We struggle to manage complexity every day. We follow intricate diets to lose weight, juggle multiple remotes to operate our home entertainment syste… - Kathleen M. Eisenhardt

Recent organizational approaches to control (e.g., Ouchi 1979) suggest two underlying control strategies. On the one hand, control can be accomplished through performance evaluation. Performance evaluation refers to the cybernetic process of monitoring and rewarding performance. This strategy emphasizes the information aspects of control. Namely, to what degree can the various aspects of performance be assessed? Alternatively, control can be achieved by minimizing the divergence of preferences among organizational members. That is, members cooperate in the achievement of organizational goals because the members understand and have internalized these goals. This strategy emphasizes people policies such as selection, training, and socialization.

Organizational design often focuses on structural alternatives such as matrix, decentralization, and divisionalization. However, control variables (e.g., reward structures, task characteristics, and information systems) offer a more flexible approach. The purpose of this paper is to explore these control variables for organizational design. This is accomplished by integration and testing of two perspectives, organization theory and economics, notably agency theory. The resulting hypotheses link task characteristics, information systems, and business uncertainty to behavior vs. outcome based control strategy. These hypothesized linkages are examined empirically in a field study of the compensation practices for retail salespeople in 54 stores. The findings are that task programmability is strongly related to the choice of compensation package. The amount of behavioral measurement, the cost of measuring outcomes, and the uncertainty of the business also affect compensation. The findings have management implications for the design of compensation and reward packages, performance evaluation systems, and control systems, in general. Such systems should explicitly consider the task, the information system in place to measure performance, and the riskiness of the business. More programmed tasks require behavior based controls while less programmed tasks require more elaborate information systems or outcome based controls.

Agency theory is an important, yet controversial, theory. This paper reviews agency theory, its contributions to organization theory, and the extant empirical work and develops testable propositions. The conclusions are that agency theory (a) offers unique insight into information systems, outcome uncertainty, incentives, and risk and (b) is an empirically valid perspective, particularly when coupled with complementary perspectives. The principal recommendation is to incorporate an agency perspective in studies of the many problems having a cooperative structure.