Research

As the share of births that rely on assisted reproductive technologies (ART) keeps growing, policies around infertility treatments remain ridden with controversy. We use population-wide Swedish administrative data with uniquely detailed information on individual-level use of ARTs, combined with quasi-experimental empirical methods, to characterize the rate of infertility burden, its private and public costs, and the role of insurance coverage in alleviating infertility. We estimate that one in eight women will experience primary infertility– the inability to have any child at all– over her fertile years. Our analysis reveals that persistent infertility causes a long-run deterioration of mental health and couple stability, with no long-run “protective” effects (of having no child) on earnings. Insurance coverage plays a central role in driving the demand for expensive infertility treatments (IVF). The rate of IVF initiations drops by half when treatment is not covered by health insurance. Our estimates imply that couples are willing to pay at most 33% of their annual disposable income for a course of IVF treatment that gives an about 40% chance of having a child. The response to insurance coverage is more pronounced at the lower end of the income distribution. We show that, as a result, coverage of infertility treatments determines both the total number of additional children as well as their allocation across the socio-economic spectrum.

The last two decades have seen a rapid growth in the development and use of precision medicine in healthcare. I study the equity and efficiency consequences of the entry of a precision medicine test in the context of breast cancer treatment. Using Medicare claims data linked to cancer registry data from the Surveillance, Epidemiology, and End Results Program (SEER), I leverage variation in the timing of a patient's breast cancer surgery relative to the technology adoption date of the provider, in an event-study framework. On average, I find that patients who saw a provider in the year after the provider adopted the technology are more likely to avoid unnecessary treatments. More specifically, patients are less likely to receive chemotherapy at no increased risk in the 5-year cancer mortality rate. However, this welfare-improving impact appears to be concentrated among white patients and patients with higher socioeconomic status. I find large differences in technology diffusion by race that exist even among patients seeing the same physicians, suggesting that differences in use are not solely driven by differences in patient access to different physicians. Moreover, using the linked cancer registry data I show that differences in medical appropriateness for the test (e.g., cancer stage, hormone receptor status) only explain about half of the within-provider Black-white gap. Overall, the results highlight the possibility of precision medicine to improve healthcare efficiency at a potential risk of widening health disparities

Technological change has been widely argued to contribute to healthcare spending growth. This is a particularly salient issue in the context of precision medicine in the U.S. I provide new evidence on the diffusion of an emerging class of precision medicine technologies knows as multi-analyte algorithmic assays (MAAA) in the Medicare population. Using MAAA testing for colorectal cancer screening as a case study, I estimate the spending implications of MAAA adoption. Leveraging variation in when individuals have their Welcome to Medicare visit relative to when their primary care provider adopts the technology in an event-study framework, I find that overall colorectal cancer spending increases by approximately 20 percent after MAAA testing adoption. The spending increase is driven by a substitution away from relatively cheaper colorectal cancer screening tests to the more expensive MAAA test (technology substitution), and an increase in overall screening (technology expansion), with the latter accounting for about 37 percent of the spending increase. Moreover, the technology expansion effect is concentrated to more vulnerable populations, suggesting federal spending may have been progressive. Overall, the results suggest the diffusion of precision medicine in the context of MAAA testing for colorectal cancer led to increases in spending, with the benefits of parts of this increase (namely that due to increased overall rates of screening) potentially far exceeding its costs.