Reduce Tax Preferences for Employment-Based Health Insurance

Overview of the Issue

The federal tax system provides preferential treatment for health insurance that people buy through an employer. Unlike cash compensation, employers’ payments for employees’ health insurance premiums are excluded from income and payroll taxes. In most cases, the amounts that workers pay for their own share of health insurance premiums is also excluded from income and payroll taxes. Contributions made to certain accounts to pay for health costs are excluded from income and payroll taxes as well. In all, that favorable tax treatment cost the federal government about $275 billion in forgone revenues in 2016, and that cost will probably rise over time as the cost of health care rises. The tax preferences will continue even after a new excise tax takes effect in 2020 and somewhat reduces their consequences.

Further reducing the tax preferences for employment-based health insurance would raise federal revenues. It also would reduce the number of people with employment-based coverage, boost enrollment in the health insurance marketplaces established under the Affordable Care Act, and increase the number of people without insurance. And it would make total spending on health care lower than it would have been otherwise.

Current Law. The federal tax system subsidizes employment-based health insurance both by excluding employers’ premium payments from income and payroll taxes and by letting employees at firms that offer “cafeteria plans” (which allow workers to choose between taxable cash wages and nontaxable fringe benefits) pay their share of premiums with before-tax earnings. The tax system also subsidizes health care costs not covered by insurance by excluding from income and payroll taxes the contributions made to various accounts that employees can use to pay for those costs. Examples include employers’ contributions to health reimbursement arrangements (HRAs), employees’ contributions to flexible spending arrangements (FSAs), and employers’ and employees’ contributions to health savings accounts (HSAs). On average, people with higher income or more expensive health insurance plans receive larger subsidies.

The favorable tax treatment of employment-based health benefits is the largest single tax expenditure by the federal government. (Tax expenditures are exclusions, deductions, preferential rates, and credits in the tax system that resemble federal spending in that they provide financial assistance to specific activities, entities, or groups of people.) Including effects both on income taxes and on payroll taxes, that exclusion is projected to equal 1.5 percent of gross domestic product over the 2017–2026 period.

The excise tax that is due to start in 2020 will effectively reduce the tax subsidy for employment-based health insurance. It will be levied on employment-based health benefits—consisting of employers’ and employees’ tax-excluded contributions for health insurance premiums and contributions to HRAs, FSAs, or HSAs—whose value exceeds certain thresholds. The excise tax will thus curtail the current, open-ended, tax exclusions. (Even when the excise tax is in effect, however, employment-based health insurance will still receive a significant tax subsidy, and that subsidy will still be larger for people with higher income.)

The excise tax will equal 40 percent of the difference between the total value of tax-excluded contributions and the applicable threshold. If employers and workers did not change their coverage in response to the tax, roughly 5 percent to 10 percent of people enrolled in an employment-based health plan in 2020 would have some tax-excluded contributions in excess of the thresholds, according to estimates of the Congressional Budget Office and the staff of the Joint Committee on Taxation (JCT). (However, CBO and JCT do expect people’s responses to the tax to reduce that share, discussed below.)

In 2020, CBO and JCT project, the thresholds will be $10,800 for individual coverage and $29,100 for family coverage. (Those thresholds will be slightly higher for retirees who are 55 to 64 years old and for workers in certain high-risk professions. Further adjustments will be made for age, sex, and other characteristics of an employer’s workforce.) After 2020, the thresholds will be indexed to the growth of the consumer price index for all urban consumers (CPI-U), which measures inflation. Because health insurance premiums will probably continue to rise faster than inflation, the excise tax will probably affect a growing number of people over time. As a result, CBO and JCT project, revenues stemming from the tax will rise from $3 billion in 2020 to $20 billion in 2026.

Effects of Current Law. The tax exclusions have effects that include encouraging the use of employment-based insurance, making it likelier that healthy people will buy health insurance (which lowers the average cost of insurance), and increasing spending on health care. Another effect is that higher-income workers receive larger subsidies than lower-income workers do.

