(γ) ``Internal Conflicts and Shocks: A Narrative Meta-Analysis,'' (2025) with Camille Laville. Journal of Peace Research, 62 (4), p. 1159–1175, first published online in 2024 (SJR: Q1 in Political Science and International Relations, Safety Research, and Sociology & Political Science)
[Direct link] [Online appendix][Earlier APSA WP version][Earlier WBG WP version] [Data and replication files][Data and replication files, alternative link] [Global dev blog post] [IGPSA blog post] [World Bank LTD blog post]
Abstract:
Do variations in local incomes influence peace and conflict in low- and middle-income countries? The present meta-regression analysis contributes to answering this question by delving into the narratives that researchers use to qualify how various shocks affect conflict risk through channels implicitly linked to income. After examining 2,464 subnational estimates from 64 recent empirical studies, we find that several publication biases related to authors’ methodological choices influence our understanding of this phenomenon. Importantly, studies that fail to uncover empirical effects that conform to researchers’ expectations on the theoretical mechanisms are less likely to be published. After accounting for publication selection bias, the analysis finds that, on average, income-increasing shocks in the agriculture sector are negatively associated with the local risk of conflict. Nonetheless, the analysis finds no average effect of income-decreasing shocks in the agriculture sector or income-increasing shocks in the extractive sector on the local risk of conflict. The article opens avenues for further study on the granular observed heterogeneity in the literature, particularly focusing on the conditional aspects of how shocks and conflicts are measured and the geographical coverage, among others.
(β) ``Has Chinese Aid Benefited Recipient Countries? Evidence from a Meta-Regression Analysis,'' (2023) with Martha Tesfaye Woldemichael. World Development, 166 (June), 106211, lead article (SJR 2023: Q1 in Development, Economics, Geography, and Sociology & Political Science)
[Direct link] [Earlier IMF WP version] [Data and replication files] [Brookings Future Development blog post]
Abstract:
This paper employs a meta-regression analysis of 1,149 estimates from 29 studies to take stock of the empirical literature on Chinese aid effectiveness. After accommodating publication selection bias, we find that, on average, Beijing’s foreign assistance is positively associated with economic outcomes in recipient countries, but correlates with deforestation and negative perceptions of China among citizens, albeit negligible in size. We also show that studies that fail to uncover empirical effects that conform to researchers’ expectations are less likely to be submitted or accepted for publication. Differences in the choice of data, estimation method, and authors’ institutional affiliation explain the heterogeneity among Chinese aid effectiveness estimates reported in the literature.
(α) ``Political Budget Cycles: Manipulation by Leaders versus Manipulation by Researchers? Evidence from a Meta‐Regression Analysis,'' (2019) with Antoine Cazals. Journal of Economic Surveys, 33 (1), p. 274-308 (SJR 2019: Q1 in Economics; A previous version won the Vincent & Elinor Ostrom Prize of the best paper, 2016)
[Direct link] [Supplementary material] [Data and replication files]
Abstract:
Despite a long history of research on political budget cycles, their existence and magnitude are still in question. By conducting a systematic analysis of the existing literature, we intend to clarify the debate. Based on data collected from 1037 regressions in 46 studies, our meta‐analysis suggests that little, if any, systematic evidence can be found in the research record that national leaders do manipulate fiscal tools in order to be reelected. However, it is much more clear that researchers selectively report that national leaders do manipulate fiscal tools in order to be reelected. The publication selection bias highlighted has nonetheless been reduced during the past 25 years of research. We also show that the incumbents' strategies differ depending on which tools they use. Finally, the nature and quality of political institutions appear to be the factors which most affect the political budget cycles.
(7) ``Beyond the AI Divide : A Straightforward Approach to Identifying Global and Local Overperformers in AI Preparedness'' (2025). Accepted for publication in Digital Business, 5 (2), 100136 (SJR: Q1 in Business & International Management, and Business, Management & International Accounting [misc.])
