research
Working Papers
“Extremism Is Self-Undermining, but Populism Is Not.” — Job Market Paper
Abstract. Does holding office strengthen or weaken the long-term appeal of right-wing extremism and populism? Although populism and extreme-right ideology are distinct concepts, they are rarely measured separately. This paper does so, using supervised text classification of candidate platforms to construct continuous candidate-level scores. It then exploits close mayoral elections in Italian municipalities, extending close-election politician-characteristic regression discontinuity to continuous treatments. Electing an extreme-right mayor is self-undermining: the next municipal election shifts toward more moderate candidates. Support for extreme-right parties also appears to decline in subsequent national and European elections. By contrast, electing a populist mayor does not weaken the subsequent appeal of populism. The most plausible explanation for these differences lies in behavior in office. Populism and extreme-right ideology identify distinct political types: extreme-right mayors behave as policy-seekers, pursuing welfare retrenchment even at electoral cost, while populist mayors appear to act as office-seekers, with welfare expansion concentrated where re-election incentives remain.
“Paying for Moderation”
Abstract. What drives political extremism? Most explanations emphasize voters. This paper shows that the financial rewards of office also play a role. The paper uses a regression discontinuity design at a constitutional population threshold in Brazil that raises councilor salary caps and, in practice, mayoral pay. It develops two measures of candidate ideology from legally required platforms: one based on supervised transformer classifiers and the other on an ensemble of large language models. Above the threshold, both the share of extreme-right candidates and the probability of electing an extreme-right mayor decline, with no comparable decline on the left. Joint estimates of wealth and ideology suggest substitution toward relatively wealthy moderate-right candidates. Supplementary U.S. evidence suggests a similar asymmetric response and is consistent with the same selection mechanism. These findings suggest that political compensation can limit political extremism through candidate selection, reshaping the alternatives available to voters.
“War, Politics, and the Birth of Modern Bureaucracy” (with Alexander Lee and Anil Menon)
Abstract. A large literature has examined the “rise of the state” in early modern Europe, focusing on the increase in the capacity of states due to war. However, state institutions were increasing not only in size but also in their quality, through the creation of “modern” or “Weberian” bureaucracies. To illustrate these distinct processes, our project examines the state in early modern Britain, often cited as the ideal type of a bureaucratic transformation. We use an original dataset of every central government employee in Britain between 1660 and 1860, capturing not just their numbers but four measures of bureaucratic rationalization: functional specialization, remuneration through salaries rather than user fees, higher salaries and increased time in office. Consistent with the existing literature, we find that war is associated with an increase in the size of the bureaucracy, though quality does not change. Quality improvements are associated with periods that combined close political competition and fiscal pressure. The results suggest that while military pressure can expand bureaucracies, their reform is a political project.
Presentations
“Extremism Is Self-Undermining, but Populism Is Not.”
- American Political Science Association Annual Meeting, Boston, MA, September 2026
- Annual Meeting of the Society for Political Methodology (PolMeth), East Lansing, MI, July 2026
“War, Politics, and the Birth of Modern Bureaucracy” (with Alexander Lee and Anil Menon)
- American Political Science Association Annual Meeting, Los Angeles, CA, 2023