2002年-世界发展银行全球_Relative_Returns_to_Policy_Reform___Evidence_from_Controlled_Cross-Country_Regressions_68页_2mb

Summary of "Relative Returns to Policy Reform: Evidence from Controlled Cross-Country Regressions"

Core Content

This working paper by Alexandre Samy de Castro, Ian Goldin, and Luiz A. Pereira da Silva investigates the relative returns to policy reforms—specifically import liberalization and fiscal policy reforms—in a sample of 54 developing countries between 1980 and 1999. The authors aim to address the dispersion in growth outcomes among countries that implemented reforms, by using a controlled cross-country regression approach that incorporates counterfactual control groups and improved policy measurement.

Main Objectives

• To analyze the impact of policy reforms on GDP growth in developing countries.
• To assess the contingent relationships between policy reforms and growth, considering country size, export profile, and governance.
• To improve the methodology of cross-country studies by using more accurate policy indicators and better timing of reforms.

Key Methodological Innovations

• Counterfactual Control Groups (CGs): The authors construct control groups of countries that did not implement the specific reforms under study, allowing for a more accurate comparison of growth outcomes.
• Improved Policy Instruments: They use Average Tariff Rates (ATR) instead of aggregate trade openness indicators (e.g., the sum of imports and exports to GDP), as ATR is a more precise measure of import protection.
• Timing of Reforms: They focus on the exact date of reform implementation, rather than fixed time intervals, to better capture the before-and-after effects of reforms.
• Differenced Differences (DD) Methodology: The analysis is based on first differences of GDP growth, with differences between reform and control groups as the dependent variable.

Key Findings

Import Liberalization (IL)

• The growth effect of import liberalization is generally smaller when using more accurate measures and timing of reforms compared to previous studies.
• There is evidence of contingent relationships:
  • Countries with a comparative advantage in manufacturing and higher GDP levels experience a greater positive effect of import liberalization on growth.
  • Country size and governance quality influence the marginal cost of reforms. Larger countries and those with worse governance tend to have lower net marginal costs (i.e., less negative impact on growth) from fiscal deterioration.
• Non-tariff barriers (NTBs) are not captured by ATR, which may limit the scope of the analysis.
• The results suggest that while import liberalization has a positive effect on growth, this effect is less significant than previously indicated in cross-sectional studies.

Fiscal Policy Reform

• The negative relationship between government consumption and GDP growth is confirmed.
• Fiscal deterioration in larger countries has a lower net marginal cost on GDP growth.
• Worse governance increases the marginal cost of a given budget deficit.
• The analysis is more complex due to:
  • Endogeneity and multicollinearity in fiscal variables.
  • Less comparable data across countries compared to trade data.

Limitations and Considerations

• Data quality issues: Missing ATR values are filled using linear interpolation, which may introduce measurement errors.
• Sample size: The fiscal analysis is limited to 54 countries due to data constraints.
• Reduced form estimation: The authors estimate reduced form equations, as they do not model the structural channels of policy reforms on growth.
• Cross-sectional heterogeneity: While cross-country differences are considered, industry-level characteristics are not included, limiting the ability to fully identify productivity effects.
• Endogeneity concerns: Variables such as Black Market Premium (BMP) are endogenous to GDP growth and may reflect broader macroeconomic conditions rather than direct effects of trade openness.

Conclusion

• The results suggest that more accurate measurement and timing of policy reforms do not necessarily strengthen the significance of their impact on GDP growth.
• The effects are weaker than previously indicated in cross-sectional studies, indicating the need for caution in drawing policy implications from such relationships.
• The contingent relationships between policy reforms and growth are influenced by country-specific factors such as size, export composition, and governance.
• The study highlights the importance of using counterfactual control groups and better policy indicators in future cross-country policy evaluations.

Policy Implications

• The policy implications derived from the relationship between reforms and growth should be interpreted with caution.
• Structural reforms may have non-uniform impacts, depending on country-specific characteristics.
• Governance and economic structure play a crucial role in determining the effectiveness of policy reforms on growth.

Summary of Key Variables

• Growth variables: GDP growth rates, investment rates, human capital (education).
• Policy variables: Import tariffs, government consumption, fiscal deficit, money growth.
• Governance variables: Rule of Law, regulatory framework, institutional quality.
• External variables: Terms of Trade, exchange rate, debt-to-exports ratio.

References and Further Reading

• The authors reference Wacziarg (1998) and Rodríguez and Rodrik (2000) for their theoretical and methodological foundations.
• They also reference Ravallion (2002) and Rosenbaum (2002) for the DD methodology and non-random assignment in policy evaluation.