Working papers

Designing Incentive Contracts to Improve the Diagnosis and Treatment of Malaria

We examine whether a diagnosis contingent incentive contract structure improves the treatment of malaria, and whether it’s best to target those incentives to patients or providers. The contract provides incentives to use rapid tests (RDTs) to diagnose patient malaria status combined with incentives to treat with antimalarial drugs (ACTs) if the patient tests positive but not if negative. Using data from a cluster randomized field experiment with 140 pharmacies in malaria endemic regions of Kenya, we find that both patient subsidies and provider incentives significantly increased RDT testing uptake. Absent incentives, 87% of suspected malaria patients purchase ACTs, of which as many as 66% are doing so unnecessarily because they do not have malaria. The incentives lead to an increase RDT test use by 25 pp, a 7 pp increase in the purchase of ACTs by malaria positive patients, and a 27 pp decline in the purchase of ACTs by malaria negative patients. The contract increases (decreases) ACTs for those who are malaria positive (negative) through both improved diagnostic information and incentives. Diagnosiscontingent contracts are highly cost effective, actually lowering the cost per malaria positive person being treated by reducing the unnecessary treatment of malaria negative patients.

Trial pre-registration:
  • AEARegistry
  • Current draft:
  • Click here


  • Ongoing projects

    Provider Agency and Health Technology Adoption: How Financial Incentives Impact the Uptake of a New Contraceptive

    We investigate how targeted subsidies for long-acting injectable contraception and provider incentives impact initial uptake, pricing and stocking decisions for contraceptive products in local markets, sales, and user health outcomes. Following prior work on the effectiveness of subsidies to promote the adoption of new technologies and experience goods, we aim to test if subsidies lead to sustained adoption of injectable contraception as well as to study possible mechanisms for continued usage (or lack of) such as learning, price anchoring, and information effects. We condcuted a market-level cluster randomized controlled trial in 140 pharmacies in Kenya to answer these questions. Patient subsidies and pharmacist incentives will be cross-randomized and compared against the status quo to evaluate their effectiveness in promoting sustained adoption. Data collection complete. Collaborators are Paul Gertler and Carlos Paramo.

    Trial pre-registration:
  • AEARegistry


  • Caregiver Decision-making and Quality of Care for Childhood Fevers in Pharmacy Settings

    Febrile illness is a leading cause of death and serious illness for millions of children across sub-Saharan Africa, despite the availability of diagnostics and appropriate medications. Unfortunately, many fevers are not treated appropriately due to incorrect or no diagnosis, and use of inappropriate medications. Inappropriate treatment has uncertain health and economic costs to individuals as well as at a population level. This study will provide visibility on this issue through an analysis of care-seeking behavior, diagnosis/treatment decisions, and illness episode outcomes in children with fevers who seek care in Kenyan pharmacies. This pilot will contribute towards a better understanding of the causes and consequences of childhood fevers and associated care-seeking trajectories, which is crucial for effective management of childhood illness in sub-Saharan Africa. Pilot study data collection complete, manuscript writing in progress.

    Extreme weather, parental investments, and child health in Madagascar

    Using panel microdata on household inputs and child health outcomes and high-resolution gridded climate data from Madagascar, we investigate how weather-induced income shocks affect parental investments and child nutrition and cognition outcomes.

    Publications

    For list of publications, see here. Please reach out if you are unable to access an article - I am happy to send you an ungated version.