What I have learned. This page distills the main findings of my research in plain language, without jargon or technicalities. Each claim links to the paper where the evidence is presented — click through for methods, data, and caveats.

Institutions and long-run development

Medieval religious institutions left an economic footprint that is measurable centuries later.

In medieval England, monasteries following the Rule of St Benedict — which prescribed work, discipline, and careful administration — fostered better economic outcomes in the areas they managed, compared to otherwise similar monastic lands. [The Journal of Economic History, 2024]

Religious persuasion can build financial institutions: Franciscan preaching drove the birth of the Monti di Pietà.

In 15th-century Italy, itinerant Franciscan preachers legitimized lending to the poor and triggered the foundation of Monti di Pietà — early charitable credit institutions — especially in communities hit by natural disasters, with lasting effects on social capital. [Working paper]

How feudal lords were monitored mattered for development.

In medieval Sicily, territories temporarily repossessed by the Crown suffered less rent extraction and fewer agency problems — and display persistently better local economic outcomes today. [Work in progress]

Democracy, institutions, and accountability

Democracy does cause growth — modestly and mostly through indirect channels.

Pooling the evidence from more than 2,000 published regressions in a meta-analysis, the effect of democracy on economic growth is positive, but much of it operates indirectly, through channels such as human capital and economic freedom. [European Journal of Political Economy, 2020]

Inclusive institutions improve food security.

Across developing countries, democratic and inclusive institutions are associated with better food security outcomes — political inclusion has very concrete material consequences. [Canadian Journal of Development Studies, 2018]

Media competition shapes whether voters can hold politicians accountable.

When incumbents can "buy the silence" of news outlets, the structure of the media market — how many outlets compete and at what price they can be captured — determines how much information reaches voters and how effective elections are as an accountability device. [European Journal of Political Economy, 2021]

Cooperation, behavior, and human–machine interaction

Noise erodes cooperation, and talking about it afterwards does not fix it.

In indefinitely repeated Prisoner's Dilemma experiments, when intended actions are sometimes reversed by chance ("noise"), mutual cooperation weakens — and ex-post communication helps only when noise is absent or low. [Work in progress]

People cooperate with robots — more so when robots look and speak like humans.

In strategic interactions, communication and human-like features of a robotic counterpart increase human cooperation, and people attribute mental states to robots that behave consistently over time. [Journal of Behavioral and Experimental Economics, 2023] [Cyberpsychology, Behavior, and Social Networking, 2024]

Playing first changes how fair you are: order of play matters in the Ultimatum Game.

Experiencing both roles in the Ultimatum Game, and the order in which subjects experience them, affects the consistency of their behavior — evidence that perspective-taking shapes strategic fairness. [Italian Economic Journal, 2026]

Behavioral interventions in the field

Prosocial preferences are not fixed: targeted programs can rebuild trust, even behind bars.

Lab-in-the-field experiments with inmates show that rehabilitation programs can measurably increase trust and prosocial behavior, and incarcerated participants in the GRIP program (Guiding Rage Into Power) report meaningful socio-emotional change. [Journal of Economic Psychology, 2018] [Journal of Community & Applied Social Psychology, 2025]

Intensive tutoring can narrow the income-achievement gap.

Experimental evidence from high-dosage tutoring in Dutch primary schools shows that well-designed, intensive educational support significantly improves outcomes for disadvantaged pupils. [Economics of Education Review, 2023]

Child sponsorship works: sponsored children in Goma (DRC) perform better at school.

Evidence from an international child sponsorship program in the Democratic Republic of Congo shows significant improvements in school performance among sponsored children. [Journal of African Economies, 2022]

For the full list of publications, see the research page; for ongoing projects, see work in progress.