A Contingency Theory of Human Resource Management Devolution Thomas zyxwvut H. Stone faye 1. smith zyxwvutsr Oklahoma State lhiversity zyxwvutsr Abstmct Historical!,: the structuring zyxwvuts of human resource manage- ment suggests that it has evolved from a market-based slructure to more centralized control in a bureaucratic- based structure. Rccent1.v the function has devolved from the bureaucratic-based structure to multiple fornis of market-based structures as r$*ell as clan-t.\pe structures such as self-directed work teams. This evolution Jdevo- lution process is e.tplained using constructs .from market fui1ure.s theorj, and organizational economics. Proposi- tions from the proposed contingency theoty of devolution can guide future re.searth, anti our e.rplanations Mdl help human resource managers deliver their services through ihe most eJective structure. zyxwvutsrq - _ _ _ ~ ~ -~- z The purpose of this paper is threefold: (I) to use concepts derived from market failures theory and organ- izational economics (Ouchi, 1980; Williamson, 1975) to explain a continuum of organizational relationships, or contracts (Mcl,ean Parks & smith, in press; Rousseau, 1989; Rousseau & McIxan Parks, 1993), which help maintain employment relationships, (2) to identify the human resource management dimensions on which each of the organizational relationships differ, and (3) to ad- vance propositions derived from our contingency theory of HRM devolution. An historical perspective of employment relationships suggests that only one or two organizational options were prevalent in the past, and organizations evolved to a dom- inant use of bureaucratic forms of employment contracts. More recently. and in the future, we believe there will be more combinations chosen, and that organizations will devolve from bureaucratic relationships to a wider variety of employment contracts. The variety of organi7ational relationships consists of four market-type employment re- lationships (daywork, subcontract, temporary agency. and leasing), the traditional bureaucratic relationships, and a clan-type relationship (selfdirected work team). In ad- - Addrehs all correspondence to zyxwvutsrqponm T’liomas H. Stone. College of Business, Oklahoma State I iniversity. Stillwater. OK 74078. U.S.A. zyxwvutsrqpo Resume Historrquenient. la gestion des ressources humaine3 est passPe dune structure reposant sur des mkcanismes de rnarchi a une structure bureaucratique a contrde cen- tralisk. RPcemnient. cette fonction en est revenue a divers tjpes de structures basks sur les micanismes lie march; et aussi a des structures reposant sur zyx la concept cie clan, comme les Pquipes tie travail autonome>. Ce proces.\us de “dP-Pvolution ” est expliquke par le hiais de construits enipntnths a la throrie de faillite des marches et a IPm- nomique des organisatioms. Diverse5 propositions son t jiites afin de guider la recherchefuture. Lles e.rplications aideront les gestionnairrs des ressource humaines offrir leurs .services par ie biais de la structure la plus efficace. dition, since each type of employment relationship differs somewhat in terms of its contractual dimensions, each will require a more complex approach to how human resource management activities are administered. We identify the six types of employment relationships and the dimensions on which they differ in Table I. In the following sections of the paper, each type of control mechanism (market, bureaucratic, clan) is discussed to establish the theoretical foundation which explains the conditions under which each type is best suited. Then we discuss the four market-type employment relationships (day worker, subcontract. temporary agency, and leasing). the bureaucratic relationship, and the clancontrol rela- tionship (selfdirected work teams) in terms of how they are similar or different along eleven dimensions, including specific human resource management dimensions. Fi- nally, we develop a set of propositions which will guide future empirical investigations of our theoretical perspec- tives. Although the frameworks are applicable to many dif- ferent kinds of transactions, we are specifically interested in employment transactions and relationships. A common criticism of the HRM literature is that researchers tend to observe phenomena, but that the work is atheoretical (Davis-Blake & U7.4 1993; Pfeffer & Baron, 1988) and descriptive (Cohen & Haberfeld, 1993; Mangum, Mayall, & Nelson, 1985; Mayall & Nelson, 1982). Empirical stud- CASK 1996 Revue canadienne des sciences de I’adniinistration I Canadian Journal of Administrative Sciences U(l). I 12