Encouraging the Use of Employment-Based Insurance. By subsidizing employment-based health insurance, the tax preferences encourage firms to offer it and workers to enroll in it. Such insurance would be attractive to employers and employees in any case, because it pools risks within groups of workers and their families and reduces the administrative costs of marketing insurance policies and collecting premiums. But the preferences give employment-based insurance additional appeal. In 2015, according to a Medical Expenditure Panel Survey, 84 percent of private-sector employees worked for an employer that offered health insurance coverage; 76 percent of those employees were eligible for that coverage (the rest were ineligible for various reasons, such as working only part time); and 75 percent of the eligible workers chose to enroll.

Reducing Adverse Selection. A major problem that can occur in insurance markets is adverse selection, in which less healthy people are likelier to buy health insurance (or to buy certain types of plans) than healthier people are. Adverse selection occurs because insurance provides more benefit to enrollees with above-average costs—and is therefore more attractive to them—and less benefit to people with below-average costs. As premiums increase to cover the less healthy enrollees, the healthier ones may stop buying insurance, which results in another price increase—a spiral that may continue until the market is very small or nonexistent. Adverse selection also can reduce markets’ efficiency by making it harder for insurers to predict costs for a group of potential enrollees.

Employment-based health insurance and the tax preferences that encourage its use limit adverse selection in several ways. Employers generally select a workforce on the basis of criteria other than health care costs, so most workforces consist of a mix of healthier and less healthy people. Pooling risks across such a workforce reduces the variability of average health care spending for the group. Also, once employers are offering health insurance, they tend to pay a large share of premiums in order to encourage employees to enroll—making the employees’ share small in relation to their expected health care costs, encouraging them to buy insurance, and reducing adverse selection. The tax exclusions also limit adverse selection by reducing the after-subsidy price of insurance, encouraging even the healthy to enroll.

Recent changes in regulations governing markets for nongroup (that is, individually purchased) health insurance—which are separate from markets for employment-based insurance—reduce the problem of adverse selection in that market. In addition, subsidies are now available in the nongroup market. Those changes have weakened the rationale for subsidizing employment-based insurance because the nongroup market now provides an alternative way of providing insurance—one that is available to people regardless of their health and that subsidizes their coverage. Nevertheless, employment-based insurance is still a relatively efficient way of providing insurance because its administrative costs are much lower than those in the nongroup market.

Increasing Health Care Spending. The tax preferences for employment-based health insurance contribute to the growth of health care spending. That occurs because the preferences encourage workers to favor health care over other goods and services that they could purchase and also because the tax exclusions encourage employers to compensate their workers with a combination of health insurance coverage and cash wages rather than entirely with cash wages (which the employees would be unlikely to spend on health care to the same extent). Furthermore, the tax exclusions are currently open-ended (and will be until the excise tax takes effect in 2020). That is, their value increases with an insurance plan’s premium, encouraging people to enroll in plans that cover a greater number of services, cover more expensive services, or require enrollees to pay a smaller share of costs. As a result, people use more health care—and health care spending is higher—than would otherwise be the case.

Concern about that effect has lessened somewhat in recent years because employment-based insurance plans that require workers to pay a higher share of health costs have become more common. For example, 29 percent of people with employment-based coverage reported enrolling in a high-deductible health plan in 2016, up from 8 percent in 2008.

Subsidizing Workers With Different Income Differently. Another concern about the tax exclusions is that they subsidize workers with different income differently. The value of the exclusions is generally larger for workers with higher income, partly because those workers face higher income tax rates (although they may face lower rates of payroll taxation) and partly because they are more likely to work for an employer that offers coverage. Because larger subsidies go to higher-income workers, who are more likely to buy insurance even without the tax exclusions, and smaller subsidies go to lower-income workers, who are less likely to buy coverage, the exclusions are an inefficient means of increasing the number of people who have health insurance, and they are regressive in the sense of giving larger benefits to people with higher income.

