Deciding to Donate: An Experimental Test of Endorsements and Political Giving

• In Event: Experimental Research on Voter Rationality, Engagement, and Turnout

Abstract

Over the past 20 years, changes in campaign finance rules and electronic communications have made it easier for individual donors to send money to political candidates and causes. Small contributions from individual donors are now the single largest source of money in presidential elections. The ability of candidates to solicit contributions from small donors at very low cost and donors’ ability to bestow money on favored candidates with the click of a button has contributed to the escalating costs of political campaigns. Scholars studying the influence of money in politics have examined the characteristics and motivations of donors using observational contributions data or surveys of individual donors. In both cases, the data are collected after decisions to contribute have been made, with little consideration of other choices that potential donors might make and no information about individuals who choose not to contribute.

We overcome these limitations in existing studies of political donations by designing an experiment where respondents allocate $5 between themselves and various education-related organizations, candidates, and ballot measures. Specifically, respondents make a series of such allocation decisions, including three involving education-related organizations (Create California, California Institute of the Arts, Inner City Arts), one involving an education-related ballot measure Committee (Yes on Prop 28) and one involving a political candidate for Superintendent of Public Instruction (Tony Thurmond). Respondents also make one allocation between themselves and a disaster-relief organization (California Fire Foundation), which serves as a baseline for their willingness to donate to charity.

To understand factors that might affect the willingness to make political donations, we randomly assign respondents to receive either 1) candidate and ballot measure endorsements from prominent spokespersons (e.g., celebrities), 2) candidate and ballot measure endorsements from issue experts (e.g., school superintendents and teacher unions), or 3) no additional information about the political candidate or education-related ballot measure. We compare respondents’ decisions of whether to donate, the amount they donate, and their choice of which donation decision to implement in our treatment groups to the control group.

We test a number of hypotheses drawn from previous studies of charitable giving and political donations. We find that many of the same characteristics associated with charitable giving (e.g., age, income, education) are also associated with donations to education-related political candidates and ballot measure committees. Consistent with previous studies of political donors, we show that donations to political candidates and ballot measure committees is greater among strong partisans, ideological extremists, citizens who rate high in political interest, political knowledge, internal efficacy, those who care about education as a political issue, and who have immediate family working in education.

While previous research leads us to expect that citizens will give little to political candidates and causes (relative to their baseline willingness to donate to charity), we find that respondents give substantial sums to politics. Our analyses indicate that respondents who receive endorsements from issue experts give significantly more money to both a political candidate and ballot measure committee. Endorsements from prominent spokespersons also modestly increase donations to the education-related ballot measure. The effects of endorsements from issue experts are greater among strong partisans and ideological extremists, as well as those who care about education as a political issue. Taken together, the results of our donation experiment shed new light on the ideological and other interests motivating political donations. Our experimental design also extends research on the effects of political endorsements. In addition to influencing citizens’ choices in candidate and direct democracy elections, we offer evidence that they stimulate political giving. Indeed, donations to political candidates and causes in our treatment groups are comparable to what citizens are willing to give to more sympathetic charities and relief organizations.

Authors

• Cheryl Boudreau, University of California, Davis

• Scott MacKenzie, University of California, Davis