2005年-世界发展银行全球_Reducing_Child_Malnutrition_in_Tanzania___Combined_Effects_of_Income_Growth_and_Program_Interventions_28页_399kb

Summary of "Reducing Child Malnutrition in Tanzania: Combined Effects of Income Growth and Program Interventions"

Core Content

This paper examines the combined impact of income growth and nutrition interventions on the reduction of child malnutrition in northwestern Tanzania, using a four-round panel dataset from the Kagera Health and Development Survey (KHDS). The study evaluates how household income and community-based nutrition programs such as feeding posts and Partage influence child nutritional status, measured through anthropometric indicators like stunting, wasting, and underweight.

The paper emphasizes that while income growth is crucial for improving nutrition, nutrition interventions are also necessary to achieve the Millennium Development Goals (MDGs), particularly the nutrition MDG which aims to halve malnutrition rates by 2015. The findings suggest that income alone is insufficient, and that complementary program interventions are required to significantly reduce malnutrition.

Main Points

1. Malnutrition and Its Determinants

• Malnutrition is associated with inadequate diet, poor health and sanitation services, and insufficient child care.
• It is measured using z-scores for stunting (low height for age), wasting (low weight for height), and underweight (low weight for age).
• Stunting is considered a long-term indicator of cumulative deprivation, while underweight is more acute and reflects short-term food shortages.
• The WHO recommends stunting as a reliable measure of overall social deprivation, while the MDG uses weight for age.

2. Data and Methodology

• The study uses panel data collected from 1991 to 1994 in Kagera, a region in Tanzania with high adult mortality due to the AIDS epidemic.
• The dataset includes 1140 children and 2871 person-wave observations, with anthropometric data and socioeconomic variables.
• OLS regression is used to estimate the impact of nutrition interventions and household income on child nutrition, with a preferred random effects model to account for unobserved heterogeneity and endogeneity.

3. Key Findings

• Income growth has a positive effect on child nutrition, with higher per capita consumption associated with better nutritional outcomes.
• Nutrition interventions (feeding posts and Partage) have a substantial beneficial effect, particularly in low-income communities.
• The OLS results show that children in communities with interventions are 0.3 SD taller than those without.
• However, the OLS estimates may be downward biased due to endogenous program placement.
• Using an instrumental variables approach, the impact of interventions is found to be higher, with a coefficient of 0.44 for the presence of a program in the village.
• Parental education, especially maternal education, has a positive impact on child nutrition.
• Parental height is also a significant predictor of child height, indicating intergenerational persistence in nutritional status.

4. Policy Simulations

• To halve malnutrition by 2015, income growth must be complemented by large-scale nutrition programs.
• A 2.5% annual income growth could reduce the fraction of underweight children by 27–34%, but only 14–16% of children in Kagera would be affected by such growth alone.
• Even a 3% per capita income growth is insufficient to achieve the nutrition MDG without program interventions.
• Random effects model is preferred over pooled regression due to the significant unobserved heterogeneity in the data.

Key Information

• Income growth is essential but not sufficient for reducing malnutrition.
• Nutrition interventions (feeding posts and Partage) significantly improve child nutrition, especially in low-income areas.
• Endogeneity of program placement is addressed using instrumental variables, including household consumption, adult height, and community shocks.
• Parental characteristics (education, height) play a crucial role in determining child nutritional status.
• MDG simulations show that income growth and program coverage must be combined to meet the nutrition targets.
• Stunting is more responsive to interventions than underweight, indicating that interventions may be more effective in addressing long-term nutritional deprivation.

Conclusion

The study underscores the interplay between income growth and nutrition programs in improving child health outcomes. While income growth contributes to better nutrition, community-based interventions are essential to achieve the MDG targets, especially in regions with persistent malnutrition. The random effects model provides a more accurate estimation of the combined impact of these factors, suggesting that policy interventions must be scaled up in conjunction with income growth to effectively reduce child malnutrition.