The forthcoming excise tax will be levied on insurers and on employers who offer their own insurance plans, but economic theory and empirical evidence suggest that the cost will ultimately be passed on to workers. CBO and JCT expect that in many cases, that will occur when employers and workers decide to avoid paying the tax by shifting to health plans with premiums below the thresholds. In those cases, the money that would otherwise have been used to pay for the more expensive premiums would generally increase either workers’ wages or employers’ profits, both of which are taxable. Because workers with higher income will pay higher marginal tax rates on those increased wages, the result will be a reduction in the tax exclusions’ regressive nature. When employers and workers do not shift to lower-cost health plans to avoid the excise tax, the costs of that tax will be spread equally among workers, JCT and CBO expect. However, workers with higher income are more likely to be enrolled in high-cost plans and thus more likely to have their subsidies reduced in the first place.

Most workers will have health benefits whose value is below the thresholds and therefore will be largely unaffected by the excise tax. Consequently, the existing tax preferences and the new excise tax will continue to subsidize employment-based health insurance and to provide larger subsidies to higher-income people.

Key Design Choices That Would Affect Savings

Lawmakers who wanted to design laws to reduce the tax preferences for employment-based health insurance could take various approaches. Those approaches would have different effects on federal revenues, on the taxes owed by people at various income levels, on employers’ and employees’ choices about health insurance plans, and on their resulting health care costs. One approach would involve modifying both the current tax exclusions and the upcoming excise tax. Another approach—one that is not examined in this volume—would replace the current tax exclusions with an income tax credit for employment-based health insurance.

In general, reducing the tax preferences for employment-based health insurance would tend to lower the number of people with such insurance. It also would increase out‑of-pocket payments by people enrolled in employment-based insurance, which would decrease spending on health care and increase the financial burden on people with substantial health problems. The precise effect, however, would depend on the specific features of any policy change.

Modifying the Tax Exclusions and the Excise Tax. Lawmakers could cancel the excise tax that is scheduled to take effect under current law and instead subject contributions for health insurance premiums, along with contributions to various health-related accounts, to income or payroll taxation. If lawmakers did that, they would have to decide whether to tax all of the contributions or only some of them. For example, the exclusions could be retained, but with an upper limit that applied to all taxpayers, or the exclusions could be phased down for higher-income people. Such limits also could be allowed to vary according to other characteristics of employees that are associated with average health costs, such as age, sex, or occupation. (The forthcoming excise tax includes several adjustments of that sort. For instance, the threshold above which health care costs are taxed is higher for some groups of people whose average costs are high because they work in dangerous occupations.)

Lawmakers also would need to decide whether to subject the contributions to income taxation, payroll taxation, or both. On average, enrollees in employment-based plans face slightly higher federal income tax rates than payroll tax rates. Specifically, CBO and JCT estimate that those workers’ average marginal income tax rate—that is, the rate that applies to the last dollar of their earnings—will be about 20 percent in 2020, whereas their average marginal payroll tax rate (including both the employer’s and the employee’s shares of payroll taxes) will be about 14 percent. Therefore, subjecting contributions to income taxation would raise slightly more revenue than subjecting them to payroll taxation, all else being equal, and doing both would raise the most revenue.

Even if the average income tax rate and the average payroll tax rate for enrollees in employment-based plans were the same, subjecting contributions to income taxation and to payroll taxation would have very different effects on the tax liability of people in different income groups. Higher-income people are likely to have higher marginal income tax rates but lower marginal payroll tax rates than lower-income people. Among people with employment-based insurance, therefore, subjecting contributions to income taxation would raise the tax liability of higher-income people more than that for lower-income people. The opposite would be true if contributions were subjected to payroll taxation.

Subjecting contributions to taxation would reduce insurance coverage, but the reduction would be smaller if the contributions were subjected to income taxation than if they were subjected to payroll taxation (provided that the same upper limit applied in each case). That difference is primarily attributable to the fact that lower-income people are more likely than higher-income people to forgo insurance when the after-tax price of their insurance goes up. (Higher-income people are more likely to stay enrolled in insurance—because they tend to have more assets to protect, higher demand for health services, and a larger penalty to pay if they forgo insurance.) Also, for lower-income people, the average marginal tax rate is smaller for income taxes than for payroll taxes. Subjecting their contributions to income taxation would not reduce their after-tax compensation (and thus increase the after-tax price of their health insurance) as much as subjecting their contributions to payroll taxation would. They would be less likely to forgo insurance, and overall reductions in insurance coverage would be smaller. At the same time, because higher-income people, on average, face a higher marginal income tax rate than marginal payroll tax rate, more higher-income people would stop enrolling in insurance if their contributions were subjected to income taxation than if they were subjected to payroll taxation. However, that reduction in insurance coverage for higher-income people would be smaller than the reduction for lower-income people because higher-income people are less responsive to price changes in health insurance.

