even be the dominant outcome requirement. A third end goal of HRM is the pursuit
of social legitimacy and the requirement to meet societal expectations and
regulations which vary from country to country. In practice, all three outcomes—
productivity,Xexibility, and legitimacy—are likely to be required to a greater or
lesser extent. ‘A superior performance in HRM... implies an outstanding mix of
outcomes across these three areas’ (Boxall and Purcell 2003 : 8 ) (see Chapter 3 ).
The second cluster of problems concerns the choice of HR practices. There is an
extraordinary lack of agreement among researchers on what practices to include and
a paucity of explanation on why particular lists are chosen. The use of ‘HR archi-
tecture’ identifying four diVerent levels of HRM provides some logic and order in the
choice of practices. The use of generic categories of practice such as KSA and AMO is
particularly helpful since diVerent combinations may have the same outcome.
Third, progress in HRM–performance research can only be made if theory is
focused on the causal chain between practice and outcome. Employees are at the
heart of all casual chains because they are the focal point of HR practices and they
deliver performance. Employee responses to HR practices have not been included
in most research designs, until recently, and this has impoverished theWeld. Once
employees are put center stage, involving very diVerent research designs and
datasets, it becomes possible to advance the theory of HRM and performance.
Social exchange theory oVers one possible avenue for development within HRM.
The focus on employees’ reactions to HR practices—their perceptions of them,
not just as practices, but in terms of fairness and justice, motivation and eVective-
ness in inXuencing attitudes and behavior—takes us back to the debate on what
HRM really is. Other features of organizational life are likely to inXuence employee
attitudes and behavior so much that they cannot be excluded. These are the role
line managers play in the delivery of HR practices and in establishing relationships
with employees, features of the operational management system which aVect job
design and job experience, and organizational culture and its sub-area, climate.
The covariance problem is so strong that either these factors need to be controlled
for or, and this is the route preferred here, they need to be incorporated in our
deWnition of HRM. Once we have a robust, theoretically sound, model of HRM
and a clear view of the causal chain, it will be possible, with research focused mainly
on employees, their responses to people management, and organizational out-
comes, to move to the next step in the search for the HRM Holy Grail.
References
Appelbaum, E., Bailey, T., Berg, P., and Kallebergh,A.( 2000 ).Manufacturing Advan
tage: Why High Performance Systems Pay oV. Ithaca, NY: ILR Press.
Arthur,J.( 1994 ). ‘EVects of Human Resource Systems on Manufacturing Performance
and Turnover.’Academy of Management Journal, 37 :670 87.
548 john purcell and nicholas kinnie