京都大学 大学院経済学研究科・経済学部

“Public Private Partnership Models in Social Health Insurance schemes – Evidence from a Pioneering Program in India”　Tanika Chakraborty

Abstract:
Maternal and child health continue to be a significant policy focus in developing countries, including India. An emerging model in health care is the creation of public and private partnerships. Since the construction of physical infrastructure is costly, governments at various levels have tried to implement social health insurance schemes where a trust calculates insurance premiums and medical payments. Typically, qualifying families get full subsidization of the premium and get access to private hospitals, in addition to low cost public hospitals, for their tertiary care needs. We analyze one such pioneering social insurance scheme in the Indian state of Andhra Pradesh (AP). The Rajiv Aarogyasri program (RA) was introduced by the Government of AP on a pilot basis in 2007 and implemented in 2008. We examine the extent to which access to reproductive health care changed. Specifically, we examine whether the RA scheme increases institutional deliveries, particularly in private hospitals. Since AP was the only state in which such a state insurance program was implemented, the neighboring states constitute a plausible control group. Combined with the policy timing, and the year of birth, we employ a difference-in-difference strategy to identify the effects of RA on the residents of AP, as a baseline strategy. We further restrict our analysis to bordering districts and compare contiguous district pairs in treatment and control states and test the sensitivity of our baseline results. We find that deliveries in private hospitals increased, and government hospitals decreased after the introduction of the program. However, even as the use of private facilities increased, we find that out-of-pocket costs declined. Our heterogeneity results suggest a substitution effect of the relative price change – a switch to private from government hospitals. We do not find an overall increase in access to tertiary care. Our findings are robust to pre-treatment trends and other sensitivity checks. Finally, we observe a modest effect of the policy on infant and child mortality.