## Section V — Concluding Remarks and Policy Implications

## Source details

**Canonical URL:** [Section V — Concluding Remarks and Policy Implications](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023220-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023220-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023220-print-pdf.pdf.json)

---

### Data and Key Measures
- Panel dataset: 135 countries over the period 1980–2020.
- Dependent variable: number of deaths per population due to natural disasters (EM-DAT).
  - EM-DAT: data on the occurrence and effects of over 22,000 large-scale natural disasters from 1900 to the present.
- Natural disaster categories used: Geophysical, Meteorological, Hydrological, Climatological, Biological.
- Main explanatory variable: corruption (ICRG corruption index).
  - Original ICRG scale: 0 (highest potential risk) to 6 (lowest potential risk).
  - Index inverted so the new variable increases with the degree of corruption.
  - ICRG coverage starts in 1984; the 1984 value is used for 1980–1983 to maximize sample size.
- Controls: real GDP per capita, trade openness (exports + imports as share of GDP), urbanization (share urban population), hospital beds per population, bureaucratic quality, democratic accountability, government stability.
- Estimation method: country-fixed effects, time-fixed effects, Driscoll-Kraay (1998) robust standard errors.

### Descriptive Statistics (selected)
- Natural disaster-related deaths per population — Observations: 7,944; Mean: 1.33; Std. dev.: 1.30; Minimum: 0.02; Maximum: 2,331.5
- Corruption — Observations: 4,737; Mean: 3.0; Std. dev.: 1.4; Minimum: 0.0; Maximum: 6.0
- Real GDP per capita — Observations: 7,126; Mean: 11,023; Std. dev.: 16,130; Minimum: 166; Maximum: 114,048
- Trade openness — Observations: 6,512; Mean: 83.8; Std. dev.: 52.6; Minimum: 0.0; Maximum: 442.6
- Urbanization — Observations: 7,913; Mean: 53.3; Std. dev.: 23.9; Minimum: 4.3; Maximum: 100.0
- Healthcare (hospital beds) — Observations: 3,415; Mean: 4.5; Std. dev.: 3.4; Minimum: 0.1; Maximum: 19.9
- Bureaucratic quality — Observations: 4,800; Mean: 2.2; Std. dev.: 1.2; Minimum: 0.0; Maximum: 4.0
- Democratic accountability — Observations: 4,800; Mean: 3.9; Std. dev.: 1.6; Minimum: 0.0; Maximum: 6.0
- Government stability — Observations: 4,800; Mean: 7.4; Std. dev.: 2.0; Minimum: 0.0; Maximum: 12.0

### Baseline Econometric Specification
- Baseline reduced-form model (log deaths per population, winsorized at 5th and 95th percentiles):
  - Deaths_it = β1 + β2 Corrupt_it−1 + β3 X_it−1 + η_i + μ_t + ε_it
- Corruption and all explanatory variables are lagged one period.
- Controls include logs of real GDP per capita and hospital beds, trade openness, urbanization, and institutional indices (rescaled where noted).

### Main Empirical Findings
- Corruption is positively and significantly associated with higher natural disaster-related fatalities:
  - Baseline estimates (Table 2): Corruption coefficients across specifications — 0.0203***, 0.0207**, 0.0204**, 0.0205** (standard errors: (0.007), (0.008), (0.008), (0.008)).
  - Interpreted as: a 1 percent increase in corruption leads to an increase of about 2.1 percent in the number of deaths per population caused by natural disaster, after controls.
  - Difference between least and most corrupt countries in sample implies a sixfold increase in deaths per population in a given year.
- Control variable findings (Table 2):
  - Real GDP per capita: coefficients -0.0675, -0.0718, -0.0988, -0.0939 (not statistically significant across specifications in Table 2).
  - Trade openness: coefficients 0.0323, 0.0158, 0.0135, 0.0039 (not statistically significant in baseline specifications).
  - Urbanization: coefficients -0.3405, -0.6330*, -0.6978*, -0.6220* (strong and statistically significant negative relationship with deaths per population).
  - Healthcare (hospital beds): coefficients -0.0695**, -0.0877*, -0.0897*, -0.0877* (strong and statistically significant negative relationship).
  - Institutional measures (e.g., bureaucratic quality coefficient -0.0714* in Table 2): generally associated with declines in deaths per population, with varying statistical significance.

