Program

Abstract: The adoption of automation technologies has the potential to transform production processes and boost productivity growth. Yet initial uptake has remained limited and largely concentrated among frontier firms, reinvigorating concerns about a slowdown in technology diffusion. Moreover, whether and how these technologies later spread across the business population has proved difficult to study empirically. In this paper, we address this challenge by studying the role of production-network linkages in the diffusion and impacts of Industry 4.0 automation technologies. We examine how exposure to automated buyers and suppliers influences both the investment decisions of connected firms and the performance of those that do not adopt automation themselves. We leverage unique data from Italy, combining  information on the timing of automation investments -- measured by take-up of the Industry 4.0 bonus depreciation scheme (2017–2019) in the manufacturing sector -- with the universe of firm-to-firm transactions and firm-level financials. We find that exposure to automated peers significantly increases the likelihood of adoption, based on linear-in-means peer effect models that are robust to several sources of firm- and network-level endogeneity. Furthermore, event-study estimates reveal increases in sales, costs, value added, employment and labor productivity for non-adopters following exposure to automated peers, a result that appears to be mainly driven by downstream buyers. Since adoption in our setting is policy-induced, we conclude by discussing the possible  direct and indirect industrial policy effects on technology diffusion patterns.

Abstract: In response to rising urban air pollution, European cities have adopted Low Emission Zones (LEZs), restricting the most polluting vehicles. While effective in improving air quality, these policies remain controversial due to concerns over fairness and acceptability. This paper examines the impact of London’s 2021 and 2023 Ultra Low Emission Zone (ULEZ) expansions on subjective well-being (SWB). Using panel data from the UK Household Longitudinal Study and a staggered difference-in-differences design with individual and year fixed effects, we compare changes in life satisfaction among residents inside and outside the affected areas. We find that the 2021 expansion led to a decline in life satisfaction by approximately 0.4 points -- which doubles for low-income households -- with no evidence of pre-existing differential trends. We do not detect statistically significant effects within the available post-treatment window on Londoners living in the expanded zone in 2023. We explore the mechanisms driving this decline and find that the well-being loss is fundamentally mediated by car dependency and transport mode availability. While the policy increased reliance on public transport, we show that a higher accessibility to public transport reduces the well-being decline of Londoners. These findings suggest that LEZs can generate short-term welfare costs despite achieving behavioural change, highlighting the need for complementary measures to enhance social acceptability.

Abstract: This paper analyzes the reduction in total labor costs induced by an increase in the minimum wage in France. Using the Ines microsimulation model developed by the French National Statistical Institute (Insee), I simulate a 2% increase in wages for all workers paid at the minimum wage. Due to the complex of exemptions of the French socio-fiscal system, an increase in the minimum wage leads to a reduction in employers’ social security contributions (SSCs) for workers earning between 1.01 and 3.5 times the minimum wage. I confirm existing results from L’horty (2000): overall, a 1% increase in the minimum wage reduces employers’ SSCs by approximately €1.67 billion.

Abstract: Preschool programs have expanded rapidly in low and middle income countries, but concerns remain about their quality and effectiveness. This study evaluates the impact of Mexico's preschool mandate on primary school performance using a staggered difference in differences design. The reform increased Grade 5 test scores by 0.05 to 0.16 standard deviations but also widened gaps between highly and weakly marginalized localities. Smaller preschool groups and longer attendance improved cognitive outcomes despite a higher share of low skilled teachers. The results highlight the importance of targeting quality improvements when expanding early childhood education.

Abstract: This paper estimates the causal effect of food insecurity on Venezuelan emigration to Colombia using municipal-level panel data for 2010–2023. Exploiting exogenous variation in crop losses via an instrument that interacts drought intensity with 2007 agricultural production, I show that a one percentage-point increase in food insecurity raises the emigration rate by 4.2 percent on average. A 10 percent drop in crop yields increases food insecurity by 0.56 percentage points, implying a 2.4 percent rise in the emigration rate. Reduced-form evidence indicates that drought shocks mattered only after the 2014 food import collapse, showing the role of local agricultural fragility in the broader political crisis. Responses are strongest in municipalities with low crop diversity, a larger oil sector, and intermediate levels of development. These findings provide the first causal evidence that hunger operates as a major push factor in contemporary emigration from developing countries and highlight that climate adaptation and food security policies are critical to reducing forced migration.

Abstract: Malaria remains a major public health challenge in Burkina Faso. To address this burden, the government implemented two major policies: the Seasonal Malaria Chemoprevention (SMC) strategy, progressively rolled out across health districts between 2014 and 2019, and the introduction of free healthcare for children under five and pregnant women in 2016. The objective of this paper is to estimate the impact of these interventions on child mortality, fertility decisions, and the access and use of insecticide-treated nets (ITN), by exploiting the staggered rollout across health districts combined with spatial variation in pre-intervention malaria burden. The empirical strategy builds on a dynamic difference-in-difference model, allowing for staggered treatment timing and heterogeneous treatment effects. The analysis draws on multiple waves of the Demographic Health Surveys for Burkina Faso (2003–2021), geo-referenced estimates from the Malaria Atlas Project (MAP), and Annual Statistical Reports (2005–2023). The results suggest substantial heterogeneity driven by pre-intervention malaria burden: mortality and fertility reductions are larger in higher-risk districts, while ITN access and use reveal behavioural heterogeneity, with historically high-burden areas showing stronger commitment to preventive practices.

