Adapting Agencies : Competition, Imitation, and Punishment in the Design of Bureaucratic Performance

This essay examines the roles of competition, comparison, imitation, and punishment in the design of bureaucratic performance. Through a series of simulations, this essay examines how these elements - alone and in combination - drive both the performance and technology search paths of adaptive agencies. I first provide a baseline model of an agency as a complex adaptive system, which experiences performance, sets goals, and searches for new technologies in a fully adaptive manner. I then provide three basic design choices for agencies, including: a competitive environment, an environment that allows agencies to imitate one another, and an environment in which lone agencies are punished at random for poor performance. These design choices help to assess ways of altering an agency's performance, and the likelihood of innovation and refinement as agencies search for new technologies. The models show the difficult tradeoffs encountered in designing agencies. Predictability is an enemy of change. Search can be as important as the outcomes of search. Stability competes with growth. Together, these tradeoffs offer a set of competing "goods" designers face when attempting to alter an agency's direction. By encouraging change we upset prediction; by encouraging innovation, we do not guarantee outcomes