### Heterogeneity and Robustness
- By income group (Table 3):
  - Full sample ALL — Corruption: 0.0207** (0.008).
  - Advanced economies (AE) — Corruption: 0.0190 (0.014) — not statistically significant.
  - Emerging markets (EM) — Corruption: 0.0267** (0.011).
  - Low-income countries (LIC) — Corruption: 0.9558*** (0.000).
  - Interpretation: Corruption has no significant effect in advanced economies but a statistically highly significant effect in developing countries; in developing countries a 1 percent increase in corruption leads to an increase of almost 2.5 percent in deaths per population (as reported in the text).
- By level of corruption (Table 4):
  - All countries, High corruption: Corruption 0.0147* (0.008).
  - All countries, Low corruption: Corruption 0.0209 (0.015).
  - Emerging markets, High corruption: Corruption 0.0225** (0.010).
  - Emerging markets, Low corruption: Corruption 0.0001 (0.018).
  - Finding: impact of corruption on disaster deaths is nonlinear and increases with the level of corruption; deleterious effects are especially pronounced in developing countries.

### Conclusions and Policy Implications
- Core conclusions:
  - Widespread corruption increases the number of disaster-related deaths, conditional on economic, demographic, healthcare and institutional controls.
  - The effect is stronger in developing countries than in advanced economies.
  - Nonlinearities: higher levels of corruption produce disproportionately larger increases in fatalities from natural disasters.
- Policy implications emphasized in concluding remarks:
  - Strengthen institutions and promote anti-corruption measures to reduce humanitarian losses from natural disasters.
  - Improve quality of buildings and infrastructure, and bolster health and risk-management systems, particularly in developing countries where weak infrastructure and widespread corruption magnify disaster fatalities.
  - Enhance transparency in governance and appropriate use of public resources to mitigate the human cost of natural disasters.

*Source: wpiea2023220-print-pdf — Section V, concluding remarks and policy implications, as provided in the supplied content.*

### Section V offers concluding remarks including policy implications.

### Section V — Concluding Remarks and Policy Implications

### Data and Key Measures
- Panel dataset: 135 countries over the period 1980–2020.
- Dependent variable: number of deaths per population due to natural disasters (EM-DAT).
  - EM-DAT: data on the occurrence and effects of over 22,000 large-scale natural disasters from 1900 to the present.
- Natural disaster categories used: Geophysical, Meteorological, Hydrological, Climatological, Biological.
- Main explanatory variable: corruption (ICRG corruption index).
  - Original ICRG scale: 0 (highest potential risk) to 6 (lowest potential risk).
  - Index inverted so the new variable increases with the degree of corruption.
  - ICRG coverage starts in 1984; the 1984 value is used for 1980–1983 to maximize sample size.
- Controls: real GDP per capita, trade openness (exports + imports as share of GDP), urbanization (share urban population), hospital beds per population, bureaucratic quality, democratic accountability, government stability.
- Estimation method: country-fixed effects, time-fixed effects, Driscoll-Kraay (1998) robust standard errors.

### Descriptive Statistics (Table 1)
- Natural disaster-related deaths per population — Observations: 7,944; Mean: 1.33; Std. dev.: 1.30; Minimum: 0.02; Maximum: 2,331.5
- Corruption — Observations: 4,737; Mean: 3.0; Std. dev.: 1.4; Minimum: 0.0; Maximum: 6.0
- Real GDP per capita — Observations: 7,126; Mean: 11,023; Std. dev.: 16,130; Minimum: 166; Maximum: 114,048
- Trade openness — Observations: 6,512; Mean: 83.8; Std. dev.: 52.6; Minimum: 0.0; Maximum: 442.6
- Urbanization — Observations: 7,913; Mean: 53.3; Std. dev.: 23.9; Minimum: 4.3; Maximum: 100.0
- Healthcare (hospital beds) — Observations: 3,415; Mean: 4.5; Std. dev.: 3.4; Minimum: 0.1; Maximum: 19.9
- Bureaucratic quality — Observations: 4,800; Mean: 2.2; Std. dev.: 1.2; Minimum: 0.0; Maximum: 4.0
- Democratic accountability — Observations: 4,800; Mean: 3.9; Std. dev.: 1.6; Minimum: 0.0; Maximum: 6.0
- Government stability — Observations: 4,800; Mean: 7.4; Std. dev.: 2.0; Minimum: 0.0; Maximum: 12.0