Abstract: I show that temporary oil price increases have persistent effects on economic activity in European economies. Real GDP remains depressed for more than a decade after oil prices have returned to their pre-shock level, indicating the presence of hysteresis. These hysteresis effects are driven by two channels: a sharp and lasting contraction in investment, leading to sustained productivity losses, and a steady decline in labour force participation. The effects are asymmetric: adverse oil price shocks cause permanent output losses, while favourable shocks do not produce sustained gains.

Abstract: This paper studies the role of career ladders — constraints that require workers to progress step by step through hierarchies of jobs — in explaining labour market outcomes. I develop a network-based method to construct career ladders from matched employer-employee data arising naturally from a structural on-the-job search model with multidimensional skills and training frictions. Using Norwegian administrative data from 2010-2024 with detailed job titles, I estimate latent skill distances and job values from observed job transitions. The results uncover around 50 distinct career ladders that explain substantial wage variation and reveal heterogeneous patterns of career progression across the labor market. In ongoing work, I use my model to quantify the positive externalities, through increased skill supply, generated by firms' job creation across granular job types. This will help design and evaluate skill policies such as subsidies for entry-level job creation.

Abstract: The transition to a low-carbon economy requires substantial investment to replace the production technologies and infrastructure reliant on fossil fuels. In addition to regulation and carbon pricing, a range of financial policies has been proposed to accelerate green investment. This article evaluates the implications of three of them — direct green investment subsidies, green public guarantees, and capital market deepening — in terms of emission reduction, economic activity, and public debt. The analysis relies on a stock-flow consistent, input-output model of the EU economy, which explicitly incorporates industries’ marginal abatement costs, intersectoral input-output linkages, and investment financing channels. Model simulations reveal that direct subsidies are the most effective tool for achieving significant emission reductions; however, they also result in substantial increases in the debt-to-GDP ratio. In contrast, public guarantees and equity market development tend to strengthen public finances and economic activity but yield only moderate emission cuts. The results further suggest that combining policies can effectively balance emission mitigation and economic activity without compromising public finance sustainability.

Abstract: This paper develops a unified framework to disentangle three forces of the energy transition: (i) capital installation, where investment in newer, more efficient processes of production triggers a change in inputs ratio; (ii) Directed Disembodied Technical Change (DDTC), corresponding to pure R&D, which shifts the production frontier and allows more output to be produced with the same inputs ratio; and (iii) adjustment frictions, which constrain how much the input intensity of alreadyinstalled capital can be modified over time in response to changing relative prices. A key contribution is the introduction of a parameter, measuring adjustment frictions, that interpolates between two corner cases studied in the literature: puttyclay and putty-putty capital. Bayesian estimation on U.S. data (1949-2024) shows that the level of such adjustment frictions matters for structural parameter values: the elasticity of substitution (EoS) between emissions and capital ranges from 0.18 to 1.16. Ultimately, this matters for optimal policy: the higher the EoS, the more the emissions transition is driven by capital installation, which advocates for subsidies to capital scrapping or to investment.

Abstract: This paper examines how changes in textbook adoption rules shaped historical narratives in American history textbooks between 1870 and 1930. During this period, most Western and Southern states transitioned from local to state-level textbook lists. We investigate whether it enabled Southern states to influence how publishers portrayed Civil War and the South. We employ large language models to systematically measure pro-Southern slant on a novel dataset of 338 editions of middle-school history textbooks. Leveraging variation in the timing of state adoption laws, we find that state-level consolidation led to 18.5% more pro-Confederate coverage of the Civil War. This increase is partly driven by more positive portrayals of Confederate army leaders which was an explicit long-run objective of the United Daughters of the Confederacy activity. We show the effect was driven by both Southern publishers revising narratives to target the Southern market and, to a lesser extent, Northern publishers adjusting their content to compete for it. Preliminary analysis using 1937-1949 Gallup survey data suggests that cohorts exposed to these biased textbooks during schooling exhibited more white supremacist attitudes later. Relative to unexposed cohorts, they are 18% more likely to approve of the Ku Klux Klan and 22% more likely to vote for Democratic candidate in 1936 elections.

Abstract: Is deplatforming effective in reducing toxicity on social media? To answer this question we study a policy change on Reddit in June 2020 which led to a simultaneous ban of thousands of forums containing hateful content, but not the users of these forums. We use data on the near universe of comments left on Reddit to examine the impact of the ban on user behaviour in a differences-in-differences design. We find that the most active users of banned subreddits comment more after the policy change and substitute to new forums in the weeks after the ban. The increase in activity persists in the long run, but is not associated with higher toxicity: instead, the comments left by affected users outside banned subreddits contain 20% fewer instances of hate speech. We do not find evidence that the policy leads to lower quality of engagement, negative spillover effects or recreation of banned subreddits elsewhere on the platform. Overall, the results suggest that moderation targeting toxic digital spaces can be effective in combating hate speech without lowering user engagement, and thus can be aligned with platforms’ incentives.

