Polls and the Problem of Strategic Information in Elections



In the analysis up to now, we have assumed that voters use available and relevant information to make rational choices. They vote insincerely, for example, because there is a reason—they can effect a better outcome by not voting for all their most-preferred or acceptable candidates. In plurality voting, in particular, the practice of voting for a second choice is not uncommon in elections wherein one’s first choice would appear not to be a serious contender.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Footnotes to Chapter 7

  1. 1.
    This chapter is based largely on Steven J. Brams, “Strategic Information and Voting Behavior,” Society 19,3 (September/October 1982), 4–11.Google Scholar
  2. 2.
    A study of the effects of vote projections on late California voters in the 1964 presidential election found this not to be true, but the effects of projections, especially on turnout, seem to have been substantial in the 1980 presidential election, perhaps in part because Ronald Reagan’s landslide victory over Jimmy Carter deviated from expectations about a close race. See Harold Mendelsohn, “Election-Day Broadcasts and Terminal Voting Decisions,” Public Opinion Quarterly 30, 2 (Summer 1966), 212–225; and John E. Jackson and William H. McGee III, “Election Reporting and Voter Turnout” (mimeographed, 1981). Turnout decline on the West Coast in the 1980 election is disputed in Laurily K. Epstein and Gerald Strom, “Election Night Projections and West Coast Turnout,” American Politics Quarterly 9, 4 (October 1981), 479–491, where other references to the literature—and continuing debate on this issue—are given. Theoretical and empirical effects of polling are reviewed in Steven J. Brams, Paradoxes in Politics: An Introduction to the Nonobvious in Political Science (New York: Free Press, 1976), pp. 65–78. Quantitative calculations making use of poll information are developed in Dale T. Hoffman, “A Model for Strategic Voting,” SIAM Journal on Applied Mathematics 42, 4 (August 1982), 751–761.CrossRefGoogle Scholar
  3. 4.
    This example was given in Steven J. Brams, The Presidential Election Game (New Haven: Yale University Press, 1978), pp. 219–220; and S. J. Brams, Comparison Voting, Innovative Instructional Unit (Washington, DC: American Political Science Association, 1978), pp. 47–48, which will appear in slightly revised form in Modules in Applied Mathematics: Political and Related Models, Vol. 2, edited by Steven J. Brams, William F. Lucas, and Philip D. Straffin, Jr. (New York: Springer-Verlag, 1982). It was originally suggested to S. J. Brams by Philip D. Straffin, Jr.Google Scholar
  4. 6.
    William C. Stratmann estimates that, under approval voting, Goodell would have won with about 59 percent of the vote (i.e., approval from about 59 percent of the voters) to about 55 percent each for Buckley and Ottinger. Personal communication to S. J. Brams, September 21, 1977, based on work reported in W. C. Stratmann, “The Calculus of Rational Choice,” Public Choice 18 (Summer 1974), 93–105.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2007

Personalised recommendations