[Direct link][Earlier WBG WP version][Data and replication files][Data and replication files, WP version]
Abstract:
This study introduces a straightforward data-driven framework to identify countries that outperform in artificial intelligence (AI) preparedness relative to their economic complexity, utilizing the IMF’s Artificial Intelligence Preparedness Index and a multidimensional Economic Complexity Index derived from trade and research data. Employing weighted least squares, the study estimates expected AIPI scores and classifies countries as global or local overperformers if their observed scores exceed predictions and surpass income-group medians. The analysis identifies 10 high-income global overperformers and 14 local overperformers across middle- and low-income groups, revealing regulation and ethics as universal drivers of overperformance, with digital infrastructure and human capital varying by economic context. Case studies elucidate diverse coordination models—state-led, market-responsive, and distributed innovation—while highlighting transferability constraints due to institutional and historical factors, among others. The replicable methodology provides policymakers and other key actors a robust tool to benchmark AI readiness and design context-specific strategies, addressing the global AI divide. The study opens avenues for future research into refined AI preparedness metrics, alternative identification techniques, and comparative analyses of national innovation systems.
(6) ``Stuck in a Fragility Trap : The Case of the Central African Republic Civil War'' (2025). Joint with Vincent Nossek and Tomi Diderot Sandjong. Defence and Peace Economics, 36 (4), p. 435-468, first published online in 2024 (SJR: Q1 in Social Sciences [misc.], and Q2 in Economics)
[Direct link][Earlier WBG WP version][Data and replication files][World Bank LTD blog post]
Abstract:
This study utilizes the synthetic control method to assess the economic consequences of the ongoing civil war in the Central African Republic since December 2012. Drawing on a donor pool of low-income and lower-middle-income countries, it constructs a synthetic counterfactual to depict the economic trajectory in the absence of conflict. The analysis reveals a significant decline in national gross domestic product (GDP) per capita, estimated between 45.3 percent and 47.8 percent over a decade of conflict, resulting in a cumulative GDP loss of US$29.7 billion to US$32.4 billion (purchasing power parity, PPP, adjusted). Two model specifications are employed, one using pre-treatment outcomes and the other integrating external covariates. Robustness checks support the findings, indicating a minimum 10-year decline of 35.3 percent in GDP per capita. Even considering the 2003 coup, this civil war has the most detrimental economic impact. The analysis remains robust when incorporating GDP data from remote sensing sources. These effects align with the fragility trap concept, portraying one of the highest economic impacts of civil conflict in terms of relative GDP per capita decline.
(5) ``Inequality, Growth Fluctuations, and Employment,'' (2024) with Burcu Hacibedel, Priscilla Muthoora and Nathalie Pouokam. Empirical Economics, 66 (2), p. 587–622 (SJR 2024: Q1 in Mathematics [misc.], and Social Sciences [misc.], and Q2 in Economics, and Statistics and Probability)
[Direct link] [Earlier IMF WP version] [Data and replication files unavailable][World Bank LTD blog post]
Abstract:
This paper analyses how short-term fluctuations in economic growth affect inequality in developing countries. Using a sample of 71 countries, we find that, on average, growth upswings are associated with reductions in inequality. This reduction is, however, largely undone by growth slowdowns. A mediation analysis framework helps identify unemployment, especially youth unemployment, as the main channel through which fluctuations in growth affect the dynamics of inequality. These findings suggest that both the quality of jobs created, and labor market policies are important to ensure that growth outcomes are conducive to inequality reduction.
(4) ``Dutch Disease and the Public Sector: How Natural Resources Can Undermine Competitiveness in Africa ,'' (2022) with James (Jim) Cust and Shanta Devarajan. Journal of African Economies, 31( S1), p. i10–i32 (SJR 2022: Q1 in Development, and Q2 in Economics)
[Direct link] [Pre-print] [Data and replication files][Brookings Future Development blog post]
Abstract:
Slow growth in manufactured and agricultural exports has been attributed to the high share of natural resources in many African economies. Not only does the resource sector draw labour and capital away from other sectors, but also the spending of resource revenues in the domestic economy bids up the price of non-tradable goods, making the tradable sectors less competitive—a phenomenon known as Dutch disease. This paper argues that an important and neglected channel for this effect is through the public sector. Since government receives a large portion of resource revenues, the public-sector booms alongside the resource sector. But the government is largely unaccountable for the spending of these revenues, since they are not raised via taxes on citizens. The result is this money might be spent in inefficient and distortionary ways, undermining competitiveness. One solution may be for government to transfer natural-resource revenues directly to citizens and then tax them to finance public expenditure. The increased accountability might improve the effectiveness of the public sector and therefore the competitiveness of the private sector.
