¹ It is worth noting that the bonus-vote reform was adopted by the Republicans after the angry divisions and charges of rigging in their 1912 convention and their subsequent schism and loss of the presidency. The Democrats, being in power at the time, did not follow suit for several decades. See David, Paul T., Goldman, Ralph M., and Bain, Richard C., The Politics of National Party Conventions (Washington, D.C.: The Brookings Institution, 1960), pp. 165–168 Google Scholar.

² The literature on representation is as vast as its intellectual puzzles and institutional mazes are difficult. I have found especially helpful the work of Pitkin, Hanna Fenichel, especially her The Concept of Representation (Berkeley and Los Angeles: University of California Press, 1967)Google Scholar. Also useful are many of the papers in Pennock, J. Roland and Chapman, John W., eds., Representation: Nomos X (New York: Atherton Press, 1968)Google Scholar.

³ Commission on Party Structure and Delegate Selection, Mandate for Reform (Washington, D.C.: Democratic National Committee, 1970), Guideline A–2, pp. 40, 26–28 Google ScholarPubMed; see also the summary in Congressional Quarterly, 11 28, 1969, pp. 2418–2419 Google Scholar.

⁴ The first proposal was made by Woodrow Wilson in his initial address to Congress in 1913. Others have since been made by Senators George Smathers (1952), Paul Douglas (1953), Estes Kefauver (1953), William Langer (1955), William Proxmire (1960), and Margaret Chase Smith (1961).

⁵ For a useful listing of the state presidential primaries and their outcomes from 1912 to 1964, see Davis, James W., Presidential Primaries: Road to the White House (New York: Thomas Y. Crowell Company, 1967), pp. 278–305 Google Scholar.

⁶ See, for example, Polsby, Nelson W. and Wildavsky, Aaron B., Presidential Elections: Strategies of American Electoral Politics, 2nd ed. (New York: Charles Scribner's Sons, 1968), pp. 238–239 Google Scholar; Davis, , Presidential Primaries: Road to the White House, pp. 269–271 Google Scholar; Pomper, Gerald, Nominating the President (New York: W. W. Norton & Company, 1966), pp. 230, 275–276 Google Scholar; and Sorauf, Frank J., Party Politics in America (Boston: Little, Brown and Company, 1968), p. 273 Google Scholar.

⁷ The most praised features of the Oregon presidential primary are the requirements that: (1) all serious presidential contenders are placed on their particular party's ballot unless they specifically state that they are not candidates, which is thought to make it difficult for any bona fide contender to duck a hard contest; and (2) the delegates are legally bound to support the preferential poll's winner for one or more convention ballots. At present the primaries in Maryland, New Mexico, North Carolina, Tennessee, and Wisconsin have similar, though not identical, provisions.

⁸ Campbell, Angus, “Surge and Decline: A Study of Electoral Change,” Public Opinion Quarterly, 24 (Fall, 1960), 397–418 CrossRefGoogle Scholar. The article has been reprinted in Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., Elections and the Political Order (New York: John Wiley and Sons, Inc., 1966), pp. 40–62 Google Scholar. The latter version is cited here, and the quotation in the text is from page 51.

⁹ Key, V. O. Jr., American State Politics: An Introduction (New York: Alfred A. Knopf, 1956), pp. 134–138 Google Scholar.

¹⁰ Wolfe found a mean turnout rate in all primary elections for governor of 32 per cent, as compared with a mean turnout rate in general gubernatorial elections of 53 per cent. The corresponding means for U.S. Senator were 30 per cent in the primaries and 54 per cent in the general elections; and for U.S. Representatives, 27 per cent in the primaries and 52 per cent in the general elections. See Wolfe, Arthur C., The Direct Primary in American Politics (unpublished Ph.D. thesis in political science, University of Michigan, 1966), Table 4.6, p. 102 Google Scholar. For comparable data on turnout in primary elections and general elections for U.S. Representatives from 1938 to 1956, see Foster, Julian F. S., Congressional Primaries (unpublished Ph.D. thesis in political science, University of California, Los Angeles, 1963), Ch. IVGoogle Scholar.

¹¹ My data show that in all election-pairs in which both parties' primaries were contested and both parties presented candidates in the general election, the mean turnouts were: gubernatorial primaries, 34 per cent, gubernatorial general elections, 57 per cent; senatorial primaries, 32 per cent, senatorial general elections, 57 per cent. A somewhat revised version of these data is presented in “Parties in State Politics,” in Jacob, Herbert and Vines, Kenneth N., eds., Politics in the American States: A Comparative Analysis, 2nd ed. (Boston: Little, Brown and Company, 1971), Table 3, 98 Google Scholar.

