Abstract: We develop a dynamic model of narrative-based political competition. Voters recall the frequency of policies and outcomes, but not their correlation. Politicians exploit such coarse memory by crafting plausible narratives that overstate their policy's effectiveness, reattributing outcomes to steal credit and shift blame. Because plausibility makes narrative optimism fall with policy implementation, incumbency erodes narrative advantage. This force generates endogenous policy cycles and, in the long run, makes tenure independent of true policy effectiveness. Extensions to analogy-based expectations, preferences for truth-telling, and heterogeneous voter feedback show how our mechanisms operate in richer settings that restore a role for policy effectiveness.

Presentations: 16th UniTo-CCA PhD Workshop; 1st Political Economy Winter Workshop; UC Davis Microeconomic Theory Seminar; EEA-ESEM 2024 Rotterdam; XVIII GRASS Workshop Turin;  2025 Lisbon Meetings in Game Theory and Applications; 12th Bounded Rationality in Choice Conference; ESSET 2026.

Awarded Best Graduate Paper at 2025 Lisbon Meetings in Game Theory and Applications.

Abstract: This paper studies causal rhetoric in politics: language that links political actions to outcomes through blame and credit. We measure its use in a corpus of 4.2 million congressional tweets spanning 2012 to 2023 and analyze its economics. We document that causal rhetoric becomes increasingly central in politicians' communication. Leveraging quasi-exogenous variation in Twitter penetration, we show that causal rhetoric---especially blame---generates political returns in small-donor contributions: blame alone mobilizes additional donors, whereas credit increases only average giving. Using fine-grained data on legislative activity, we find that politicians strategically reallocate blame and credit as policy ownership shifts their relative credibility. Finally, economic causal rhetoric appears only weakly grounded in mainstream macroeconomics, especially blame.

Presentations: Munich Workshop on Beliefs Narratives and Memory; 2nd Bocconi-CCA-Cornell Political Economy Workshop; Princeton Junior PE Seminar; Workshop on Text-as-Data in Economics at Lancaster; 2nd Verona Early Career Workshop in Economics; Potsdam Text-as-Data in Behavioral Economics Workshop; XIV Alghero IBEO Workshop; 1st Berlin Micro Theory and Behavioral Economics PhD Conference; 10th Monash-Paris-Warwick-Zurich-CEPR Text-As-Data Workshop; 1st CEPR Future of Democracy Conference at WZB; Ludwig Erhard ifo Research Workshop; 18th UniTo-CCA PhD Workshop in Economics; Applied Young Economist Webinar; XXIII RIDGE Political Economy Workshop; 9th Economics of Media Bias Workshop; AI and Society Conference; 2nd Berlin Applied Microeconomics PhD Conference.

Abstract: This paper studies persuasion through memory. I use a theory-driven experiment in which an advisor and an investor share a common history of past news, but the advisor's only action is to re-report one previously observed item when the investor forms his belief. The experiment randomly varies both advisors' incentives and the association between contexts and signals across realized histories. Re-reporting past information creates scope for persuasion: investors' beliefs move with advisors' incentives, and advisors strategically choose which item to re-report. In this nuanced strategic setting, however, the data do not support the specific associative-memory mechanism emphasized by the framework. The findings identify memory as a distinct margin of persuasion while leaving open which features of memory drive persuasion in strategic environments.

Recipient of the Behavioral Finance e.V. Research Grant 2025

Abstract: How does policy implementation change the rhetoric elected officials use to connect policy to economic outcomes? We study Medicaid expansion under the Affordable Care Act, whose staggered state rollout changed the local policy environment facing members of the U.S. House. Linking implementation timing to 1.6 million tweets from 2012 to 2019, we measure insurance premium mentions and cross-party semantic similarity in ACA rhetoric within states. Before implementation, both parties linked the ACA to premiums, but with opposite valence. During implementation, however, supporters moved away from that link: Democrats reduced ACA--premium co-mentions by about 3.5 percentage points over two years, while Republicans did not; moreover, cross-party semantic similarity rose by about one-half of a standard deviation. These shifts occur even though the ACA is associated with lower premium growth in expansion states. The evidence suggests that implementation can narrow supporters’ rhetorical possibilities even when a policy is associated with its intended outcomes.

Abstract: Biased, overly pessimistic beliefs represent an important obstacle to the adoption of cost-effective policy instruments, such as carbon taxes. With a large survey experiment, we investigate the effectiveness, on impact and over time, of information provision about the functioning of carbon taxes. We aim to simulate the potential of an actual information campaign, including interactions among peers about the content of the campaign. Such information campaigns are widely used in European countries. We focus on one of the few such countries without its own domestic carbon pricing scheme, Italy.