Riker, William H. (1920–1993)

William H. Riker was one of the most important of the postwar U.S. political scientists, influencing several generations of scholars in the direction of rational choice theory. His work spanned the issues of federalism, coalition theory, democratic theory, and his invention of heresthetics. He also contributed several important articles on power and was the first to bring the Shapley-Shubik power index to bear upon the issue of coalition formation in legislatures. His work on democratic theory founded what became known as the Rochester school (after Rochester University, where he taught most of his life).

Riker's articles on power made a distinction between ego-oriented power—the ability to increase your own utility by manipulating outcomes—and other-oriented power—the ability to decrease another's utility by manipulating others. It is a variant of the more standard distinction between outcome power and social power. His work here is important for its recognition that for ego-oriented definitions power always exists in social relations because it refers to outcomes; so for any outcomes we can always assign power to someone.

Riker applied the Shapley—Shubik Index of power to coalition formation. In his book The Theory of Political Coalitions he argued that politics was essentially zero sum; that is, the spoils of government mean that whatever one party gains the other loses. Thus if there is no single party that commands a majority in a legislature, the coalition that forms must share the spoils between them. In his model, the spoils of winning is office. Parties will seek to form the smallest winning coalition—that is, the smallest coalition that still commands a majority in the legislature—because that way each party will be able maximize the number of cabinet seats it enjoys. In this model, as in the Shapley-Shubik power index, small parties that are required for a coalition to win will have as much power as larger parties in the coalition. This counterintuitive result ensured the book's success and its controversy. For well over 30 years political scientists measured actual coalitions against Riker's predictions. They demonstrated that (a) not all coalitions are minimum winning ones, and (b) the spoils of office tend to be split up among winning parties in proportion to the number of seats each holds in the parliament. However, minimum winning coalitions form more often that one would expect by chance, which suggests there is some truth in Riker's simple model. Coalition theory has advanced enormously from Riker's argument, produced in 1962. Far more sophisticated and empirically accurate models now exist, but these could not have been achieved without the inspiration provided by Riker's original contribution.

Riker's work on the social choice problems of democracy has been equally influential, and the Rochester school equally controversial. Riker introduced the social choice problems that bedevil our understanding of democratic results to many generations of political scientists, even if his own arguments about what these mean do not receive universal assent. It has been said of Riker, as of Thomas Hobbes, that he is more interesting when wrong than most of us when right.

...