2024-12-16-IMF-国际货币基金组织支持的项目对性别不平等的影响（英）_49页_5mb

Summary of "Effects of IMF-Supported Programs on Gender Inequality"

Core Content

This working paper evaluates the impact of IMF-supported programs on gender inequality by employing Synthetic Control Methods (SCM) to construct credible control groups. The study focuses on over 150 IMF-supported programs implemented between 1994 and 2022, analyzing how these programs affect various gender inequality indicators such as female labor force participation, education parity, maternal mortality, and adolescent fertility.

The paper argues that previous studies on the effects of IMF programs on gender inequality often suffer from contamination of estimates, conflating the effects of economic crises with those of the programs themselves. By using SCM, the authors aim to isolate the effects of the programs from the broader context of economic crises, allowing for a more accurate assessment of whether the programs have a significant impact on gender inequality.

Main Findings

• No systematic impact: The study finds no systematic differences in the evolution of gender inequality indicators between countries that adopted IMF-supported programs and those that did not.
• Statistical insignificance: In 87-90% of the cases, the impact of IMF-supported programs on gender inequality is statistically insignificant compared to the control group.
• Mixed effects: In the remaining cases with statistically significant results, the effects on gender inequality are mixed, with both positive and negative impacts observed.
• Crisis as primary driver: The results suggest that the primary driver of changes in gender inequality is the economic crisis itself, not the IMF-supported programs.
• Robustness: The findings are consistent across gender indicators, countries, income levels, and time periods.

Key Indicators Analyzed

The following gender inequality indicators are analyzed in the study:

• Female labor force participation
• Ratio of female to male labor force participation
• Adolescent fertility rate
• Maternal mortality rate
• Gender parity indices for primary, secondary, and tertiary education

Methodology Overview

Synthetic Control Method (SCM)

• Purpose: To construct custom-tailored control groups that closely match the pre-crisis gender and economic trends of program countries.
• Data Matching: Control groups are created by combining data from non-program countries that share similar economic and gender trends.
• Time Horizon: The study uses a 4-year pre-crisis and 4-year post-crisis period to assess the impact of the programs.
• Bias Correction: The method includes bias correction techniques to enhance the accuracy of the results.
• Inference: The p-values are used to assess the statistical significance of the observed gender impact gaps.

Comparison with Other Methods

• Program dummies: Previous studies often used program dummies (yes/no) to estimate the effects, but these are imprecise and not tailored to individual countries.
• Difference-in-Difference (DiD): This method assumes parallel trends in pre-crisis data, which is often not valid in this context.
• Propensity Score Matching: This method is limited when pre-crisis characteristics differ significantly across countries.

Key Implications

• Need for credible control groups: The study emphasizes the importance of using credible control groups to assess the impact of IMF-supported programs on gender inequality.
• Avoiding generalizations: It cautions against generalizing from individual country experiences or small sample studies due to substantial variability in significant outcomes.
• Policy design matters: The findings suggest that attention to policy design can lead to better outcomes for gender inequality, even in the context of economic crises.

Conclusion

The study concludes that IMF-supported programs do not systematically impact gender inequality, and that economic crises are the primary driver of observed changes. The use of SCM provides a more accurate and nuanced understanding of the effects of these programs on gender outcomes, highlighting the necessity of tailored control groups and robust methodologies in policy evaluation.

Key Information

• Authors: Theo Eicher, Reina Kawai Eskimez, and Monique Newiak
• Date: December 2024
• JEL Classification Numbers: E60; F40, J16; H51, H52; O40
• Keywords: Gender Inequality; Female Labor Force Participation; Education; Adolescent Fertility; Maternal Mortality; IMF-supported programs; Economic Crisis
• Data Sources: The study uses gender and macroeconomic time series data from various countries, focusing on pre- and post-crisis trends.
• Sample Size: Over 150 IMF-supported programs from 1994 to 2022.
• Control Group Construction: Control groups are built using non-program countries with similar economic and gender trends.
• Robustness: The results are robust across multiple indicators, countries, and time periods.

Tables and Figures

• Table 1: Literature review on the impact of IMF-supported programs on gender inequality, including control group methodologies used in previous studies.
• Figures:
  • Figure 1: Female labor force participation rate in the Dominican Republic and its control group.
  • Figure 2: Normalized and bias-corrected data for female labor force participation.
  • Figure 3: Normalized gender impact gap for female labor force participation.
  • Figure 4: Event study of the effects of IMF-supported programs on gender inequality.

Annexes

• Annex A: Country-by-country results on the effects of IMF-supported programs on gender inequality.
• Annex B: Data sources, variables, and codenames used in the study.
• References: Includes citations of prior studies and methodologies used in the evaluation of IMF-supported programs.