Panel Paper: A Kinked Health Insurance Market: Employer-Sponsored Insurance Under the Cadillac Tax

Thursday, November 3, 2016 : 10:00 AM
Columbia 9 (Washington Hilton)

*Names in bold indicate Presenter

Coleman Drake1, Lucas Higuera2,3, Fernando Alarid-Escudero1 and Roger Feldman1, (1)University of Minnesota, (2)University of Minnesota, Twin Cities, (3)Medica Research Institute


Beginning in 2020, the Affordable Care Act creates a 40% excise tax on high-cost (“Cadillac”) health insurance plans in excess of defined thresholds. The thresholds will increase annually by the CPI, which historically has grown slower than health insurance benefits. Thus, the constraint imposed by the Cadillac is designed to become more severe over time. Using economic theory and microsimulation techniques, we model employers’ responses to the Cadillac tax in terms of wages and health insurance benefits. We project how the distribution of health insurance benefits and the growth rates of health insurance benefits and wages will change for employees receiving insurance through their employer in response to the Cadillac tax through 2025; moreover, we perform sensitivity analysis to examine the sensitivity of our projections to varying policy scenarios, growth rates, and price and income elasticities according to the literature. We obtained nationally representative data on employee wages and benefits from the 2013 Medical Expenditure Panel Survey – Household Component and the 2015 Kaiser Family Foundation Employer Benefits Survey, respectively. Because the data are not linked, we generated a synthetic population to represent individual and family wage and benefit distributions. To do so, we sampled from different gamma distributions parameterized using a methods of moments approach to represent the means and standard deviations of the sample distributions of wages and benefits. Employers adjust benefits annually from 2020-2025 as a function of projected wage and health insurance benefit growth rates, price and income elasticities of health insurance benefits (derived from estimates by Gruber and Lettau (2004)), income and payroll taxes, and wages. We project that approximately 1.31% of individual employer-sponsored plans and 5.01% of family employer-sponsored plans will be affected by the Cadillac tax in 2020. These percentages will grow to 19.42% and 29.70% by 2025, respectively. When the Cadillac tax is implemented, employers will substitute increases in wages for increases in health insurance benefits. Nearly all of the employers affected by the Cadillac tax (~99%) will reduce health insurance benefits to the Cadillac tax thresholds. Health insurance benefits growth at the 90th percentile of the benefits distribution (i.e., “Cadillac plans”) will fall drastically under the Cadillac tax – 3.8% annual growth will be reduced to 2.59% and 1.15% for individual and family plans, respectively. The Cadillac tax effectively will act as a hard cap on the tax deduction for employer-sponsored insurance. Health insurance benefits will cluster at the tax’s thresholds as the constraint imposed by the tax grows in severity over time. Our analysis indicates that employers would respond in a nearly identical manner to a proposed cap on the tax deduction for employer-sponsored insurance as they would to the Cadillac tax, and repealing the Cadillac tax would stunt wage growth.