodhue and Thompson 1995). Con-tingency theory researchers use the concept of fit tolook at the inter-relationships of constructs and theirimpacts on various variables. Related to telecommut-ing, in a generic model of outcomes in distributed workarrangements, Belanger and Collins (1998) usedVenkatraman’s conceptualisation of fit as gestalt toconsider the internal coherence between a largenumber of attributes (e.g. individual, organisational,work and technology characteristics) whose interac-tions could not be precisely formulated. In thedevelopment of our multi-level STS telecommutingframework, we propose that fit represents the interac-tion between the STS subsystems and their environ-ment across levels of analysis.

The framework depicts high-level relationshipsbetween the theoretical constructs of the telecommut-ing environment. In the framework, dashed linesportray the individual, group and organisational levelsof analysis and represent the permeability of thesecomponents across levels of analysis. In most complexorganisational settings (a.k.a., the real world), thetelecommuting work environment is an amalgam ofindividual, group and organisational contexts. Eachindividual telecommuter will therefore experience thiscombination of contexts differently. Conceptually,however, the telecommuting subsystems should bedepicted as existing at different levels of analysis(Rousseau 1985, Klein et al. 1994, Hendrick andKleiner 2001). In the model, subsystems are shown asinterconnected, illustrating the principles of jointcausation with outcomes depicted at the individual,group and organisational levels of analysis.

Figure 2. A multi-level STS telecommuting framework.

1270 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

At all levels of analysis, the framework specifiestelecommuting personnel, technical and organisationalstructure subsystems that result in telecommutingwork outcomes. Outcomes of telecommuting propen-sities, motivations, behaviours, processes and struc-tures at the individual level of analysis impactoutcomes across levels of analysis. The same is trueabout the outcomes of telecommuting propensities,behaviours, processes and structures at other levels ofanalyses. These outcomes, over time, will affect themulti-level STS factors that impact future multi-leveltelecommuting behaviours and outcomes.

The framework can help researchers distinguishmulti-level antecedents and outcomes of telework toidentify areas of importance to be investigated andidentify gaps where significant topics are not beingaddressed. In Table 3, we illustrate the use of theinternal subsystems of STS to show how antecedentsand outcomes can differ across multiple levels ofanalysis in telecommuting. For example, when investi-gating the performance of work activities by telecom-muters, the framework helps differentiate antecedents ofperformance at the individual-level from the group andorganisational levels, and it also makes clear thatoutcomes at the individual-level will only partiallyexplain overall performance implications.

5. Illustrating the multi-level telecommuting

framework

An example of the applicability of the multileveltelecommuting framework can be derived from datawe collected using semi-structured interviews to under-stand communication patterns and media choices inwork groups with distributed team members. The datawere collected from two, Fortune 100 high-tech ICTsales organisations where employees were given theoption to telecommute. For detailed explanations seeWatson-Manheim and Belanger (2007).

5.1. Firm A

Firm A has a culture that relies heavily on emailcommunication, and has a loose organisational struc-ture with sales people and technical workers interfacingon multiple virtual project teams. There is a general lackof clarity in roles and responsibilities on these virtualteams. Workers depend on their social networks toknow who, where and how to access necessaryinformation and to share knowledge. The amount ofrelationship development (referring to the personnelsubsystem) varies widely depending on what project aworker is involved in at the time. The structuralsubsystem includes high task interdependence due toindividuals working on multiple project teams. Thus,

the number of meetings and the project managementprocesses vary widely. Training on general companyissues is provided, but little training exists for howworkers can best access information and share whiletelecommuting (technical subsystem). Few workerspossess permanent offices. Instead, hotelling facilitiesare set up in remote locations where workers canschedule an office or work from home (resourceenvironment). At the organisational level, the onlineemployee directory and Firm A’s intranet are reportedas good sources of general information. In addition, thevariety of ICT options and support are reported aspositive motivators at the group and organisation level.

Despite organisational support for telecommuting,workers in Firm A report a general lack of desire andfrustration with this arrangement. Employees com-plain that the formal training received was not useful.Telecommuting was also reported to result in poorinformation gathering, lack of knowledge sharing andlittle social network development. There was a perceivedlack of training at the group and organisation levels onhow to access and share information effectively. At thegroup level, lack of interaction with peers resulted inlimited knowledge sharing within groups, and difficultyin finding and talking to people when needed sometimesresulted in inefficient teamwork. At the individual level,most issues stemmed from individuals’ lack of ability todevelop relationships and be ‘plugged in’ to the networkso that they could do their work effectively andefficiently. Working remotely is perceived to createbarriers to building social networks needed to get thework done. In addition, a perceived lack of clarity ofroles and responsibilities in groups is reported to makework more difficult. Finally, there were positive out-comes for individuals in Firm A. Workers enjoyed theflexibility to work at home and reduced travel distancesand costs.

