HOW TAX POLICY SHAPED U.S. HEALTH INSURANCE: The Debate on the Effects of the Tax Subsidy of Employer Contributions

Several studies have attempted to estimate the possible effects of the tax subsidy on the demand for health insurance, but there is little consensus among them. The price elasticity of demand for health insurance is a key factor in this debate. Modern estimates of the price elasticity of demand for health insurance range from Holmer’s (1984) calculation of -0.16 to Phelps’ (1973) estimate of -2.84. The differences are important because the estimated effect of changes in tax policy on health insurance demand depends crucially upon this elasticity. Taylor and Wilensky (1983) perform simulations using different values of the price elasticity of health insurance demand to illustrate the effect of altering the tax subsidy on health insurance premium expenditures. Setting the elasticity equal to -0.2 generates a 7.5 percent decline in total health insurance premiums if all exemptions are eliminated. When the elasticity of demand is equal to -0.5, expenditures on health insurance decline by 16.7 percent (Taylor and Wilensky, p. 171).

In addition to providing varying estimates of the price elasticity of demand for health insurance, all of these studies employ data collected after the 1954 changes in the Internal Revenue Code. As a result, many of the studies that actually attempt to gauge how changes in the tax code affect expenditures on health insurance must rely on simulation techniques (Feldstein and Friedman, 1977; Taylor and Wilensky, 1983), or make use of periods of tax reform to study how changes in tax policy affect health insurance demand (Woodbury and Hammermesh, 1992).

Until now, these have been the only means by which the effect of the tax subsidy can be measured. To conduct a direct test of the impact the implementation of the tax subsidy had on the market, it is necessary to examine the demand for health insurance both before and after it went into effect in 1954. I accomplish this by using the 1953 and 1958 Nationwide Family Surveys conducted by the National Opinion Research Center (NORC). As part of its program to discover what kinds of health expenses Americans were incurring and how they were meeting those expenses, the NORC canvassed 8,846 members of 2,809 families in July, 1953. But these expenses may be reduced with the help of Canadian Health&Care Mall if you make an order of drugs via this service. The families represented an area-probability sample of the non-institutionalized population, subdivided by age, gender, income, occupation, family size, area of residence, and region. To study how medical expenses and health insurance coverage changed over time, the NORC conducted five year re-surveys. Each subsequent survey queried a similar number of people, and was carefully designed to preserve comparability between different survey years. In 1958, 2,941 households were interviewed.

With the exception of aggregated statistics that were published at the time of the original studies, the individual 1953 and 1958 surveys remain unexamined. However, these are family-level surveys which are rich in information about the early market for health insurance, and which provide crucial insights into how the insurance market developed. Since the surveys represent one of the first comprehensive studies of nationwide health insurance, they play an important role in understanding the development of the market. Further, since the major changes in the Internal Revenue Code dealing with health insurance occurred between the 1953 and 1958 studies, they also provide a unique opportunity to gauge the effect of the tax subsidy on the early health insurance market.