The article is an examination of the recent reform process (1968-1988) which changed the American system of selection of the presidential candidates. This examination is done on the basis of a fundamental point of view: the change of the selection system is an example (one of the few in democratic societies, indeed) of institutional change. Before entering this point of view, the article discusses three preliminary questions. The first is concerned with the historical evolution of the selection system. The second, with the processes of political change (along with the schools which conceptualized it) which underlined the forces who pushed for the recent change of that system. The third question is concerned with the specific characteristics of the change, both of the selection system and of the financing of the political campaign. After these questions are discussed, the article examines the consequences of the reform process, at the level both of the party organization and of the party in the government. In this respect, the article draws this conclusion: what characterizes the reform process is its high level of impredictability and contradiction. That process did not produce linear consequences, for example pro- or against- the parties. Of course, there were many unintended effects, but the general trend was one of confusion. Why did that happen? The article tries to argument this point: the reform, which was probably necessary, has been carried out on the basis of an inadequate political theory. The limit was in the behavioral approach to political institutions adopted by the reformers. Do we have an alternative (and reliable) approach to adopt for the institutional change?