Firm A illustrates the impacts that organisational,group and individual level factors can have on individualworkers choosing to telecommute, as shown in Figure 3.This organisation illustrates that the real barrier totelecommuting in Firm A is the lack of ability to buildknowledge and social networks that allow workers toaccess and share information in this highly taskinterdependent environment. This effect occurs overtime as individuals realise how difficult it is to developtheir networks of experts when they are not physically inthe office. The overall result is that, although telecom-muting is supported by the culture and ICT, Firm Aworkers generally preferred not to telecommute.

5.2. Firm B

Firm B sells total ICT solutions to primarily one client.Therefore, many of the sales and technical workers’

Behaviour & Information Technology 1271

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

3.

Sample

telecommutingantecedents

andoutcomes

across

levelsofanalysis.

STSsubsystem

Individuallevel

ofanalysis

Grouplevel

ofanalysis

Organisationallevel

ofanalysis

Personnel

(Factors

relatedto

personal,socialandcultural

characteristics

orissues

affectingtelecommuters)

Antecedents:

.Individualcharacteristics

such

as

introversionorextraversion

.Suitabilityofhome-life

situation

fortelework

.Understandingofhow

toaccess

the

inform

ation

and

knowledge

needed

toperform

work

.Individual

satisfaction/dissatis-

factionwithsocialfactors

Antecedents:

.Work

relationshipsthat

facilitate

groupwork

outcomes

.Groupculturalnorm

s.Groupknowledgenetworking

.Understandingofwhoisdoing

what

.Understandingofwho

knows

what

Antecedents:

.Perceptionsoftelecommuters

inorganisation

.Organisationalculturalnorm

s

Outcomes:

.Work

vs.home-life

balance

issues

.Professional

relationship

developmen

t.Feelingsofisolation

.Individualtask

perform

ance

Outcomes:

.Professionalrelationship

developmentopportunities

.Groupcommunication

.Groupinform

ationand

knowledgeexchange

.Groupperform

ance

Outcomes:

.Professional

relationship

de-

velopmentopportunities

.Socialnetwork

impacting

ac-

cess

tonecessary

inform

ation

andrecognition

.Organisationalperform

ance

Technicalandwork

(Factors

describinghow

work

isperform

edandthetypes

and

availabilityofIC

Tandother

resources

needed

toperform

work)

Antecedents:

.Individualsatisfactionanddissa-

tisfactionwithtask

design

.Individualtask

design

.Managerialcontrolpractices

.Types

andavailabilityofIC

T(i.e.

ITartefacts)

.Characteristics

or

properties

of

theIT

artefacts(i.e.IC

T)

.Level

ofuse

ofIC

Trequired

.Rew

ard

andcompensation

strategies

.Availability

ofpermanentoffice

space

Antecedents:

.Grouptask

design

.Use

ofIC

Tto

facilitate

group

work

andcollaboration

.Groupdecision-m

aking

processes

.Managerialcontrolpractices

.Group

ICT

available

(i.e.IT

artefacts)

.Characteristics

orproperties

of

theIT

artefacts(i.e.IC

T)for

groupwork

.Groupreward

and

compensation

Antecedents:

.Accessibilityto

largeamounts

ofinform

ation

.Security

policies

.IC

Toperatingprocedures

.Value-addingbusiness

processes

.OrganisationalIC

Tavailable

(i.e.IT

artefacts)

.Characteristics

orproperties

of

theIT

artefacts(i.e.IC

T)that

support

oraffectthe

organisationduring

telecommuting

(continued)

1272 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Table

3.