Replacing the Tax Exclusions With a Tax Credit. Another approach to reducing tax preferences for employment-based health insurance would be to replace the current tax exclusions with an income tax credit. If the credit was a fixed dollar amount for everyone and was refundable—so that people could receive money back from the government if their credit exceeded the amount of federal income tax that they owed—all workers would receive the same value from the credit, regardless of their tax bracket or their health care costs. If the credit was a fixed dollar amount but was nonrefundable, low-income workers, who have little or no income tax liability, would benefit much less. Alternatively, the credit’s value might not be a fixed dollar amount; it could be phased out for people with higher income. In any of those designs, the credit would have a set dollar value for a given worker, so that the worker could not increase it by purchasing more extensive or more costly insurance.

Lawmakers would face various trade-offs as they set the value of such a tax credit. A larger credit would increase the number of people who obtained health insurance, but would reduce the amount of tax revenues collected. Phasing down the credit for people with higher income would focus it on people who would be less likely to obtain insurance otherwise, but that approach also would raise effective income tax rates for people whose credit was being phased down, potentially distorting their decisions about how much to work.

One disadvantage of switching to a refundable tax credit is that administering it would be substantially more complex than administering the current tax exclusions. A potential drawback of a flat tax credit is that it would offer the same benefit to everyone, regardless of their health status. The current tax exclusions, by contrast, offer an extra benefit to people who are less healthy, because those people tend to use more health services and to enroll in plans with higher premiums.

Specific Alternatives and Estimates

CBO and JCT analyzed three alternatives for reducing the tax preference for employment-based health insurance. Each alternative would take effect in 2020, and all would follow the first approach outlined above, replacing the excise tax on high-cost plans with a limit on the tax exclusions. Two alternatives would limit the exclusions from income and payroll taxation; the third would limit the exclusion from income taxation but continue the unlimited exclusion from payroll taxation. Those policy changes would increase the tax liability and affect the behavior of people with high premiums for employment-based health plans, but the specific increases in taxes and changes in behavior would be different under each approach.

Replace the Excise Tax With a Limit on the Income and Payroll Tax Exclusions Set at the 50th Percentile of Premiums. The first alternative would eliminate the excise tax and instead impose a limit on the extent to which employers’ and employees’ contributions for health insurance premiums—and to FSAs, HRAs, and HSAs—could be excluded from income and payroll taxation. Specifically, starting in 2020, contributions that exceeded $7,700 a year for individual coverage and $19,080 for family coverage would be included in employees’ taxable income for both income and payroll taxes. Those limits, which are equal to the estimated 50th percentile of health insurance premiums paid by or through employers in 2020, would be indexed for inflation after 2020 by means of the CPI-U. The same limits would apply to the deduction for health insurance available to self-employed people. Because the limits would be lower than the thresholds scheduled to take effect for the excise tax—for example, $10,800 for individual coverage in 2020—federal tax subsidies would be lower as well.

This alternative would decrease cumulative federal deficits by $429 billion by 2026, CBO and JCT estimate. By reducing the appeal of employment-based health insurance, it also would cause about 4 million fewer people to have such coverage in 2026 than would have it under current law. Of those people, about 2 million would buy coverage through the health insurance marketplaces, fewer than 500,000 would enroll in Medicaid, and about 1 million would be uninsured. (Those numbers do not add up to the total because of rounding.)

