2009年-世界发展银行全球_The_Health_Impact_of_Extreme_Weather_Events_in_Sub-Saharan_Africa_34页_649kb

The Health Impact of Extreme Weather Events in Sub-Saharan Africa

Core Content

This paper investigates the health impact of extreme weather events in Sub-Saharan Africa, focusing on their effects on children under the age of three. It uses a panel dataset combining Demographic and Health Surveys (DHS) and climate data from the Africa Rainfall and Temperature Evaluation System (ARTES) to analyze the relationship between extreme weather and health outcomes such as diarrhea, malnutrition, and infant mortality.

Main Findings

• Extreme Weather Events: The study quantifies the impact of extreme rainfall and temperature events on child health.
• Health Outcomes: Diarrhea and weight-for-height malnutrition are found to be significantly affected by extreme weather events, while long-term health indicators like height-for-age malnutrition and under-five mortality rates show less sensitivity.
• Climate Change Impact: The projected health cost of increased diarrhea due to climate change in 2020 is estimated to be between 0.2 to 0.5 percent of GDP in Africa.
• Sub-National Analysis: The use of sub-national data allows for a more detailed understanding of health impacts, as these data reveal strong regional variations in health outcomes and environmental conditions.
• Socioeconomic Factors: Socioeconomic variables such as access to water, sanitation, and health services are included to assess how these factors moderate the health effects of extreme weather events.

Key Variables and Data

• Dependent Variables:

  • Diarrhea: Proportion of children under three affected by diarrhea in the survey period.
  • Malnutrition: Measured by weight-for-height (wasting) and height-for-age (stunting) indicators.
  • Infant Mortality: Under-five mortality rate, which is difficult to link directly with climate variables.

• Explanatory Variables:

  • Weather Variables: Defined by monthly rainfall and temperature thresholds. Excess rainfall is identified as 20% above the 20-year average in the wet season, and extreme temperatures are based on the highest and lowest six months of the year.
  • Climate Zones: Classified using the Köppen-Geiger system, with most regions falling into the equatorial savannah with dry winter (Aw), warm temperate with dry winter and hot summer (Cwa), and steppe or desert (Bsh) categories.
  • Socioeconomic Variables: Include access to piped water, toilet facilities, education of adult females, and health care access, measured through principal components analysis.

Methodology

• Model Specification: A reduced form model is used to test the statistical relationship between health outcomes and extreme weather events, controlling for socioeconomic factors.
• Estimation Methods: Both Ordinary Least Squares (OLS) and panel data estimators are applied. The Hausman test determines whether a fixed effects (FE) or random effects (RE) model is more appropriate. The RE model is preferred for its ability to predict health impacts under different climate zone assumptions.
• Data Sources:
  • DHS Data: From 108 regions in 19 Sub-Saharan African countries between 1992 and 2001.
  • ARTES Data: Provides rainfall and temperature data from 1980 to 2001.

Policy Implications

• The findings highlight the importance of sub-national data in understanding health impacts of climate change.
• Public health interventions targeting access to water, sanitation, and health services can mitigate the adverse effects of extreme weather events.
• The study provides a statistical basis for more precise health projections in conjunction with localized climate models.

Conclusion

The paper emphasizes that extreme weather events, particularly excess rainfall and extreme temperatures, have a significant short-term impact on child health in Sub-Saharan Africa. These impacts are more pronounced on indicators like diarrhea and weight-for-height malnutrition. The use of a panel dataset enables a better understanding of the independent effects of weather shocks, and the results can be used to simulate future health burdens and guide policy interventions.