During the period in which our article (APSR vol. 68 [June 1974]) circulated in manuscript form it provoked an unusual amount of collegial reaction. Of course, we were quite prepared for a reaction from those who use decision-theoretic models in their research—they were our intended audience. More surprisingly, we also received comments from less directly involved bystanders—a medieval historian for example. All this correspondence indicates to us that nearly everyone has his own theory of how voters behave, and that most such theories do not agree with the one presented in our article. The comments of Professors Tullock, Beck, Mayer and Good, and Stephens further support this conclusion.

In an appendix to this note we have responded to the imaginative point raised by Tullock. As for the traditional questions raised by our other critics, however, we adopt a different line of rebuttal. Rather than conduct an unfruitful debate over the a priori plausibility of the minimax regret model we will do something that theorists too seldom do: examine some data. Before doing so we will make an important distinction between using a model prescriptively and using it descriptively. (Decision-theoretic types tend to move a bit too easily from one usage to the other.) Then, after reviewing the major point of our article we will turn to the data.

This analysis of open-ended interviews conducted in 1969–70 with small samples of English, French, and white and black American children focuses on orientations toward the heads of state of the three nations and the Prime Ministers of Britain and France. The English children exhibit remarkably positive views of the Queen. Many of them believe her to be the nation's effective leader rather than a figurehead. Any political animus they express is directed toward the Prime Minister. The French children tend to describe the President of the Republic positively when they express feelings toward him at all but expect him to behave harshly and arrogantly in actual situations. Their descriptions of political leaders exude authoritarian imagery and perceptions; they perceive the President of the Republic in an impersonal, undifferentiated manner, and are only barely aware of the Premier. The general descriptions of the President of the United States by white American children interviewed in 1969–70 are remarkably similar to the benevolent-leader perceptions of children in the Eisenhower-Kennedy years, but the 1969–70 children exhibit much less idealized views of a president depicted as a law-breaker. A post-Watergate white American comparison group interviewed in June 1973 is generally aware of, but puzzled by, the Watergate events. At this early stage in the Watergate revelations, white children were only slightly less likely than the 1969–70 respondents to idealize the President, but were substantially more likely to perceive the president depicted as a law-breaker in terms implying that the President is “above the law.” The American black comparison group, which is too small and special in its geographical circumstances to offer more than suggestive findings, is the most negative of the four groups in general responses to the head of state, but is more like the white American group than like the English or French children in expectations about the actual behavior of the leader. Even though the black and white American children seem to be similar in their expectations about how certain political encounters would ensue, interpretations of encounters are strikingly different.

No rational egoist should bother voting because, as Skinner's Dr. Frazier notices, the probability of any one man casting the decisive ballot “is less than the chance he will be killed on the way to the polls.” No matter how deeply he cares about the electoral outcome, a man must realize that his vote is only one among very many. The larger the electorate, the smaller the probability of any one vote's changing the outcome, so in most modern polities the politically rational thing to do is to conserve on shoe leather.

Real-world voters, of course, do flock to the polls, which is usually explained in terms of a feeling of “civic duty.” The fact that men get some satisfaction from discharging their civic duty by voting might answer the question of why it is rational for them to go to the polls. Unfortunately, it leaves another—how do they vote once they get there? Presumably, they go ahead and vote according to the dictates of their egoistic interests.

In an article, “The Paradox of Vote Trading,” (APSR 67 [December, 1973]) William H. Riker and Steven J. Brams have argued that systematic logrolling among all members of a legislature produces a paradox: While each trade is individually rational, the effects of externalities offset the potential gains from exchanging votes and each voter finds himself worse off than he would have been by voting sincerely. We extend the results of Riker and Brams to a unanimity decision rule and find that a paradox of vote trading holds for that decision rule as well as for simple majority rule. Under a unanimity rule, however, trades which would be collectively rational (i.e., which would produce a Pareto optimal result) are not individually rational; the non-trader is the beneficiary under such a decision rule. Finally, we pose the question Riker and Brams suggested: Is the paradox of vote trading inescapable? Except under very restrictive conditions, we find that it is. However, given certain assumptions about the distributions of individual utilities, we present proofs of the necessary and sufficient conditions for the Pareto optimality of vote trading and argue that in actual legislative situations, when vote trading is Pareto optimal, learning behavior should serve to extricate the members from the paradox of vote trading.

This paper examines William Riker's thesis that only minimum winning coalitions form in n-person zero-sum symmetric games. It demonstrates that Riker's conclusion is false by identifying the conditions under which larger than minimum winning coalitions can form. Since these conditions are quite general it indicates that Riker's conclusion is valid only for a highly restricted class of games. This class of games is identified as those in which players not in a minimum winning coalition have no incentive to form any coalitions among themselves. These games are characterized as games inessential over coalitions of losers. Only in these games can minimum winning coalitions be expected to form exclusively. In all other games, larger than minimum winning coalitions are possible.

The centenary of Merriam's birth provides the opportunity to reappraise the consequences of his prophetic advocacy of a more scientific expression and systematization of political knowledge. The vehicle for this appraisal is a comparison of Merriam's ’activist” epistemology wjth the more self-limiting methodology of Max Weber who, perhaps among all twentieth-century social scientists, stated most explicitly and experienced most poignantly the tensions among the requirements of acquiring objective knowledge about politics and exercising responsibility in political action. Notwithstanding their many points of difference, Merriam and Weber are interpreted as sharing common grounds of disbelief that the disjunction between science and politics will be removed by the development of a unifying, paradigmatic world-view, either within political science or between the several sciences of man, nature, and society. The political context and role of scientists are visualized by the author as consisting in: (1) mastering the personal temptations and obstacles to achieving their own peculiar brand of political competence, (2) securing public recognition and respect for the factual-scientific component of controversial situations involving their sphere of expertness, and (3) acting upon the assumption of joint skills and contributions, along with other scientists, philosophers, technicians (including politicians), and participating citizens in improving the utilization of scientific research in the formulation of public policy and reform of governing institutions.

The Shapley value of the presidential election “game” is approximated by the method of multilinear extensions; the likely error in this approximation is computed by studying the error in the electoral college game. A comparison with the Banzhaf ratio is also given. The several biases inherent in the electoral system are analyzed from a mathematical point of view. It is shown that the system incorporates a strong bias in favor of the larger states, and a small bias in favor of the smallest states.