2006年-世界发展银行全球_Making_Conditional_Cash_Transfer_Programs_More_Efficient___Designing_for_Maximum_Effect_of_the_Conditionality_29页_322kb

Summary of "Making Conditional Cash Transfer Programs More Efficient: Designing for Maximum Effect of the Conditionality"

Core Content

Conditional cash transfer (CCT) programs are designed to encourage poor parents to invest in their children's human capital by linking financial support to specific behaviors, such as school attendance and health practices. These programs are increasingly used globally, including in Mexico (Progresa/Oportunidades), Brazil (Bolsa Escola), Nicaragua, Jamaica, Bangladesh, and Chile. While effective in improving education and health outcomes, they can be costly and inefficient, particularly when transfers are given to children who would have attended school regardless of the program.

The article by Alain de Janvry and Elisabeth Sadoulet focuses on improving the efficiency of CCT programs by refining targeting and calibration rules. It argues that the key to efficiency lies in designing rules that are simple, observable, and non-manipulatable, while also ensuring that the program's conditions have a significant impact on behavior. The authors use data from the Progresa experiment in Mexico to demonstrate that efficiency gains can be substantial without increasing inequality among poor households.

Main Views

• Dual Objective of CCTs: CCT programs aim to reduce poverty both immediately through cash transfers and long-term through increased human capital investment.
• Efficiency Challenges:
  • Inefficiency 1: Paying for behaviors that children would already perform without the transfer.
  • Inefficiency 2: Offering transfers that are either too high or too low relative to the opportunity cost of the behavior.
• Progresa as a Case Study: The program provides cash transfers to children who meet certain conditions, such as school attendance and health check-ups. It is evaluated for its efficiency in increasing school enrollment.
• Efficiency Gains: The analysis shows that efficiency gains of 29–44% can be achieved without increasing inequality among poor households by optimizing targeting and transfer rules.

Key Information

• Progresa Overview:
  • Introduced in 1997, it targets poor rural children in Mexico.
  • Includes three components: education transfers, basic healthcare, and nutritional supplements.
  • By 2000, it had reached 2.6 million families with a budget of $950 million.
  • The average increase in household income was 22%.
• School Enrollment Impact:
  • Progresa increased the probability of secondary school enrollment by 13 percentage points.
  • The program's effect on primary school continuation was smaller, suggesting it is less efficient in this area.
  • Transfers to primary school children accounted for 55.4% of the education budget, which may be unnecessarily costly.
• Efficiency Leaks:
  • 64% of poor children who graduate from primary school would enter secondary school without a transfer.
  • This indicates that a significant portion of the program's budget is spent on children who are already likely to enroll.
• Budget Constraints and Optimization:
  • The program must be designed to maximize enrollment gains within a fixed budget.
  • An optimal model is defined by equations (1), (2), and (3), which balance the cost of transfers against the benefit of increased enrollment.
  • The optimal conditional transfer and eligibility rules depend on the ratio of a child's baseline enrollment probability to the marginal effect of the transfer.
• Implementation Considerations:
  • Simple and transparent indicators are crucial for effective program implementation.
  • The authors propose using a linear index based on observable characteristics (e.g., age, gender, mother's literacy, household education level, etc.) to simplify the targeting and calibration process.
  • A scoring system similar to those used in other welfare programs is suggested for practical application.

Simulation Results

• The marginal effect of the conditional transfer is high, with a 1.42 percentage point increase in enrollment per $10 of transfer.
• Key factors influencing enrollment include:
  • Child's age (negative effect)
  • Mother's literacy and household education level (positive effect)
  • Number of agricultural workers (negative effect)
  • Total household expenditure (positive effect)
  • Distance to school (negative effect)
• State effects are also significant in determining enrollment outcomes.

Conclusion

• By focusing on the transition from primary to secondary school and using a predictive model, the authors show that efficiency can be improved significantly.
• They argue that the current design of Progresa is inefficient in primary school but more effective in secondary school.
• A more efficient program can be achieved by simplifying the targeting and calibration rules, using observable and verifiable indicators, and ensuring that transfers are calibrated to the opportunity cost of the behavior.
• These improvements can be done without increasing inequality among poor households, making the program both equitable and efficient.