Teamwork, Skill Development and Employee Welfare

Teamwork, Skill Development and Employee Welfare
Duncan Gallie, Ying Zhou, Alan Felstead and Francis Green
August 2009
Duncan Gallie is Professor of Sociology at the University of Oxford.
Ying Zhou is a Consultant at International Survey Research, Towers Perrin.
Alan Felstead is Research Professor of Sociology at Cardiff University.
Francis Green is Professor of Economics at the University of Kent.
Word Count (including Abstract) 10, 297 words
1

Abstract
There has been a sharp divergence in the literature about the benefits of the consequences
of teamwork. Some have claimed that it is solely in the interests of management, others
that it is beneficial for employees and yet others that it makes little difference to either
productivity or well-being. A feature of this debate is the lack of high quality
representative data on employee experiences. This paper draws upon the British Skills
Survey Series which provides a particularly rich source of evidence. It shows that, while
teamwork did expand between the early 1990s and 2006, this was due primarily to the
growth of the type of teamwork that allowed employees little in the way of decision-
making power. Indeed there was a decrease in the prevalence of self-directive teamwork.
At the same time our evidence shows that the benefits of teamwork, in terms of both
enhancing productivity and employee welfare, are confined to self-directive teams, while
non-self-directive teams suppress the use of personal initiative and discretion at work.
2

Introduction
Teamwork has been at the centre of debates about whether new forms of work
organization are emerging in advanced capitalist societies. The growth of teamwork has
been depicted as a major factor breaking down the hierarchical and conflictual nature of
traditional Taylorist forms of work organization by promoting an organizational design
that enhances both managerial objectives of increased productivity and employee self-
realization and well-being. It has been widely suggested that organizations have moved
towards more decentralized patterns of responsibility, which offer employees greater
initiative and control over their jobs, and which thereby better engage their creative
potential and productive capacities. However, other researchers, while agreeing that
teamwork is increasingly prevalent, have developed sharply contrasting perspectives on
its implications, with some arguing that it has negative effects for experiences of work
and others sceptical about whether it makes any significant difference either way.
A number of studies provide some empirical support for the generally positive effects of
teamwork. For instance, Cohen and Ledford (1994), examining more than eighty self-
managing teams at an American telecommunications company, found that self-managing
teams had significantly better job performance and higher employee job satisfaction than
traditional working groups or departments. Hamilton et al. (2003) found that the adoption
of teams at the plant level improved worker productivity even after taking into account
the selection of high-ability workers into teams. Batt (2004) showed that self-managed
3

teams were associated with significantly higher levels of perceived discretion,
employment security and satisfaction for workers and were effective in improving
objective performance measures. In a wider European study, Benders et al. (2001) also
found a positive effect of group delegation for reducing employee absenteeism rates and
improving organizational performance. A review of survey based research over the last
decade concluded that the great majority of studies had found positive effects on
operational measures of organizational performance (Delarue et al. 2007).
A central argument for linking teamwork to higher productivity is that it gives employees
a sense of empowerment, by increasing the control they can exercise over their
immediate work environment (Goodman et al. 1988; Harley 1999). Workers with higher
control over their jobs are likely to feel more committed to their organizations and more
satisfied with their jobs. As a result, they will be more willing to deploy discretionary
effort, thereby enhancing organizational performance (Cohen et al. 1996; Dunphy and
Bryant 1996; Pil and MacDuffie 1996). This assumption also underpins theories of ‘high
commitment’ and ‘high performance’ management systems where teamwork is viewed as
one of a set of structural features that enhances organizational effectiveness by raising
employee motivation (Ramsay et al 2000).
It has also been argued that teamwork enhances performance through the increased scope
it gives employees to use their knowledge, skills and abilities. This raises motivation,
thereby reducing shirking and enhancing employee retention (Huselid, 1995; Batt and
Appelbaum 1995; Benders et al. 2001; Dunphy and Bryant 1996; Janz et al. 1997;
4

Spreitzer et al. 1999). At the same time, it facilitates employee learning and skill
acquisition, as well as information sharing, which may be particularly important in
conditions of growing economic uncertainty (Wagner et al. 1997; Vaskova 2007;Wall et
al. 2002). In an economy in which employee expertise and specialist knowledge are
increasingly important to corporate performance, teamwork can facilitate employees’
accumulation of task-specific human capital by encouraging mutual and collective
learning processes. This is particularly likely to be the case for diagnostic skills in
complex systems where on-the-job learning is a prerequisite to obtaining the necessary
knowledge and for the acquisition of tacit skills, where learning from others is likely to
be the most effective source of skill development (Barrett 2001).
In sharp contrast to these relatively optimistic perspectives on teamwork, other writers
have cast doubt on the view that it implies a qualitative break with the hierarchical logic
of the past and argue instead that it is a continuation of the rationale of Taylorism
(Berggren 1992; Dohse et al. 1985; Thompson and McHugh 1995; Vidal, 2007). It has
been argued that teamwork systems replace supervisory control with a less visible but
equally constraining form of normative control, that encourages employees to internalize
managerial definitions of organizational goals (Barker 1993; Graham 1995; Grenier
1988). The rhetoric of greater worker autonomy and empowerment therefore masks a
strategy of heightened managerial control. Combined with group norm and peer pressure,
the deployment of new technologies can make teams take on the responsibility for
intensifying their own work activities in a process that amounts to ‘Team Taylorism’
(Sewell 1996). These accounts then suggest that employees working in the team context
5

