Public administration has been influenced substantially in recent years by the public policy cycle literature--a fact which is illustrated by the rich contributions of the growing number of policy implementation studies. This research strain can be said to assume a functionalist approach to the discipline. By contrast, another recent strain in the literature, the growing body of policy issue network studies, is nudging the discipline in the direction of a structuralist approach. Using definitions from structural/functional analysis (Harmon and Mayer, 1986; 168-174; Katz and Kahn, 1978; 73-93, 174-198; Parsons, 1951; Benson, 1975; Marin and Mayntz, 1991b; Kenis and Schneider, 1991), structures are defined as patterns of routinized interactions among participants in a common enterprise while are seen as outcomes of these interactions necessary for survival of the enterprise. Although many social scientists see structure and function as related phenomena in a unified body of theory, public administration theorists have not developed a research orientation which facilitates a nexus of these two promising research strains (Burstein, 1991; Benson, 1975). Contributing to this shortcoming in the current literature is the lingering influence of the politics/administration dichotomy on public administration. As a result, empirical studies have accumulated without a common theoretical base that might facilitate weaving the threads of research findings into a more meaningful tapestry. In the following four sections, I (1) explain the nature of structural/functional analysis as it relates to public administration, (2) argue that Wilson's one-sided functional dichotomy of politics and administration distorts our understanding and inhibits the accumulation of knowledge, (3) review two major bodies of literature from the field to illustrate the loss of opportunities that results because researchers are not working from a unifying theoretical base, and (4) propose a structural/functional matrix as a better framework for understanding the role of public administration in contemporary public policy. A Structural-Functional Model for Public Administration It is truly unfortunate that the policy cycle literature and the policy issue network literature have developed largely independently when they are related so clearly in the context of broader social science theory (Kingdon, 1984; 1-22; Benson, 1975; Burstein, 1991; Kenis and Schneider, 1991). The policy cycle literature suggests that the public policy process consists generally of a set of four major functional stages--agenda setting, policy formulation, implementation, and evaluation (Jones, 1970; 10-14; Anderson 1990; 18-22; Palumbo, 1988; 17-22; Dye, 1992; Rosenbaum, 1991; 66-70; Burstein, 1991). In structural/functional terms, these functions are outcomes of policy subsystems that must be accomplished for policy to be produced. As Jones (1970; 13) states, Within each [policy cycle stage], processes can be identified which function to achieve the goals of the system. Viewed together, the four functions can be said to comprise a functional framework within which the activities of the policy subsystem can be located and categorized. Various writers have used this approach to develop concepts explaining the behavior of actors throughout the policy process. The strength of this literature is its deconstruction of the policy process--an analytical separation into functional stages and identification of roles performed in various stages. One of its major weaknesses is the absence of common definitions of the various roles and knowledge of which participants perform which roles. For public administration, it is important to know the extent to which professional, merit-selected employees are free to perform the various roles and to participate across all the functional stages. Obviously in our system of governance they are not completely free to do so. …