(3) ``Political Cycles: What Does a Meta-Analysis Reveal about?'' (2019) with Antoine Cazals, in the collective handbook Routledge Advances in Applied Financial Econometrics— International Financial Markets, Volume 1, 1st Edition, Chapter 11
Abstract:
Despite a long history of research on political budget cycles, their existence and magnitude are still in question. We conduct a meta-regression analysis of this literature with the intention to clarify the debate. Based on data collected from 1,037 estimates, results suggest that manipulation of national public accounts by political leaders before elections is in average limited at best and over-estimated by researchers. In addition, through a Bayesian model averaging and a multiple meta-regression analyses, we identify some factors, such as country characteristics and methodological choices affecting magnitude of cycles estimated by scholars.
ISBN 9781138060920
(2) ``Give a Fish or Teach Fishing? Partisan Affiliation of U.S. Governors and the Poverty Status of Immigrants,'' (2018) with Sékou Keita. European Journal of Political Economy, 55 (1), p. 65-96 (SJR 2018: Q1 in Economics, and Political Science)
Abstract:
This paper investigates how governors' partisan affiliation affects the poverty status of immigrants to the U.S. To this end, we compare the poverty outcomes of immigrants in states ruled by Democratic governors relative to the outcomes for those in states ruled by Republican governors. We employ a regression discontinuity design using the re-centered Democratic margin of victory as a running variable, to overcome the identification challenge posed by confounding factors. Consistent with the literature on partisan affiliation, we find that immigrants are more likely to get out of poverty in states with Democratic governors than states with Republican governors. Our results are submitted to a variety of robustness checks and sensitivity tests, to assess the validity of the identification strategy, and highlight conditional lame-duck effects. A formal mediation analysis reveals that the empirical results are mediated through better access to the labor market and possibly through higher wages and labor earnings for immigrants. Last but not least, we check for alternative hypotheses and potential detrimental effects for native populations.
Additional info:
We assess the validity of the RD design and look at the alternative story of immigrants’ ``voting with their feet'' by moving between states following Democratic victories. We also check a selection bias in state location for immigrants according to their schooling attainment. Last but not least, we also find a positive effect of Democratic governors both on the poverty reduction of white natives and black natives, despite different labor market outputs for these populations. Hence, our results do not suggest that Democratic governors improve labor market conditions for the whole resident population, as effects are ambiguous for majority groups of white natives. [...] Last but not least, as there is no free lunch, the global fiscal solvency may be worse in states ruled by Democrats.
(1) ``Flexible Fiscal Rules and Countercyclical Fiscal Policy,'' (2017) with Martine Guerguil and René Tapsoba. Journal of Macroeconomics, 52 (1), p. 189-220 (SJR 2017: Q2 in Economics)
[Direct link] [Earlier IMF WP version] [Data and replication files unavailable]
Abstract:
This paper assesses the impact of different types of flexible fiscal rules on the procyclicality of fiscal policy with propensity scores-matching techniques, thus mitigating traditional self-selection problems. It finds that not all fiscal rules have the same impact: the design matters. Specifically, investment-friendly rules reduce the procyclicality of both overall and investment spending. The effect appears stronger in bad times and when the rule is enacted at the national level. The introduction of escape clauses in fiscal rules does not seem to affect the cyclical stance of public spending. The inclusion of cyclical adjustment features in spending rules yields broadly similar results. The results are mixed for cyclically-adjusted budget balance rules: enacting the latter is associated with countercyclical movements in overall spending, but with procyclical changes in investment spending. Structural factors, such as past debt, the level of development, the volatility of terms of trade, natural resources endowment, government stability, and the legal enforcement and monitoring arrangements backing the rule also influence the link between fiscal rules and countercyclicality. The results are robust to a wide set of alternative specifications.
(1) « L' Approche Historique du Développement Comparé : Des Visions Complémentaires ? », (2015) [in French]. Revue d’Économie du Développement 23 (1), p. 97-128 (SJR 2015 : Q4 in Development, and Economics [misc.]; CNRS 2015, section 37 : rang 3; HCERES: B)
Abstract:
Dans cet article nous montrons que la théorie des dotations biogéographiques de Diamond (1997) et que la théorie du renversement de la fortune d’Acemoglu, Johnson et Robinson (2002), loin d’être substituables, sont complémentaires pour expliquer les écarts de développement économique dans le monde, depuis le début de l’holocène, en 11000 av. J.-C., jusqu’à aujourd’hui. Empiriquement, nous montrons la complémentarité de ces deux approches théoriques à l’aide d’un modèle structurel et de l’emploi de l’estimateur des triples moindres carrés, à information complète, sur un échantillon le plus large possible de pays de tout niveau de développement et sur un échantillon d’anciennes colonies Européennes. Ces résultats, relativement optimistes, confirment que les institutions agissent sur le niveau de développement. Par ailleurs ces résultats insistent sur l’aspect complémentaire de théories du développement qui sont trop souvent jugées concurrentes, à première vue.
