Beginning in 1978, with the support of the National Science Foundation, the National Election Studies (NES) launched a series of large scale and ambitious studies of elections to the U.S. House of Representatives. It is no exaggeration to say that these studies have revolutionized our understanding of voting in House contests, and have energized the analysis of the House of Representatives as an institution as well. But in the meantime, our understanding of Senate elections languished. How voters choose their Senators, how and how well Senators represent the interest of their constitutents, what kinds of campaigns Senators and their opponents wage: on such basic questions as these, we had little to say. The problem was not that Senate elections are theoretically innocuous; the study of Senate elections provides an excellent and in some respects optimal entree to fundamental questions for American democratic politics. Nor was it a case of indifference on the part of the research community: there was in fact considerable and growing interest in the study of Senate elections. The problem, rather, was the absence of suitable data. The important theoretical and substantive questions occasioned by Senate elections could not be resolved with the evidence at hand. With this problem in mind, and with the support of the National Science Foundation, the Principal Investigators and the NES Board of Overseers initiated in 1988 and 1990 the first two parts of a three-part study of Senate elections. The current project supports completing the Senate election study in 1992. The justification for completing the study rests on the substantive and theoretical advances that the 1992 study makes possible, for four themes especially: (1) political representation, of interest not least because contrary to original intentions, the contemporary Senate appears to be more responsive to popular sentiment than the House; (2) divided government, now the norm in the American system of separated powers, one that may menace coherent legislation and erode political accountability; (3) the nature and consequences of political campaigns, motivated by the recognition that Senate elections provide an opportunity to examine the interplay among candidates, the political institutions governing elections, and voters in a particularly powerful way; and (4) the "simple act of voting," developing and testing comparative theories of voter choice. Provisional analysis of the first two phases of the NES Senate election Study suggest that the complete study proposed here will trigger the same kind of explosion of research that the 1978 National Election Study set off for our understanding of the House. Indeed, the gains are likely to be greater here. The Senate Election study builds upon research on House and Presidential elections, and is carefully coordinated, theoretically and operationally, with ongoing NES studies. The research is comparative, contributing broadly to theories of voting, representation, campaigns, and, ultimately to our understanding of American electoral institutions.
