Where the Votes Come From: An Analysis of Electoral Coalitions, 1952-1968* ROBERT AXELROD University of California, Berkeley Where do the Democrats get their votes from? Where do the Republicans get their votes from? Because of the great interest of scholars, politicians, and the public in the changing shape of electoral politics in America, these questions have been asked over and over. V. O. Key has argued that one party or the other usually dom- inates American politics for decades by putting together a stable winning coalition and that since the Depression the dominant party has been the Democrats. 1 He has, however, also pointed to the existence of slow but steady secular trends in electoral coalitions. 2 In 1968 the rise of a third party challenge and the election of a Republican President led to nu- merous reassessments of the future of the party coalitions. Kevin Phillips's assertion of The Emerging Republican Majority 3 is perhaps the best known of these recent prognostications, but a critical review by Nelson Polsby 4 has already undermined much of Phillips's argument. The Republican Ripon Society has given a liberal interpretation of the Lessons of Victory, 5 while the less partisan interpretation of Scammon and Wattenberg has emphasized the size of the mid- dle groups in American politics. 6 The long term cyclical pattern in American electoral politics has been reported by Sellers, 7 and Burnham 8 * I gratefully acknowledge the help of Nicholas Miller, my research assistant. I also wish to thank the Institute of Governmental Studies and the Committee on Research of the University of California, Berkeley, for their financial support of this study, and the Inter- University Consortium for Political Research for making the survey data available. Finally, I wish to thank William Cavala, Jack Citrin, Merrill Shanks, and Aaron Wildavsky who read drafts of this paper. J V. O. Key, "A Theory of Critical Elections," Journal of Politics, 17 (February, 1955), 3-18. 2 V. O. Key, "Secular Realignment and the Party System," Journal of Politics, 21 (May, 1959), 198-210. ' Kevin Phillips, The Emerging Republican Majority (New Rochelle, New York: Arlington House, 1969). * Nelson W. Polsby, "An Emerging Republican Majority? Review Essay," The Public Interest, No. 17 (Fall, 1969), 118-126. 5 Ripon Society, The Lessons of Victory (New York: Dial Press, 1969). * Richard M. Scammon and Ben J. Wattenberg, The Real Majority (New York: Coward-McCann, 1970). 7 Charles Sellers, "The Equilibrium Cycle in Two- Party Politics," Public Opinion Quarterly, 29 (Spring, 1965), 16-38. 'Walter Dean Burnham, "American Voting Be- havior and the 1964 Election," Midwest Journal of and Converse et al. 9 have done excellent studies of some of the factors in recent presidential elections. Yet, a number of basic questions about the sources of partisan support remain. Surprisingly enough, even professional poli- ticians have only a rough and ready idea of how their own national coalitions are formed. For example, many do not know whether Blacks contribute more or less than union members to the vote totals of the Democratic party. As another example, we may ask whether the common wisdom is really correct that the Republicans do worse than the Demo- crats among the young and the poor. Questions such as these require measurement of groups and their voting behavior. Measuring where the parties have been and where they are now is the first step toward a sophisticated analysis of where they are going. This paper has two purposes. The first is ana- lytic—to specify the components of a group's contribution to a party's electoral coalition. The second is empirical—to measure the actual magnitudes of the contributions that have been made by selected groups in each of the last five presidential elections. The Meaning of a Contribution to a Coalition What does it mean to ask how large a contri- bution a group makes to the electoral total of a party? Clearly, a large group can contribute more votes to a party than can a small group. It is also true that a group with a high turnout can contribute more votes than a slightly larger group with a lower turnout. And yet it is also true that a relatively small group with a poor turnout can contribute votes to a party quite out of proportion to its size if the group loyally gives overwhelming support to that party. Thus measurement of the contribution of a group to a party's total vote must depend on three fac- tors: the size of the group, its turnout, and its loyalty to the given party. Political Science, 12 (February, 1968), 1-40. See also his Critical Elections (New York: Norton, 1970). •Phillip E. Converse, Warren E. Miller, Jerrold G. Rusk, and Arthur C. Wolfe, "Continuity and Change in American Politics: Parties and Issues in the 1968 Election," American Political Science Review, 53 (December, 1969), 1083-1105. 11