Research Agenda

My research agenda centers around how American political elites, such as candidates for office or interest groups, publicly communicate, with a focus on their use of new technologies like social media. My work has been accepted at PS: Political Science & Politics, the Journal of Law & Courts, and the Journal of Women, Politics, & Policy.

I study political communication by political elites and its effects. In my work, I find that with social media, a broader range of actors (i.e. those who are under-resourced or non-traditional) are able to participate in the political process than were able to otherwise. Specifically, I have shown that congressional campaigns are able to use Twitter to raise money from organized interests and people nationwide, by directly reaching these audiences and by amplifying their message through the media. I have also found that campaigns strategically use more negative sentiment when they are disadvantaged in the House, and that the Facebook posts of those candidates who use more negative sentiment are shared more than the posts that do not contain negative sentiment. In work studying congressional campaigns and interest groups, I argue that social media offers an inexpensive communication tool that political elites are able to use to achieve their political goals and receive tangible benefits.

I incorporate aspects of both political behavior and political institutions as I explore these questions using big data and modern machine learning methods. Though my main area of substantive focus is congressional campaigns, these questions apply to a wide range of political actors such as interest groups, police departments, and other levels of American government. At best, these new technologies may lower the barrier for entry into the political process; at worst, they incentivize more polarizing rhetoric and negative sentiment. My goal in studying political communication is twofold. First, I use social media data to test long-standing questions in political science research. Second, I seek to understand how the impact of who a political audience is (voters, politicians, the media, etc.), combined with the means of communicating, affects how elite political actors behave and what issues they emphasize.

Ongoing Projects

To Moderate, Or Not to Moderate: Strategic Domain Sharing by Congressional Campaigns (with Megan A. Brown, Joshua A. Tucker, and Jonathan Nagler). Under review.

We test whether candidates move to the extremes before a primary but then return to the center for the general election to appeal to the different preferences of each electorate. Incumbents are now more vulnerable to primary challenges than ever as social media offers a viable pathway for fundraising and messaging to challengers, while homogeneity of districts has reduced general election competitiveness. To assess candidates' ideological trajectories, we estimate the revealed ideology of 2020 congressional candidates (incumbents, their primary challengers, and open seat candidates) before and after their primaries, using a homophily-based measure of domains shared on Twitter. This method provides temporally granular data to observe changes in communication within a single election campaign cycle. We find that incumbents did move towards extremes for their primaries and back towards the center for the general election, but only when threatened by a well-funded primary challenge, though non-incumbents did not.

Reaching Across the Political Aisle: Overcoming Challenges in Using Social Media for Recruiting Politically Diverse Respondents (with Megan A. Brown, Nejla Asimovic, Rajeshwari Majumdar, Lena Song, Laura Huber, Sarah Graham, Abby Budiman, Joshua A. Tucker, & Jonathan Nagler). Under review.

A challenge for public opinion surveys across modes today is achieving representativeness of respondents across relevant political characteristics such as attitudes and demographics. We focus here on overcoming known challenges in online recruiting of Republican respondents for surveys on American politics by testing recruitment strategies using Facebook ads. While the use of Facebook ads for recruitment has increased and offers potential benefits, such a recruitment strategy suffers from a lack of Republican engagement relative to Democrats and Independents. In this project we focus on one feature which may influence a potential participant to click on an ad, consent to begin, and complete a survey: the institution listed as fielding the survey. We argue that a potential participant’s trust, or lack of trust, in an institution will affect whether they choose to participate. We focus on the relevant case of academic institutions. Respondents are recruited through two otherwise identical Facebook pages and ads listing: New York University (NYU, a university from a more liberally-perceived state) and the University of Mississippi (Ole Miss, a university from a less liberally-perceived state). We predict that Republicans will be most likely to both begin and complete surveys from Ole Miss, followed by CSMaP Research Center, and least likely to do so for surveys from NYU. Additionally, we will assess whether the Republicans who agree to participate under each treatment differ from each other in terms of easily observed characteristics such as education and region, and also, more importantly, in their political attitudes. Our findings will be a useful step forwards in overcoming existing challenges in recruiting Republicans for research.

