Data and Code for Different Strokes for Different Folks

Despite recent increases in institutional deliveries in India, birth outcomes have been slow to improve. One likely explanation is low quality medical care, which is common in many parts of India. Low quality medical care in India cannot simply be attributed to insufficient medical knowledge among providers. Instead there is evidence that, even among knowledgeable clinicians, provider effort is very low – suggesting poor alignment between clinician incentives and patient health. <br><br>To address low quality medical care directly, we conducted a policy experiment in Karnataka to evaluate the effectiveness of financial incentives to maternity care providers for (a) the clinical quality of services (e.g., assessing the mother’s blood pressure on arrival during labor) provided to women from their patient lists and catchment areas, and (b) rates of adverse maternal and neonatal health outcomes (e.g., rates of post-partum hemorrhage) among women from their patient lists and catchment areas. Performance bonuses for clinical quality hold some promise for better aligning physician incentives with patient needs. Moreover, directly rewarding lower rates of adverse maternal and neonatal health outcomes flexibly identifies a socially desirable outcome without rigidly prescribing how it should be achieved – allowing local providers to use their local knowledge of what health improvement strategies are likely to work best. We study effect of input and output contracts on providers with varying levels of skills.