(Continued)

STSsubsystem

Individuallevel

ofanalysis

Grouplevel

ofanalysis

Organisationallevel

ofanalysis

.Availabilityofsupport

and

training

.Availabilityofother

workers

when

needed

.Satisfaction/dissatisfactionwith

compensationandresources

available

.Telew

ork

training

.Level

ofavailability

ofgood

workerswhen

needed

.Support

andtrainingforIC

T.Trainingandsupport

forhow

towork

intelework

environment

Outcomes:

.Im

pact

ongainingaccessto

required

task

inform

ation

.Individualtask

perform

ance

.Accessto

organisational

applications

.Perceived

quality

ofIC

Tsupport

.ProblemsusingIC

T.Individualtask

perform

ance

.Satisfaction/dissatisfaction

with

compensation

and

resources

available

Outcomes:

.Structuringofwork

.Availability

ofworkerswhen

needed

.Availabilityofinform

ation

togroupwhen

telecommuting

.Quality

ofcommunication

invirtualteams

.Grouptask

perform

ance

.Accessspeedforgroup

application

.Number

ofform

almeetings

.Tim

espentin

form

almeetings

.Grouptask

perform

ance

Outcomes:

.Ability

tolocate

people

with

specificknowledgeoftask

in-

form

ationneeded

.Ability

toaccess

specific

todata,

inform

ation,

wisdom,

andknowledgeneeded

.Im

pact

onoffice

costs

.Im

pactsonIC

Tsupport

costs

.Im

pactsonIT

infrastructure

Organisationalstructure

(Factors

describing

organisationalstructure

and

roles)

Antecedents:

.Role

clarity

.Number

ofwork

projectsto

work

on

.Number

ofdirect/indirectreport-

ingrelationships

.Role

variability

.Number

ofform

almeetings

.Number

of

inform

al

meetings/

interruptions

Antecedents:

.Task

interdependen

ce.Variabilityin

managem

ent

practices

.Existence

ofrole

clarity

within

thegroup

.Number

ofdirectreports

.Number

ofform

almeetings

Antecedents:

.Variability

inmanagem

ent

practices

over

multipleprojects

thatuse

thesamepeople

.Levelsofmanagem

ent (continued)

Behaviour & Information Technology 1273

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

offices are located at the client’s location. Firm B isorganised around an embedded matrix structure.Telecommuting was implemented in Firm B with thegoal of decreasing costs and providing better service tothe client. Yet, the culture at Firm B emphasisescustomer satisfaction and spending face time with thecustomer. This results in a misalignment of telecom-muting and reward and compensation plans.

Sales and technical employees are expected to lookfor opportunities to sell the customer on new ICTsolutions while working on existing projects. Selectionand promotion in Firm B is based on ‘visibility andwho you know’. For the structural subsystem, workersare assigned to multiple projects and managersnegotiate to get people they know assigned to theirprojects. Firm B’s reward and compensation strategiesfor most sales and technical personnel are based on asplit percentage (e.g. 65/35, 80/20, etc.) between a basesalary and revenue generation (based on customersatisfaction and sales quotas). The worker, in yearlyreview meetings with his superior, chooses thispercentage. Account managers are given a base salary,but bonuses are based entirely on customer satisfac-tion. The technical subsystem provides workers withreimbursement for an extra phone line at home.Management control is primarily based on face-to-face meetings and organisational communication isheavily reliant on email and face-to-face meetings.

At the individual level, sales personnel reportsatisfaction with Firm B’s reward and compensationplan (it is possible to earn up to 150% of their basesalary), while most technical personnel are dissatisfiedwith the plan. However, at the organisation level, FirmB’s reward and compensation plan reinforces existingfunctional silos and discourages groups from workingtogether. Individual access to necessary informationfrom remote locations is restricted due to organisa-tional level security concerns. Therefore, workerscannot access many files they need to work fromhome or remote locations. Some ICT applications areavailable to facilitate project approvals but are notuser friendly. These applications reportedly take toolong to access information, schedule meetings and aredifficult to operate because they are ‘over engineered’.

At the group level, workers feel pressure frommanagers to be in front of the client or in face-to-facegroup meetings (even if these meetings are sponta-neously called and ICT is available), which greatlydiscourages telecommuting. Further, group morale isreported as being low due to a lack of socialinteraction, and a general perception that a lack ofvisibility make it difficult to get on good projects, get toknow key people in the organisation, and ultimately toget promoted. As one person pointed out, ‘it’s hard tostand out in an email crowd’. Workers also reportT

able

3.

