A chief criticism of the American party system is the lack of party responsibility. In the view of some students, one characteristic of our political system that contributes to this irresponsibility is the practice in some states of allowing individuals to vote in primaries without regard to their partisan allegiances. In such an open primary Republicans may, if they wish, vote in the Democratic primary, and vice versa. The contrasting, and more common, practice is the closed primary, in which participation is restricted to party “members.” Some political scientists think that the closed primary, by subjecting legislators to the presumed discipline of periodic scrutiny by their party's members, induces a greater measure of party regularity than the open primary, in which the official has to satisfy a more motley clientele. This position was taken in the best-known statement of the “party government” school, the 1950 report of the APS A Committee on Political Parties:
The closed primary deserves preference because it is more readily compatible with the development of a responsible party system…. on the other hand, the open primary tends to destroy the concept of membership as the basis of party organization.
Other political scientists have expressed doubts about this presumed relationship between primaries and party responsibility, but there has been no systematic empirical evidence on the point. This paper will examine the relationship between primaries and party responsibility by comparing the party regularity of senators from open primary and closed primary states.