Abstract: We study how tasks, formality, and sectoral reallocation jointly shape task-based labor-market polarization in Colombia’s urban labor market from 1984 to 2019. Using harmonized task groups built from standardized occupational classifications, consistent measures of formality, and within–between decompositions, we document a robust hollowing-out of routine work. Between 1995 and 2019, the routine share falls by about 8.6 percentage points, while manual and abstract shares rise by roughly 3.9 and 4.7 points; close to four-fifths of the routine decline reflects within-industry substitution rather than shifts across industries. Adjustment is disproportionately absorbed by the formal sector, whose employment share increases from 55.6% to 63.8%, alongside a formal-sector tilt toward high-skill services. We then show that macro shocks can amplify task reallocation: city-level reduced-form estimates indicate that commodity-price-driven terms-of-trade shocks consistent with Dutch Disease are linked to larger increases in polarization in cities that were initially more routine-intensive, with effects concentrated during the 2003–2014 boom window. While employment shares display a pronounced U-shaped polarization pattern across tasks, wages do not; instead, the ranking of wage growth across task groups shifts across decades. Composition-adjusted (residual) wages reveal a rising abstract premium and a late improvement of manual relative to routine. Overall, the evidence is consistent with within-industry technological adjustment interacting with boom-driven relative-price movements that reallocate employment away from routine-intensive tradables toward service activities combining abstract and manual tasks.

Abstract: We develop an asset-pricing model in which investors search for return predictability in high-dimensional data using a LASSO-based learning rule that mirrors empirical practices. Although returns are unpredictable under full-information rational expectations, investors test this hypothesis by regressing returns on a vast array of signals. We show that this behavior induces a nonlinear feedback from beliefs to prices that prevents convergence to rational expectations and instead generates transient, sparse episodes of predictability. The key implication is that the search for predictability itself endogenously generates short-lived windows of predictability. We characterize these dynamics analytically and document their empirical relevance in the cross-section of U.S. stock returns.

Abstract: This paper provides causal evidence that NGO campaigns influence land-use decisions in Brazil’s soy and beef supply chains—two of the leading drivers of tropical deforestation globally. Using a shift-share design that exploits quasi-exogenous variation in campaign timing and firm–municipality trade linkages, I find that municipalities that trade with firms that are targeted by NGOs significantly reduce deforestation, primarily by deterring forest-to-cattle conversion. Campaigns also promote reallocation from cattle grazing to soy cultivation, indicating a shift toward more intensive and economically productive land uses. These transitions are accompanied by measurable gains in land productivity, suggesting that NGO pressure not only curbs extensive deforestation but also enhances the efficiency of existing agricultural areas. The results underscore the persistent influence of civil society in shaping environmental outcomes, even after formal supply-chain commitments are in place, and highlight the broader potential of reputational governance to complement state regulation in managing global environmental externalities.

Abstract: Women’s economic and social roles changed substantially over the course of the twentieth century, but the extent to which war contributed to this transformation remains unclear in the German case. This project examines how the First World War affected female labour force participation in Germany, with a current empirical focus on Bavaria. Exploiting regional variation in military death rates, the analysis asks whether areas more severely affected by wartime male losses experienced stronger increases in women’s employment. The project combines occupational and population censuses with military casualty records and a unique 1917 survey of more than 700 Bavarian firms covering over 89,000 women employed in the war economy. This allows me to study both overall changes in female labour force participation and the mechanisms driving them. In particular, I examine how wartime labour shortages and household financial necessity following the loss of male breadwinners shaped women’s entry into paid work. Preliminary results indicate that the relationship between wartime male losses and female labour force participation was heterogeneous: in particular, rural counties with higher military death rates saw stronger increases in female employment.

Abstract: Asymmetric access to consumer data is common in many digital markets and online platforms. I study a Bertrand duopoly in which one firm possesses richer data than its rival and can commit to a segmentation policy for personalized pricing. In equilibrium, the optimal information structure coarsens consumer data to soften price competition, thereby reducing each consumer's surplus.

Abstract: Do corporate bond markets stabilize the economy when banks collapse, or do they contract alongside bank credit during systemic crises? This paper provides the first long-run cross-country evidence on primary bond issuance dynamics around banking crises, assembling a new sector-level dataset covering 21 countries since 1835. I document a persistent reallocation of bond market activity following crisis onsets. Event-study and long-difference local-projection estimates show that government bond issuance rises sharply and remains elevated, financial-sector issuance increases only temporarily, and non-financial corporate issuance declines persistently. There is no evidence of an aggregate corporate “spare tire” expansion offsetting bank distress. I then examine whether within-crisis heterogeneity in corporate issuance dynamics is systematically related to post-crisis recovery. Crisis episodes with stronger corporate issuance growth exhibit significantly smaller subsequent investment declines and moderately stronger output recoveries, particularly when bank credit conditions are weak. These findings suggest that bond markets matter for macroeconomic adjustment not through aggregate corporate expansion, but through selective access and heterogeneous financing resilience during financial shocks.

