Tax Breaks for Employer-Sponsored Health Insurance

One of the most-discussed issues during the recent health care reform debate was the proposal to cap the tax exclusion for employer-sponsored health insurance. Currently, employers' spending on health insurance premiums is exempt from taxation for both employers and employees. Premiums paid by employees are exempt as well if the firm has established a Section 125 cafeteria plan; roughly 80 percent of employees with insurance have such a plan. This tax exclusion is extremely costly - it reduces federal and state tax revenues by $260 Billion per year and is the government's third largest expenditure on health care, after Medicare ($400 Billion) and Medicaid ($300 Billion).

What are the arguments for and against the tax exclusion for employer-sponsored insurance (ESI)? How would reducing or eliminating this exclusion affect tax revenues and insurance coverage? These questions are the subject of a recent working paper by researcher Jonathan Gruber, The Tax Exclusion for Employer-Sponsored Health Insurance (NBER Working Paper 15766).

ESI is the dominant form of insurance for the non-elderly population in the U.S. More than sixty percent of non-elderly individuals receive their insurance through their own employer or that of a family member. By contrast, only six percent purchase insurance privately through the non-group market. The remainder of the population is either insured by the government (19 percent) or uninsured (17 percent).

The author begins by noting that the primary argument for the tax exclusion is that it may be the "glue" holding the ESI system together. ESI provides an important pooling mechanism, that is, a way to create large pools of individuals with predictable distributions of risk. Workers at high risk of having large health expenditures are pooled together with other healthier workers, allowing them to access insurance at a reasonable price. In the non-group market, by contrast, insurers worry that those seeking coverage may be high-risk individuals. Prices are high and variable, and in most states individuals can be excluded from coverage based on their health status. Without the tax exclusion, employers might cease to offer ESI and individuals might have to turn to the non-group market, where affordable coverage may not be available, particularly for sicker individuals.

The author notes that there are also numerous costs of the ESI exclusion, starting with the revenue cost. Furthermore, the benefits of the tax exclusion flow disproportionately to those at the top of the income distribution, since the value of the exclusion rises with the individual's income tax rate and richer individuals are more likely to have insurance and to have more generous plans. The tax exclusion may also bias individuals towards buying excessively generous insurance, as it makes insurance cheaper relative to other goods. Finally, it may distort workers' decisions with respect to job changes and retirement.

In his primary analysis, the author uses a microsimulation model of health insurance to simulate the effect of repealing or capping the tax exclusion. The model takes a sample of individuals from the Current Population Survey and matches this to information on health insurance premiums and health costs. The model incorporates behavioral responses by firms and employees to changes in the price of insurance. Specifically, firms make decisions about whether to offer insurance, how to divide premiums between the firm and employees, and what the level of insurance spending should be; workers decide whether to take up insurance offered by the firm and whether to buy in the non-group market. To determine the magnitude of these behavioral responses, the best available empirical evidence is used. The author notes, however, that his findings "must be interpreted with considerable caution, as they are using the price elasticity estimated from existing variations in the tax price to estimate the impact of a much more radical change in policy."

The author first simulates the effect of repealing the tax exclusion for ESI. He estimates that this policy will lead to a one-third reduction in employer spending on health insurance, due in part to a reduction in firms' contributions to premiums. Importantly, there is also a decrease in the number of individuals with ESI of 15 million, or roughly 10 percent of the number with ESI prior to the repeal. Some of those losing ESI purchase non-group coverage or move to government insurance, so the net increase in the number of uninsured is 11 million, or 22 percent relative to the baseline number of insured. Those who leave ESI are similar in health status to those who stay, mitigating concerns that there might be a further unraveling of ESI due to changes in the composition of the pool.

Next, the author simulates a policy of removing the tax exclusion for employers but continuing to allow an exclusion for section 125 spending. This policy is estimated to raise $184 Billion, versus $263 Billion from full repeal. There is a large rise in employee spending, as spending migrates from employers to the section 125 accounts. As the author notes, "this highlights the leakages in revenue raising that can occur from partial reform." The number of uninsured rises by 10 million under this policy.

The complementary policy - keeping the tax exclusion for employers but removing deductability of section 125 accounts - raises only $42 Billion and increases the number of uninsured by only one million. Relative to full repeal, eliminating the exclusion for the income tax but maintaining it for the payroll tax raises less revenue and has a smaller effect on the number of uninsured, as one might expect, but also places more of the burden of the policy on those at the top of the distribution, since they benefit most from the income tax exclusion.

Finally, the author simulates the policy of capping the tax exclusion rather than eliminating it. Specifically, he caps the exclusion as the median level of premiums - roughly $5,200 for individual plans and $13,700 for family plans. This policy raises only $47 Billion, but also increases the number of insured by only one million.

The author concludes that the ESI exclusion is costly and highly regressive, so "repealing or capping the exclusion could result in significant increases in government revenues and an improvement in revenue raising progressivity." However, he also cautions that these policies could lead to a significant reduction in insurance coverage. He suggests "when considering changes to the tax treatment of ESI, policy makers may simultaneously wish to examine other policies that affect the availability of non-ESI coverage."