Abstract
Volunteer-based voter contact presents multiple potential principal-agent problems for political campaigns. Conflicting potential solutions to these principal-agent problems generate two opposing expectations about campaigns’ preferences for ideological types of volunteers. Concerns about volunteers substituting their own ideological messages for the moderate and noncommittal ones campaigns prefer should make moderate volunteers more desirable; concerns about maximizing volunteer work-hours should lead to preferences for volunteers whose ideology matches the candidate’s. Using interviews with campaign operatives, a conjoint experiment, and a correspondence experiment, we show campaigns prefer volunteers whose views align with the candidate – interpreted by campaign operatives as a signal of likely enthusiasm and dedication – rather than moderate volunteers. However, even without resource constraints, these preferences are weak and fade in the presence of stronger indicators of commitment. They are absent in real-world volunteer recruitment. Overall, campaigns are more concerned with volunteers shirking responsibilities than they are with volunteers going off-message.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data Availability
Replication data for this project is available at https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/VNUJM8.
Notes
Previous work examining the 2012 Obama campaign has even suggested campaigns “use recruitment to…offset the principal-agent problem” as there were very minor ideological differences (neither substantively nor statistically significant) between the Obama volunteers who were “recruited by the campaign” and volunteers who had “asked to volunteer” (Enos and Hersh 2015, 264–66). However, that work concluded the Obama campaign’s recruitment of moderate volunteers was ineffective (264). That work, however, does not provide clear reasons why there is no substantive difference between recruited and non-recruited volunteers: were campaigns unconcerned about ideological misrepresentation and prioritized solving other principal-agent problems which led them to recruit more ideologically extreme volunteers, or did they prefer moderate volunteers, but were constrained in their recruitment efforts?
In addition, if the main use of volunteers is for turnout, moderate volunteers might be less effective at turning out supporters whose views are more ideologically extreme. Although the prioritization of mobilization over persuasion (or vice versa) could be a result of the type of volunteers that campaigns have available to them, our evidence suggest that campaigns are skeptical about the effectiveness of door-to-door persuasion (see interviews below for more details). In line with this view, research shows that even the most effective canvassing has little to no effect on persuasion (Kalla and Broockman 2018) but has large substantive effects on turnout (Green and Gerber 2019). However, we find no evidence that campaigns consider volunteer ideology with the explicit purpose of being more effective at turning out supporters.
It is important to emphasize that we do not dispute the potential for principal-agent problems to arise on ideological dimensions and affect the effectiveness of messaging to persuadable voters. Campaigns, however, do not appear to prioritize these concerns.
The research was approved by IRBs at Florida State University and Brigham Young University-Idaho.
Indeed, voters use information about a candidate’s supporters as a heuristic for the candidate’s positions and likely behavior in elected office (Popkin 1994, Chapter 3).
However, if campaigns aim to use volunteers more for turnout rather than persuasion, volunteers with more extreme ideologies might be more likely to turnout supporters. As we note further on, campaign practitioners view direct voter contact as inefficient for persuasion because of their perception that persuasion requires repetition which is hard to implement through direct voter contact. Previous work has also shown substantive effects of volunteer voter turnout efforts (Green and Gerber 2019) and limited direct voter contact persuasion effects (Kalla and Broockman 2018).
The only examination of a campaign’s perspective we are aware of is Maisel (1982) who discusses tangentially the principal-agent problems of shirking and misaligned goals he dealt with in his own 1978 congressional primary campaign in Maine.
When asked specifically if there were differences in the volunteer recruitment or deployment processes between presidential and congressional campaigns, one GOP operative with experience in presidential campaigns responded, “No. Categorically no. There isn’t a difference in who presidential campaigns and congressional campaigns want to recruit to volunteer or how they think about deploying their volunteers. There’s just more of them.” A Democratic operative said nearly the same thing indicating preferences for volunteers was the same “whether it’s a school board race or the Biden campaign.”.
One campaign manager on a long-shot campaign even indicated he “would have loved to have been misrepresented because it would (provide a) chance to go on the media and clarify.”.
Although message consistency was an aspect of volunteer monitoring, much of this monitoring was done to ensure, as one operative put it, “people are doing the work that they say they’re doing.”.
In contrast, television advertising is repetitive and largely used for persuasion (Sides et al., 2022).
Three campaigns responded twice.
Limited survey roll-off meant 192 individuals assessed 1,150 volunteer profiles in the conjoint experiment.