(iii) ``Ballots and Guns: Is Compliance with Swiss Military Service Correlated with Extensive Use of Initiative Rights?'' (2025). Joint with Clément Anne. CERDI Études et Documents, 2025 #05
Abstract:
This article investigates the potential correlation between compliance with military service and the extensive use of initiative rights in Switzerland. While political rights are often believed to influence civic duties, this study examines whether such a relationship exists in the Swiss context, where both direct democracy and a citizen army are prominent. Based on multivariate regressions using panel data from 26 cantons between 2010 and 2022, we do not find significant correlations between the extensive use of initiative rights and effective fitness for military service once linguistic and cultural factors are controlled for. Our expanded analysis reveals that German-speaking cantons consistently show higher military service fitness rates, particularly those that integrated earlier into the Old Swiss Confederacy. Additionally, we find that religious composition and Swiss citizenship percentage are strongly associated with military service compliance. These findings suggest that civic duties, specifically military service obligations, are closely linked to cultural heterogeneities related to the tradition of classical republicanism rather than extended political rights per se.
(ii) ``Why Some Countries Can Escape the Fiscal Pro-Cyclicality Trap and Others Cannot?'' (2019). Joint with Santiago Herrera and Wilfried Anicet Kouamé. WBG WPS 8963
[Direct link] [Brookings Future Development blog post][World Bank Africa Can End Poverty blog post]
Abstract:
This paper analyzes the procyclicality of fiscal policy on the tax and spending sides in a sample of 116 developing countries between 2000 and 2016. About 20 percent of the countries in the sample switched from procyclical to countercyclical policy stance. In Sub-Saharan Africa, 30 of 39 countries remained caught in the procyclicality trap and the region has the highest degree of procyclicality. The Middle East and North Africa region switched from a countercyclical policy stance to a procyclical one over time. The Europe and Central Asia and Latin America and the Caribbean regions significantly reduced the degree of procyclicality. The main economic variables that affect procyclicality are financial depth, tax base variability, and natural resource dependence. In line with the political economy literature, the perception of corruption, social fragmentation, and inequality in resource distribution are positively associated with procyclicality. The findings also show that the quality of fiscal institutions is associated with procyclicality; countries with fiscal rules have smaller procyclical bias, but the effect is not homogeneous; and higher degrees of expenditure rigidity are associated with lower procyclical bias. The study finds asymmetric policy stances along the business cycle, with procyclicality being more pronounced during recessions. Similarly, the political cycle affects procyclicality, as procyclical bias increases in electoral years. From the tax management perspective, procyclical bias is still present, but there are significant changes: most of the political economy variables lose significance; the resource-dependence variable is not significant; external credit availability reduces procyclicality; tax base variability increases procyclical bias; and expenditure rigidity is no longer significant, but fiscal space becomes determinant of procyclical bias.
(i) ``Forms of Democracies and Financial Development'' (2015). Joint with Clément Mathonnat. CERDI Études et Documents, 2015 #23
Abstract:
The political economy of finance literature emphasizes the critical role of political institutions in promoting financial development. Related empirical findings highlight a robust positive effect of democratic regimes on financial development compared to dictatorships. However, no study focused so far on identifying the precise political institutions explaining the financial development enhancing effect of democracies. In this paper, we study the effects of disaggregated political institutions on financial development along three institutional dimensions, namely forms of government, electoral rules and state forms. Using a large panel of 140 countries over 1984-2007, we show that institutional details are of crucial importance, since the positive effect of democracies on financial development clearly depends on the precise institutional dimensions at work, namely: parliamentary governments and, to a lesser extent federal states. Thus, our study contributes to the institutional design debate, by showing that the simple promotion of democratic regimes might not be sufficient to foster financial development.