194 ANALYSING HUMAN RESOURCE MANAGEMENT
Lepak and Snell have applied this idea to the study of HRM and employ-
ment relations predicting alternative bundles of HRM practice according to
whether the employment relationship is focused on: the core employees in
the organization or a job-based approach to employees or an organizational
alliance approach to employment or a short-term contract-based approach to
employment. One of the central conclusions of their recent empirical work
exploring the theory was that employers are somewhat inconsistent in their
approach to HRM and employment relations. This problematic finding of
Lepak and Snell (2002) is consistent with a large volume of studies within
the literature on IR and HRM. Indeed, where consistency in HRM vis-à-
vis organizational performance has most often been found, is in the broad-
based tendency of organizations to limit their implementation of HRM during
problems in organizational performance (Truss et al. 1997).
A group of HRM researchers then, first in the USA and more recently in
the UK, have claimed over the last ten years that bundles or configurations
of HRM practices can lead to superior organizational performance (Delery
and Doty 1996; Huselid 1995; Ichniowski and Shaw 1999; Ichniowski,
Shaw, and Prennushi 1995; Macduffie 1995). This research finding has
been somewhat constrained by the noticeable tendency (Arthur 1994;
Becker and Gerhart 1996) of ‘universalist’ (one size fitsall) approaches to
HRM being more frequently linked to high organizational performance
than ‘contingent’ studies (e.g. ‘cost leadership strategiescausally imply a
different bundle of HRM practices to quality differentiation strategies’).
One major explanation for this discrepancy between theoretical prediction
(use consistent bundles or configurations of HRM) and empirical reality
(this somewhat arbitrary universal list of HR practices are reported to be
statistically associated with high performance more so than theoretically
predicted contingent clusters/bundles) are the overly simplistic typifications
of universalist and contingent HRM (Boxall and Purcell 2003; Purcell
1999; Wood and Albanese 1995). The research studies in particular on
contingent HRM have tended to rely on very bland, somewhat a-historical,
non-processual academic stereotypes of strategy. For example, they have
been based on rather stereotypical concepts of cost/quality/innovation
derivative from Porter (1985) or alternatively draw from Miles and
Snow’s typifications (1984) of ‘defender, prospector, and analyser’
strategies.
Another problem with this whole line of configurationist research is that
it seeks to reduce issues of organizational and employee outcomes to bench-
marks driven primarily by an interest in economic capital. Its focus is on gen-
eral financial indicators (e.g. profit per employee), financial indicators specific
to the industry (e.g. scrap/waste measures, productivity per employee, quality
management measures), and financial indicators specific to HRM (employee
turnover rates, proportional cost of employees). Such measures are central