![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Research on political participation has long emphasized differences between Americans who donate money to political campaigns and Americans who perform other kinds of political activity. It’s unclear, however, if there are distinctions in how campaigns use advertisements to solicit donations and mobilize grassroots voters. Analyzing Facebook ad data from thousands of ads, donations, and e-mail sign-ups from a recent statewide gubernatorial campaign, we show that campaigns deploy ad themes that vary in their success at attracting donations and mobilizing grassroots activists. Ad effectiveness varies by region of the state as well. Ideological extremity boosts small donations and campaign sign-ups. However, less ideological campaign messages are also effective, perhaps because recipient lists are already curated to include the candidate’s likely supporters.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/19331681.2024.2350493.
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Notes
1. 96.2% (206,201 of 214,386) of e-mail sign-ups with an identifiable zip code were from Texas, as were 96.7% (3,701 of 3,829) of donors. The campaign’s Facebook ads were microtargeted ads aimed at various Texas resi- dents, but nonetheless reached some people motivated to give nonresident contributions, which have become a more common part of American political donations (Gimpel, Lee, & Pearson-Merkowitz, Citation2008).
2. In a way somewhat similar to Cho and Gimpel (Citation2010), we look at someone who signs up for campaign e-mails about mobilization and volunteering as a “volunteer,” even though we don’t know for sure if that person did so or not.
3. https://www.pewresearch.org/politics/2019/12/17/in-a-politically-polarized-era-sharp-divides-in-bot
5. https://www.pewresearch.org/social-trends/2019/01/17/generation-z-looks-a-lot-like-millennials-on-k
6. Microtargeting means different things to different people. We define microtargeting as the use of political and consumer data to specifically identify people who are likely to respond in a desired way by a campaign. The gubernatorial campaign in question had a variety of microtargeted ads running at the same time, aimed at different groups of people, so it is difficult to describe its microtargeting strategy in detail. The campaign revealed, however, that its microtargeting strategy used a combination of factors to determine which individ- uals were likely to respond to specific kinds of ads. The campaign specifically targeted both those who were predicted to be likely to vote in general, and those were predicted to be likely to vote for the candidate. Roughly 8 million Texans voted in the 2018 gubernatorial election, and over 225,000 Texans were microtargeted with Facebook ads by the campaign.
7. These ads were microtargeted, and not randomly assigned. The campaign also does not have data on individuals who received a Facebook ad but didn’t respond by donating or signing up for volunteering and mo- bilization e-mails. Therefore, we avoid making strong cause-and-effect claims about the relationship between ads and participation. While we know who responded, and to which ads, we can’t be certain that those people would not have responded similarly to a different ad theme. As a result, we are not doing any direct causal inference, and we unfortunately can’t rule out the possibility that the relationships we find are all statistical accidents. We would love to do a small experiment that directly tests the causal effects of these ad themes on donations and volunteering, and plan to do so in the future. However, any experiment that we do now will only be able to provide insight into contemporary relationships between these ad themes and political activism. The data that we use in this study is from 2018, and a new experiment will not be able to provide direct insight about this past political environment.
8. For details on these 20 ad themes & examples, see the supplemental Online Appendix.
9. About 91% of all ads in the file were successfully tagged with a theme using this method (8,142/8,968).
10. This includes prominent state and national politicians of the other party, as well opposed ideology groups.
11. Our regional designations are designed to roughly match those in Cho and Gimpel (Citation2010), with updates to account for a considerably different political landscape and regional political leanings since their analysis of a 2006 gubernatorial campaign. Our regional coding also takes into account the increased importance of the urban-rural divide (Gimpel, Lovin, Moy, & Reeves, Citation2020) in Texas politics. Region is a factor in political behavior that is especially important in Texas (Cho & Gimpel, Citation2010; Owens, Wink, & Bryant, Citation2022). Cho and Gimpel (Citation2010) coded Texas regional metropolitan areas as the county or counties that hold the principal city or cities along with adjacent counties with sufficient overlap that include suburbs and commuter areas for that urban core. To pinpoint regional and metropolitan-rural distinctions, we leave many counties in Central and South Texas uncategorized. This allows us to isolate the political and cultural similarity of metropolitan areas like the DFW Metroplex and Austin Area, along with smaller regions of recent political change like the Rio Grande Valley.