Abstract: This paper examines how human capital accumulation influenced England's long-run economic transformation (1200-1800). Using new microdata on university scholars within a general equilibrium framework, it estimates agricultural and academic productivity and changing preferences for urban, skill-intensive goods. After the Black Death, higher real wages spurred innovations in both city and countryside: urban demand for knowledge and leisure advanced printing and education, while reduced food needs enabled selective land use, raising rural productivity by 55% (13001600). Later, population growth drove technological adoption, doubling agricultural productivity by 1800. Interlinked rural gains and urban knowledge flows underpinned England's path to modern growth.

Abstract: The design of French income tax provides an approach for estimating tax elasticities of all kinds. Unlike natural experiment designs that analyse between-income variation (such as arising from tax reforms affecting certain income groups), the variation examined in this paper stems from differential within-income shifts to the tax liability function. This makes it possible to control for income without absorbing identifying variation, it eases testing of exogeneity and makes monotonicity less likely to be violated. It also means that the complier population is larger, allowing for heterogeneity analyses across income distribution, gender and a range of socio-economic characteristics. Applying this strategy to data covering the universe of French income tax returns from 2003 to 2023 reveals considerable heterogeneity along all these dimensions.

Abstract: This paper studies how the introduction of no-excuse mail voting impacted partisan voting among couples in the United States. By leveraging its staggered introduction by US states, I show that mail-in voting has reduced the gender gap in partisan voting among married voters by at least 1.1 percentage point while it did not affect that of single voters. In particular, this effect is driven by increased Republican voting among lower-educated women and a converse moderate decline among men. Studying mechanisms, I show that turnout accounts only for a fraction of the result. Instead, this paper documents that mail-in voting significantly increases political interactions among couples. Shifts in voting are unequally followed in terms of ideological feeling across demographics, suggesting increased interactions might enable ideological convergence, but that it is only partial and greatly intersects with within-couple agency.  Finally, studying the universe of reported violent events between partners in 28 US states over 10 years, I show that daily intimate partner violence incidence at the county level increases by 2.5% during mail voting periods. This result provides strong evidence that coercion, enabled by mail voting, can be driving part of the unaccounted-for results.

Abstract: In this article, I quantify two main channels through which droughts impact industrial production. First, I assess the physical transmission of water scarcity to the production growth of firms that are more exposed to river water variations. Second, I use water restrictions imposed on the most water-intensive plants to identify the medium-term semi-elasticity of industrial production to water. I assume that the difficulties in consistently estimating the impact of droughts on industrial production rely in two major characteristics of droughts: their highly localised nature and the specific characteristics of firms in their use of water for their production processes. To account for the high localisation of droughts, I rely on local drought alerts and locally imposed restrictions. To reduce the sample of firms to those dependent on water use, I first use the distance to the nearest river as a proxy for river water dependency. In a second approach, I limit the sample to a small number of highly water-intensive firms that are subject to specific restrictions. I centre the study on the period 2019-2023 which includes the last two years of severe droughts, 2019 and 2022, with the latest trends of water scarcity in a world of climate change. I show that each additional day of severe drought in France reduces the production growth of firms close to rivers by -0.010pp to -0.014pp. Using restrictions applied to the most water-intensive firms, I estimate that a 1pp increase in annual water restrictions reduces production growth by 0.07 percentage points.

Abstract: Urban heat is an important topic in scientific research and public debate. However, little is known about how cities might evolve in relation to urban heat. This paper aims to fill this gap by using a quantitative spatial model, which accounts for numerous spatial interactions between economic agents and allows for the simulation of credible changes in the city. Urban heat is modeled as an endogenous disamenity, which is generated by the share of each ward covered by built-up and affects household location choices through the model's amenities. I base this analysis on Ho Chi Minh City, a Vietnamese megacity highly exposed to urban heat. To estimate the effect of urban heat on ward's amenities, I rely on an instrumental variable approach that exploits differences in soil composition. The 2SLS estimation yields that an increase by 1% in urban heat decreases ward's amenities by 1.05%. The quantification of the model and the computation of counterfactual equilibria allow for the examination of long-term effects of urban policies. Notably, heat mitigation policies tend to favor densification of the city center, whereas the intensification of heat leads to the dispersion of residents across the province, which would then involve other urban costs.

