Research

This paper investigates how hurricanes affect U.S. bank stability and the financial transmission of physical climate risk. We construct a novel gridpoint hurricane exposure dataset (1994–2023) based on IBTrACS storm tracks and physically grounded wind-field modeling combining Willoughby (2006) radial profiles and Chen (1994) asymmetry corrections. From 15-minute interpolated wind fields at 9km resolution, we derive maximum sustained winds, power dissipation indices, and duration above Saffir–Simpson thresholds, matched to geolocated bank branch networks.  We merge these exposure measures with granular U.S. bank balance-sheet data to quantify the causal impact of hurricanes on insolvency risk, credit quality, and profitability. Our empirical strategy relies on a Difference-in-Differences framework comparing banks differentially exposed to hurricane shocks before and after landfall, controlling for bank and time fixed effects. This design exploits the quasi-exogenous spatial variation in storm paths to identify the financial effects of extreme weather.

This project investigates how climate risk—specifically flash drought episodes—affects U.S. banking stability through the agricultural sector, and whether federal crop insurance mitigates these effects. As droughts become more frequent and severe under climate change, farms face heightened yield volatility and income uncertainty, which may impair loan repayment capacity and increase banks’ credit risk. We conceptualize agriculture as the key transmission channel linking physical climate shocks to financial instability.  To test this mechanism, we construct a novel county-year dataset combining flash drought indicators, detailed bank-level stability measures, and granular crop insurance data from the Federal Crop Insurance Program (FCIP). We develop coverage indices capturing the share of insured agricultural acreage and assess whether insurance dampens the impact of droughts on non-performing loans and bank risk indicators. By integrating physical climate measures with financial and insurance data, the project provides new evidence on how public risk-sharing schemes can enhance financial resilience in a warming economy.

Climate change increasingly manifests through short-lived but intense shocks rather than gradual shifts in long-run averages. This paper studies the financial stability implications of flash droughts—rapid-onset episodes of severe soil moisture depletion that develop within weeks and often escape standard drought monitoring systems. While droughts are among the costliest climate hazards worldwide, their effects on banks at the institutional level remain poorly understood. We combine high-resolution meteorological data with granular U.S. banking data covering more than three decades, over 10,000 banks, and tens of thousands of branches. Flash drought exposure is identified using soil-moisture-based indicators and aggregated to the bank level using deposit-weighted branch exposure. Exploiting the plausibly exogenous and localized nature of flash droughts, we estimate their dynamic effects on bank stability using local projections. We find that flash droughts significantly and persistently reduce bank stability, operating primarily through declines in profitability and a gradual buildup of credit risk, with no evidence of a liquidity channel. Effects are strongly seasonal and markedly larger for spring and summer droughts. These results identify flash droughts as a distinct and underappreciated source of physical climate risk for the banking system.

This study investigates the impact of climate shocks on U.S. bank stability, using the El Niño–Southern Oscillation (ENSO) as a natural experiment. As the most significant annual climate variation on Earth, ENSO provides an ideal setting to examine exogenous climate shocks. Combining over 800,000 bank-quarter observations from 1994 to 2023 with state-of-the-art ENSO signal and teleconnection metrics, we document that La Niña events—more than El Niño phases—exert substantial destabilizing effects on U.S. banks. Our baseline estimates suggest that strong La Niña shocks reduce distance to default by 20%. This increased vulnerability under La Niña conditions is further evidenced by increased credit risk—reflected by a 17% rise in non-performing loans and a 40% increase in foreclosure rates—reduced bank profitability, and lower solvency. Dynamic panel local projections reveal that these disruptions operate primarily through banks’ loan portfolios and persist over multiple quarters, peaking around 7 to 11 quarters after the shock. The adverse effects are most severe for banks with high real estate exposure, limited geographic diversification, and large balance sheets. Our findings underscore the role of local temperature anomalies as key drivers of bank instability and highlight the importance of integrating granular climate risk into prudential policy framework, especially as climate change is expected to intensify ENSO patterns.

We study the dynamic effects of floods on local housing markets using comprehensive transaction-level data for France over the period 2010–2024. Combining near-universe real estate transactions with administrative records of officially recognized flood events, we estimate how housing prices and market activity respond to climate-related shocks at the municipality level. Our empirical strategy addresses the key challenge posed by heterogeneous housing markets and transaction-level recomposition effects. We first estimate a hedonic price model to construct a quality-adjusted measure of housing prices that is invariant to changes in the composition of properties sold. We then embed this outcome in a local projection framework to trace the dynamic causal response of housing prices and transaction volumes to flood shocks. We document economically significant and persistent effects of floods on local housing markets, with substantial heterogeneity across property types and depending on whether floods are anticipated or unexpected. Accounting for municipalities’ historical exposure to flooding further reveals evidence of adaptation and learning. Our findings highlight the importance of climate risk for real estate valuation, local public finance, and the spatial distribution of climate vulnerability.