### Baseline Econometric Specification
- Baseline reduced-form model (log deaths per population, winsorized at 5th and 95th percentiles):
  - Deaths_it = β1 + β2 Corrupt_it−1 + β3 X_it−1 + η_i + μ_t + ε_it
  - Corruption and all explanatory variables are lagged one period.
  - Controls include logs of real GDP per capita and hospital beds, trade openness, urbanization, and institutional indices (rescaled where noted).

### Main Empirical Findings
- Corruption is positively and significantly associated with higher natural disaster-related fatalities:
  - Baseline estimates (Table 2): Corruption coefficients across specifications — 0.0203***, 0.0207**, 0.0204**, 0.0205** (standard errors: (0.007), (0.008), (0.008), (0.008)).
  - Interpreted as: a 1 percent increase in corruption leads to an increase of about 2.1 percent in the number of deaths per population caused by natural disaster, after controls.
  - Difference between least and most corrupt countries in sample implies a sixfold increase in deaths per population in a given year.
- Control variables:
  - Real GDP per capita: inversely correlated with disaster deaths; coefficient larger in magnitude than corruption but statistically insignificant across specifications in Table 2 (Income coefficients: -0.0675, -0.0718, -0.0988, -0.0939; standard errors noted in table).
  - Trade openness: not statistically significant in baseline specifications (Openness coefficients: 0.0323, 0.0158, 0.0135, 0.0039).
  - Urbanization: strong and statistically significant negative relationship with deaths per population (Urbanization coefficients: -0.3405, -0.6330*, -0.6978*, -0.6220*; standard errors noted).
  - Healthcare (hospital beds): strong and statistically significant negative relationship (Healthcare coefficients: -0.0695**, -0.0877*, -0.0897*, -0.0877*).
  - Institutional measures (bureaucratic quality, democratic accountability, government stability): generally associated with declines in deaths per population, with varying statistical significance (e.g., bureaucratic quality coefficient -0.0714* in Table 2).

### Heterogeneity and Robustness
- By income group (Table 3):
  - Full sample ALL — Corruption: 0.0207** (0.008).
  - Advanced economies (AE) — Corruption: 0.0190 (0.014) — not statistically significant.
  - Emerging markets (EM) — Corruption: 0.0267** (0.011).
  - Low-income countries (LIC) — Corruption: 0.9558*** (0.000) — reported coefficient and standard error as in source.
  - Interpretation: Corruption has no significant effect in advanced economies but a statistically highly significant effect in developing countries; in developing countries a 1 percent increase in corruption leads to an increase of almost 2.5 percent in deaths per population (as reported in the text).
- By level of corruption (Table 4):
  - All countries, High corruption: Corruption 0.0147* (0.008).
  - All countries, Low corruption: Corruption 0.0209 (0.015).
  - Emerging markets, High corruption: Corruption 0.0225** (0.010).
  - Emerging markets, Low corruption: Corruption 0.0001 (0.018).
  - Finding: impact of corruption on disaster deaths is nonlinear and increases with the level of corruption; deleterious effects are especially pronounced in developing countries.

### Conclusions and Policy Implications
- Core conclusions:
  - Widespread corruption increases the number of disaster-related deaths, conditional on economic, demographic, healthcare and institutional controls.
  - The effect is stronger in developing countries than in advanced economies.
  - Nonlinearities: higher levels of corruption produce disproportionately larger increases in fatalities from natural disasters.
- Policy implications emphasized in concluding remarks:
  - Strengthen institutions and promote anti-corruption measures to reduce humanitarian losses from natural disasters.
  - Improve quality of buildings and infrastructure, and bolster health and risk-management systems, particularly in developing countries where weak infrastructure and widespread corruption magnify disaster fatalities.
  - Enhance transparency in governance and appropriate use of public resources to mitigate the human cost of natural disasters.

