No CrossRef data available.
Article contents
The Political Scientist Decides: An Examination of the 1969 APSA Ballot
Published online by Cambridge University Press: 28 November 2022
Extract
I always vote for the man, not the party.
–Trad.
In its election for offices in 1969, the American Political Science Association, apparently for the first time in its rarely turbulent history, found the nominees of its Official Nominating Committee challenged by an insurgent group. In order to handle this unprecedented situation, it was decided at the annual meeting to carry out the election by mail ballot and the American Arbitration Association was engaged to administer the operation.
Ballots were mailed to the 13,061 members of the Association in October, 1969. Accompanying them were materials containing statements of belief and biographies for each of the candidates. The response rate was 64 percent.
The ballots carried the contestants indicated in Table 1. For each office the candidates are listed in the Table in the order of their vote result (they were listed in alphabetical order on the ballot) and for each candidate the group endorsements, as they were presented on the ballot, are indicated. Except for the group endorsements, no identifying information accompanied the names of the candidates on the ballots.
- Type
- Research Article
- Information
- Copyright
- Copyright © The American Political Science Association 1970
Footnotes
*
Data analysis in this report was carried out at the Computing Center of the University of Rochester which is supported in part by NSF Grant GJ-828. Keypunching was paid for by a University of Rochester research grant. Helpful comments were forcefully delivered by Michael Magie, Henry Manne, and Kenneth Shepale.
References
1 The Election Committee's Report on their procedures and other headaches is in PS, Winter 1970, pp 28–33.
2 The Ad Hoc Committee's letter is reprinted in PS, Fall 1969, pp. 703–04.
3 The circumstances appear to be explained in the Election Committee's Report, op. cit.
4 At least one voter has complained about this limitation of options. See PS, Winter 1970, p. 80.
5 The official results are given in PS, Fall 1969, p. 670. The comparison suggests that the missing ballots might be marginally more favorable to Ad Hoc candidates. Hence the present analysis may understate slightly the strength of the Ad Hoc endorsement. In one case, Robert Clarke's vote for Treasurer, the ballots suggest a higher vote than the official results. Gremlins, no doubt.
6 In fact, as seen below, there were many who voted for Elden and otherwise selected an Ad Hoc slate except to withhold a vote from Prestage. If these are added in, Elden received more votes from Ad Hoc partisans than from Caucus supporters.
7 These “invalid patterns” represent, of course, voters who voted for more than eight candidates. That only 0.1 percent of the voting members of the Association actually made this mistake may be taken by some to be a new monument to the quantitative revolution in the profession. Things are even better on the Vice Presidential race where only 4 members voted for more than three candidates. And near-perfection is reached on the Presidential race: only one member voted for both candidates.
8 For similar findings in quite a different election, see Mueller, John E., “Voting on the Propositions: Ballot Patterns in California,” 63 American Political Science Review 1197–1212 (December 1969).CrossRefGoogle Scholar
9 Those who selected patterns 4 and 7 were also likely to withhold their vote from Tobe Johnson in the unopposed Council race. This was also true for the pattern 21 voters, but to a much lesser degree.
10 It should be made clear that the characterizations in Table 4 are the most reasonable that could be made in each case. For example, pattern 6 is declared to represent Ad Hoc voters who gave Kariel their eighth vote. This is a far more parsimonious interpretation than to assume they are APSA partisans with deviations, for the pattern represents three deviations from a pure APSA ticket: the addition of McClosky and Sindler and the elimination of Lipsitz. For some cases such as pattern 12, however, parsimony dictates no single solution.
11 Some effort was made to assess the order effect in this election through multiple regression analysis of the aggregate results. The effect, however, was too small and the number of candidates too limited for the effect to be adequately separated out in the analysis – which is not to say, of course, that the effect isn't there, simply that it is difficult to measure. For a case where such an approach did work, see Mueller, John E., “Choosing Among 133 Candidates.” Public Opinion Quarterly, Fall 1970.CrossRefGoogle Scholar See also Bain, Henry M. and Hecock, Donald S., Ballot Position and Voters Choice (Detroit: Wayne State University Press, 1957).Google Scholar
12 One might also add to the Ad Hoc group those selecting patterns 20 and 23 and to the APSA group pattern 17, making the difference even smaller. Pattern 12 can be declared a standoff.