(Continued)

STSsubsystem

Individuallevel

ofanalysis

Grouplevel

ofanalysis

Organisationallevel

ofanalysis

Outcomes:

.Change

inavailable

individual

timedueto

changein

thenumber

ofform

almeetings

.Im

pactsoninform

almanagem

ent

feedback

.Diffi

culty

on

managing

adis-

persedteam

orgroup(m

anager)

.Individualtask

perform

ance

Outcomes:

.Im

pactsongroupperform

ance

.Im

pactsonmanagerialcontrol

.Number

of

form

al

group

meetings

.Grouptask

perform

ance

Outcomes:

.Inform

ationsecurity

impacts

.Diffi

culty

measuring

impacts

oftelework

.Organisationalperform

ance

1274 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

difficulties understanding common group goals whenworking away from the office. Finally, telecommu-ters report there is not enough time to do all thetasks assigned to them. Many reported feelings ofisolation and a lack of connection to the organisa-tion and co-workers since choosing to telecommute.Time plays a vital role as individuals experienceincreased frustration with the work environment andwith their group interactions as they try workingremotely. Eventually, they conclude that they haveto be physically present to be able to function in theorganisation.

Figure 4 illustrates the compounding, cross-levelimpacts of telecommuting in Firm B caused bymisaligned organisational culture, management prac-tices and reward and compensation for telecommuters.In Firm B, telecommuting was implemented toimprove customer and worker satisfaction and increaseorganisational performance and sales. However, re-ward, promotion and compensation norms of thecompany based on face-to-face interaction (i.e. pro-fessionalism factors in the personnel subsystem) andsecurity measures prevented workers from accessingnecessary information when telecommuting (i.e. fac-tors associated with the technical subsystem andexternal environment of the telecommuting worksystem). As a result, few people chose to telecommuteand morale plummeted among the workforce.

6. Implications and contributions

The multilevel telecommuting framework offers possi-bilities of conceptualisations to address prior issuesidentified in the literature, including the need toaccount for time (through feedback loops), to accountfor telecommuting being embedded in work practices(through the STS subsystems) and to account fortelecommuting as a multi-level concept.

The telecommuting framework has importantimplications for research and practice. First, the variedand often conflicting findings in prior literature withrespect to the effects of telecommuting on workoutcomes (e.g. better technology leads to betterproductivity) can potentially be explained by introdu-cing multiple levels of analysis and the effect of time ontelecommuting. Impacts of telecommuting factorsacross levels of analysis are inherent and of vitalimportance in organisational research (Rousseau1985), but multi-level issues have not been expresslyconsidered in telecommuting research (Pearlson andSaunders 2001, Bailey and Kurland 2002). Examples ofresearch questions this framework can help addressinclude: Do the factors in the telecommuting workenvironment that affect individual telecommuters alsoaffect virtual teams? What are the characteristics ofICT artefacts at different levels of analysis that affecttelecommuting behaviours and work outcomes?

Figure 3. Illustrating the multilevel telecommuting framework in firm A.

Behaviour & Information Technology 1275

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

It is also possible for individual, group andorganisational level motivators to be in conflict witheach other, affecting the decision to telecommute andmaking it difficult for practitioners and academics tounderstand ultimate outcomes of telecommuting. Forexample, an individual may want to work at home toincrease productivity by reducing interruptions. At thesame time, he may experience group-level incentives tocollaborate with co-workers, creating a ‘‘distance’’with colleagues. Thus, telecommuters may experienceindividual satisfaction with telecommuting while at thegroup level there may be negative impacts on memberrelationships, which may ultimately lead to reducedperformance. Therefore, it is recommended that futureinvestigations by both academics and practitionersconsider the multi-level nature of telecommuting whenexploring its various impacts.

Another key contribution is how the frameworkstresses the importance of looking at the passing oftime in studies of telecommuting. In prior literature,few, if any, studies have attempted to address thecompounding effects of telecommuting over time,which the proposed theoretical framework takes intoconsideration. It suggests that outcomes will not bestatic but change as the experiences of each telecom-muter change. Moreover, telecommuting behavioursand work outcomes at one point in time are expectedto be influenced by past experiences and in turn affectfuture telecommuting experiences. Potential research

questions include: How do telecommuting workoutcomes change over time when workers telecom-mute? Do telecommuters report the same satisfactionand dissatisfaction after a single experience withtelecommuting as they do after several months? Howdo properties of telecommuting ICT artefacts impactfuture telecommuting behaviours and work out-comes? Over time, what are the adjustments to theirwork environments that organisations need to makeafter implementing telecommuting to improve workoutcomes? Only a truly longitudinal study couldattempt to evaluate these effects, and it is recom-mended that such studies should be conducted in thefuture.

Finally, the framework shifts telecommuting fromthe static context in which the work is carried out to anembedded aspect of the work itself. For example,investigating the implications of changes in commu-nication patterns, e.g. as found by Belanger andAllport (2008), and new forms of connectivity thatare likely to emerge over time, becomes more salientunder this framework. These changes may haveprofound implications for organisations, which arenot likely to emerge when telecommuting is treatedsimply as a work setting.