Abstract: Outside options are central to theories of wage determination, yet fundamentally unobservable. This paper addresses this problem by constructing a new proxy for workers’ outside options. I exploit job-to-job mobility networks and partition workers into types via clustering, then use machine learning methods to estimate transition probabilities from each worker’s current firm to all potential employers. These probabilities are then summarized into a single measure rooted in discrete choice theory - the inclusive value - that captures the attractiveness of a worker’s opportunity set. I implement this approach using matched employer-employee data from the Veneto region of Italy (1995–2001). I validate the measure using mass layoffs: displaced workers with better pre-layoff outside options experience shorter unemployment spells, earn higher re-employment wages, and sort into higher-paying firms. I find that outside options predict job switching and future wages, even conditional on current wages. Turning to current wages, which may themselves result from bargaining over match surplus, I estimate that, in my preferred specification with firm and worker-type fixed effects, a one-standard-deviation increase in outside options is associated with 2 percent higher wages. This finding is corroborated by specifications with individual worker fixed effects and by an instrumental variables strategy that exploits large firm openings as exogenous shocks to workers’ outside options. Taken together, these results show that wages depend not only on the current employer but also on potential employment opportunities.

Abstract: Leveraging an Italian labour market reform known as the Jobs Act (JA), I study the effect of employment risk on household consumption and labour supply. The JA reduced protection against unlawful individual termination only for workers hired after March 6, 2015, by firms with at least 15 employees. Using this time-based discontinuity in employment protection as a source of exogenous variation in employment risk, I find that workers subject to the reform consume 8% less than workers hired before March 6, 2015. The effect is stronger among individuals younger than 40 and for those living in Northern Italy, the wealthiest region of the country. There is, on the other hand, no sizable effect on labour supply. Finally, I show that a variant of the Bewley–Huggett–Aiyagari model, augmented with ex-ante employment risk heterogeneity, matches the empirical result, and that risk accounts for a sizable part of the effect

Abstract: How does collective bargaining impact the wage gap between natives and immigrants? While unions typically compress wage distributions, their effects on native-immigrant wage differentials remain theoretically ambiguous: unions may reduce inequality or reinforce insider-outsider divisions. Using Current Population Survey data aggregated at the state-industry-year level for the United States over the period 1994-2021, I examine how variation in union density affects residual wages and the wage gap between natives and immigrants. To estimate its causal effect, I employ an instrumental variable strategy that leverages the differential impact of Right-to-Work laws across industries. The results show that a 1 percentage point increase in union share raises native residual wages by 0.28% but has no significant effect on immigrant wages, thereby widening the wage gap by approximately 0.24%. These findings suggest that collective bargaining institutions transmit wage-setting power unequally across worker groups, with unions favoring labor market insiders and potentially reinforcing wage disparities between natives and immigrants.

Abstract: I study the optimal operation and capacity of grid-scale storage in a strongly wind-dominated electricity system with or without fossil backup. Wind generation follows a continuous-time stochastic process calibrated to German wind production data. A social planner with isoelastic preferences chooses charge and discharge rates to maximize intertemporal welfare subject to storage constraints. The model delivers a simple feedback policy and enables efficient numerical solution. Absent any fossil backup, the price elasticity of demand is a primary driver of the marginal value and optimal size of storage.

An optimal wind-plus-batteries system without backup in Germany features more than ten times the 2023 installed wind capacity and storage energy covering hours to days of average German demand, with a levelized cost of more than twice current wholesale electricity prices. By contrast, an optimal wind-plus-hydrogen system features less extreme wind capacities and storage energy covering weeks to months of average demand, despite storage losses. The wind-plus-hydrogen levelized cost is well below gas or coal levels when discounting with 2% annually. A deterministic benchmark confirms the first-order relevance of uncertain wind output: storage needs are driven by the risk of prolonged low-wind spells. Hydrogen's far lower energy capacity cost dominates efficiency losses and power capacity costs in this setting.

Allowing for a fossil backup sharply reduces optimal storage. Marginal abatement costs are initially negative, then spike when near zero emissions: it is very costly to avoid using an already present fossil backup even in the rarest events of prolonged low renewable production.

Abstract: This paper studies how multinational funding structures reshape monetary transmission in open economies. We build a tractable heterogeneous-firm model in which firms differ in productivity and access to foreign credit. Foreign-owned subsidiaries combine domestic borrowing with intra-group lending, while domestic firms rely on local credit and access foreign funding only at high productivity levels. A central friction is that leverage capacity depends on the effective borrowing rate, so that when rates rise, collateral values are discounted more heavily and credit limits tighten. A domestic tightening therefore squeezes purely domestic firms through both higher funding costs and tighter leverage, whereas globally funded firms partially cushion the shock by shifting borrowing abroad. A foreign tightening propagates selectively through internationally exposed balance sheets, generating cross-border monetary spillovers even absent trade linkages. Because wealth shares adjust slowly, temporary interest-rate shocks produce persistent effects on capital, output, and the composition of firm wealth. The framework remains tractable: firm decisions are closed-form and the transition dynamics of the wealth-share distribution are characterised explicitly.