¹² Cf. the discussions of the presidential primaries' role in determining who gets the nominations in David, Goldman, and Bain, The Politics of National Party Conventions, Ch. 10; Polsby, and Wildavsky, , Presidential Elections, pp. 83–86 Google Scholar; Davis, , Presidential Primaries: Road to the White House, Chs. 4, 10 Google Scholar; and Scammon, Richard M. and Wattenberg, Ben J., The Real Majority (New York: Coward-McCann, Inc., 1970), Chs. 7–10Google Scholar.

¹³ The criterion of not more than 80 per cent of the vote for the winner excluded most primaries: in 1948 either Dewey or Truman or both won over 80 per cent of the vote in one or both contests in every state, and the same was true for Eisenhower in 1956, Nixon in 1960, and Johnson in 1964.

¹⁴ As the terms are used in this paper, “demographic representativeness” denotes the degree of likeness between the distributions of socioeconomic characteristics among the representatives (in this case, voters in the primaries and other low-turnout elections) and the corresponding distributions among the represented (in this case, the eligible but nonvoting citizens). “Preferential representativeness” denotes the degree of likeness between the distributions of preferences for candidates and attitudes on public issues among the representatives and the corresponding distributions among the represented.

¹⁵ Campbell, , “Surge and Decline …,” pp. 53–57 Google Scholar. See also Campbell, Angus and Cooper, Homer C., Group Differences in Attitudes and Votes (Ann Arbor: Survey Research Center, University of Michigan, 1956), Ch. IVGoogle Scholar.

¹⁶ Ranney, Austin and Epstein, Leon D., “The Two Electorates: Voters and Nonvoters in a Wisconsin Primary,” The Journal of Politics, 28 (08, 1966), 598–616 CrossRefGoogle Scholar; and Ranney, Austin, “The Representativeness of Primary Electorates,” Midwest Journal of Political Science, 12 (05, 1968), 224–238 CrossRefGoogle Scholar.

¹⁷ The New Hampshire study was conducted by Audits and Surveys, Inc., of New York City. A state-wide probability sample of registered Democratic and Republican voters was interviewed, yielding a total of 1143 usable interviews. The data are made available by the Inter-University Consortium for Political Research, and the Social Science Data and Program Library Service of the University of Wisconsin. The Wisconsin study was conducted by the Wisconsin Survey Research Laboratory, using a statewide clustered area probability sample of persons of voting age and yielding a total of 573 usable interviews. The data are made available by the WSRL.

¹⁸ Cf. White, Theodore H., The Making of the President 1968 (New York: Atheneum Publishers, 1969), Ch. 4Google Scholar; and Chester, Lewis, Hodgson, Godfrey, and Page, Bruce, An American Melodrama: The Presidential Campaign in 1968 (New York: Dell Publishing Company, 1969), pp. 1–10 Google Scholar.

¹⁹ Both studies used the Survey Research Center's familiar battery of questions for eliciting party identification.

²⁰ All New Hampshire respondents were asked before the primary, “As of now, how certain are you that you will vote in the primary this year?” The replies were distributed as follows:

Since the election returns showed that 67.0 per cent of New Hampshire's registered voters voted in the primary, it seemed reasonable to classify those answering “very certain” as intenders, and those giving the other answers as nonintenders. After removing the respondents whose party identifications were not ascertained, there remained 769 intenders and 329 non-intenders.

²¹ Some colleagues have suggested that the representativeness of presidential primary electorates would be better tested by comparing the party identifiers who vote in a state's primary with all who identify with that party in the state: for example, RI/RI + RN rather than RI/RN. They correctly point out that this approach would narrow the participant/nonparticipant differences reported in this study, and thus enhance the primaries' claim to be truly representative. However, I have chosen the straight participant/nonparticipant comparison because I believe it more accurately operationalizes the concept of representation held by most party leaders who regard the primaries as the nominating system's most representative element. From what they say and do, on the McGovern-Fraser Commission and elsewhere, it seems clear that most leaders assume a sharp distinction between the “representatives” (the convention delegates, the primary voters) and the “represented” (the party's grass-roots supporters). If, as I believe, this concept of representation permeates most current political dispute about the proper organization and role of primaries and conventions, then accepting it rather than substituting a concept which regards the representatives as part of the represented seems likely to make our empirical findings more relevant to those disputes.

²² Mandate for Reform, pp. 27–29.

²³ Milbrath, Lester W., Political Participation (Chicago: Rand McNally & Company, 1965), p. 52, emphasis in the originalGoogle Scholar.

