Do Global Financial Ties Make Countries More Resilient to Natural Disasters?

The study asks whether countries that are more connected to global financial markets recover more quickly from natural disasters. The authors use quarterly data from 61 advanced and emerging economies between 1970 and 2018, including detailed information on floods, storms, earthquakes, and other sudden disasters. They compare economic outcomes—GDP, consumption, and investment—between countries with high versus low levels of cross-border financial assets and liabilities. They find that highly connected advanced economies experience stronger recoveries: GDP is about 2 percentage points higher and investment roughly 4–5 points higher within two years. Emerging economies show weaker and less consistent gains.

The Policy Scientist's Perspective

This article addresses a broadly significant policy question: how global financial ties shape countries’ ability to withstand increasingly frequent and costly natural disasters. As climate-related shocks intensify, understanding the structural drivers of economic resilience is essential, and this study contributes timely evidence. The dataset used in the paper is unusually rich, spanning nearly five decades and 61 countries, which enhances both empirical depth and cross-national generalizability. Although the methods rely on dynamic panel regressions, the identification strategy is transparent and grounded in exogenous shock variation. The article extends prior work on international risk sharing and offers a meaningful contribution to the literature.

Full Citation and Link to Article

Bremus, F., & Rieth, M. (2024). Integrating out natural disaster shocks. American Economic Journal: Economic Policy, 16(1), 1-38. https://doi.org/10.1257/pol.20220801

Central Research Question

The article asks whether a country’s degree of international financial integration alters the way it recovers from sudden, unexpected natural disasters. The authors focus on whether economies with larger cross-border asset and liability positions experience systematically stronger rebounds in GDP, consumption, and investment following exogenous shocks such as floods, storms, earthquakes, and landslides. The authors also examine whether these responses differ between advanced economies and emerging markets and whether particular types of external positions—such as external debt assets versus external debt liabilities—play distinct roles in shaping recovery dynamics. More broadly, the research investigates whether global financial linkages provide a meaningful channel for international risk sharing in the context of climate-driven shocks that are growing more frequent and costly.

Previous Literature

The article contributes to several connected strands of research. First, the international risk-sharing literature has long documented gaps between theoretical and empirical predictions: although financial globalization should, in theory, smooth idiosyncratic country-specific shocks, empirical evidence often finds limited or uneven benefits. Prior studies show that consumption risk-sharing is substantially stronger in advanced economies than in emerging markets, and that the composition of cross-border positions—equity, long-term debt, short-term debt, or interbank exposures—matters for the degree of smoothing achieved. A second literature focuses on the macroeconomic consequences of natural disasters. These studies generally find that disasters destroy capital, depress short-run output, and raise prices, although long-term effects vary across contexts. A third group of studies examines how financial institutions respond to disasters using micro-level data, identifying mechanisms such as credit reallocations or balance-sheet constraints. However, these micro-oriented studies do not provide aggregate, cross-country evidence on recovery trajectories. Finally, Ramcharan’s work using natural disasters as sources of exogenous variation to study exchange rate regimes provides the methodological foundation for using disasters as instruments for real shocks. The present article integrates these literatures by examining whether international financial integration—rather than banking sector structure or macro-policy regime—conditions the country-level recovery path after externally generated shocks. It thereby isolates a specific mechanism (global financial positions) while leveraging the inherent exogeneity of natural disasters.

