Luca Bellodi
Research Fellow, Hoover Institution; Postdoctoral Scholar in Political Science, Stanford University
Project: The Anti-Democrat Penalty: The Political Consequences of Regulatory Enforcement in the United States
This project examines how environmental regulatory enforcement shapes political behavior and attitudes among employees at sanctioned firms. Leveraging large-scale administrative data on Environmental Protection Agency penalties, employment histories, and voter files, the project investigates whether workers respond to enforcement actions by blaming firms for environmental misconduct or by directing backlash toward the Democratic Party, which has traditionally championed stricter environmental regulation and enforcement. The grant supports a large-scale original survey of employees at regulated firms to examine mechanisms underlying these responses, including awareness of enforcement actions, perceptions of economic and workplace consequences of penalties, and exposure to employer framing, offering new evidence on how regulatory enforcement affects trust in government agencies and regulatory politics.
Christopher Buckley
PhD Candidate, American Politics and Comparative Politics, Stanford University
Project: (New) Political Economy of Disaster Relief
This project studies changes in the nature and significance of political influence in distribution decisions using presidential disaster declarations as a case. The goal is to identify when and how changes occurred. Doing so will illuminate how public goods distribution in other areas is affected by political elites, under what conditions incentives align to encourage better governance, and whether and how voters respond to these behavioral changes.
Jennifer Depew
PhD Candidate, Philosophy in History, Stanford University
Project: Spirits of ’76: Memories of the American Founding in American Politics, 1976–2016
This project is an excavation of how Americans have talked about the American founding from the Bicentennial to the rise of President Donald Trump. Using archival research and oral histories, the project traces the ways in which Americans have used the memory of the founding for their own ends. Particular attention is paid to moments of tension within the political Right and Left and between them. The research analyzes where Americans succeeded in civic education on the founding and brings to the fore moments in which those efforts fell short. The project concludes with a new chronology of post-1970s American political development and recommendations for how the American Revolution can be used to help build a stable twenty-first-century future.
Vasiliki Fouka
Bing Professor of Human Biology and Associate Professor of Political Science, Stanford University; Senior Fellow at the Stanford Institute for Economic Policy Research
Project: Migration, Sorting, and the Educational Divide
This project examines how immigration contributes to the growing education-based divide in political attitudes. It focuses on how individuals with different education levels experience immigration differently, potentially shaping their views and behavior. The study explores patterns of residential sorting, social interaction, and cultural proximity between immigrants and natives and how these vary across groups. Using large-scale, address-level data, the project aims to document these patterns and assess their broader implications for social cohesion and political polarization.
Shanto Iyengar
William Robertson Coe Professor of American Studies and Professor of Political Science, Stanford University
Project: Does the Message Hurt the Messenger? Investigating the Decline in Public Trust in the News Media
This project tests whether transformations in journalistic style have contributed to the long-term decline in Americans’ trust in the news media. Using a sample of more than 4,500 television news broadcasts covering US presidential campaigns from 1968 to 2024, the longest-spanning dataset to date in the literature, the research first measures whether news coverage has become more interpretive and subjective over time. It then fields a survey experiment to test the immediate effects of journalistic subjectivity on trust in media and couples this with time series analyses of long-run survey data to assess whether such short-term effects can plausibly account for the broader decline in trust.
Joseph S. Mernyk
PhD Candidate, Sociology, Stanford University
Project: Leveraging Large Language Models to Inform Citizens About Elections
Millions of Americans already turn to AI chatbots for election information, yet general-purpose large language models may hallucinate facts, reflect partisan biases, or surface outdated information. This project develops and rigorously evaluates an AI voter guide grounded in verified, nonpartisan data. Building on a preregistered survey experiment conducted during the 2024 general election, in which the tool increased turnout intentions, improved alignment between voters’ preferences and candidate choices, and reduced partisan animosity, the next phase will scale the guide nationwide ahead of the 2026 midterms and link outcomes to verified voter records.
Jennifer Pan
Sir Robert Ho Tung Professor of Communication and Professor, by Courtesy, of Political Science and Sociology, Stanford University; Senior Fellow at the Freeman Spogli Institute for International Studies
Project: How Partisanship Affects Preferences for Content Moderation in Large Language Models
As generative AI systems become a primary information source for millions, their content moderation decisions carry growing implications for public trust. This project investigates how partisanship shapes perceptions of large language model (LLM) outputs that vary in political alignment and degree of moderation. The research tests whether Americans exhibit a “preference gap” in their views about how LLMs should respond and whether they engage in “party promotion,” strategically favoring suppression of content misaligned with their own political views. The study further examines how these moderation choices affect trust in LLMs and the companies that build them.
Sergey Sanovich
Hoover Fellow, Hoover Institution
Project: Authority Model Collapse: Information Commons after AI
Over the past decade, access to social media data and increasingly powerful machine learning tools has enabled social scientists to make significant progress in addressing key empirical questions about the political impact of digital communication, from filter bubbles and algorithmic polarization to digital censorship and propaganda bots. Much of this progress, however, has depended on the availability of benchmarks for factuality, neutrality, and verifiability. Generative AI threatens the information commons on which this work relies. This project will use empirical political science methods and AI detection tools from computer science to comprehensively evaluate the integrity, reliability, and resilience of key information commons in the age of AI, beginning with social media.
Mitchell Stevens
Professor, Graduate School of Education, Stanford University
Project: Mining History for Models of Adult Learning Opportunities in an Era of Lengthening Lives
This project examines how colleges and universities can adapt their civic mission to serve broader and more diverse adult populations amid declining public trust and shrinking traditional student pipelines. Drawing on historical surveys and contemporary landscape analyses of nontraditional instructional models developed to serve adults across the life course, the project lays the groundwork for future publications and institutional strategies.
Janine Zacharia
Carlos Kelly McClatchy Lecturer, Department of Communication, Stanford University
Project: Restoring Trust in Journalists and Journalism by Having People Who No Longer Trust the Media Take Part in the Reporting Process and Meet Local Reporters and Editors
This project explores whether participation in the reporting process and in-person interaction with journalists can increase trust in the media among people who currently distrust the press. A follow-on to a preliminary intervention conducted with The Denver Post, this next phase is being carried out in partnership with The Arizona Republic to solicit story pitches from people across the state and hold several in-person pitch sessions with editors. Surveys are being conducted to assess whether the experience changes attitudes toward the media. The long-term goal is to build an intervention that can be replicated in newsrooms across the country.
