Forced to Migrate in a Warming World: Labor Migration Responses to Drought-Induced Shocks in Tunisia, joint with Federica Alfani, Minh-Phuong Le, and Mohamed Ali Marouani. Revise and Resubmit at the Journal of Environmental Economics and Management. Working paper. Abstract
This study investigates two key strategies for coping with drought in Tunisia: agricultural adaptation and migration, using panel data on net migration rates, agricultural production, and weather conditions at a detailed administrative level. Our results indicate that farmers expand irrigated land to cope with droughts. However, since this is not sufficient, migration becomes a prominent alternative strategy. Indeed, less favorable weather is associated with higher out-migration, particularly among males, less educated individuals, and informal or agricultural workers. Migration, however, is predominantly accessible to wealthier households due to its associated costs. Furthermore, we provide also strong evidence of climate-induced international migration, but primarily to neighboring countries.
Creating Better-Paid Jobs Thanks to Regional Trade Agreements? Evidence from Indonesia, joint with Florian Mayneris, William Parienté, and Gonzague Vannoorenberghe. CEPR Discussion Paper 19899 · PDF · VoxEU column. Revised September 2025. Abstract
The creation of stable, well-paid jobs by private firms is a key issue for low- and middle-income countries. It is not clear whether further trade integration would help to achieve this goal. In this study, we examine the impact of the ASEAN–China Free Trade Agreement on the likelihood of becoming a wage worker in Indonesia, which serves as a proxy for better-paid employment in the Indonesian context. Our findings suggest that reductions in import and export tariffs did not, on average, increase the probability of becoming a wage worker, except among those with more than a primary education. The estimated impact on various subgroups and sectors varies considerably across regions, being stronger in Java. In line with these limited average effects at worker level, we find no significant changes in employment or value added in large and medium-sized manufacturing firms, the main providers of salaried jobs. Our results suggest that the agreement had a regressive impact on access to better-paying jobs.
Firm Resilience and Political Cycles in Pandemic Time, joint with Lisa Chauvet and Mohamed Ali Marouani. SSRN working paper. Abstract
In this article, we explore the relationship between electoral proximity and the impact of the COVID-19 pandemic on firm performance. Using an event study approach based on health restriction measures and election dates across 138 countries, we find that restriction stringency is significantly lower in the months preceding an election compared to the election month or the post-election period. This effect is more pronounced in parliamentary regimes, but it also holds for presidential systems and autocracies. Furthermore, we observe that the volatility of restrictions is higher in the pre-election period, particularly in democratic countries. Second, using an instrumental variable strategy and data from the World Bank Enterprise and COVID-19 Follow-up Surveys for 32 countries, we show that proximity to elections mitigates some of the negative outcomes for firms stemming from the pandemic-induced demand shock.
Good or Bad Timing? The Effects of Productivity Shocks on Education and on Schooling Performance. IRES Discussion Paper 2019/21. Abstract
It is often argued that negative productivity shocks have adverse effects on education in developing countries. In this paper, I assert that positive productivity shocks can also come at the expense of education. I present a theoretical model to predict the mechanisms when shocks occur in early childhood and in school-age. To capture exogenous shocks in productivity, I exploit variation in intensity of climate and prices across location and over time. The empirical part provides evidence that in early childhood, positive productivity shocks have persistent positive consequences on schooling performance. In contrast, the relationship becomes counter-cyclical when children are of school age. Current positive shocks increase child labor, reduce schooling performance, and decrease educational attainment when shocks become recurrent.
From climate shock to resettlement: what are the effects on the socio-economic conditions of the host communities?, joint with Mélody Mourlhion–Ostrowski
Hidden attitudes towards environmental conservation around a protected area in Madagascar, joint with Thomas Calvo, Mireille Razafindrakoto, and François Roubaud.
Mind the Gap: Climate Perception Bias across Africa, joint with Alessandro Giacardi and Thomas Calvo.
Migration Response to Drought in Mali: An Analysis Using Panel Data on Malian Localities over the 1987–2009 Period, joint with Dimitri Defrance and Flore Gubert. Environment and Development Economics, 28(2), 171–190, 2022. Published article
Returns to Farm Child Labor in Tanzania, joint with Pierre André and Christelle Dumas. World Development, 138, 105181, 2021. Published article
The Effect of the Universal Primary Education Program on Consumption and on the Employment Sector: Evidence from Tanzania. World Development, 142, 105345, 2021. Published article · NBER working paper