Fred DeVeaux

I'm a PhD candidate at UCLA studying American politics and political methodology with a focus on elections, representation, and class. I study how politicians' class backgrounds influence their election campaigns and behavior in office. I combine original data on the occupational backgrounds of congressional candidates from their campaign websites with modern empirical techniques for causal inference and recent advances in computational text analysis. I am currently on the job market.

CV

Research

• Do Working-Class Candidates Activate Class-Based Voting?
  With Jared Abbott. Electoral Studies, 2024. [Replication materials]
  Abstract

  After steadily leaving the Democratic Party, working-class voters are increasingly seen as pivotal in US elections. What type of candidates should parties nominate to win over working-class voters? Parties often nominate candidates based on characteristics they think will appeal to certain groups of voters. These typically include campaign messages, such as policy positions or rhetoric, but can also include descriptive characteristics. In this paper we use a conjoint experiment to test whether candidates’ class background can activate class-based voting. Overall, we find that a candidate’s occupation has a substantial effect on voter perceptions: working-class respondents are 6.4 percentage points more likely to prefer a candidate with a working-class occupation over one with an upper-class occupation. This effect is not driven by inferences that respondents make about candidates’ policy positions or group-based rhetoric. Instead, we find that working-class voters perceive working-class candidates as more understanding of their problems. Our results suggest that candidates’ class background is an underappreciated yet effective mechanism for activating class-based voting.

• What Happens When Working-Class Candidates Run For Office?
  With Jared Abbott. Revise & Resubmit
  Abstract

  Working-class Americans are significantly underrepresented in politics. Yet it remains unclear whether elections contribute to this underrepresentation, and if so, at which stage. To address this question, we collect the full occupational histories of over 8,000 congressional candidates who ran in contested primaries or general elections between 2010 and 2024. Comparing candidates under similar electoral conditions, we find that working-class candidates perform substantially worse in primaries but nearly as well as other candidates in general elections. In primaries, working-class candidates receive about 3 percentage points less vote share and are 7 percentage points less likely to win, whereas parties nominating working-class candidates receive only 0.6 percentage points less general-election vote share, with no detectable effect on victory. To explain this difference, we compare working-class and non-working-class candidates across election stages. Working-class general-election nominees remain less likely to hold elite credentials, raise less money, and receive relatively greater support from working-class than upper-middle-class voters. Because these differences persist after nomination, the absence of a meaningful general-election penalty is difficult to attribute to a screening account in which primaries select only working-class candidates who resemble other nominees. Instead, the pattern is consistent with party labels attenuating the electoral consequences of class-based disadvantages in general elections.

• Do Legislators’ Occupations Predict Their Behavior in Office?
  Work in Progress
  Abstract

  Compared to the population at large, Congress has few workers and tradespeo- ple and many lawyers and business executives. This raises concerns about unequal representation. Yet it remains unclear how much legislators’ prior careers predict behavior in office, and whether differences appear mainly in the issues they work on, the positions they take, or both. I compile occupational histories for 3,600 members of Congress, 1947–2024. Comparing same-party legislators representing the same dis- trict over time, I find that legislators focus more on issues connected to their prior oc- cupations in speeches and sponsored bills, serve on relevant committees, and advance occupation-related legislation more effectively. But they do not vote differently on issues related to their occupation than on other issues. Instead, occupation predicts voting along traditional class lines, especially between business- and working-class legislators on employment and labor. Congress’s occupational composition therefore shapes which issues receive attention and whose economic interests are represented.

• Occupational Donor Networks in Congressional Elections
  Work in Progress
  Abstract

  Members of Congress are drawn from a narrow set of occupations whose members are also major sources of campaign money. How much do occupational donor networks shape fundraising in congressional elections? I assemble occupational histories for 9,100 candidates for U.S. Congress between 2010 and 2024, and classify the self-reported occupations of more than 14 million unique individual donors who contributed $14.4 billion to federal House and Senate campaigns. I find that candidates capture a larger share of contributions from donors who share their occupation than from other donors. These occupational donor networks operate in both general and primary elections, but are stronger in primaries. They vary both in the share of each occupation's donor pool that candidates capture and in the size of the donor pool itself. As a result, candidates from elite professional occupations gain tens of thousands of dollars from donors who share their occupation, while candidates from working-class occupations gain little. Occupational donor networks therefore provide the greatest fundraising advantages to candidates from occupations already overrepresented in Congress.

• Anti-Establishment Rhetoric in Congressional Campaigns
  With Jared Abbott. Work in Progress
  Abstract

  The past decade has witnessed a sharp rise in anti-establishment politics in the United States, but its origins remain poorly understood. Is anti-establishment sentiment primarily an ideological phenomenon of the far right and far left, or does it represent an independent dimension of political conflict? Does it emerge through the entry of new outsider candidates, or through adaptation by established politicians to changing voter demands? To address these questions, we construct a new dataset of more than 12,000 congressional candidates running in primary and general elections between 2010 and 2024, drawing on over one million sentences scraped from their campaign websites. Using large language models and supervised machine learning, we classify the degree of anti-elite rhetoric in each candidate’s platform. We then examine how the prevalence of anti-elite messaging has evolved over time and the extent to which it reflects candidates’ ideological positions, occupational and socioeconomic backgrounds, district characteristics, or strategic responses to challengers. This approach provides the first systematic portrait of the rise of anti-elite rhetoric in U.S. congressional elections and the types of candidates and contexts in which it emerges.

Teaching

• PS40: Intro to American Politics (2022, 2023, 2024). Teaching Assistant
• PS167: Political Economy of Development. Teaching Assistant
• PS116: Marxism. Teaching Assistant

Contact

Email: fdeveaux@ucla.edu