2011年-世界发展银行全球_Indonesias_PNPM_Generasi_Program___Final_Impact_Evaluation_Report_168页_3mb

Indonesia's PNPM Generasi Program Final Impact Evaluation Report Summary

Core Content

The PNPM Generasi Program (National Community Empowerment Program – Healthy and Smart Generation) was a large-scale community-driven development initiative in Indonesia, launched in 2007 to improve health and education outcomes in rural areas. The program was implemented in five provinces: West Java, East Java, North Sulawesi, Gorontalo, and Nusa Tenggara Timur (NTT). It combined community block grants with performance incentives to encourage villages to focus on specific health and education targets. The evaluation aimed to assess the program's effectiveness after 30 months of implementation, using a randomized controlled trial (RCT) design.

Main Objectives

• Improve maternal and child health through services such as antenatal care, postnatal care, and child immunizations.
• Increase primary and junior secondary school enrollment and attendance.
• Enhance health and education infrastructure and community participation in service delivery.
• Provide incentives for villages to achieve targets, based on their performance in health and education indicators.

Key Findings

1. Program Impact on Targeted Indicators

• After 30 months, Generasi had a statistically significant positive impact on the 12 targeted indicators.
• Health improvements were more pronounced than education ones:
  • Increased frequency of weight checks for young children.
  • Higher number of iron sachets received by pregnant mothers through antenatal care.
  • Greater participation in village health posts (Posyandu).
• Education indicators showed improvements in the final evaluation, reversing earlier negative or neutral findings. The most notable improvement was in primary school participation rates.

2. Long-Term Outcomes

• Malnutrition reduction was the main long-term impact:
  • Childhood malnutrition decreased by 2.2 percentage points (about a 10% reduction).
  • In NTT, underweight rates fell by 8.8 percentage points (20% reduction), severe underweight by 5.5 percentage points (33% reduction), and severe stunting by 6.6 percentage points (21% reduction).
  • In Java, there was a negative impact on stunting, which requires further investigation.
• Child mortality and learning outcomes did not show sustained improvements in the final evaluation, despite some gains in the interim period.

3. Incentive Effectiveness

• Performance incentives were more effective for health than education:
  • On average, 50–75% of health improvements were attributed to the incentive component.
  • Education indicators did not show significant differences between incentivized and non-incentivized groups.
  • Possible reasons:
    • Time lags in education implementation.
    • Health targets were easier to achieve due to lower baseline levels.
    • Community preferences for direct school assistance over out-of-school children.
    • Complexity of incentive rules may have hindered understanding.

4. Heterogeneity in Impact

• Largest impacts were observed in areas with low baseline health and education indicators.
• Geographic disparities were significant, with rural and remote areas benefiting more.
• Temporal variations showed that the program had more pronounced effects in the second year of implementation.
• Poor vs. non-poor individuals showed different responses, with poorer communities benefiting more from the program.

5. Program Design and Implementation

• The program used a randomized control trial (RCT), with subdistricts randomly assigned to either receive Generasi or be in a control group.
• Two versions of the program were tested:
  • Incentivized (Treatment A): Based on performance in health and education indicators.
  • Non-incentivized (Treatment B): No performance-based incentives.
• The block grant was allocated annually, with performance-based adjustments in subsequent years.

Policy Implications

1. Generasi is most effective in areas with low health and education status, not necessarily in the poorest areas.
2. Community incentives have mixed results—health outcomes responded more positively than education.
3. The existing PNPM framework was a useful platform for implementing the program.
4. Target indicators should be regularly reviewed to ensure they reflect community needs and government priorities.
5. Future evaluations are recommended to assess the long-term sustainability of the program and its impacts on learning and health outcomes.

Conclusion

The PNPM Generasi Program demonstrated that performance-based community block grants can effectively improve health outcomes in rural Indonesia. However, the educational impacts were less consistent and may require more time or different design elements to be realized. The program's success highlights the importance of community participation, targeted indicators, and flexible implementation in development initiatives. It also underscores the need for continuous monitoring and evaluation to adapt and improve future interventions.