Data

The authors construct a large quarterly panel dataset covering 61 advanced and emerging economies from 1970Q1 to 2018Q4. The dataset integrates several major international sources. Natural disaster shocks are drawn from EM-DAT, which provides event-level information on the timing, type, and estimated economic damage associated with disasters that meet established reporting thresholds. To ensure comparability across countries, the authors standardize damage by nominal GDP one year prior to the disaster and weight damage by the onset month to account for the fact that events early in the quarter have a larger contemporaneous effect on economic activity. They restrict attention to sudden-onset events—floods, storms, landslides, and earthquakes—and winsorize extreme values at the 97.5th percentile. The resulting dataset includes 526 non-zero disaster events. Macroeconomic variables (GDP, private consumption, and private investment) are drawn from the OECD and IMF national accounts, supplemented by national statistics for missing values. Financial integration measures come from the External Wealth of Nations database, which provides detailed information on external asset and liability positions, including portfolio equity, portfolio debt, FDI, other investment, and foreign reserves. Additional variables include GDP per capita (1995 baseline), institutional quality indicators, and measures of capital account openness. Importantly, the distribution of shocks is balanced across high-integration and low-integration states, and across advanced and emerging economies, which supports internal validity. The long time horizon and global coverage enable the authors to identify patterns that are not idiosyncratic to specific countries or specific decades.

Methods

The empirical strategy relies on dynamic panel regressions that exploit the exogeneity of natural disasters to identify causal effects. The core specification is an autoregressive distributed-lag model that relates changes in GDP, consumption, or investment to the interaction between disaster shocks and indicators for high versus low financial integration. Financial integration is operationalized as above-median external asset positions, external liability positions, or total external positions relative to GDP. The model includes lags of the dependent variable, country fixed effects, year fixed effects, and quarter-of-year dummies to absorb global and seasonal variation. Country fixed effects control for time-invariant characteristics such as geographic exposure or long-run institutional factors. By comparing the dynamic responses of economies in high-integration states to those in low-integration states, the model isolates how integration conditions recovery trajectories. The authors conduct extensive robustness checks: alternative shock definitions (e.g., top-10-percent disasters, alternative weighting schemes, no winsorization), alternative integration measures (continuous values, country-specific trends, time-invariant states), and alternative estimators (local projections, GLS estimators, and specifications with additional lags). They also test assumptions regarding strict exogeneity using Granger-causality tests and show that disaster timing is not predictable based on prior economic conditions. Collectively, the empirical design aims to isolate the causal relationship between external financial positions and post-disaster macroeconomic dynamics.

Findings/Size Effects

The results show that financial integration significantly strengthens the post-disaster recovery of advanced economies. When an advanced economy is in a state of high financial integration, GDP is roughly two percentage points higher two years after a disaster than when the same economy is in a low-integration state. Private consumption rises by approximately two to three percentage points relative to low-integration states, while private investment exhibits even larger differences, often exceeding four to six percentage points in cumulative terms over eight quarters. These effects occur alongside more favorable current-account adjustments, indicating that advanced economies draw on foreign income flows to finance reconstruction and maintain consumption. When decomposing integration by instrument, external debt assets—such as foreign bond holdings or reserves—are most strongly associated with stabilization, consistent with their steady income streams. External liabilities also help stabilize output, likely because they allow domestic losses to be shared with foreign investors. Emerging markets show more limited and less statistically consistent benefits. Although consumption responses are sometimes stronger under high integration, standard errors are wide, and investment and output typically do not show statistically significant differences. The authors identify a sharp asymmetry between debt assets and debt liabilities in emerging markets: debt assets buffer shocks, while debt liabilities amplify them. This asymmetry likely contributes to the muted aggregate effects of integration in this group. The overall pattern suggests that financial integration facilitates international risk sharing for advanced economies, but institutional quality and portfolio composition matter greatly for emerging markets.

Conclusion

The article demonstrates that international financial integration plays a meaningful role in shaping recovery dynamics following natural disasters, especially for advanced economies. By combining long-run quarterly data with plausibly exogenous shocks, the authors provide evidence that highly integrated economies are better able to smooth the macroeconomic effects of severe weather-related events. This finding is increasingly relevant as climate change raises the frequency and scale of natural disasters. The study also highlights important heterogeneity: advanced economies benefit consistently, while emerging markets do so only under specific conditions—strong institutions, favorable external asset composition, and limited reliance on external debt liabilities. The paper thereby advances the literature by establishing a causal and policy-relevant mechanism linking global financial positions to macroeconomic resilience.