My current work examines how electoral rules and norms (1) influence political discussion, and (2) impact who is successful at winning a seat in Congress. To answer these complex questions, I pair new data on congressional primaries with cutting-edge methodological techniques like text analysis, matching, regression discontinuity, and machine learning.


“No Experience Required: Early Donations and Amateur Candidate Success in Primary Elections.” with Tyler Steelman. Conditionally Accepted at Legislative Studies Quarterly. [Paper]

The electoral dominance of “quality” candidates—political insiders with a history of holding office is well-established. However, research on the recent rise in successful political neophytes is less studied. Despite longstanding trends in the predominance of experienced candidates in primary elections, nearly half of all quality candidates who ran in non-incumbent races lost to a candidate without prior electoral experience in 2018. In this article, we investigate the success of political newcomers by examining a topic often overlooked in the growing literature on primaries: campaign finance. We show that, from 2016-2020, political newcomers saw (1) greater success in future fundraising, and (2) an increased likelihood of primary election victory when they garnered more early contributions from outside their district. This contrasts prior elections where early money from inside a candidate’s own congressional district served as the strongest predictor for future fundraising and electoral success.

“Running as a Woman?: Candidate Presentation in the 2018 Midterms” with Maura McDonald and Sarah Treul. (2020) Political Research Quarterly, 73(4): 967-987 [Paper] [Supplemental Information]

The record high number of women who ran for the U.S. Congress during the midterm elections led many journalists proclaim 2018 as another “Year of the Woman.” Although not every female candidate was successful, this large number of women running for of- fice provides the opportunity to advance our understanding of the ways in which women present themselves to their voters. Using the “Biography” pages of over 1,500 2018 con- gressional campaign websites, we utilize a structural topic model to examine how these candidates present themselves to their constituencies. In doing this, we find great vari- ance in the presentation styles of women running for Congress in 2018. We also find that prior political experience, more so than gender, is the primary driver in influencing how candidates (both men and women) present themselves. Experienced candidates utilize similar styles that highlight their past political work while amateur candidates are more likely to use “values-driven” language.

Book Chapters

“The Nationalization of Congressional Elections” with Jamie L. Carson and Jason M. Roberts in New Directions in Congressional Politics, 2nd Edition: Routledge.

“Irregular Order: Examining the Changing Congressional Amending Process” with Michael S. Lynch and Anthony J. Madonna in Party and Procedure in the United States Congress, 2nd Edition: Rowland and Little.

Working Papers

“Policy Facts or Partisan Friction?: Explaining Political Communication on America’s Opioid Epidemic” [Working Paper]

Party messaging has become a central component of today’s political dialogue. However, not all issues can be so easily employed to facilitate partisan messaging goals. To assess how politicians define and dramatize party distinctions over issues for which party stances are not immediately clear, I turn to America’s opioid epidemic. Assessing the topical content of opioid issue text from an original collection of congressional campaign platforms, I find the local salience of the opioid crisis to be highly predictive of politicians’ messaging behavior. Candidates from districts with a relatively high rate of opioid deaths tend to focus their opioid issue positions on CDC-endorsed public-health solutions for the crisis; alternatively, candidates from districts with a relatively low rate of opioid deaths use opioid messaging opportunities as a vehicle to purport party-defining issues that have only loose ties to America’s opioid epidemic. Pairing a novel method for multi-corpora topic modeling with text from bill summaries, floor speeches, and press releases about opioid issues, I find that incumbents carry forward their campaign messaging behaviors to Congress. These results suggest that rhetoric from political elites plays a pivotal role in perpetuating harmful stigmas about addiction, and discourages treatment-seeking behaviors.

“Descriptive Candidacies & Position-Taking in Campaigns for the U.S. House of Representatives” with Sarah A. Treul and Maura McDonald. [Working Paper]

Although the consequences of electing descriptively unique candidates to Congress have been well-explored, scant attention has been paid to the impacts that descriptive candidacies have on campaigns and elections. Employing comprehensive text data on the campaign platforms for U.S. House of Representatives candidates from 2018, we seek to determine if and how a candidate’s descriptive identity impacts the kinds of issues adopted by her primary election opponents. We find that Democratic male, Democratic white, and Republican civilian candidates are significantly more likely to take up women, veteran, and Black-associated issues when a candidate who possess this identity runs in their primary. In a content analysis using text-as-data, we show that identity-holders and candidates who do not share this identity are remarkably consistent in how they talk about issues of interest. Finally, in our comparison of primary and general election platform text, we demonstrate that this website content remains largely unchanged over the course of elections. Our findings indicate that position taking in the primary will have long-term consequences on campaign strategy.

“Estimating the Ideological Extremity of Primary Electorates.” Under Review. [Working Paper], [Supplemental Information]

Contemporary accounts of elections characterize voters who participate in primaries as the extreme subset of their party. However, previous methodological limitations have impeded efforts to directly measure the ideological extremity of primary voters at the congressional district level. Using multilevel regression with synthetic poststratification (MrsP), I create a direct measure of primary constituency ideology for both parties in each congressional district. My analysis demonstrates that there is significant variation in constituency extremity across districts and that this variation is dependent on district characteristics. Examining the effect of primary constituency extremity on polarization, I show that (1) ideological candidates are more likely to emerge in districts with extreme primary electorates, and (2) incumbents are less representative of primary voters when the general election is competitive for both parties. The estimates developed here offer numerous opportunities for future work to explore patterns in candidate emergence and success at the primary election level.

“Reevaluating Experience in Congressional Primary Elections” with Sarah A. Treul. Under Review. [Working Paper], [Supplemental Information]

The dynamics of congressional primaries have changed dramatically in recent years. Today, political amateur are more likely to defeat incumbents and experienced ``quality” candidates than ever before. These trends lead us to question whether traditional theories of candidate success translate to today’s primary electoral landscape. Utilizing new data on candidate political experience for all primary candidates from 1980 to 2018, we show that past political experience is no longer the overriding predictor for primary election success. We test three potential explanations for this change: 1) voters’ preferences for inexperienced candidates; 2) the ideology of the candidates; and 3) the role of PAC money. We demonstrate PAC contributions to amateurs have skyrocketed in tandem with inexperienced candidate success. Finally, we evaluate whether nominating a politically inexperienced candidate puts parties at a disadvantage in the general election.