In addition to the recommendations above, severalavenues for future research can be explored with theframework. First and foremost, the theoretical frame-work needs to be tested in a variety of telecommuting

Figure 4. Illustrating the multilevel telecommuting framework in firm B.

1276 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

environments. A first step might be to evaluate theexplanatory power of the framework with datacollected over time, from more than one organisation,across levels of analysis and analysed and integratedusing the framework to allow comparisons. Additionalresearch can also explore other forms of technology-mediated work for which the framework could apply.For example, how can the framework help explainfindings related to flexitime? Do organisations withmandatory telecommuting policies exhibit similarantecedents and consequences as organisations thatallow individuals the choice to telecommute? Theframework can provide a lens to look at these andother research questions in the distributed workdomain. Finally, the framework could be extendedaccording to STS to include the external environment.Such an extension could help explore legal, cultural, orpolitical impacts of telecommuting on organisations,groups, and individuals.

7. Conclusion

Building on STS theory, a multi-level telecommutingframework was developed, which proposes theoreti-cal relationships that address the conceptualisationissues found in telecommuting research: namely, (1)telecommuting and its ICT artefacts as the contextor environment in which work is performed insteadof just as an aspect of the actual work itself; (2)telecommuting as a multi-level concept whose im-pacts are often realised at the individual level ofanalysis but also have influences and outcomesacross levels of analysis and (3) telecommuting as aconcept whose antecedents and outcomes are affectedby the passing of time. The theoretical frameworkcan be used as a lens for evaluating past telecom-muting research, and developing new areas ofinquiries in telecommuting.

Notes

1. The context of interest in this article is work away froman office but related to the organisation as opposed tofull-time home work. In the review, we also focused onempirical studies since the purpose of the review ismainly to identify conceptualisation issues.

2. We thank an anonymous reviewer for pointers to this setof papers.

3. We thank an anonymous reviewer for this insight.

References

Araujo, E.R., 2008. Technology, gender and time: acontribution to the debate. Gender, Work and Organiza-tion, 15 (5), 477–503.

Axtell, C., Hislop, D., and Whittaker, S., 2008. Mobiletechnologies in mobile spaces: findings from the contextof train travel. International Journal of Human–ComputerStudies, 66 (12), 902–915.

Bailey, D.E. and Kurland, N.B., 2002. A review of teleworkresearch: findings, new directions, and lessons for thestudy of modern work. Journal of OrganizationalBehavior, 23 (4), 383–400.

Baker, E., Avery, G.C., and Crawford, J., 2006. Home alone:the role of technology in telecommuting. InformationResources Management Journal, 19 (4), 1–22.

Belanger, F., 1999. Worker’s propensity to telecommute: anempirical study. Information and Management, 35 (3),139–153.

Belanger, F. and Allport, C., 2008. Collaborative technolo-gies in knowledge telework: an exploratory study.Information Systems Journal, 18 (1), 101–121.

Belanger, F., Collins, R., and Cheney, P.H., 2001. Technol-ogy requirements and work group communication fortelecommuters. Information Systems Research, 12 (2),155–176.

Belanger, F. and Collins, R.W., 1998. Distributed workarrangements: a research framework. The InformationSociety: An International Journal, 14 (2), 137–152.

Belanger, F. and Watson-Manheim, M.B., 2006. Virtualteams and multiple media: structuring media use to attainstrategic goals. Group Decision and Negotiations, 15 (4),299–321.

Belanger, F., Watson-Manheim, M.B., and Jordan, D.H.,2002. Aligning IS research and practice: a researchagenda for virtual work. Information Resources Manage-ment Journal, 15 (3), 48–70.

Brodt, T.L. and Verburg, R.M., 2007. Managing mobilework – insights from European practice. New Technol-ogy, Work & Employment, 22 (1), 52–65.

Brown, O., Jr. 2002. Macroergonomics methods: participa-tion. In: H.W. Hendrick and B.M. Kleiner, eds. Macro-ergonomics: theory, methods, and applications, Mahwah,NJ: Lawrence Erlbaum Associates, xi, 412.

Cousins, K.C. and Robey, D., 2005. Human agency in awireless world: patterns of technology use in nomadiccomputing environments. Information and Organization,15 (2), 151–180.

DeGreene, K.B., 1973. Sociotechnical systems. EnglewoodCliffs, NJ: Prentice-Hall.