This study introduces a novel ex ante approach to assess the short-term impact of climate shocks on banking sector stability by examining the effect of El Niño Southern Oscillation (ENSO) on banking sector distance-to-default. Using dynamic panel data econometric modeling, we investigate the macroeconomic implications of ENSO-induced climate shocks, such as El Niño and La Niña events, on banking sector stability in 51 countries across three regions particularly exposed to the consequences of ENSO oscillations (East Asia and Pacific, Latin America and the Caribbean, and Sub-Saharan Africa) during the period 2000–2020. Our findings show that the adverse effects of these climate shocks on banking sector stability are unevenly distributed among countries, with more pronounced and robust adverse effects of El Niño events in the short-term, particularly in Latin America and the Caribbean and, to a lesser extent, Sub-Saharan Africa. We also document the short-term adverse effects of La Niña events for Latin American and the Caribbean countries. Further estimates suggest that the increase in non-performing loans is a key transmission channel linking El Niño events to banking sector stability. As global warming should intensify the frequency and magnitude of ENSO's cyclical pattern, these findings can help estimate the potential adverse effects of climate change-related natural disasters on banking sector stability and inform future mitigation and adaptation policies. 

Private credit and public debt levels are drifting to uncharted waters. Are such high levels concerning? This paper suggests bridging private and public borrowing to study their interaction. This investigation relies on an array of econometric techniques and builds on disaggregated data on debt recipients covering 1970-2018. Overall, this paper demonstrates that the separate narratives that have emerged regarding private credit and public debt with output growth conceal meaningful and substantial interactions. In particular, evidence shows that private credit exerts a robust adverse effect on subsequent growth both in the short-run and long-run. The expansion of household credit and the triggering of financial crises account for most of this effect. This study also shows that it is not the level of public debt but rather its trajectory that affects economic growth. Public debt tends to be a drag on growth, not initially, but in the aftermath of a crisis if policymakers are unable to pursue a counter-cyclical fiscal policy. From a policy perspective, economies should closely monitor private credit expansions and maintain some fiscal space in normal times to buffer the fall-out from financial crises.

Since the global financial crisis of 2008, a strand of the literature has documented a threshold beyond which financial development tends to affect growth adversely. The evidence, however, rests heavily on internal instrument identification strategies, whose reliability has received surprisingly little attention so far in the finance-growth literature. Therefore, the present paper conducts a reappraisal of the non-linear conclusion twofold. First, in light of new data, second, by a thorough assessment of the identification strategy.  Evidence points out that a series of unaddressed issues affecting the System GMM setup results in spurious threshold regressions and overfitting of outliers. 

Looking back at half a century of financial expansion together with a host of financial crises, one can wonder if it has done more harm than good? No straight answer has emerged. The finance-growth literature and the banking crisis literature have left many researchers with conflicting and contradictory findings. Undoubtedly, there are both pros and cons to financial development, giving support to investigating the prevailing outcome of financial deepening over time. The present paper thereby contributes to analyzing the finance-growth nexus by providing new estimates supporting an overall damaging influence of financial deepening on economic growth. The evidence suggests that the negative influence of financial depth is stronger in high-income countries, with possibly growth-enhancing effects at the early stages of development. 

Local public investment is a key driver of infrastructure provision and long-term growth. Hence, understanding investment financing choice is paramount. Unlike central governments, most local authorities face a narrow choice between debt and tax funding, subject to fiscal rules and financial market access constraints. To analyze this choice, we transpose Williamson (1988)’s corporate finance framework, which views financing as a governance decision aimed at minimizing transaction costs rather than as market neutral or an agency conflict. In this adaptation to public finance, debt corresponds to rules-based, market governance, while tax funding represents hierarchical, discretionary governance. In stable environments, debt dominates; under instability, the ranking reverses. We test these propositions using France’s local public debt crisis that began in 2010, triggered by municipalities’ exposure to toxic structured loans. Using a panel of 35,000 municipalities (2000–2018) and a distance-based instrumental-variable strategy, we find that before the crisis, structured loans did not alter investment behavior, but after the crisis, exposed municipalities sharply curtailed investment and re-ranked their financing choices away from borrowing toward higher self-financing. These results highlight the governance nature of local financing decisions and extend Williamson’s corporate finance framework to public finance.