*Source: wpiea2023220-print-pdf — Section V, concluding remarks and policy implications, as provided in the supplied content.*

### REFERENCES

### REFERENCES

### Works on corruption and economic growth, governance, and public sector
- Aidt, T., J. Dutta, and V. Sena (2008). “Governance Regimes, Corruption and Growth: Theory and Evidence,” Journal of Comparative Economics, Vol. 36, pp. 195–220.  
- Alesina, A., and B. Weder (2002). “Do Corrupt Governments Receive Less Foreign Aid?” American Economic Review, Vol. 92, pp. 1126–1137.  
- Brinkerhoff, D. (2008). “Building Governance and Anticorruption in Disaster-Prone Countries: Policy Lessons from Nepal.” Public Administration and Development, Vol. 28, pp. 327–341.  
- D’Agostino, G., J. Dunne, and L. Pieroni (2016). “Government Spending, Corruption, and Economic Growth,” World Development, Vol. 84, pp. 190–205.  
- Gründler, K., and N. Potrafke (2019). “Corruption and Economic Growth: New Empirical Evidence,” ifo Working Papers No. 309 (Munich: Leibniz Institute for Economic research at the University of Munich).  
- Gupta, S., H. Davoodi, and E. Tiongson (2002). “Corruption and the Provision of Healthcare and Education Services,” In G. T. Abed and S. Gupta (Eds.), Governance, Corruption, and Economic Performance (Washington, DC: International Monetary Fund).  
- Habib, M., and L. Zurawicki (2002). “Corruption and Foreign Direct Investment,” Journal of International Business Studies, Vol. 33, pp. 291–307.  
- Heywood, P. (2007). “The Ethical Foundations of Public Administration,” Public Administration Review, Vol. 67, pp. 96–102.  
- Hodge, A., S. Shankar, P. Rao, and A. Duhs (2011). “Exploring the Links Between Corruption and Growth,” Review of Development Economics, Vol. 15, pp. 474–490.  
- Klomp, J., and J. de Haan (2013). “Corruption and the Effectiveness of Foreign Aid,” European Journal of Political Economy, Vol. 29, pp. 84–101.  
- Lehoucq, F., and J. Molinas (2002). “Stuffing the Ballot Box: Fraud, Electoral Reform, and Democratization in Costa Rica,” Comparative Political Studies, Vol. 35, pp. 749–775.  
- Machado, J., and J. Santos Silva (2019). “Quantiles via Moments,“ Journal of Econometrics, Vol. 213, pp. 145–173.  
- Mauro, P. (1995). “Corruption and Growth,” Quarterly Journal of Economics, Vol. 110, pp. 681–712.  
- Mo, P. (2001). “Corruption and Economic Growth,” Journal of Comparative Economics, Vol. 29, pp. 66–79.  
- Rose-Ackerman, S. (2006). International Handbook on the Economics of Corruption (Cheltenham, UK: Edward Elgar).  
- Shah, A. (2006). “Corruption and Decentralized Public Governance,” World Bank Policy Research Working Paper No. 3824 (Washington, DC: World Bank).  
- Tanzi, V. (1998). “Corruption Around the World: Causes, Consequences, Scopes, and Cures,” IMF Working Paper No. 98/63 (Washington, DC: International Monetary Fund).  
- Uberti, L. (2022). “Corruption and Growth: Historical Evidence, 1790–2010,” Journal of Comparative Economics, Vol. 50, pp. 321–349.  
- Alam, M., and H. Vennemo (2014). “Corruption and Catastrophic Disaster: A Review,” Ecological Economics, Vol. 108, pp. 49–58.  
- Chang, C., and Y. Hao (2018). “Environmental Performance, Corruption and Economic Growth: Global Evidence Using a New Dataset,” Applied Economics, Vol. 49, pp. 498–514.  
- Cieślik, A., and L. Gozcek (2018). “Control of Corruption, International Investment, and Economic Growth: Evidence from Panel Data,” World Development, Vol. 103, pp. 323–335.  
- Farzanegan, M., and S. Witthuhn (2017). “Corruption and Political Stability: Does the Youth Bulge Matter?” European Journal of Political Economy, Vol. 49, pp. 47–70.  
- Huang, C. (2016). “Is Corruption Bad for Economic Growth? Evidence from Asia-Pacific Countries,” North American Journal of Economics and Finance, Vol. 35, pp. 247–256.  
- Iqbal, M. (2018). “Political Economy of Building Regulations Enforcement: Empirical Insights from Pakistan,” Land Use Policy, Vol. 73, pp. 435–444.  

