2002年-世界发展银行全球_Returns_to_Investment_in_Education___A_Further_Update_36页_1mb

Returns to Investment in Education: A Further Update Summary

Core Content

This working paper by George Psacharopoulos and Harry Anthony Patrinos provides an updated analysis of the returns to investment in education, focusing on both private and social rates of return. The paper is part of a broader effort by the World Bank to document the benefits of education investments in Latin America and the Caribbean.

Main Views and Key Information

1. Historical Context and Theoretical Framework

• Returns to education have been estimated since the late 1950s, based on human capital theory.
• Over 40 years of research have established patterns of returns, with several reviews contributing to the understanding of these trends.
• The rise in earnings inequality during the 1980s and 1990s sparked renewed interest in education returns.
• Education is more productive in times of technological volatility, as argued by earlier studies like Nelson and Phelps (1966) and Griliches (1969).

2. Latest Patterns in Returns to Education

• The classic pattern of decreasing returns to education with higher levels of economic development is still observed.
• Private returns to higher education have increased in the updated data set.
• Average global returns to another year of schooling are around 10%.
• Low and middle-income countries have the highest returns to education.
• Latin America and the Caribbean and Sub-Saharan Africa show the highest average returns to schooling.
• High-income OECD countries have lower returns to education.
• Women receive higher returns to their schooling investments, but men benefit more from primary education (20% vs. 13%) and women benefit more from secondary education (18% vs. 14%).

3. Selective Approach to Estimating Returns

• There is criticism of indiscriminate rate of return compilations due to non-comparable data and methodological issues.
• Representative samples are ideal, but often not available.
• Firm-based surveys are highly selective and focus on urban areas, leading to biased estimates.
• Civil service wages do not reflect market wages, which affects the accuracy of return estimates.
• Instrumental variable (IV) estimates based on family background are higher than classical OLS estimates.
• Twins studies and natural experiments provide more reliable estimates, as they control for ability bias.

4. Social Returns to Education

• There is interest in estimating social returns, which include externalities or spillover benefits.
• However, these are difficult to measure and inconclusive in empirical evidence.
• Micro studies and cross-country regressions are used to estimate social returns.
• Heckman and Klenow (1997) suggest that social returns become similar to private returns when technology differences are considered.
• Some studies in Africa show positive externalities in agricultural productivity, but quantifying them remains a challenge.

5. Policy Implications

• Returns to education are now seen as indicators, not just prescriptive tools.
• The OECD has adopted returns to education as a key indicator in its reports.
• Governments and agencies increasingly use return studies to guide education reforms and policy decisions.
• Quasi-experimental studies are used to evaluate specific education programs, such as the Indonesia school building program and India's blackboard project.
• Public policy should consider the productivity and incentives for individual investment in human capital, especially for low-income families.

6. Research Gaps and Future Directions

• A major research gap exists between micro and macro evidence on education returns.
• More research is needed on social benefits of schooling, especially in developing countries.
• Quasi-experimental designs offer greater opportunities for accurate estimation.
• The evidence on externalities is inconclusive, with some studies suggesting negative or very high returns.

Conclusion

• Education investment behaves similarly to physical capital investment, with returns that are measurable and significant.
• The marginal and average returns to education are equated in advanced industrial countries.
• Further research is necessary to understand the social returns and externalities of education, especially in developing economies.
• Policy design should be informed by these returns to ensure equitable and efficient investment in education.