2004年-世界发展银行全球_Poverty_Alleviation_through____________Geographic_Targeting__How_Much_Does_Disaggregation_Help__42页_301kb

Summary of "Poverty Alleviation through Geographic Targeting: How Much Does Disaggregation Help?"

Core Content

This paper investigates the effectiveness of geographic targeting in poverty alleviation using recently developed poverty maps for Cambodia, Ecuador, and Madagascar. The authors simulate the impact of distributing a given budget to different geographic sub-groups based on their relative poverty status, comparing the results to a uniform transfer scheme. The goal is to assess how much more effective targeting can be when using fine geographic data, and whether combining such data with other targeting mechanisms could further improve poverty reduction outcomes.

Main Points

• Poverty Maps: These are used to estimate consumption or income-based welfare outcomes at the local level, using regression models from household survey data. The estimates are imputed to the population census and aggregated into geographic regions.
• Targeting Schemes: The study compares various targeting strategies, including:
  • A benchmark uniform transfer scheme, where the budget is distributed equally to all households.
  • A naive targeting scheme, which ranks geographic areas by estimated poverty and transfers the same amount to all households in the poorest areas.
  • An optimal geographic targeting scheme, which minimizes the squared poverty gap by equalizing the poverty gap across regions, subject to a budget constraint.
• Performance Measure: The squared poverty gap is used to assess the impact of targeting schemes. This measure is sensitive to both the number of people below the poverty line and the distance between their income and the poverty line.
• Key Findings:
  • Geographic targeting improves poverty reduction, especially when using more detailed data (e.g., village or district-level estimates).
  • The gains from geographic targeting are less significant when the targeting scheme uses only crude local poverty estimates.
  • The effectiveness of geographic targeting is lower in countries with high poverty lines or large budgets.
  • Despite the gains from geographic targeting, the inability to target households directly means that targeting performance is still far from perfect.
  • Combining geographic targeting with other mechanisms, such as individual or household-level means-testing, may further enhance poverty reduction outcomes.

Key Information

• Countries Studied: Cambodia, Ecuador, and Madagascar, which are geographically and socio-politically diverse.
• Methodology:
  • First-stage estimation: Per-capita consumption is predicted using household survey data, with a focus on demographic, occupational, and educational variables.
  • Simulation approach: The authors simulate the distribution of a budget across different geographic units, incorporating uncertainty in poverty estimates.
  • Error components: Prediction errors are composed of:
    • Idiosyncratic error: Deviations due to unobserved household-specific factors.
    • Model error: Variance in the first-stage parameter estimates.
    • Computation error: Errors from the simulation process, which can be minimized by increasing the number of simulation draws.
• Challenges and Caveats:
  • Geographic targeting may not account for local political-economy dynamics, which could influence the distribution of transfers within communities.
  • The presence of community-specific public goods may alter the effectiveness of targeting.
  • Behavioral responses (e.g., migration or misrepresentation) could undermine the accuracy of geographic targeting.
  • Administrative costs may increase with more detailed targeting, potentially reducing the overall amount available for transfers.
  • The assumption that geographic targeting implies a willingness to sacrifice horizontal equality is a critical limitation of the analysis.

Implications for Policy

• The results suggest that geographic targeting can significantly improve poverty reduction when based on detailed poverty maps.
• However, perfect targeting remains elusive, and policymakers must consider the trade-offs between targeting efficiency and equity.
• Combining geographic targeting with other methods (such as means-testing or self-selection mechanisms) may yield better results.
• The study emphasizes that geographic targeting should not be seen as a standalone solution, but rather as a tool that can complement other targeting approaches.

Conclusion

While geographic targeting using poverty maps can lead to substantial improvements in poverty reduction, it is not a perfect solution. The effectiveness of such targeting depends on the level of disaggregation, the accuracy of poverty estimates, and the context in which the transfers are implemented. Policymakers should evaluate the potential benefits and costs of geographic targeting in the context of local conditions and consider integrating it with other targeting mechanisms to achieve more efficient and equitable poverty alleviation.