* Since my conclusions may appear controversial, I shall indicate the major sources of the statements used here. Those concerning the normal operations of the party system are adapted from such standard works as Pendelton Herring, The Politics of Democracy; Austin Ranney and Wilmoore Kendall, Democracy and the American Party System; V. O. Key, Politics, Parties and Pressure Groups, and Robert Dahl, Preface to Democratic Theory. The material on the goals of convention delegates comes from Nelson Polsby and Aaron Wildavsky, Presidential Elections. Generalizations on voting behavior are drawn primarily from Campbell, Converse, Miller, and Stokes, The American Voter. I have also relied heavily on the only systematic work comparing convention delegates and ordinary voters — McClosky, Herbert et al. , “Issue Conflict and Consensus Among Party Leaders and Followers,” The American Political Science Review, LIV (06, 1960), 406–27CrossRefGoogle Scholar. (See also McClosky's, “Consensus and Ideology in American Politics,” American Political Science Review, LVIII (06, 1964), 361–82)CrossRefGoogle Scholar.

After the 1956 conventions, McClosky distributed, and then analyzed, lengthy questionnaires filled out by large numbers of Republican and Democratic delegates and by samples of voters who identified with these parties. Using McClosky's work, one can make reliable statements about the characteristics and issue preferences of party leaders (represented at the conventions) and party followers (as they are found throughout the nation). Finally, I have made substantial use of some 150 interviews with Goldwater delegates held at the Hotel Fairmont in San Francisco, where the California, Illinois, and New Jersey delegates were housed. The interviews were conducted in the hotel lobby by myself, my sister, Judy Gordon, a friend, Maralyn Millman, and two graduate students — James Payne and Joseph Paff. Gerry Bass helped by monitoring the press for us. While these unstructured, probing interviews in no way represent a systematic sample of the delegates, they were undertaken because they have one great advantage over the usual mail questionnaire: the interviewer can try to insist on answers and pound away until some kind of response is forthcoming.

¹ The American Political Science Review, LIV (06, 1960), 423Google Scholar.

² New York Times, October 27, 1964.