This paper examines the Dexia crisis through the lens of political cycles, which are more deeply rooted at the local than at the national level. The project develops two complementary approaches.  First, we analyze how political contestability may influence financing decisions, in particular the choice to issue structured loans. This perspective allows us to reassess so-called “toxic” loans in light of electoral timing and local political competition.  Second, we investigate how exposure to these structured loans shaped municipal policy choices in the aftermath of the crisis. Beyond the detailed Dexia portfolio of structured loans, we use data from the National Registry of Elected Officials to construct political variables (including contestability and Herfindahl–Hirschman concentration indices), and combine them with local public procurement data to provide a granular analysis of post-crisis municipal decisions.

Faced with the municipal toxic loans crisis, the French state has created a support fund conditioned on local governments abandoning legal proceedings against these structured credits. This article econometrically studies the state’s strategy by retrieving the factors explaining the fund’s assistance to the municipalities. The overhead ratio of these toxic loans stands out as the main factor in granting aid, ahead of the municipalities’ financial situation. This indicator was used as a basis for the legal authorities to rule in favor of the municipalities. The fact that it predicts the fund’s support reflects the state’s strategy to limit litigation. This contribution thus discusses the strategic choices for settling the crisis. 

Ce Focus présente les hypothèses et paramètres qui sous-tendent le calcul des indices d’efficacité des dépenses publiques (EDP) appliqués aux politiques éducatives analysées dans la Note du CAE n° 84. L’EDP, ou Marginal Value of Public Funds en anglais, mesure le bénéfice social généré pour chaque euro net investi par l’État. Cette méthode permet d’évaluer de manière homogène et rigoureuse une grande diversité de politiques publiques, en intégrant les effets différés sur les trajectoires individuelles, les retours fiscaux et les économies budgétaires potentielles. Le présent document explicite l’ensemble des hypothèses empiriques, des paramètres économiques et des choix méthodologiques mobilisés pour construire ces indicateurs. Il vise à garantir la transparence de l’exercice d’évaluation, à éclairer l’interprétation des résultats et à favoriser l’appropriation de l’outil par les décideurs, les chercheurs et le grand public.

Reducing class size is a well-documented policy tool to improve student learning, particularly in primary education. This Focus applies the methodological framework of the Marginal Value of Public Funds (MVPF) to this policy in order to assess its social cost-benefit ratio. The analysis focuses on class splitting in priority education areas, while also considering the potential effects of a general reduction in class size in both primary and secondary education. The MVPF index measures the social return of a public policy by comparing the gains it generates to its net cost for public finances. In the case of class size, it relies on a transmission chain linking a reduction in the number of students per class to improved academic skills, and then to increased future income. The parameters used are drawn from rigorous evaluations based on experimental or quasi-experimental variations, as well as on French administrative data on educational and salary trajectories. The results highlight a strong return on investment for the policy in primary education, where the MVPF index suggests that class splitting is self-financing in the long run. In middle school, the estimated effect on learning is more uncertain, but the index remains well above 1, indicating a net social gain for every euro invested. These results support the relevance of continuing efforts in primary education, while also calling for targeted experiments to better document the effects in secondary education.

This Focus provides an estimate of the effect of improvements in school skills on future earnings. Educational policies are essential to strengthening students’ competencies and enhancing the overall performance of the education system, thereby contributing to the formation of human capital for future generations. In France, education represents the second-largest item of public expenditure, which underscores the importance of carefully assessing its effectiveness.  Among the many benefits of education, its impact on individuals’ future earnings is particularly important, given the close relationship between academic skills and labor market outcomes. However, measuring this effect is challenging, notably because of the time lag between the implementation of educational policies and their impact on professional trajectories.  Using results from the 1995 national assessments of sixth-grade students, we analyze their influence on entry wages through a Mincer earnings equation. Although rarely documented in France, this estimate is crucial for guiding public investment in education. It also contributes to calibrating the Marginal Value of Public Funds (MVPF) indicator, enabling an evaluation of the impact of educational policies on beneficiaries’ future earnings. Our results show that a one standard deviation increase in school skills is associated with an approximately 10% rise in future wages, a magnitude comparable to findings in the international literature. Despite some methodological limitations, particularly related to restricted wage tracking over time, this approach provides a first estimate of the economic return to education in France.