Abstract: Governments have increasingly been creating institutions to encourage the reporting of violent crimes against women. This is also true in societies where such crimes are generally tolerated. Can the creation of such institutions impact how people view these violent acts themselves? Using a difference-in-difference approach, I analyze the impact of opening a Women Police Station (WPS) in Indian districts over the period 2015-2018 on the acceptance and reported incidence of Intimate Partner Violence (IPV) by married couples. The attitudes of wives become more accepting of violence in the districts where a WPS was established. Conversely, the attitudes in control districts improve. The attitudes of husbands seem to be stable over time except for a general increase in the acceptability of IPV if the wife is suspected of being unfaithful. Wives are more likely to report having experienced emotional IPV over time in both treated and control districts. However, wives in treated districts become more likely to report having experienced severe physical IPV after the opening of a WPS. The treatment effects are not driven by interviews taking place after the COVID-19 pandemic and are stronger for households closer to the WPS. The treatment effect on severe physical IPV is associated with high household wealth but low financial empowerment and low education of wives. Although it is not possible to disentangle the exact mechanisms given the data limitations, the treatment effects on attitudes and reported experience of violence seem to stem from different subpopulations. Evidence from administrative crime data suggests that WPS do lead to an increase in reporting of IPV and domestic violence related crimes 2-3 years after opening.

Abstract: In many countries, students choose between academic and vocational tracks at the end of comprehensive schooling, a decision with lasting consequences for educational and labor market trajectories. Using administrative data on the universe of French students, I estimate middle school value-added to academic track enrollment, controlling for baseline achievement, demographics and primary school characteristics. Moving up one standard deviation in the value-added distribution raises academic track enrollment by 8 percentage points, around half the conditional gap between the highest and lowest socioeconomic groups. To understand how schools generate these effects, I estimate separate value-added measures for student requests and teacher validation. Both margins exhibit substantial variation: a one standard deviation increase in request value-added raises requests by 6 percentage points, and a one standard deviation increase in validation value-added, estimated among schools that generate similar request behavior, raises validation by 4 to 6 percentage points. Learning gains are positively but moderately correlated with these effects. Schools that direct more students toward the academic track also show lower dropout, lower grade repetition, and more lenient grading, suggesting that a motivational school climate matters beyond academic preparation.

Abstract: This paper studies the impact of granting labour rights to domestic workers on the labour supply of mothers. To do so, I exploit a 2012 Spanish reform that formalized the household services sector, mandating Social Security contributions and extending minimum wage protection to domestic workers. The reform significantly increased the cost of outsourcing household production and childcare, particularly in provinces where the domestic workers sector was more heavily informal prior to the reform. I exploit this heterogeneous spatial distribution of informality across Spanish provinces as a source of quasi-experimental variation in treatment intensity. Using rich administrative data from the Continuous Sample of Working Lives, I implement a dynamic triple difference strategy that compares trends in labour outcomes between mothers and non-mothers across provinces with different pre-reform levels of informality. Results show that the reform reduced the labour supply of mothers of young children in high-skilled occupations on the extensive margin. On average, mothers in high-informality provinces became 1.6% less likely to be employed relative to the pre-reform baseline, and 9.98% more likely to be unemployed, pointing to a transition from employment to unemployment rather than inactivity. I also find a reduction in the probability of holding a temporary contract, suggesting that the mothers most affected were those with weaker labour market attachment. Taken together, these results suggest that access to affordable informal household services played a meaningful role in sustaining the labour force participation of mothers in high-skilled occupations in Spain.

Abstract: When import prices fall, firms may raise markups due to imperfect pass-through, but lower foreign input costs also induce more firms to start importing, intensifying competition and increasing pass-through. A monopolistic competition model links this extensive margin to pass-through and provides estimable parameters. French data show a 10% rise in cost savings raises importing firms by 3.2%, explaining about 15% of the aggregate price decline.

Abstract: Mandatory human rights due diligence (HRDD) laws are increasingly used to regulate global value chains. This paper examines how such regulation affects firms’ import decisions, focusing on France’s 2017 Duty of Vigilance Law. The law requires large firms to identify, prevent, and mitigate human rights risks throughout their supply chains. We combine French administrative data with a measure of human rights risk based on the U.S. Trafficking Victims Protection Reauthorization Act. Our identification strategy exploits variation in firms’ pre-law exposure to human rights risks. Our results indicate that covered firms at risk were already reducing imports from high‑risk markets prior to the law, and this trend continues thereafter. This suggests that changes along the intensive margin reflect pre-existing adjustments rather than a discrete regulatory effect. In contrast, the probability of importing from a risky market declines starting in 2018, indicating a potential policy effect. We further show that this adjustment spills over to other risky products imported from the same origin country and is not accompanied by diversion toward non-risky countries.

Abstract: Prosociality is a fundamental skill in human societies and is linked to many long-run positive economic outcomes. Many factors shape prosociality during early childhood - a key stage in skill development - but the interactions among those factors remain understudied. In this paper, we study how ethnic identity and parental influence are associated with social preferences. To do so, we elicit social preferences across 1,000 parent-child pairs from both ethnic majority and minorities in Bulgaria using survey- and task-based items. Our results indicate that prosocial behavior and in-group preferences are common in our context and are transmitted between generations. We also find that in-group preferences are driven by a relatively small number of individuals across groups. Notably, the intergenerational transmission process differs by ethnic group: contact with the out-group - in ethnic Roma households - and positive out-group stereotypes - in ethnic Bulgarian households - appear to be distinct mechanisms in that process.