Desrochers, S., Hilton, J.M., and Larwood, L., 2005.Preliminary validation of the work-family integration-blurring scale. Journal of Family Issues, 26 (4), 442–466.

Gajendran, R.S. and Harrison, D.A., 2007. The good, thebad, and the unknown about telecommuting: meta-analysis of psychological mediators and individualconsequences. Journal of Applied Psychology, 92 (6),1524–1541.

Golden, T., 2006. The role of relationships in understandingtelecommuter satisfaction. Journal of OrganizationalBehavior, 27 (3), 319–340.

Golden, T. and Viega, J.F., 2005. The impact of extent oftelecommuting on job satisfaction: resolving inconsistentfindings. Journal of Management, 31 (2), 301–318.

Golden, T.D., 2007. Co-workers who telework and theimpact on those in the office: understanding the implica-tions of virtual work for co-worker satisfaction andturnover intentions. Human Relations, 60 (11), 1641–1667.

Golden, T.D., Veiga, J.F., and Dino, R.N., 2008. The impactof professional isolation on teleworker job performanceand turnover intentions: does time spent teleworking,interacting face-to-face, or having access to communica-tion-enhancing technology matter? Journal of AppliedPsychology, 93 (6), 1412–1421.

Behaviour & Information Technology 1277

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Goodhue, D.L. and Thompson, R.L., 1995. Task-technologyfit and individual performance. MIS Quarterly, 19 (2),213–235.

Hartig, T., Kylin, C., and Johansson, G., 2007. The teleworktradeoff: stress mitigation vs. constrained restoration.Applied Psychology, 56 (2), 231–253.

Hendrick, H.W., 2002. An overview of macroergonomics. In:H.W. Hendrick and B.M. Kleiner, eds. Macroergo-nomics: theory, methods, and applications. Mahwah, NJ:Lawrence Erlbaum Associates, xi, 412.

Hendrick, H.W. and Kleiner, B.M., 2001. Macroergonomics:an introduction to work system design. Santa Monica, CA:Human Factors and Ergonomics Society.

Hendrick, H.W. and Kleiner, B.M., eds. 2002. Macroergo-nomics: theory, methods, and applications. Mahwah, NJ:Lawrence Erlbaum Associates.

Hill, E.J., et al., 2010. Workplace flexibility, work hours, andwork-life conflict: finding an extra day or two. Journal ofFamily Psychology, 24 (3), 349–358.

Hislop, D. and Axtell, C., 2009. To infinity and beyond?:workspace and the multi-location worker. New Technol-ogy, Work & Employment, 24 (1), 60–75.

Holden, R.J. and Karsh, B., 2009. A theoretical model ofhealth information technology usage behaviour withimplications for patient safety. Behaviour & InformationTechnology, 28 (1), 21–38.

Hornung, S., Rousseau, D.M., and Glaser, J., 2009. Whysupervisors make idiosyncratic deals: antecedents andoutcomes of i-deals from a managerial perspective.Journal of Managerial Psychology, 24 (7–8), 738–764.

Hunton, J.E. and Norman, C.S., 2010. The impact ofalternative telework arrangements on organizationalcommitment: insights from a longitudinal field experi-ment. Journal of Information Systems, 24 (1), 67–90.

Igbaria, M., 1999. The driving forces in the virtual society.Communications of the ACM, 42 (12), 64–70.

Jarvenpaa, S.L. and Lang, K.R., 2005. Managing theparadoxes of mobile technology. Information SystemsManagement, 22 (4), 7–23.

Joli, B., 2009. Telecommuting trends in the 2009 economy.Bright Hub [online], Available from: http://www.brighthub.com/office/home/articles/22829.aspx[Accessed 2 July 2012].

Kakihara, M. and Sørensen, C., 2004. Practicing mobileprofessional work: tales of locational, operational, andinteractional mobility. The Journal of Policy, Regulationand Strategy for Telecommunication, Information andMedia, 6 (3), 180–187.

Katz, D. and Kahn, R.L., 1966. Common characteristics ofopen systems. In: D. Katz and R.L. Kahn, eds. The socialpsychology of organizations. New York,: Wiley, 14–29.

Khalifa, M. and Davison, R.M., 2008. Explaining theintended continuance level of telecommuting. Interna-tional Journal of Internet and Enterprise Management, 5(3), 264–294.

Klein, K.J., Dansereau, F., and Hall, R.J., 1994. Level issuesin theory development, data collection, and analysis.Academy of Management Review, 19 (2), 195–229.