### Studies on natural disasters, economic losses, and resilience
- Akhtaruzzaman, M. (2011). “Post-Disaster Relief Operations in Bangladesh: Context and Challenges,” Disasters, Vol. 35, pp. 803–818.  
- Albala-Bertrand, J. (1993). Political Economy of Large Natural Disasters (New York: Oxford Univeristy Press).  
- Anbarci, N., M. Escaleras, and C. Register (2005). “Earthquake Fatalities: The Interaction of Nature and Political Economy,” Journal of Public Economics, Vol. 89, pp. 1907–1933.  
- Cavallo, E., S. Galiani, I. Noy, and J. Pantano (2013). “Catastrophic Natural Disasters and Economic Growth,” Review of Economics and Statistics, Vol. 95, pp. 1549–1561.  
- Escaleras, M., N. Anbarci, and C. Register (2007). “Public Sector Corruption and Major Earthquakes: A Potentially Deadly Interaction,” Public Choice, Vol. 132, pp. 209–230.  
- Haque, C. (2003). “Perspectives of Natural Disaster in East and South Asia, and the Pacific Island States: Socioeconomic Correlates and Need Assessment,” Natural Hazards, Vol. 29, pp. 465–483.  
- Kahn, M. (2005). “The Death Toll from Natural Disasters: The Role of Income, Geography, and Institutions,” Review of Economics and Statistics, Vol. 87, pp. 271–284.  
- Kellenberg, D., and A. Mobarak (2008). “Does Rising Income Increase or Decrease Damage Risk from Natural Disasters?” Journal of Urban Economics, Vol. 63, pp. 799–802.  
- Klomp, J. (2016). “Economic Development and Natural Disasters: A Satellite Data Analysis,” Global Environmental Change, Vol. 36, pp. 67–88.  
- Loayza, N., E. Olaberria, J. Rigolini, and L. Christiaensen (2012). “Natural Disasters and Growth: Going Beyond the Averages,” World Development, Vol. 40, pp. 1317–1336.  
- Noy, I. (2009). “The Macroeconomic Consequences of Disasters,” Journal of Development Economics, Vol. 88, pp. 221–231.  
- Padli, J., and M. Habibullah (2009). “Natural Disaster and Socioeconomic Factors in Selected Asian Countries: A Panel Analysis,” Asian Social Science, Vol. 5, pp. 65–71.  
- Raschky, P. (2008). “Institutions and the Losses from Natural Disasters,” Natural Hazards and Earth System Sciences, Vol. 8, pp. 627–634.  
- Schumacher, I., and E. Strobl (2011). “Economic Development and Losses due to Natural Disasters,” Ecological Economics, Vol. 72, pp. 97–105.  
- Skidmore, M., and H. Toya (2007). “Economic Development and the Impact of Natural Disasters,” Economic Letters, Vol. 94, pp. 20–25.  
- Taghizadeh-Hesary, F., N. Yoshino, A. Mortha, and T. Sarker (2019). “Quality Infrastructure and Natural Disaster Resilience,” ADBI Working Paper No. 991 (Manila: Asian Development Bank Institute).  
- Tol, R., and F. Leek (1999). “Economic Analysis of Natural Disasters,” In T. Downing, A. Olsthoorn, and R. Tol (Eds.), Climate Change and Risk (London: Routledge).  

### Methodology, econometrics, and related empirical techniques
- Im, K., M. Pesaran, and Y. Shin (2003). “Testing for Unit Roots in Heterogeneous Panels,” Journal of Econometrics, Vol. 115, pp. 53–74.  
- Koenker, R., and G. Bassett (1978). “Regression Quantiles,” Econometrica, Vol. 46, pp. 33–50.  
- Machado, J., and J. Santos Silva (2019). “Quantiles via Moments,“ Journal of Econometrics, Vol. 213, pp. 145–173.  

*References list from wpiea2023220-print-pdf - REFERENCES*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023220-print-pdf.pdf_