This Focus presents a tool for evaluating public policies known as the Marginal Value of Public Funds (MVPF). This tool measures the social benefits generated per net euro invested in a given policy. It is thus a useful metric to inform public choices: by enabling the comparison of different policies accor- ding to the value they create for society, the MVPF provides a rigorous framework for prioritizing resource allocation. As an illustration, the analysis focuses on the case of the Sourdun boarding school of excellence, a program aimed at offering better schooling conditions to youth from disadvantaged backgrounds.

The European Union must deepen and integrate its capital markets to finance strategic priorities such as the green transition, artificial intelligence, and economic resilience to financial shocks. Stronger market-based finance is essential to revive long-term growth and counter Europe’s declining growth potential. Yet the EU financial system remains largely bank-based and fragmented along national lines, limiting cross-border investment and risk-sharing.  We propose five measures to advance a growth-oriented Capital Markets Union. First, expand the European Single Access Point (ESAP) to private firms to improve transparency and asset valuation. Second, harmonize insolvency regimes to reduce costs and foster cross-border investment. Third, reform and strengthen ESMA, including streamlined governance and levy-based funding. Fourth, enhance funded pensions and supervisory integration, notably via EIOPA reform, to boost equity investment. Finally, expand public co-financing of venture capital and introduce EU-funded child investment accounts to broaden household participation.

The US Inflation Reduction Act (IRA) entered into force at the start of 2023. Its combined objectives of easing domestic inflation and tackling climate change will take an estimated budget of anything up to $900 billion over nine years, making it one of the largest areas of US government spending. Since the IRA surfaced, EU policymakers have expressed concerns about its impact on the EU and asked whether the EU should be taking similar measures or even seeking to mitigate its effects. The authors of this article, Maxime Fajeau, Associate Professor at the University of Lille and Scientific Advisor at the French Council of Economic Analysis, Niklas Garnadt, Secretary General at the German Council of Economic Experts, Camille Landais, Professor of Economics at London School of Economics and Political Science, Co-Chair of the Franco-German Council of Economic Experts, and Monika Schnitzer, Professor in the Department of Economics at Ludwigs-Maximilians Universität München and Co-Chair of the Franco-German Council of Economic Experts, discuss the macroeconomic implications of the IRA for the EU and recommend actions which the EU could consider in response.

The US Inflation Reduction Act, aimed at promoting the production and adoption of clean energy, came into effect at the beginning of 2023. This column assesses the impact of the Act on the EU. The main conclusion is that the volume of the Act appears small, as will be its aggregate macroeconomic effects, especially for the EU. The EU should focus on rethinking its industrial policy doctrine and, given the sizeable energy price differential with the US, reducing the costs for energy in Europe.

This paper provides grounds to conduct an in-depth assessment of whether the System-GMM estimates meet reliability requirements. After a didactic overview of the System-GMM estimator, this guide spotlights two overlooked issues: instrument proliferation and weak instruments. These concerns often remain unexplored in System-GMM setups because of a lack of standardized tests. Indeed, most empirical applications simply focus on the generic AR(2) and Hansen tests, which are not an absolute shield safeguarding against spurious conclusions. Consequently, this paper presents a host of ready-to-use tools that are both simple and informative, encouraging researchers to open the black box and provide more convincing evidence of the estimates' reliability. 

This paper provides a rigorous treatment of the between dimension in panel data analysis. Using projection operators, it formally decomposes panel data into orthogonal within and between components, showing that the standard OLS estimator can be expressed as a weighted average of these two dimensions. The analysis highlights an important methodological issue: in cross-country regressions, collapsing data into means before applying nonlinear transformations (logs or squares) does not correspond to a strict between estimator. Instead, transformations should be applied prior to averaging. Using cross-country finance–growth data, the paper demonstrates that this distinction materially affects threshold results, with the strict between approach failing to support the widely cited inverted U-shaped relationship between finance and growth.

This paper presents a discrete-time monetary model of exchange rate determination under both flexible and short-run sticky price regimes. Starting from money market equilibrium and uncovered interest parity, the model is log-linearized and solved under naïve and rational expectations. Under flexible prices, the exchange rate depends on current and expected future monetary fundamentals, highlighting its forward-looking nature. The framework is then extended to incorporate short-run price rigidity, generating a two-dimensional dynamic system in prices and exchange rates. Using phase diagram analysis and eigenvalue characterization, the equilibrium is shown to be saddle-path stable. A permanent monetary expansion produces exchange rate overshooting: the currency initially depreciates beyond its long-run level before gradually converging as prices adjust.