Kuldeep, K., van Fenema, P.C., and Glinow, M.A., 2009.Offshoring and the global distribution of work: implica-tions for task interdependence theory and practice.Journal of International Business Studies, 40 (4), 642–667.

Lee, H.G., Shin, B., and Higa, K., 2007. Telework vs. centralwork: a comparative view of knowledge accessibility.Decision Support Systems, 43 (3), 687–700.

Leonard, B., 2000. Telework has many advantages. HRMa-gazine, (45), 23–24.

Lim, E.T.K., Pan, S.L., and Tan, C.W. 2005. Managing useracceptance towards enterprise resource planning (ERP)systems – understanding the dissonance between userexpectations and managerial policies. European Journalof Information Systems, 14 (2), 135–149.

Madsen, S.R., 2006. Work and family conflict: can home-based teleworking make a difference? InternationalJournal of Organization Theory and Behavior, 9 (3),307–350.

Markham, S.E., 1988. Pay-for-performance dilemma revis-ited: empirical example of the importance of groupeffects. Journal of Applied Psychology, 73 (2), 172–180.

Mayo, M., et al., 2009. Why some firms adopt telecommutingwhile others do not: a contingency perspective. HumanResource Management, 48 (6), 917–939.

Mazmanian, M.A., Orlikowski, W.J., and Yates, J., 2005.Crackberries: the social implications of ubiquitouswireless e-mail devices. In: C. Sørensen, Y. Yoo, K.Lyytinen and J.I. DeGross, eds. Designing ubiquitousinformation environments: socio-technical issues and chal-lenges. New York: Springer, 337–343.

McCloskey, D.W. and Igbaria, M., 1998. A review of theempirical research on telecommuting and directions forfuture research. In: M. Igbaria and M. Tan, eds. Thevirtual workplace. Hershey: Idea Group Pub., 338–358.

Morrison, D., et al., 2005. Job design, opportunities for skillutilization, and intrinsic job satisfaction. EuropeanJournal of Work and Organizational Psychology, 14 (1),59–79.

Mortensen, M., Caya, O., and Pinsonneault, A., 2009.Understanding virtual team performance: a synthesis ofresearch on the effects of team design, emergentprocesses, and emergent states. In: Research Paper No.4738-09. Cambridge, MA: MIT Sloan School ofManagement.

O’Neill, T.A., et al., 2009. Predicting teleworker success: anexploration of personality, motivational, situational, andjob characteristics. New Technology, Work & Employ-ment, 24 (2), 144–162.

O’Leary, M.B. and Mortensen, M., 2010. Go (Con)figure:subgroups, imbalance, and isolates in geographicallydispersed teams. Organization Science, 21 (1), 115–131.

Orlikowski, W. and Barley, S.R., 2001. Technology andinstitutions: what information systems research andorganization studies can learn from each other? MISQuarterly, 25 (2), 145–165.

Orlikowski, W. and Scott, S., 2008. Sociomateriality:challenging the separation of technology. Work, andOrganization. AoM Annals, 2 (1), 433–474.

Paez, A. and Scott, D.M., 2007. Social influence on travelbehavior: a simulation example of the decision totelecommute. Environment and Planning A, 39 (3), 647–665.

Pasmore, W.A., 1988. Designing effective organizations: thesociotechnical systems perspective. New York: Wiley.

Pearce, J.A., 2009. Successful corporate telecommuting withtechnology considerations for late adopters. Organiza-tional Dynamics, 38 (1), 16–25.

Pearlson, K. and Saunders, C., 2010. Managing and usinginformation systems. 4th ed. New York, NY: John-Wiley.

Pearlson, K.E. and Saunders, C.S., 2001. There’s no placelike home: managing telecommuting paradoxes. Acad-emy of Management Executive, 15 (2), 117–128.

1278 F. Belanger et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014

Perez, M.P., Sanchez, A.M.L.C., and Pilar Jimenez, M.J.V.,2004. A technology acceptance model of innovationadoption: the case of Teleworking. European Journal ofInnovation Management, 7 (4), 280–291.

Peters, P. and Heusinkveld, S., 2010. Institutional explana-tions for managers’ attitudes towards telehomeworking.Human Relations, 63 (1), 107–135.

Pinsonneault, A. and Boisvert, M., 2001. The impacts oftelecommuting on organizations and individuals: areview of the literature. In: N.J. Johnson, ed. Telecom-muting and virtual offices: issues and opportunities.London: Idea Group Publishing, 163–185.

