problems within the HRM architecture and the respective HRM system(s). With-
out this complex alignment, the HRM system may not be embedded within the
organization and cannot create the ‘invisible capability’ that enhances organiza-
tional eVectiveness (Becker and Huselid 1998 ). This multidimensional view further
highlights the fact that identical HRM policies but diVerent HRM practices or the
same HRM practices but distinct HRM processes may result in the same outcome if
they are aligned, which is consistent with the conWgurational perspective (Delery
and Doty 1996 ). Such a complex interactive view is not new to SHRM, although
these complexities have traditionally been scantly treated. Recently, Colbert ( 2004 )
extended the current perspectives in SHRM by bringing complexity theory into the
Weld. We do not wish to simply repeat Colbert’s complexity perspective. Instead, we
believe it is important to discuss the types of internalWt in more detail, and to
highlight the complexity of the HRM architecture and the need for more rigorous
research in SHRM.
- 3 Internal Fit and Possible EVects (Additive
and Synergistic EVects)
After discussing the diVerent types of internalWt, it is necessary to describe the
diVerent ways in which HRM activities could ‘Wt’ together. This discussion is spe-
ciWcally related to the two most discussed types ofWt in the literature: intra-HRM and
inter-HRM activity areaWt. In this literature, it is rare for much attention to be given
to precisely describing how practices actuallyWt together and the consequences of
non-Wt. Often, researchers simply list a number of ‘best’ HRM activities and suggest
that, because they are ‘best’ activities, they mustWt together. Delery ( 1998 ), however,
provided an in-depth overview of internalWt and the diVerent relationships bet-
ween HRM practices. Here, we want to brieXy review this conceptual work linking
various HRM activities with each other.
The internal relationships between HRM activities (both intra-HRM and inter-
HRM activity area) can take many diVerent forms (see Delery 1998 ). First, simple
additive eVectsare possible. Here, the combination of two HRM activities on the
same level of abstraction results in the sum of the separate eVects of each individual
HRM activity (e.g. a personality and a work sample test in the staYng area at the
process level). Additive eVects come about if the individual HRM activities have
independent, non-overlapping eVects on the outcome. HRM activities that simply
have additive eVects do not actually show ‘Wt’ in the way that term has been used in
the SHRM literature. It is wise for organizations to combine such practices,
however, because the combination results in better performance than using either
activity alone.
Second,interactiveorsynergistic eVectsare possible. These are more consistent
with the way many SHRM researchers have presented the idea of ‘Wt’ in that
hrm systems and the problem of internal fit 393