J. Tobin Grant

An Attributional Model of Economic Voting: Evidence from the 2000 Presidential Election

In this article we propose and test an attributional model of economic voting. Exploiting an innovative responsibility instrument to analyze the 2000 U.S. presidential election, we find that the incumbent partys can- didate did benefit from the belief that economic conditions had improved, but only among voters who attributed responsibility for that improvement to the president. Moreover, we find that by explicitly mod- eling the effects of responsibility attributions instead of presuming a homogeneous and automatic sanctioning process, the attributional model outperforms the classical reward-punishment model in explaining presidential vote choice. Our results provide new insights into the out- come of the 2000 presidential election and demonstrate the relative importance of issues and the economy.

TO GIVE OR NOT TO GIVE: Modeling Individuals' Contribution Decisions

Employing data from a recent national survey on campaign finance, we examine the contribution behavior of individual citizens in the 2000 election. By disaggregating types of contributions, our model enables us to observe potential heterogeneity in the determinants of giving money to parties and candidates. We find that for both types of contributions, the effects of informational resources and solicitation on the decision to contribute outweigh those of financial resources. In addition, we propose both a theoretical and an empirical distinction between the selection effects of solicitation and the stimulus effects of solicitation. By distinguishing between these dual dimensions of solicitation, our analysis provides new insight into the causal linkages between income, solicitation, and contributions. We find that while solicitation increases the likelihood of contributing through selective targeting or rational prospecting, it also does so through a stimulus mechanism.

Campaign Effects on the Accessibility of Party Identification

This study uses response latency, the time required for a survey respondent to formulate an answer upon hearing a question, to examine the accessibility of partisan self identifications over the course of a political campaign season. Although the aggregate distribution of partisanship remains fairly stable during the campaign, party identifications become more accessible to individuals with weaker party identifications as the election approaches. Consistent with theoretical expectations, the authors find that partisan orientations are more useful in forming political judgments when those orientations are more accessible to the voter. The effect of partisanship on vote choice is a third greater for voters with highly accessible party identifications than for those with less accessible party identifications.

Demand for God and Government: The Dynamics of Religion and Public Opinion

We explore the relationship between religiosity and public support for greater government services. We theorize that increases in religiosity and public opinion both reflect demands from citizens in the face of insecurity. We argue that religiosity is comprised of two factors: responses to insecurity; and long-held preferences for religion, or secularity. We show that previous studies that have observed increased religiosity leading to decreased support for government spending have not distinguished among religiosity as driven by secularity versus insecurity. To test our theory, we first estimate a series of simulations, and we then turn to the dynamics of aggregate religiosity and public opinion in the United States over the past fifty years, an environment where long-held preferences for religious goods have remained relatively stable. Consistent with our theory, religiosity and public opinion respond to insecurity; the series are positively correlated, move together through time, and react in similar ways to changes in GDP per capita. Our findings indicate that during times when there is greater insecurity, both religiosity and demand from government increase.

The Referendum Conundrum: Referenda or Referendums ?

Scholars often use referenda as the plural for referendum . This choice is a hypercorrection—it may sound like proper Latin, but it is not. Referendums is always the correct choice. However, we maintain that there is value in using referendums for multiple events and referenda for multiple propositions.

Tax Policy Attitudes and the 1996 Presidential Election

This study examines whether voters employ a collectively based evaluation of tax policies when voting. We model public opinion toward Bob Dole’s 15% income tax cut proposal and assess the issue’s impact on presidential vote choice in 1996. A preelection survey experiment offered two versions of a question about tax cuts: One presents the issue in terms of what is right or wrong for the country (collective version), whereas the other poses a “favor or oppose” question (individual version). We find that support for the tax cut is far greater in the individual version than in the collective version, and asking voters to consider the potential fiscal consequences of the tax cut drives down its popularity. However, logit models of vote choice show that support for the tax cut has a greater effect on vote choice when judgment of the policy is based on its collective consequences.