Abstract: This paper examines how parental consent requirements and access frictions jointly shape teenage reproductive decisions. Exploiting the Spanish 2015 reform that mandated parental consent for 16–17-year-olds, together with Spanish administrative microdata on all registered abortions and births, I find that the reform led to declines in both abortions and pregnancies among affected teenagers. Consistent with a two-stage decision framework, most of the reduction in abortions operates through a decrease in pregnancies, indicating behavioral responses before pregnancy. A simple model of teenage abortion decisions is used to interpret these findings and to clarify how legal and access barriers interact. Using data on proximity to abortion centers and local religiosity, I show how these access frictions operate in the context of parental consent requirements. Where travel costs are high, parental involvement is effectively required even in the absence of formal consent laws, limiting the impact of the reform. When parental consent does bind, local norms shape the margin of adjustment: in more traditional municipalities, the reform primarily affects abortion decisions conditional on pregnancy.

Abstract: This paper examines a large quasi-exogenous expansion of Brazil’s flagship conditional cash transfer program, Bolsa-Família, which increased beneficiary enrollment by roughly 13% in affected municipalities. While prior research often finds that social spending strengthens incumbent support, we show that the expansion boosted voter turnout but did not produce lasting gains in incumbent vote share. Evidence of short-term electoral returns appears only in the presidential election immediately following the expansion, with no significant effects in subsequent presidential, gubernatorial, or mayoral contests. We find no indication that political polarization explains the weak long-run response. To test the asymmetry of these effects, we analyze a 2013 eligibility review that removed over 700,000 families from the program. These removals are correlated with higher rates of blank voting and reduced support for the incumbent party, suggesting that losses in social benefits may carry sharper electoral consequences than equivalent gains.

Abstract: Far-right voting is stigmatized, yet rising worldwide. Do signals of its newfound popularity embolden voters to support the far-right, even in the secrecy of the voting booth? French two-round elections provide a natural experiment to answer this question. When far-right candidates narrowly rank first in round one---a salient but purely symbolic victory---this brings them more votes in round two, held merely one week later. At its peak, this effect explains 88% of far-right round two victories in narrowly treated districts. Evidence aligns with voters attending more to salient signals when they update second-order beliefs about the far-right's acceptability. As predicted if voters attach greater weight to more salient signals, more unusual wins have larger effects. Leveraging a large corpus of newspaper articles I scraped, I further show that these unusual wins trigger a spike in media coverage. Consistent with stigma erosion, the vote effect is specific to the far-right, larger where stigma is likely stronger, and persists in the long-term. Using administrative records on campaign funding, I document similar patterns for campaign donations---acts of support that are less secret than votes and may therefore carry larger reputational costs. Lastly, I show that these wins normalized the open expression of anti-immigration views.

Abstract: Expiring budgets often lead organizations to spend money on low-value projects at the end of the fiscal year. Allowing unused funds to carry over (“rollover”) reduces this spending pressure but creates accumulated balances that may be politically or institutionally costly. This paper studies a decentralized alternative: permitting units to transfer unused funds to other units before budgets expire. A simple model shows that transfers weakly dominate expiring budgets and can outperform rollover when accumulated balances are costly. A laboratory experiment documents large reductions in low-value spending, recovering most rollover efficiency gains. Survey evidence indicates that such mechanisms are institutionally feasible.

Abstract: Quantile treatment effects (QTE's) provide a natural framework for capturing heterogeneity in causal effects across different parts of the outcome distribution. This paper develops an estimator for semiparametric conditional quantile models with endogenous and continuous treatment variables. Identification is achieved within a structural triangular equations model using a control function approach. Estimation proceeds via spline series methods. The proposed framework delivers estimates of the entire QTE process and functionals thereof like treatment effects on tail- or interquantile expectations. To allow for inference on these objects, I formally derive the uniform asymptotic properties (across quantile levels) of the estimator for the semiparametric QTE's. The finite-sample performance of the estimator is assessed through Monte Carlo simulations. In an empirical illustration, I revisit the original U.S. Census data and compulsory schooling instrument of Angrist and Krueger (1991) to estimate heterogeneous returns to education across the wage distribution.

