Impact Evaluation of Community Clinic Intervention on Maternal Health Care Utilization in the Slums of Delhi: A Propensity Score Matching Model

Raj Kumar, Jawaharlal Nehru University

This study use impact evaluation methodology to estimate Community Clinic intervention impact on antenatal care and postnatal care. Data is collected by household level randomized survey in eight different slum clusters in Delhi. Survey is conducted in areas with and without Community clinic to construct treatment and control groups. Conditions of inclusion, exclusion, recall period, intervention period are specified to minimize bias in the survey. Three different Propensity Score Matching models are used; nearest neighbour matching, kernel matching, and radius matching models. Results are robust as the estimated ATT is quite close in all three matching models. The estimated ATT is significant and negative for complete antenatal care under Nearest Neighbour (-0.206), Radius (-0.172) and Kernel (-0.174) matching methods. Results also show a strong positive impact on postnatal care utilization for both women (0.093, 0.107, 0.106) and children (0.211, 0.232, 0.235). The negative impact on antenatal care is associated with lack of systematic and planned antenatal care provisions. Rubin’s ‘R’, the ratio of treated to (matched) non-treated variances is also checked, satisfying the condition. Likelihood ratio test of the joint significance is insignificant suggesting that matching quality is good. Density distribution of the propensity score is checked.

Keywords: Sexual and Reproductive Health and Rights, Health and Morbidity, Children, Adolescents, and Youth, Inequality, Disadvantage and Discrimination

See extended abstract.