performance of individualWrms. Variations of such results have been found in
a number of diVerent manufacturing industry sectors including apparel (Appelbaum
et al. 2000 ), auto assembly (MacDuYe 1995 a), semiconductors (Appleyard and
Brown 2001 ), and steel (Arthur 1994 ; Ichniowski et al 1997 ). As Batt and Doellgast
( 2005 : 146 ) observe, these studies may diVer on certain details but eachWnds support
for the systems’ approach of lean manufacturing with the performance eVects of
bundles of management practices measured through cluster analysis or additive
indices. The details of these studies, however, reveal important context-speciWc
elements within each sector that suggest that precisely what aVects performance
varies. Moreover, Appelbaum’s ( 2002 : 148 ) recent review of the survey evidence on
the impact of HPWSs on workers ends with the conclusion that the ‘jury is still out.’
The evidence on the impact of lean manufacturing and HPWSs derived from
detailed case research has generally been more consistent, and negative, in its
conclusions on the impact for the shopXoor workforce.^8 Batt and Doellgast
( 2005 ) suggest that this is reXected in the divergent traditions of European and
US research but, while it is true that the European tradition is more consistently
critical in perspective, case researchers from North America have concluded simi-
larly. A number of qualitative studies in the motor industry (Fucini and Fucini
1990 ; Graham 1995 ; Rinehart et al. 1997 ) found high degrees of worker dissatisfac-
tion with new and more demanding and stressful working conditions that were
ascribed to moves toward the adoption of the lean manufacturing model. These
cases argued that the primary outcome for workers was a combination of a loss of
autonomy and inXuence with a concomitant increase in the demands placed upon
them both in terms of physical eVort and expectations of contributions to con-
tinuous improvement programs. More generally, the impact for employees of new
workplace practices continues to be open to debate with a mixed picture emerging
(Geary and Dobbins 2001 ; Godard 2001 ; Harley 2001 ; Anderson-Connolly et al.
2002 ). Understanding what lies behind this ‘mixed picture’ needs research that
accesses the ‘black box’ of employee perceptions and expectations while retaining
the prospect of workplace-level comparisons. While quantitative, sector-level
studies are helpful (Appelbaum et al. 2000 ), this suggests qualitative research will
continue to play a particularly important role. The impacts of lean manufacturing
on workers’ health (Landsbergis et al. 1999 ) and, in particular, stress levels (Conti
and Gill 1998 ) have also been the subject of recent studies but good-quality data
remains rare and this is another area for further research. The extent to which the
inherent demands of a tightly coupled and fragile production system may be
oVset by the support of teamworking and complementary HR practices remains
uncertain (Ramsay et al. 2000 ; Delbridge 2005 ).
A recurring problem faced in the implementation of the lean manufacturing
model is that the daily demands of the production process drive out the space and
(^8) For a fuller discussion of the evidence on the impact of lean manufacturing for workers, see
Delbridge 2005.
hrm and contemporary manufacturing 415