Robert S. Erikson

Measuring State Partisanship and Ideology with Survey Data

The study of state politics has suffered because of the lack of good data on major political orientations at the state level. This paper briefly discusses this problem and the need for survey-based measures, and then presents sets of estimates of state partisanship and ideology. These are derived from aggregating CBS News-New York Times polls at the state level. Using fifty-one polls taken from 1974 through 1982, our estimates are based on over 76,000 respondents. The estimates are shown to have good overall validity and reliability, and should prove valuable in studies of comparative state elections and policymaking.

Economic Conditions and the Presidential Vote.

This analysis demonstrates that the relative growth of per capita income change is an important determinant of post-World War II presidential election outcomes. Per capita income change is even a better predictor of presidential election outcomes than the electorates relative attraction to the Democratic and Republican candidates as calibrated in National Election Study surveys. The significance of this finding is discussed.

Constituency Opinion and Congressional Behavior: A Reexamination of the Miller-Stokes Representation Data

Reexamining Miller and Stokess 1958 representation data, this article demonstrates that Miller and Stokes may have underestimated the extent of congressional representation in their classic study. It is argued that the correlations between sampled constituency opinion and congressional behavior are more seriously attenuated by sampling error than has been recognized. Unlike sampled constituency opinion, measures of constituency opinion based on simulation are shown to correlate relatively highly with congressional behavior. Evidence is presented which indicates that elections are a major source of this representation.

Public Opinion and Policy Liberalism in the American States

This paper examines the effect of public opinion on public policy in the American states. We use a new measure of state public opinion, liberal-conservative ideological identification of state electorates, derived from aggregating CBS News/New York Times national opinion surveys. Regression and LISREL are used in the analysis to demonstrate that state opinion is a major determinant of state policy. Citizen preferences are markedly more important than state social and economic characteristics in accounting for patterns of policy liberalism in the states. These results constitute a major challenge to economic development as an explanation of state policy. Unless mass views have some place in the shaping of policy, all the talk about democracy is nonsense. -V. 0. Key (1961, p. 7) Popular control of public policy is a central tenet of democratic theory. Indeed, we often gauge the quality of democratic government by the responsiveness of public policymakers to the preferences of the mass public as well as by formal opportunities for, and the practice of, mass participation in political life. The potential mechanisms of democratic popular control can be stated briefly. In elections, citizens have the opportunity to choose from leaders who offer differing futures for government action. Once elected, political leaders have incentives to be responsive to public preferences. Elected politicians who offer policies that prove unpopular or unpleasant in their consequences can be replaced at the next election by other politicians who offer something different. Of course, this picture describes only the democratic ideal. A cynic would describe the electoral process quite differently: Election campaigns

Equilibria in Campaign Spending Games: Theory and Data

We present a formal game-theoretic model to explain the simultaneity problem that makes it difficult to obtain unbiased estimates of the effects of both incumbent and challenger spending in U.S. House elections. The model predicts a particular form of correlation between the expected closeness of the race and the level of spending by both candidates, which implies that the simultaneity problem should not be present in close races and should be progressively more severe in the range of safe races that are empirically observed. This is confirmed by comparing simple OLS regression of races that are expected to be close with races that are not, using House incumbent races spanning two decades.

Economic Conditions and the Congressional Vote: A Review of the Macrolevel Evidence

It is widely believed that variation in the national congressional vote is largely determined by the health of the national economy. However, scholars have yet to agree on a satisfactory microlevel explanation of the macrolevel relationship. The present study suggests the solution that there may not exist any macrolevel relationship to explain. With the proper specifications, per capita income growth is not significantly related to the congressional vote. This result varies sharply with the evidence at the presidential level.

Campaign Spending and Incumbency: An Alternative Simultaneous Equations Approach

This paper estimates the effects of incumbent spending and challenger spending in U.S. House elections in the 1970s and 1980s. The paper employs FIML simultaneous equations analysis involving instrumental variables as vote predictors, and zero-covariance restrictions for the vote-spending disturbances. This procedure allows the estimation of spending effects given plausible assumptions about the effects of unobserved causes of the vote on candidate spending. The results are that incumbent spending matters even with only modest amounts of simultaneity. Evidence is presented to suggest that the effectiveness of new incumbent spending declines with seniority but accumulates to the incumbents long-term advantage.

State Political Culture and Public Opinion

Do the states of the United States matter (or are they of no political consequence)? Using a data set with over 50 thousand respondents, we demonstrate the influence of state political culture on partisanship and ideology. For individuals, we find that the state of residence is an important predictor of partisan and ideological identification, independent of their demographic characteristics. At the aggregate level, state culture dominates state demography as a source of state-to-state differences in opinion. In general, geographic location may be a more important source of opinion than previously thought. One indication of the importance of state culture is that state effects on partisanship and ideology account for about half of the variance in state voting in recent presidential elections.

Why Do People Vote? Because They Are Registered

This article separately investigates who registers to vote and who votes among the registered. Standard predictors of participation are much better predictors of registration than of voting among the registered. It is speculated that voting is stimulated by the act of registration.

Policy representation of constituency interests

This paper examines two aspects of congressional representation: the responsiveness of House candidates to constituency opinion and the effect of candidate positions on House election outcomes. For both Democratic- and Republican-held seats, constituency liberalism (as measured by the constituencys vote for McGovern in 1972) is strongly related to the liberalism of incumbent candidates but less so to the liberalism of challengers. House election outcomes are visibly influenced by the positions of incumbent candidates but not those of nonincumbent candidates. The paper argues that elections contribute significantly to achieving congressional representation. Candidate positions are measured from a 1974 CBS survey of all major House candidates.