GOVERNMENT TERMINATION.

This paper reviews a range of approaches to the analysis of government termination, by any account a very important substantive concern for political science. One essential preliminary matter is the distinction between government duration and government durability--the former an essentially empirical concept, the latter essentially theoretical. It is also important to note that empirical research into government termination is heavily conditioned by the precise definition of what marks the end of one government and the beginning of the next. Approaches to analyzing government termination can be divided into those that are fundamentally empirical and those based on a priori modeling. Both research traditions are reviewed. The empiricist approach has evolved into a body of work that applies increasingly sophisticated event-history models to a dataset that has become to a large extent common within the profession. The a priori approach has developed within the traditions of noncooperative game theory to model the responses of key actors to new information, for example, about the likely results of an election that might be called if the government were to fall. There is clearly unfulfilled potential to merge these two research traditions into a single more comprehensive account of government termination. [ABSTRACT FROM AUTHOR]