Pollster_YouGov_UK. 2007. SW2000 Telework studies [on-line]. Available from: http://www.noelhodson.com/in-dex_files/statistics2006to2008.htm#_Toc229386682 [Ac-cessed 8 January 2009].

Ramsower, R.M., 1985. Telecommuting: the organizationaland behavioral effects of working at home. Michigan:UMI Research Press.

Rousseau, D.M., 1985. Issues of level in organizationalresearch. In: L.L. Cummings and B.M. Staw, eds.Research in Organizational Behavior, Vol. 7. Greenwich,CT: JAI Press, 1–37.

Ryan, S.D. and Harrison, D.A., 2000. Considering socialsubsystem costs and benefits in information technologyinvestment decisions: a view from the field on anticipatedpayoffs. Journal of Management Information Systems, 16(4), 11–40.

Sanchez, A.M., et al., 2007. Teleworking and workplaceflexibility: a study of impact on firm performance.Personnel Review, 36 (1), 42–64.

Sanchez, A.M., et al., 2008. Telework adoption, changemanagement, and firm performance. Journal of Organi-zational Change Management, 21 (1), 7–31.

Schweitzer, L. and Duxbury, L., 2006. Benchmarking the useof telework arrangements in Canada. Canadian Journalof Administrative Sciences, 23 (2), 107–109.

Siha, S.M. and Monroe, R.W., 2006. Telecommuting’spast and future: a literature review and researchagenda. Business Process Management Journal, 12 (4),455–482.

Smith, R.L.J., 2004. Work at home grows in past year by7.5% in US [online]. Available from: www.workingfro-manyywhere.org/news/pr090204.htm [Accessed 2 July2012].

Sørensen, C. and Al-Taitoon, A., 2008. Organisationalusabilityof mobilecomputing-Volatility and control inmobile foreign exchange trading. International Journal ofHuman–Computer Studies, 66 (12), 916–929.

Sørensen, C., et al., 2008. Enterprise mobility: lessons fromthe field. Information Knowledge Systems ManagementJournal – Special issue on Enterprise Mobility 7 (1 & 2),243–271.

Towers, I., et al., 2006. Time thieves and space invaders:technology, work and the organization. Journal ofOrganizational Change Management: Space and Timeand Organisation Change, 19 (5), 593–618.

Trist, E.L. and Bamforth, K.W., 1951. Some social andpsychological consequences of the longwall method ofcoal-getting. Human Relations, 4 (1), 3–38.

Venkatraman, N., 1989. The concept of fit in strategyresearch: toward verbal and statistical correspondence.Academy of Management Review, 4 (3), 423–444.

Vessey, I. and Galletta, D., 1991. Cognitive fit: an empiricalstudy of information acquisition. Information SystemsResearch, 2 (1), 63–84.

Virick, M., DaSilva, N., and Arrington, K., 2010. Modera-tors of the curvilinear relation between extent oftelecommuting and job and life satisfaction: the role ofperformance outcome orientation and worker type.Human Relations, 63 (1), 137–154.

Wajcman, J., Bittman, M., and Brown, J.E., 2008. Familieswithout borders: mobile phones, connectedness andwork-home divisions. Sociology, 42 (4), 635–652.

Ward, N. and Shabha, G., 2001. Teleworking: an assessmentof socio-psychological factors. Facilities, 19 (1/2), 61–70.

Watson-Manheim, M.B. and Belanger, F., 2002. Support forcommunication-based work processes in virtual work. e-Services Quarterly, 1 (3), 61–82.

Watson-Manheim, M.B. and Belanger, F., 2007. Commu-nication media repertoires: dealing with the multiplicityof media choices. MIS Quarterly, 31 (2), 267–293.

Westfall, R., 2004. Does telecommuting really increaseproductivity? Communications of the ACM, 47 (8), 93–96.

Whittle, A. and Mueller, F., 2009. I could be dead for twoweeks and my boss would never know: telework and thepolitics of representation. New Technology, Work andEmployment, 24 (2), 131–143.

Workman, M., Kahnweiler, W., and Bommer, W., 2003. Theeffects of cognitive style and media richness on commit-ment to telework and virtual teams. Journal of VocationalBehavior, 63 (2), 199–219.

WorldatWork. 2009. Telework trendlines 2009 [online].Available from: http://www.worldatwork.org/waw/adimLink?id¼31295&from¼presshome [Accessed 2 July2012].

Behaviour & Information Technology 1279

Dow

nloa

ded

by [

Uni

vers

ity o

f T

asm

ania

] at

23:

21 2

7 N

ovem

ber

2014