I am a researcher at NOVAFRICA, Nova SBE. I am also a research affiliate of the LEO, Orléans University, and an invited lecturer at Paris 1 Panthéon-Sorbonne University where I defended my PhD.

My research is in Development Economics and Political Economy. It relies on applied micro-econometric approaches to deal with topics related to inequalities, violence, and natural resources.




Don't Touch my Road. Evidence from India on Affirmative Action and Everyday Discrimination. World Development, 2018

This article investigates whether affirmative action, in the form of electoral quotas, affects group-based discrimination. The redistributive effect of quotas is subject to debate, and their ultimate target is discrimination. To identify the effect of electoral quotas, I take advantage of their rotation across space and over time in India. To proxy discrimination, I use a measure of caste-based exclusion from a public infrastructure (namely, streets). In 2006, 44.5% household members of the marginalized castes labeled Scheduled Castes (SC) still suffered from caste-based exclusions. I document that ongoing SC quotas reduce the likelihood of caste-based exclusion for members of the SCs by about one fifth. The results also imply that the effect is not persistent: it disappears with the end of the SC quota. From a policy-maker’s perspective, these results are mixed since electoral quotas do affect everyday discrimination, even if the effect does not last. These results are consistent with a temporary change in the behavior of members of the dominant castes after a one-shot electoral quota. These results are inconsistent with either a change in the stereotypes held by members of the dominant castes, or a change in the aspirations of members of the lower castes. > Download paper or > WP

Media coverage:, Ideas4India

Does inequality matter for conflicts? Evidence from religious conflicts in India. Revue d'économie politique, 2017

This article documents the link between conflicts and income inequality, and shows the benefits of dis-aggregating inequality indexes by groups. More precisely, it tests the link between three inequality measures and religious conflicts of low intensity in Indian villages. The three inequality measures are the overall (or vertical) inequality, and its decomposition for religious groups, into a between group component (horizontal) and a within group component. The results show that the two components of aggregated inequality have opposite links with conflicts. Indeed, on the one hand, within-group inequality appears to increase the incidence of conflicts between religious groups. On the other hand, when between-group inequality plays a significant part, it decreases the incidence of conflicts. Income distribution is thus a correlate of conflict incidence, even for low intensity conflicts. Moreover, the documented empirical relation is consistent with the most recent theories, which underline the importance of the mobilization constraint for conflict incidence. On the contrary, results are inconsistent with an interpretation of conflicts as triggered by greed or grievance. > Download paper

The Spread of Rational Choice in the Social Sciences: A Sign of Strength or Weakness? On Rafael Wittek, Tom A.B. Snijders, Victor Nee (eds), The Handbook of Rational Choice Social Research, with Camille Chaserant and Antoine Pietri, Revue française de sociologie, 2016

The Handbook of Rational Choice Social Research provides an overview of the rational choice paradigm’s achievements. This review article shows how the rational choice approach has gradually become a paradigm by overcoming sociology’s traditional divisions in three ways. > Download paper

Working papers

The gold digger and the machine. Evidence on the distributive effect of the artisanal and industrial gold rushes in Burkina Faso, with Rémi Bazillier NOVAFRICA Working Paper, 2019 - R&R Journal of Development Economics

This paper uses a quasi-natural experiment, the recent gold boom in Burkina Faso, to document the local impact of two alternative mining techniques: artisanal and industrial mines. Artisanal mines employ 90% workers in the extractive sector to produce 20% of minerals, are a traditional activity managed as a common, and often compete for land with modern industrial mines. Our identification strategy exploits two sources of variation. The spatial variation comes from the exposure of households to different geological endowments, and the temporal variation comes from changes in the global gold price. We are the first to document the economic impact of artisanal mines. We show that a 1% increase in the gold price increases consumption by 0.15% for households neighboring artisanal mines. Opening an industrial mine, in contrast, has no impact on local consumption. > Download paper

Media coverage: World Bank blog, RES media brief, GDN, The Conversation, L'OBS

Stabbed in the back. Is sabotage taking place after mandated political representation? LEO Working Paper, 2017 - R&R Social Choice and Welfare

This article documents that the beneficiaries of an affirmative action policy may at the same time suffer of spiteful sabotage, triggered by the implementation of affirmative action. I exploit the timing of the implementation of caste quotas across states of India to document the effect of these quotas on crimes against members of the lower castes. The implementation of caste quotas appears to increase both caste-based crimes and murders targeting members of the lower castes. I complement these administrative crime statistics with a nationally representative household survey. Households’ answers reveal an increase in inter-caste tensions during caste quotas. Results from both the administrative data and the survey data are consistent with sabotage targeting members of the lower castes. These results are the first country-wide evidence that an affirmative action policy may induce sabotage. Moreover, I document that the increase in crime is tied to the principle of existence of the quota rather than to its implementation modalities. (part of this paper circulated as a UNU WIDER paper in 2016, under the title Mandated political representation and crimes) > Download paper

Video summary : UNU-WIDER talk

work in progress

Rent sharing and conflicts. Evidence from Kazakhstan, with Gerhard Toews and Alma Kudebayeva, available upon request

Between 2001-2011 the oil rich districts of Kazakhstan experienced a surge in violent labor conflicts. We take advantage of time, sectoral and spatial variation to identify the impact of the recent oil boom on individuals’ economic conditions and their satisfaction with income to explain the surge in conflicts. We start by showing that workers in the oil rich districts are economically better off, in absolute and in relative terms. We proceed by showing that reported real household income and the employment status of the household head remain unaffected by changes in the oil price. Nevertheless, we show that income satisfaction reported by household heads employed in the private sector decreases during the boom (while remaining unaffected for those outside the private sector). We suggest that the drop in satisfaction is due to the increase in worker’s aspirations to benefit from the rents received by profit maximizing companies in the private sector during the boom. We conclude that the drop in satisfaction with income during the boom provided a fertile ground for violent labor conflicts and, thus, should be kept in mind by policy makers during a resource boom.

Natural resources and ethnic identity, with Nicolas Berman and Matthieu Couttenier

We investigate how households feelings evolve toward a more ethnic or national identity in response to a local economic shock: the exploitation of natural ressources.

The local health effect of artisanal mines

We use a quasi-natural experiment, the recent gold boom in Burkina Faso, to document the local health impact of artisanal mines.

Ethnicity and Conflict: A Microeconomic Empirical Study , available upon request

This paper relates conflict intensities in Indian villages to measures of social divisions, using a specification based on Esteban and Ray (2011 AER). Esteban and Ray (2011 AER) show that conflict intensity can be modeled as a weighted average of the Polarization, Fractionalization, and Greenberg-Gini indexes. Esteban, Mayoral, and Ray (2012 AER) provide empirical support to that theory for civil conflicts incidence. This paper provides further empirical support, with data that is closer to the theory, notably because the data provides (1) a detailed conflict record and (2) an innovative way to measure between-group distance. Results are consistent with village level conflicts being triggered by jealousy rather than within-group altruism.