and Sherman 1999 ; Wright and Snell 1998 ). For example, Gerhart (in press), in
a review of HRM systems, concluded that research has produced only ‘weak or, at
best, mixed’ support for the internalWt perspective.
The purpose of this chapter is to revisit the theoretical foundation of internalWt
and provide a review of the literature addressing this issue. SpeciWcally, we set out
to explore the theory and research behind the internalWt perspective in an attempt
to summarize and advance our knowledge behind HRM systems and internalWt. In
doing so, we address the theory behind ‘Wt’, the interplay between external and
internalWt, issues of the level of abstraction (e.g. focus on HRM philosophies,
policies, or practices in measuring HRM systems), diVerent types of internalWt,
problems stemming from levels of analysis issues, and the empirical evidence,
before summarizing and concluding.
19.2 The Theory Behind Internal Fit
.........................................................................................................................................................................................
The idea ofWt or alignment and synergistic eVects of internalWt has been explicit in
almost all SHRM research (e.g. Delery and Doty 1996 ; Delery and Shaw 2001 ;
Ichniowski et al. 1997 ; MacDuYe 1995 ). Delery and Doty ( 1996 ) outlined three
theoretical frameworks that were being used to describe the relationship between
HRM practices and organizational performance. The universalistic approach,
which was described as arguing for a universal relationship between individual
‘best’ HRM practices andWrm performance, was perhaps the only framework that
explicitly did not argue for positive ‘Wt’ eVects. A close examination of the work
most often attributed as arguing for universalistic relationships between single
HRM practices andWrm performance reveals that there was an acknowledgement
even there that practices might be more eVective when combined with supporting
practices (e.g. PfeVer 1994 ). For instance, PfeVer ( 1994 : 31 ), when discussing sixteen
best HRM practices, stated that ‘it is important to recognize that practices are
interrelated—it is diYcult to do one thing by itself with much positive result.’
The other perspectives identiWed by Delery and Doty ( 1996 ), the contingency and
conWgurational, have dealt more directly with at least two forms ofWt. The SHRM
literature distinguishes between what has been termed vertical or externalWtand
horizontal or internalWt. While externalWt denotes the alignment between HRM
practices and the speciWc organizational context (e.g. organizational strategy), internal
Wt refers to the coherent conWguration of individual HRM practices that support
each other (Becker et al. 1997 ; Delery 1998 ). When researchers talk about ‘internal
Wt,’ they are usually referring to an arrangement of HRM activities (e.g. HRM
policies or practices) that work in concert, whether they call these arrangements
386 sven kepes and john e. delery