On tolerating policy and organizational termination: Some design considerations

This essay explores some factors that have led us to expect perpetuity rather than termination as normal in dealing with policies and organizations. After suggesting that this is a dangerous assumption where change and uncertainty are sufficient to require substantial adaptive capacity, it is proposed that a model tolerant of acting to make termination either more or less likely is preferred. Given that we already know how to invest to make termination less likely, the next two sections explore some strategies that may allow termination to be more easily tolerated, and the learning incident to its existence increased. The first explores the use of matrix models of organization to institutionalize the capacity for initiating and terminating temporary projects within organizations. The second section suggests several external mechanisms (savings banks, insurance policies, receivership referees, trust officers, salvage specialists, and marriage/divorce/ escrow brokerages) that could operate to make termination of policies and organizations less difficult.I wish to express my appreciation especially to Arnold Meltsner, a colleague in no way responsible for either the obvious or subtle flaws of this essay. Several years ago we undertook an empirical investigation of several score terminated organizations, and I know my thinking has greatly benefited from his effort in setting both me and the subject straight. His comments on this essay have been very helpful. Eugene Bardach and Aaron Wildavsky also provided helpful comments in the drafting of this paper.