are subject to at least as intense scrutiny and monitoring as in earlier work systems. Such
claims can take support from the fact that a number of studies have found that it
aggravates job strain and work-life conflict (De Dreu and Van Vianen 2001; Findlay et
al. 2000; Stewart and Barrick 2000; White 2003; Vaskova 2007).
A third, less frequently developed position, maintains that teamwork has no significant
effect one way or the other. Vallas (1999) points out that, from a neo-institutional
perspective, employee involvement programs may be primarily designed to legitimate the
firm by conforming to the demands of its institutional environment. He suggests that,
although professional employees may have acquired significant powers of self-direction,
this would not be the case for manual employees. His case studies (Vallas 2003) in the
paper and pulp industry found that in general the introduction of teamwork led to no
substantial change in hierarchical relationships or employee attitudes. He attributed this
to a management orientation that was concerned with efficiency and the reduction of
costs, but made little effort to encourage higher levels of normative commitment. Its
preference for a centrally driven plant improvement process undercut the reality of, and
any potential benefits from, ‘self-directed’ team organization. Vidal (Vidal 2007) found
in case studies of six US manufacturing plants that teamwork made little difference to
substantive employee empowerment, because this was not necessary for management to
achieve its performance objectives.
Another possibility, given the diversity of the conclusions of case studies, is that there is
no general pattern because the implications of teamwork differ very substantially
6

between economic sectors. In particular, it could be that the introduction of teamwork in
industry may have been much more closely linked to strategies for intensifying work,
leading to notably higher costs for employees in terms of work pressure. In the service
sector, where quality of work performance is closely linked to social interaction with
customers, clients and the general public, it may have been introduced with primary
concern for employee motivation and welfare.
A notable point about these broad perspectives on teamwork is that they are based on
very limited quantitative evidence about the key mechanisms that are held to underlie its
effects. As Harley (2001) has pointed out, both optimistic and pessimistic views of
teamwork have at their heart the issue of employee task discretion. For the former, its
positive effects on individual task discretion are a crucial factor in ensuring high
organizational commitment and hence higher productivity. For the latter, its tendency to
reduce employee autonomy undermines employee well-being at work. However, there is
little direct empirical support for the nature of the crucial link between teamwork and task
discretion.
Further, existing research on teamwork has been predominantly based on case study
evidence or surveys of limited sectors of industry, leaving unknown the extent to which
specific findings can be generalised. One of the rare exceptions - a national survey by
Godard of Canadian employees (2001) - found little effect of team autonomy as such,
once controls were introduced for other work involvement practices (for instance offline
involvement through quality circles, information sharing meetings and group bonuses).
7

The main attempt to provide a broader picture for the UK is Harley’s (2001) study based
on the 1998 Workplace Employee Relations Survey. Comparing the experiences of team
workers and non-team workers, he found no significant effect of teamworking on
employee task discretion, organizational commitment, intrinsic job satisfaction or job
stress. On the basis of this he rejected both ‘positive’ and ‘critical’ views of teamwork as
unfounded.
Neither study is ideal in terms of data. Godard’s national survey involved only 508
employees, which is likely to have put considerable pressure on sample numbers given
the relatively low proportions meaningfully involved in team practices. Harley’s study
raises issues of data coverage. The classification of employees as members of particular
types of teams has to be based in WERS upon information from management about
whether members of the largest occupational group in the workplace worked in teams.
The need to focus down on employees in the largest occupational group, together with
the decision to exclude ‘managerial’ employees, restricted the sample to only 19.7% of
all valid employee responses. It is notable that other work using the WERS series has
come to different conclusions. Green (2008), analysing the 2004 Workplace Employment
Relations Survey, found that the impact of teamworking depended on the character of the
team. Where team members were not able to jointly decide about work matters (about
half of cases) teams were associated with reduced individual discretion, whereas teams
with joint decision making were neutral in their effects. There remains then a need for an
examination of the effects of teamwork based upon a genuinely representative survey of
the workforce and, as Delarue et al (2007) conclude in the most recent overview of the
8