²⁴ Statements of the underlying causes for the partisanship-participation relationship are given by Bowen, Don R., Political Behavior of the American Public (Columbus, Ohio: Charles E. Merrill Publishing Company, 1968), p. 129 Google Scholar; and in Epstein's, Leon D. discussion of the general role of party labels in “structuring the vote,” in Political Parties in Western Democracies (New York: Frederick A. Praeger, Publishers, 1967), Ch. IVGoogle Scholar. The Survey Research Center's findings on the relationship in presidential elections are set forth in Campbell, Angus, Gurin, Gerald, and Miller, Warren E., The Voter Decides (New York: Harper and Row, 1954), pp. 107–110, esp. Table 7.11Google Scholar; and in Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter (New York: John Wiley & Sons, Inc., 1960), pp. 142–143 Google Scholar. The SRC's findings on the corresponding relationship in off-year congressional elections are given in Campbell and Cooper, Group Differences in Attitudes and Votes, Ch. IV.

²⁵ See the studies mentioned in footnote 16.

²⁶ For two somewhat different analyses of the interaction of institutional requirements, including registration laws, and levels of psychological motivation as they affect voting turnout, see The American Voter, Ch. 11; and Kelley, Stanley Jr., Ayres, Richard E., and Bowen, William G., “Registration and Voting: Putting First Tilings First,” American Political Science Review, 61 (06, 1967), 359–379 CrossRefGoogle Scholar.

²⁷ Chester, Hodgson, and Page state that private polls commissioned from Quayle Associates by the Johnson organization showed a substantial surge for McCarthy in the final three weeks of the campaign: An American Melodrama, pp. 110–112. White cites other private polling data showing a steady rise in McCarthy support from January on: The Making of the President 1968, p. 89.

²⁸ In 1968 the New Hampshire Republican primary ballot carried the names of Nixon, Romney (who withdrew from the national contest on February 28, thirteen days before the primary, but whose name remained on the ballot), Harold Stassen, Willis Stone, Herbert F. Hoover, David Watumull, William Evans, Elmer Coy, and Don Dumont. There were also Republican write-in votes for Rockefeller, McCarthy, and Johnson. The Democratic primary ballot carried the names of McCarthy, John Crommelin, Richard Lee, and Jacob Gordon; and write-in votes were recorded for Johnson and Nixon. Under Wisconsin's new Oregon-style law, the state placed on the Republican ballot Nixon, Reagan, Stassen, Romney, and Rockefeller (who withdrew but received write-in votes). McCarthy and Johnson were placed on the Democratic ballot, and there were also write-in votes for Kennedy, Humphrey, and George Wallace. See Polsby and Wildavsky for an illuminating discussion of the strategy of write-in campaigns: Presidential Elections, pp. 84–85.

²⁹ In our study of the 1966 Wisconsin gubernatorial primary we asked our respondents if they favored changing the open primary. 82 per cent favored keeping it as it is, and only 9 per cent favored converting it into a closed primary.

³⁰ Cf. Sorauf, , Party Politics in America, pp. 205–206 Google Scholar. For a more detailed account of crossing over in Wisconsin primaries, see Epstein, Leon D., Politics in Wisconsin (Madison: University of Wisconsin Press, 1958), pp. 94–95 Google Scholar; and Penniman, Howard R., The American Political Process (Princeton, N.J.; D. Van Nostrand Company, Inc., 1962), pp. 85–86 Google Scholar.

³¹ Sorauf, , Party Politics in America, p. 206 Google Scholar.

³² Key, V. O. Jr., Politics, Parties & Pressure Groups, 5th ed. (New York: Thomas Y. Crowell Company, 1964), p. 391 Google Scholar. See also Lawson, Kay, Political Parties and Democracy in the United States (New York: Charles Scribner's Sons, 1968), p. 134 Google Scholar.

³³ They included twenty-seven Republicans for McCarthy, three for Johnson, three for Kennedy, and one for Wallace; and three Democrats for Nixon and one for Rockefeller.

³⁴ It is worth recalling, in this regard, that our study of the 1966 Wisconsin gubernatorial primary turned up only ten crossovers. Of these, four were Republicans who left their own uncontested primary to vote in the hotly-contested Democratic primary, but six were Democrats who made the opposite move! See Ranney, , “The Representativeness of Primary Electorates,” pp. 227–228 Google Scholar. Both the 1966 and 1968 Wisconsin findings support Sorauf's, view that “as for organized ‘raiding,’ there is precious little evidence to suggest it is more than a worrisome myth”: Party Politics in America, p. 206 Google Scholar.

³⁵ Leading discussions of the problem posed for theories of democracy and representation by differing intensities with which opinions are held by different members of the community include: Dahl, Robert A., A Preface to Democratic Theory (Chicago: University of Chicago Press, 1956), Ch. 4Google Scholar; Mayo, Henry B., An Introduction to Democratic Theory (New York: Oxford University Press, 1960), pp. 178–179, 206–212 Google Scholar; and Kendall, Willmoore and Carey, George W., “The ‘Intensity’ Problem and Democratic Theory,” American Political Science Review, 62 (03, 1968), 5–24 Google Scholar.