The development of social choice theory over the past three decades has brought many new insights into democratic theory. Surprisingly, the theory of representation has gone almost untouched by social choice theorists. This article redresses this neglect and provides an axiomatic study of one means of implementing proportional representation.

The distinguishing feature of proportional representation is its concern for the representativeness of deliberations as well as decisions. We define a representative in a way that is particularly attentive to this feature and then define a method of selecting representatives (a variant of the Borda rule) which selects a maximally representative body. We also prove that this method of selection meets four social choice axioms that are met by a number of other important social choice functions (including pairwise majority decision and the Borda rule).

This article presents a method for analyzing the extent and strength of coattail voting in presidential elections. This method allows the authors to estimate the magnitude of coattail voting and then to decompose this estimate into more “basic” elements. Estimates are given for presidential elections beginning with 1956.

The determination of the coattail vote and its decomposition depend on the theory of the voting decision that is assumed. In this article we present a model of vote determination that is similar in most respects to the traditional SRC model; the vote for congressional representation in a presidential election year is determined jointly by partisan affiliation, attitudes toward the presidential candidates, and local forces unique to the congressional race (such as may be captured by an incumbency variable). This model permits the separate estimation of the strength of short-term forces and of the efficiency of the presidential coattails.

Application of the model to survey data since 1956 indicates that efficiency of presidential coattails has declined during this period. Furthermore, the 1980 election does not appear to be an exception to this trend. On the other hand there has not been any particular trend in the strength of short-term forces during this period; instead events peculiar to the context of a specific election generate short-term forces at the level of the presidential election, but the degree to which these forces are carried over to local races seems to have declined.

Political candidates embody possibilities as well as preferences. To some extent, their possibilities vary systematically and inversely with their preferences. Political systems often have built-in “stops” preventing left-wing candidates from doing things too far to the left of their declared positions and vice versa. The preferences of right-wing candidates make them unlikely to want to pursue policies on the far left, but at least the opportunity to do so is available to them in a way that may be denied to left-wingers themselves. Taking differential possibilities into account, it might prove instrumentaily rational to vote perversely for the right-wing candidate if you really want left-wing results and vice versa. This article sketches conditions under which that proposition will hold true, shows those conditions are empirically plausible, and suggests that politicians themselves are alive to these possibilities. This awareness should make us much more cautious in reading any policy-specific mandates into electoral outcomes. It may even make us doubt the model of instrumental rationality itself.

The responsiveness of government policies to citizens' preferences is a central concern of various normative and empirical theories of democracy. Examining public opinion and policy data for the United States from 1935 to 1979, we find considerable congruence between changes in preferences and in policies, especially for large, stable opinion changes on salient issues. We present evidence that pubic opinion is often a proximate cause of policy, affecting policy more than policy influences opinion. One should be cautious, however, about concluding that democratic responsiveness pervades American politics.

Pluralist political theory identifies certain patterns of political preferences as promoting the “stability” of democratic political systems and others as threatening to such stability. Social choice theory likewise identifies certain patterns of political preferences as leading to “stability” in social choice under majority rule and related collective decision rules, and other patterns as leading to “unstable” social choice. But the preference patterns identified by pluralist theory as promoting stability are essentially those identified by social choice theory as entailing instability. Thus the notions of stability and the implicit normative criteria associated with the two theories are very close to being logically incompatible. This incompatibility suggests that the social choice ideal of collective rationality may not be one that we should endorse. Indeed, the generic instability of the pluralist political process and its consequent collective irrationality may contribute to the stability of pluralist political systems.

When issues (i.e., dimensions) are voted on one at a time, a voter whose preferences are not separable across issues must forecast the outcome of later issues in order to know how to vote in the present. This is the problem of expectations. In this article, we develop a general theory designed to handle this problem. Assuming that voters are risk averse and maximize expected utility, we demonstrate that a random variable forecast of how later issues will be decided reduces to a point forecast, which is the mean of the multidimensional random variable. We also show that single-peaked preferences are induced on each issue, and consequently there exists an equilibrium across issues.