12. This is the sum total of 67,638 e-mail sign-ups matched to a Facebook ad and from Texas zip codes.
13. For us, this includes the counties of Wise, Denton, Collin, Parker, Tarrant, Dallas, Rockwall, Johnson, Ellis, and Kaufman.
14. This includes Lamar, Delta, Hopkins, Rains, Van Zandt, Henderson, Anderson, Houston (which does not contain the city of Houston), Trinity, Walker, San Jacinto, Liberty, and Chambers Counties, and all counties east of them.
15. Montgomery, Waller, Harris (which contains the city of Houston), Fort Bend, Galveston, and Brazoria Counties.
16. Williamson, Bastrop, Travis, and Hays Counties.
17. Webb, Jim Wells, Duvall, and Kleberg Counties, and all counties south of them.
18. El Paso County.
19. This combines the counties that traditionally make up West Texas along with those in the Texas Panhandle region around Amarillo (Potter County).
20. These numbers would be even more illustrative if we had a measure of the ad’s impressions, and could weight each ad’s average number of donations/e-mail sign-ups accordingly. Unfortunately, we do not have this data.
21. All of the Education ads that inspire donations in the matched data set also have the Guns theme, so it is impossible to analyze the Education theme’s independent effect at soliciting donations. These “Education” ads are all about concealed carry of firearms on public university campuses.
22. This analysis would be improved if we had data for those who saw the gubernatorial campaign’s Facebook ads and DIDN’T contribute, which would allow us to more accurately model the true recurring-donation- generating process, but unfortunately we do not have this data.
23. In the case of Guns, the gubernatorial campaign did not purchase any Facebook ads with Guns themes that attempted to solicit e-mail sign-ups (presumably intentionally).
Additional information
Notes on contributors
Grant Ferguson
Grant Ferguson is an Instructor II and Director of Outreach and Public Service Internships in the Department of Political Science at Texas Christian University. His research focuses on public opinion, voting, elections, the historical Congress, political psychology, Universal Basic Income (UBI) policies, the politics of cryptocurrency, and the politics of emerging technology and ideas. He has published research in American Politics Research, Social Science Computer Review, Electoral Studies, Research & Politics, Congress & the Presidency, and Social Policy & Administration.
James G. Gimpel
James G. Gimpel is a Professor in the Department of Government at the University of Maryland, College Park. His areas of interest include political geography, voting and opinion, political campaigns, and the politics of immigration policy. He has written recent articles on the mobilization of volunteers and donors in political campaigns, as well as on the redistribution of the electorate through population mobility. His ongoing research involves the study of campaign contributions and new uses of geographic analysis tools (software) to study politics and policy. He has published research in American Political Science Review, American Journal of Political Science, Quarterly Journal of Political Science, Political Geography, Political Communication, American Politics Research, Journal of Political Marketing, and Political Behavior, among others. He served as editor of American Politics Research for 8 years.
Mark Owens
Mark Owens is an Assistant Professor of Political Science at The Citadel, The Military College of South Carolina. His expertise includes survey research, Congress, redistricting, and Southern politics. He has published research in American Politics Research, BMJ Open, Legislative Studies Quarterly, Congress & the Presidency, Social Science Quarterly, Journal of Political Institutions and Political Economy, Journal of Affective Disorders, and PS: Political Science and Politics.
Daron R. Shaw
Daron R. Shaw is a Professor and Frank. C. Erwin, Jr. Chair of State Government in the Department of Government at the University of Texas at Austin. His research interests include American government, campaigns and elections, political parties, public opinion and voting behavior, and applied survey research. He has also published numerous articles in professional journals such as the American Political Science Review, American Journal of Political Science, Journal of Politics, British Journal of Political Science, Political Research Quarterly, Political Behavior, Political Communication, PS: Political Science and Politics, Election Law Journal, Party Politics, Presidential Studies Quarterly, and American Politics Research.