We estimate models with a forced choice operationalization, a binary variable indicating whether the respondent selected a volunteer with that attribute/demographic profile. Rescaled to a [0,1] interval, we interpret this as the probability of taking the action in question.
As noted previously, the online appendix contains three additional models from a more limited sample of viable primary candidates only as measured by 1) the presence of a campaign Twitter account, 2) a minimum of $50,000 in fundraising, and 3) primary success. Those results show similar results.
Roughly 75% of Enos and Hersh’s (2015) sample of volunteers come from Democratic volunteers for presidential campaigns. We assume, as they do, that the trends they show apply to congressional campaign volunteers. We do not have similar data on Republican volunteers. As such, we do not know what type of volunteer might be under-represented relative to Republicans’ target voter population and thus, theoretically, in higher demand.
Through their comments at the end of the survey, some respondents appeared to guess (incorrectly) that we were primarily interested in race and gender. Indeed, our results from the correspondence experiment show Democrats actually are more responsive to potential volunteers who are men. While conjoint experiments are supposed to allow respondents to reveal biases (Hainmueller, Hopkins, and Yamamoto 2014), the small number of characteristics may have inadvertently led to demand effects. However, other research has suggested this should not have any effect on outcomes (Mummolo and Peterson, 2019).
We also tested the effects of the respondent’s ideology and found similar results which are available in the online appendix. The Pearson’s correlation coefficient between respondent’s and candidate’s ideology when scaled for ideological strength is .70.
Some of the 1,976 campaigns for whom we had emails were excluded because the contact information we had was duplicated for another campaign (often because of a shared consultant acting as campaign manager or because the candidate ran in two separate races using the same email contact information) causing us to accidentally send two emails.
For the results presented here, we estimate ideal points using the tweetscores package, which infers partisan ideal points based on which elite accounts a user follows (Barberá 2015). Analysis using an alternate method based on users’ full follow profiles are included in the online appendix and produces substantively similar results.
We use official campaign Twitter accounts. When the candidate did not have an official campaign account, we used their personal Twitter account if available.
In addition to candidate extremity, we also considered potentially moderating effects of candidate party affiliation, the office the candidate is seeking, incumbency status, and prospective volunteer gender. None were important for predicting conditional average treatment effects.
References
Bansak, K., Hainmueller, J., Hopkins, D. J., & Yamamoto, T. 2020. “Using conjoint experiments to analyze elections: the essential role of the average marginal component effect (AMCE).” Stanford University Working Paper.
Barberá, P. (2015). Birds of the same feather tweet together: bayesian ideal point estimation using twitter data. Political Analysis, 23(1), 76–91.
Bawn, K., Cohen, M., Karol, D., Masket, S. E., Noel, H., & Zaller, J. (2012). A theory of political parties: Groups, policy demands and nominations in American politics. Perspectives on Politics, 10(3), 571–597.
Garcia Bedolla, L., & Michelson, M. R. (2012). Mobilizing Inclusion: Transforming the Electorate through Get-Out-the-Vote Campaigns. Yale University Press.
Berinsky, A. J. (1999). The two faces of public opinion. American Journal of Political Science, 43(4), 1209–1230.
Bleich, E., & Pekkanen, R. (2013). How to report interview data. In L. Mosley (Ed.), Interview Research in Political Science (pp. 84–105). Cornell University Press.
Butler, D. M., & Crabtree, C. (2021). Audit studies in political science. In D. P. Green & J. Druckman (Eds.), Advances in Experimental Political Science (pp. 42–55). Cambridge University Press.
Butler, D. M., Karpowitz, C. F., & Pope, J. C. (2012). A field experiment on legislators’ home styles: service versus policy. Journal of Politics, 74(2), 474–486.
Collier, D., Brady, H. E. & Seawright, J. (2010). A sea change in political methodology. In: H. E. Brady and D. Collier (Eds.), Rethinking Social Inquiry (pp. 1–11). Rowman & Littlefield.
Costa, M. (2017). How responsive are political elites? A meta-analysis of experiments on public officials. Journal of Experimental Political Science, 4(3), 241–254.
Druckman, J. N., Kifer, M. J., & Parkin, M. (2009). Campaign communications in U.S. congressional elections. American Political Science Review, 103(3), 343–366.
Enos, R. D., & Hersh, E. D. (2015). Party activists as campaign advertisers: the ground campaign as a principal-agent problem. American Political Science Review, 109(2), 252–278.
