Research

Working Papers

How Insurer Subsidies Affect Market Equilibria

 Abstract: This paper studies how the cancellation of a large insurer subsidy affected market equilibria in the U.S. individual health insurance markets. Leveraging variation across state-level markets, I measure effects on premiums, competition, and enrollment. For each dollar withdrawn, premiums rose by $1.26. The withdrawal caused 1.2 insurers to exit in the median state, and plan quality declined. Enrollment dropped by 16% in the median state, and medical spending per enrollee also fell. These results suggest that producer subsidies can affect insurer entry decisions and what products they bring to market.

Workplace Stratification and Racial Health Disparities (with Long Hong, Kurt Lavetti, and Trevon Logan), available on request

Abstract: To what extent is a worker’s relative rank within their workplace a determinant of health status, conditional on income?  We provide the first US-based evidence on the aggregate relationship between relative workplace rank and health disparities for the near population of workers in one US state. Using a new linkage of commercial all-payer health insurance data to administrative earnings records for workers in Utah from 2013-2015, we quantify the impact of relative workplace rank on racial health disparities and the incidence of specific health conditions that drive these disparities. We show that about 75% of the conditional income-health gradient is explained by relative rank, and only one-quarter is explained by income levels. For an average Black, Hispanic, or American Indian worker, moving from the 90th to the 10th  percentile of within-firm rank holding fixed income, age, location, and health insurance policy characteristics is associated with a 23%  increase in the average number of diagnosed chronic conditions. This stronger association for racial minorities than for White workers amplifies aggregate inequality: the racial segregation of jobs in the US leads minority workers to be overrepresented in lower-ranked jobs within firms,  which in turn exacerbates racial health disparities.

Insurer Innovation and Health Care Efficiency: Evidence From Utah (with Ben Handel, Jon Kolstad, and Kurt Lavetti), available on request

Abstract:  Private health insurers play a central role in determining the cost and quantity of health care in the United States. Despite this centrality, there has been limited empirical work studying how insurers differentially affect spending and consumption, especially for expensive and pervasive chronic conditions. We use hundreds of natural experiments involving employers switching their primary health insurer, together with a movers design, to estimate these causal effects. We find meaningful differences in total cost and prices for medical and pharmaceutical spending. Within-person changes in spending caused by forced reassignment across pairs of insurers, holding fixed plan generosity, are as large as 32% of total medical spending and 40% of drug spending. We also find substantial dispersion in insurer causal effects on health care spending and quantities for patients with diabetes, hypertension, and other chronic conditions. We find strong evidence for drug offsets as insurers who causally increase drug utilization also reduce medical costs and quantities overall and by condition. Finally, we find that employers do not select plans that reduce costs for their employees based on the match between employee conditions and estimated plan treatment effects, leaving significant cost savings on the table.

Research in Progress 

The Earnings Incidence of Employer-Provided Health Insurance (with Kurt Lavetti and Tamar Oostrum)

Abstract: We estimate the compensating earnings differential and distributional earnings incidence of employer-provided insurance. We provide the first US-based estimates of earnings differentials for health insurance derived from administrative all-payer health insurance claims and insurance enrollment records linked to earnings records. Our estimates allow for arbitrary nonrandom assignment of workers to jobs based on static differences in worker productivity, health status, employer insurance offerings, employer pay policies, and job match quality. 

 Labor Market Effects of Shifts in Employer-Provided Health Insurance Costs Among Older Workers (with Kurt Lavetti)

Abstract: We study the impact of Medicare eligibility on employment and wages among elderly Americans at firms that offer health insurance. To identify the effect of Medicare eligibility on wages, we use a firm size cutoff that dictates which firms continue paying health insurance costs after employees turn 65. For firms with over 20 employees, the employer health insurance plan acts as the primary payer even if the employee is eligible for Medicare. For firms under 20 workers, Medicare always acts as the primary payer regardless of whether the employer offers health insurance. As a result, firms with over 20 workers that offer health insurance pay substantially more to hire employees aged 65+ than smaller firms. The data we use is a novel linkage between the Utah All-Payer Claims Data and earnings derived from Unemployment Insurance records in Utah from 2013-2019. Using a difference-in-differences event study model that controls for fixed worker-firm job match effects, we find that approximately half of the cost savings in small firms are passed through to earnings when workers age into Medicare. We also find that small firms are more likely to hire and retain elderly workers, which leads the fraction of workers employed by small firms to increase by 40\% between the ages of 60 and 69.

Publications

Is This Time Different? The Slowdown in Health Spending, with Amitabh Chandra and Jonathan Skinner
Brookings Papers on Economic Activity, Vol 47, No. 2 (2013)

Large Increases In Spending On Postacute Care In Medicare Point To The Potential For Cost Savings In These Settings with Amitabh Chandra and Maurice Dalton
Health Affairs, Vol 32, No. 5 (2013)