Abstract: Climate change is expected to intensify the frequency and severity of droughts in developing countries, where rural livelihoods remain highly dependent on rainfed agriculture. Yet the political consequences of such shocks are theoretically ambiguous: dry spells may heighten concern for environmental protection, but they may also depress rural incomes and thereby reduce support for parties advocating more stringent – and potentially costly – environmental policies. This paper examines how growing-season droughts shape voting behavior in Brazil, a large developing democracy characterized by strong climate vulnerability, deep agrarian inequalities, and polarized environmental politics. Combining high-resolution weather data with a panel of ideology- and environment-weighted vote shares for all presidential elections between 2002 and 2022, I exploit within-municipality variation in hydrological stress aligned with crop-specific growing seasons. I find that drier growing seasons systematically reduce support for green and left-leaning coalitions. Exploration of potential mechanisms reveals that these effects operate through lower agricultural revenue per hectare, with off-season droughts having little explanatory power. Heterogeneity analyses reveal stronger electoral penalties in smallholder-dominated areas and substantially weaker reactions in agribusiness regions, with social protection and rural organization mitigating these effects in smallholder areas. Taken together, the results indicate that in rural Brazil, the adverse economic impacts of drought dominate any countervailing rise in environmental concern, ultimately shifting votes away from environmental and redistributive platforms.

Abstract: When, how, and why do campaign donors respond to political scandals? This paper studies donor reallocation using a stacked difference-in-differences design applied to federal campaign contribution records (2000–2024). Exploiting within-donor variation across politician portfolios, it documents responses on both the extensive and intensive margins, with substantial heterogeneity across scandal types and donor characteristics.

Abstract: How do changes in the administrative hierarchy of cities impact their development? This paper focuses on the loss of regional capital status, using the context of the 1999 administrative reform in Poland. Exploiting variation in administrative status, I compare former regional capitals to control cities to construct a causal estimate of the loss of capital status. I find that treated cities experienced a persistent decline in both public and private sector activity, female employment, fertility, and local public good provision, despite receiving higher central government transfers relative to control cities. These results are consistent with a simple theoretical model that shows path-dependence in labor markets and "sticky feet" in migration: a decrease in administrative capacity does not lead to immediate adjustments in employment and migration. The findings highlight that administrative status is crucial for city-level development and that the loss of such status has negative consequences, even when accompanied by increased fiscal autonomy.

Abstract: Public procurement is increasingly used as a public policy instrument, especially with respect to environmental objectives. Since green public procurement stimulates innovation and contributes to the development of sustainable production processes, it is crucial to determine whether it is affected by natural disasters. This paper assesses the resilience of green public procurement to environmental hazards for the first time thanks to original data on municipal procurement in Japan between 2012 and 2020. The identification is based on a difference-in-differences design considering earthquakes as exogenous treatments. The paper identifies a negative impact of earthquakes on municipal green purchasing, particularly in sectors where sustainable practices are already less prevalent. The evidence tends to reveal that post-disaster repair expenditures crowd-out the implementation of sustainable procurement practices. In the context of the increasing frequency and intensity of natural disasters, these results are of significant interest, as they highlight how natural hazards can force a trade-off between reconstruction needs and local environmental governance.

Abstract: Corporate Social Responsibility (CSR) refers to firm-level initiatives that improve environmental performance, social outcomes, or governance and now underpins a global asset market exceeding $30 trillion. I argue that an important, understudied driver of CSR’s rise is that sustainability has become a valued job attribute, turning CSR into a strategic tool for attracting and retaining talent. To quantify this mechanism, I develop a labor market model in which workers’ utility depends on firms’ CSR choices and estimate it using a matched employer–employee dataset covering all French firms combined with detailed CSR metrics. Importantly, my framework accounts for unobserved worker and firm heterogeneity in productivity, mitigating biases in simple wage comparisons. Two-way fixed effect estimates show that a deviation increase in a firm’s sustainability score is associated with a 2.5 to 8% reduction in hourly wages. The estimated relationship varies significantly over time and across firms, suggesting heterogeneous incentives to engage in CSR. This also implies that policies aimed at stimulating CSR investment can affect wage-setting and workforce composition, and should be designed taking firm labour-related motives into account.

Abstract: In developing countries, missing credit markets can push workers towards jobs which facilitate consumption smoothing at the cost of lower returns. One such arrangement is the pay-in-advance contract, under which workers receive an upfront sum before starting work, later repaid out of earnings. This paper shows that workers choose high-advance contracts over alternatives offering lower upfront pay but higher piece rates and therefore future earnings, particularly if subject to liquidity constraints. However, cognitive biases distort this intertemporal trade-off: respondents significantly underestimate the negative effects of both debt repayments and lower piece rates on future earnings, causing them to overvalue contracts offering large upfront payments. I measure the impact of advances on interest in work opportunities and distinguish between these rational versus behavioural drivers of demand through an experiment with seasonal brick kiln migrants in India, an industry where advances are commonplace. I advertise real vacancies at randomly assigned contracts and collect an incentivized measure of interest in applying. When presented with a choice, 30% prefer a high advance contract compared to only 17% preferring low advance contracts attached to a piece rate. Preferences for advances are even stronger among those without alternative borrowing opportunities. However, respondents consistently fail to sufficiently adjust their estimated future earnings upwards when presented with contracts attached to higher piece rates and lower advances. This suggests that advances may aid consumption smoothing, but at a cost obscured by cognitive limitations. Shifting labour supply towards higher paying but advance-free jobs may therefore depend on both relieving credit constraints and numeracy programmes.