Green, D. P., & Gerber, A. S. (2019). Get Out the Vote: How to Increase Voter Turnout (4th ed.). Brookings Institution Press.
Grossmann, M. (2012). What (or Who) makes campaigns negative? American Review of Politics, 33, 1–22.
Hainmueller, J., Hopkins, D. J., & Yamamoto, T. (2014). Causal inference in conjoint analysis: understanding multidimensional choices via stated preference experiments. Political Analysis, 22, 1–30.
Hassell, H. J. G. (2020). It’s who’s on the inside that counts: campaign practitioner personality and campaign electoral integrity. Political Behavior, 42(4), 1119.
Hassell, H. J. G., Holbein, J. B., & Miles, M. R. (2020). There is no liberal media bias in the news political journalists choose to cover. Science Advances. https://doi.org/10.1126/sciadv.aay9344
Holbein, J. B., & Carnes, N. (2019). Do public officials exhibit social class biases when they handle casework? Evidence from multiple correspondence experiments. PLoS ONE, 14, e0214244.
Issenberg, S. (2012). The Victory Lab : The Secret Science of Winning Campaigns. Crown.
Kalla, J. L., & Broockman, D. E. (2018). The minimal persuasive effects of campaign contact in general elections: evidence from 49 field experiments. American Political Science Review, 112(1), 148–166.
Klein, E. 2021. “David Shor is telling democrats what they don’t want to hear.” New York Times. October 8. https://www.nytimes.com/2021/10/08/opinion/democrats-david-shor-education-polarization.html.
LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13(2), 230–237.
Leeper, T. J., Hobolt, S. B., & Tilley, J. (2020). Measuring subgroup preferences in conjoint experiments. Political Analysis, 28(2), 207–221.
Lynch, J. F. (2013). Aligning sampleing strategies with analytical goals. In L. Mosley (Ed.), Interview Research in Political Science (pp. 31–44). Cornell University Press.
Maisel, L. S. (1982). From Obscurity to Oblivion: Running in the Congressional Primary. University of Tennessee Press.
Martin, G. J., & Peskowitz, Z. (2018). Agency problems in political campaigns: media buying and consulting. American Political Science Review, 112(2), 231–248.
McGuire, B. (2019). Scaling the Field Program in Modern Political Campaigns. Harvard University Kennedy School of Government.
Miao, H. 2021. “Democrats’ historic Georgia senate wins were years in the making thanks to local grassroots.” CNBC, January 9.
Mosley, L. (Ed.) (2013). Just talk to people? Interviews in contemporary political science. In: Interview Research in Political Science (pp. 1–28). Cornell University Press.
Mummolo, J., & Peterson, E. (2019). Demand effects in survey experiments: an empirical assessment. American Political Science Review, 113(2), 517–529.
Panagopoulos, C. (2016). All about that base: Changing campaign strategies in U.S. presidential elections. Party Politics, 22(2), 179–190.
Popkin, S. L. (1994). The Reasoning Voter. University of Chicago Press.
Sides, J. (2006). The origins of campaign agendas. British Journal of Political Science, 36(3), 407–436.
Sides, J., Vavreck, L., & Warshaw, C. (2022). The effects of television advertising in United States elections. American Political Science Review, 116(2), 702–718.
Ward, I. 2021. “The Democrats’ privileged college-kid problem.” Politico. Oct. 9. https://www.politico.com/news/magazine/2021/10/09/david-shor-democrats-privileged-college-kid-problem-514992.
Weller, N., & Barnes, J. (2014). Finding Pathways: Mixed-Method Research for Studying Causal Mechanisms. Cambridge University Press.
Wilson, J. Q. (1974). Political Organizations. Basic Books.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We are grateful to the many campaign practitioners who provided their insights into how campaigns use volunteers and the types of volunteers they want to recruit. A previous version of this paper was presented at the 2021 APSA Conference. We are also grateful to Ryan Enos, Eitan Hersh, and Kevin Reuning for their comments and suggestions and to Quintin Beazer, Daniel Butler, and Adam Dynes for the conversations at SPSA in Puerto Rico in 2020 helped shape this project. All errors, of course, remain our own. Replication data for this project is available at https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/VNUJM8.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chewning, T.K., Green, J., Hassell, H.J.G. et al. Campaign Principal-Agent Problems: Volunteers as Faithful and Representative Agents. Polit Behav 46, 405–426 (2024). https://doi.org/10.1007/s11109-022-09836-9
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11109-022-09836-9