Research
My research focuses on how information and financial incentives influence consumers' provider choices and the corresponding provider responses to changes in consumer behavior.
In Progress
Premium Service: Comparing Cost and Quality for Colorectal Cancer Screening
Abstract: Despite emphasis on paying for value, few studies have identified the relationship between price and quality for health care services. One reason why the relationship between price and quality remains uncertain is inherent selection bias—the same patients who are likely to receive care from high-priced providers are more likely to be at higher risk for complications. This paper uses an exogenous source of variation that influences the supply of low-priced providers, state Certificate of Need (CON) laws, as an instrumental variable for receiving care at low-priced providers, ambulatory surgical centers (ASCs), for an important cancer screening test–colonoscopies. Patients who live in a state with a CON law are 22% less likely to receive care from an ASC. Using this source of variation, receiving care from an ASC reduces procedure costs by 39% and patient cost sharing by 44%. For quality outcomes, the results are more ambiguous–use of ASCs leads to a 1.9 percentage point reduction in surgical complication rates but a 8.1 percentage point reduction in potentially cancerous polyp removal. Using the same sources of variation, I find that higher colonoscopy prices, lower market share of ASCs, and increased market concentration reduce colonoscopy utilization, but increase the use of substitute cancer screening tests. Due to this substitution, higher prices, lower ASC use, and increased provider market concentration lead to lower overall cancer screening spending.
Provider Responses to Online Price Transparency (Updated version!)
Abstract: Price transparency initiatives have recently emerged as a solution to the lack of health care price information available to consumers. This paper uses the staggered and nationwide diffusion of a leading internet-based price transparency platform to estimate the effects of price transparency on provider prices. Consistent with the consumer responses, the price responses are much larger for laboratory test providers than they are for office visit providers. The results imply that access to the price transparency tool has reduced laboratory test prices by 1-4% and that provider price responses are driven by active consumer use of price information. In combination with the consumer analysis, this paper demonstrates how reducing consumer search costs can spur firm price competition in health care markets.
Searching for Health: The Effects of Online Price Transparency
Abstract: Recent technologies have provided consumers with access to price information for medical services, a market with uncharacteristic price variation and opaque prices. This paper examines the effects of an online price transparency platform on consumer choices of providers and provider prices. I find that consumer use of price transparency information leads to a 16% reduction in prices for lab tests, 13% for advanced imaging, and about 1% for physician office visits. Consistent with a a model on how search costs reduce consumer prices, there are larger effects in markets with high price dispersion. For physician office visits, the value of non-price information, such as provider gender, length of practice, and location, far exceeds the value of price information. I estimate a consumer willingness to pay of $4-$10 for non-price physician information.
Abstract: Colorectal cancer is the second leading cause of cancer-related mortality, but the cost-sharing environment for colonoscopies, the preferred form of screening, is complex. This paper combines a machine learning-based double-selection algorithm to perform principled covariate selection with differential exposure to the Affordable Care Act’s requirement that insurers fully cover cancer screening services as an instrumental variable to estimate the effect of cost sharing on colonoscopy utilization and price-shopping. The reduced form results show that the ACA’s requirements increased the use of colonoscopies by 1.7% and the IV results imply a price elasticity of 0.09. I do not find a price shopping response.
Doped up or Dropped Out: The Effects of Opioid Use on Workplace Absences (with Armando Franco and Zachary Wagner)
Abstract: While many studies have examined the effects of opioid use on health outcomes, few studies have examined how using prescription opioids impacts other relevant behaviors. We use differential adoption of state-level prescription drug monitoring laws to estimate the effect of opioid use on workplace absences. We find that every 10 percent increase in opioid use leads to a 15.6 percent increase in workplace absences, which is equivalent to one missed day every four months. A back of the envelope calculation suggests that the opioid epidemic has led to an unaccounted-for $3.3 billion annual decrease in workplace productivity.