Three Essays in Health and Development.

Abstract

The quality of healthcare is low in developing countries. In this dissertation, I examine two different proposals to improve healthcare quality: improve customer information regarding healthcare choices, or enroll individuals in health insurance.

In Chapter 1, I present results from an audit study conducted in Uganda. I compare the price paid and the drug quality received between shoppers in the same village who either ask for a diagnosis (or declare the patient has malaria) or ask for a drug recommendation (or ask for a specific product. I find that shoppers who present information about either the diagnosis or recommended treatment pay approximately $0.18 (5 percent) less. Counter-intuitively, I find that customers who present information about either the diagnosis or the recommended treatment are 3.4 percentage points more likely to be sold a substandard drug. I develop a conceptual model to justify my findings and conclude that improved information will not improve quality in a market if information and detection of low quality are not sufficiently related.

In Chapter 2, coauthored with Esther Atukunda, we present descriptive analyses from the same data collected in Uganda. We combine data from drug outlets, covert shoppers, and real customers to test hypotheses of how low quality drugs enter a market. We estimate that only 3.4 percent of purchased drugs are substandard: a much higher drug quality than found in previous studies. We develop three stylized facts: substandard medicines are typically diluted high-quality medicines; customers cannot tell which drugs are low quality; and vendors are complicit in the sale. We end with a discussion of policy interventions.

In Chapter 3, coauthored with Rebecca Thornton, I present results from an experiment conducted in Nicaragua that randomly allocated health insurance subsidies to parents. We specifically examine differential effects among children who were part of an insured household, but ineligible for health insurance themselves due to an age restriction. Our results indicate that the health insurance significantly increases access to higher-quality providers and altered the entire family’s health demands. In particular, eligible and insured children substantially increase healthcare utilization, while ineligible children in insured households decrease healthcare visits.