The Democratizing and Polarizing Impact of Fundraising on Twitter: Easy Money, Viral Incentives, and the Catalyzing Role of Mainstream Media (with Joshua A. Tucker and Jonathan Nagler)

What is the effect of social media on campaign fundraising? We propose two pathways through which campaigns can raise money through their tweets: directly via Twitter users seeing their tweets and indirectly through retweets by media and journalists, who spread a campaign's message beyond the users who would have otherwise seen the tweet. Previous work suggests that a main audience for campaigns on Twitter are journalists and the media. However, we do not know the extent to which this is true, or for which candidates. We find that on the days congressional campaigns received more retweets, they also received more contributions. Further, we find that hundreds of campaigns were retweeted thousands of times by national news media and receive higher predicted contributions than when they were not. This difference is largest when the campaigns tweeted about nationally salient and polarizing topics such as Donald Trump or abortion-- overall, and separately for candidates in safe and competitive seats and for incumbents and non-incumbents. Campaigns in competitive seats and non-incumbents further receive higher predicted contributions when they do not tweet about nationally salient topics and are retweeted by local news media. Our results suggest that Twitter can offer campaigns a pathway to build up resources and support nationwide, mediated by the traditional news media, but incentivizes politicians to tweet about nationally polarizing and salient topics to gain media coverage and appeal to national, partisan individuals.

Tweeting to the Top (book project)

Who do politicians speak to when they post on social media sites such as Twitter? The extant literature theorizes that these platforms are a democratizing space, where candidates for Congress provide information directly to the public and the public respond. I evaluate this claim using interviews with congressional candidates and staff, descriptive trends in Twitter follows and connections, as well as the content of tweets. I find no evidence that Twitter enhances the democratic connection between elected officials and their constituents. Rather, Twitter is used by congressional campaigns as a tool to communicate to fellow elites to gain offline benefits.

Bureaucratic Responsiveness in Times of Crisis: The 2020 Mass Protests and Police Department Social Media (with Anna Gunderson). Under review.

In 2020, cities across the country experienced mass protest in the aftermath of fatal police violence against Black Americans. An open question, however, is whether the departments listened to and engaged with the conversations around police reform. To explore how responsive police agencies are to local protests, we connect two datasets: the Crowd Counting Consortium data on 2020 protests and data on the Facebook activity of all police departments in the country in that same year. We investigate differences in police department responses to protest and hypothesize that these posts will be more likely in cities that are more racially diverse and more Democratic. We also argue that cities that experience protest should engage more with social justice terms and the demands of those movements. We find mixed evidence for these expectations, however. More Republican cities are less likely to post about all kinds of protests, but racial demographics do not appear to make a difference. Moreover, places with protests are not more likely to post about social justice. This paper provides initial evidence that social media is a necessary, but not sufficient, tool for the police to engage in meaningful, transformative conversations with their communities about race, justice, and equity in policing.

Testing the Interest Group Connection: Campaigns on Twitter

Politicians seek to reach several audiences through their tweets. One such audience is interest groups, whose money and support are desirable to congressional campaigns as they seek electoral goals. In this paper, I assess the extent to which interest groups observe and interact with congressional campaigns on Twitter and the relationship between their tweets and contributions from relevant industries. I show that incumbents and non-incumbents are frequently followed and mentioned by nationally active interest groups on Twitter; a cheap signal that interest groups observe politicians online. I then examine if, when a candidate chooses to tweet about certain topics, this public signal of their preferences results in a change in campaign contributions that she receives from the groups for whom that topic is important. Using campaign tweets and contributions from relevant groups to 2018 House campaigns and synthetic controls, I find that for these policies, there is a substantively small but positive causal relationship between a candidate's tweets and the contributions their campaign receives. Rather than a fundraising boost after signalling to interest groups on Twitter, I find that the platform primarily offers campaigns a tool to maintain and build relationships.

Interest Groups and Social Media Usage: How Do Organizations' Posts Differ Across Facebook and Twitter? (with Kirsten Widner & Anna Gunderson). R&R at JQD.

Despite the wide use of social media in America, political science is only beginning to understand its role in politics and policymaking. We investigate how, if at all, interest groups' social media activity differs across Facebook and Twitter using a novel dataset of nearly 13 million posts on the two platforms from 2016-2020. We find suggestive evidence that interest groups use the two platforms differently. They post more frequently on Twitter, but receive more page likes and post likes on Facebook. This research note argues that future analyses ought to consider the differences in usage across the social media platforms, and encourages scholars to think critically about how Facebook and Twitter may provide distinct affordances and opportunities for interest groups -- and political actors more generally -- to reach social media users.