2005年-世界发展银行全球_Microfinance_and_Poverty______________Evidence_using_Panel_Data_from_Bangladesh_24页_189kb

Summary of "Microfinance and Poverty: Evidence Using Panel Data from Bangladesh"

Core Content

This article by Shahidur R. Khandker examines the poverty reduction effects of microfinance in Bangladesh using panel data from household surveys conducted in 1991/92 and 1998/99. It addresses the limitations of previous studies that used cross-sectional data and aims to provide a more accurate assessment of microfinance's impact on both individual and aggregate poverty levels.

Main Points

• Microfinance Focus: Microfinance primarily supports informal activities with low returns and market demand. It is designed to help small- and medium-scale producers and businesses, particularly the poor and women.

• Poverty Impact: The study finds that microfinance contributes to poverty reduction, especially for female participants. It also suggests that microfinance has a positive effect on the local economy, particularly at the village level.

• Policy Relevance: Understanding the extent of poverty reduction through microfinance is crucial for policy decisions. The question is whether microfinance's benefits are sustainable and large enough to justify its continued support over alternative poverty reduction initiatives.

• Gender Disparity: Women are disproportionately affected by poverty and constitute the majority of microfinance beneficiaries. Microfinance helps women gain assets and influence household decision-making.

• Limitations of Microfinance: While microfinance can lift some individuals out of poverty, its overall impact on aggregate poverty may be limited due to the small scale of transactions and potential income redistribution rather than growth.

• Methodological Approach: The article uses a household-level fixed-effects model with panel data to estimate the effects of microfinance on consumption and poverty, which allows for the resolution of time-invariant endogeneity issues.

• Endogeneity and Instruments: The study acknowledges the endogeneity of program participation and the potential for measurement error. It suggests that instrumental variable estimation may be necessary to address these issues, though the fixed-effects method is sufficient for some analyses.

Key Information

• Data Source: Panel data from the Bangladesh Institute of Development Studies (BIDS) and the World Bank, covering two survey periods: 1991/92 and 1998/99.

• Sample Size: The study uses data from 1,638 households that were interviewed in both periods. The number of households increased to 2,599 in the 1998/99 survey, but only those interviewed in both periods were included in the analysis.

• Program Participation:

  • In 1991/92, 25.8% of households were program participants, 38.0% eligible nonparticipants, and 36.2% nontarget households.
  • By 1998/99, these proportions changed to 52.7% participants, 20.1% eligible nonparticipants, and 27.3% nontarget households.
  • Over 95% of the 1991/92 participants remained in the program in 1998/99.

• Landholding and Participation:

  • Program participation is higher among households with less land (target households), indicating that microfinance reaches those in need.
  • The proportion of nontarget households participating increased from about 25% in 1991/92 to 31% in 1998/99, suggesting potential mistargeting.

• Poverty Reduction:

  • The marginal impact of microfinance on consumption was estimated to be 18% for women and 11% for men.
  • About 5% of borrowers may lift themselves out of poverty each year, but this could result in less than 1% of the population being lifted out of poverty due to the limited reach of microfinance.

• Spillover Effects:

  • The study investigates whether microfinance benefits nonparticipants through spillover effects.
  • It also assesses the aggregate poverty impact, questioning whether microfinance leads to broader societal benefits or merely redistributes income.

• Econometric Model:

  • The model assumes that household consumption depends on both current and past characteristics, including borrowing.
  • Equations are used to estimate the credit demand function and consumption function, accounting for time-invariant and time-varying unobserved factors.

• Attrition and Bias:

  • The attrition rate between the two survey periods was 7.4%, which is considered low.
  • The study formally tests for attrition bias and finds it can be ignored in most outcomes.

Conclusion

The article concludes that microfinance has a significant impact on poverty reduction at the individual level, particularly for women, and contributes to the local economy. However, its aggregate poverty impact is limited, and further research is needed to assess its long-term sustainability and broader societal effects. The use of panel data provides a more robust method for evaluating microfinance impacts compared to cross-sectional analysis.