The reduction in the deficit would stem from several changes in revenues and outlays that partially offset each other. Income and payroll tax revenues would rise by $547 billion through 2026 because the number of people with employment-based coverage would decline and because many of those who retained such coverage would receive a smaller benefit from the tax exclusion. (For example, in 2026, the capped tax exclusions would reduce the combined federal income and payroll tax liability of people with employment-based coverage by an average of $1,420; that reduction would be $5,280 under current law.) Additional penalty payments by certain employers and individuals resulting from changes in health insurance coverage also would increase revenues, although only by a small amount. However, additional tax credits for coverage purchased through the marketplaces would reduce revenues, as would the repeal of the excise tax. In all, revenues through 2026 would be $476 billion higher than under current law. The alternative also would boost federal outlays by $47 billion through 2026, primarily because of increased spending on Medicaid and on subsidies for insurance purchased through the marketplaces.

Replace the Excise Tax With a Limit on the Income and Payroll Tax Exclusions Set at the 75th Percentile of Premiums. Just as the first alternative would, the second alternative would eliminate the excise tax and impose limits on the extent to which contributions could be excluded from income and payroll taxation. In this alternative, however, the limits would be higher: $9,520 a year for individual coverage and $23,860 for family coverage. Those limits are equal to the estimated 75th percentile of health insurance premiums paid by or through employers in 2020. Again, they would be indexed for inflation by means of the CPI-U after 2020.

The second alternative would decrease cumulative federal deficits by $174 billion by 2026, CBO and JCT estimate. Specifically, it would increase revenues by $193 billion and outlays by $19 billion. Also, like the first alternative, this one would reduce the appeal of employment-based health insurance, causing about 2 million fewer people to have it in 2026 than would have it under current law. In that year, about 1 million more people would buy coverage through the marketplaces, fewer than 500,000 more people would enroll in Medicaid, and about 1 million more people would be uninsured.

Replace the Excise Tax With a Limit on Only the Income Tax Exclusion Set at the 50th Percentile of Premiums. The third alternative would eliminate the excise tax and impose a limit on the extent to which contributions could be excluded from income taxation; exclusions for payroll taxation would remain unlimited. Specifically, starting in 2020, contributions that employers or workers made for health insurance—and for health care costs through FSAs, HRAs, and HSAs—that exceeded $7,700 a year for individual coverage and $19,080 for family coverage would be included in employees’ taxable income for income taxes. Those are the same limits as the ones in the first alternative, and once again, they would be indexed for inflation in subsequent years by means of the CPI-U. As the discussion above explained, limiting the tax exclusion for income taxes only would raise more revenue, and reduce insurance coverage less, than limiting the exclusion for payroll taxes only would (so long as the same limit applied in each case).

The third alternative would decrease cumulative federal deficits by $254 billion by 2026, CBO and JCT estimate: Revenues would be $283 billion higher, and outlays would be $29 billion higher. That alternative would cause about 3 million fewer people to have employment-based insurance in 2026 than would have it under current law. Of those people, about 2 million would buy coverage through the health insurance marketplaces, fewer than 500,000 would enroll in Medicaid, and about 1 million would be uninsured.

Other Considerations

Reducing tax preferences for employment-based health insurance would affect many aspects of health care in the United States, including the growth of health care costs, the health of the population, the decisions that employers and workers make about insurance coverage, and the number of people without health insurance.

Effects on Health Care Costs. Replacing the excise tax with a limit on the tax exclusions that is lower than the excise tax thresholds would make health care spending lower than it would be under current law. The current tax preferences for employment-based insurance give health insurance plans an incentive to cover more services, to cover more expensive services, and to require enrollees to pay a smaller share of the costs than would be the case otherwise. The excise tax will effectively scale back those tax preferences. The alternatives examined here would increase taxes for a larger share of employment-based plans than the excise tax will—giving employers and their workers less incentive to buy expensive health insurance, reducing upward pressure on the price and use of health care, and encouraging greater use of cost-effective care.

Effects on People’s Health. By reducing the incentive to buy expensive coverage and increasing the incentive to buy insurance plans in which people pay more out of pocket, all three of the alternatives analyzed here would reduce the amount of care received and worsen some people’s health. That conclusion is supported by an experiment conducted by the RAND Corporation from 1974 to 1982 in which nonelderly participants were randomly assigned to health insurance plans. The experiment showed that plans requiring more out-of-pocket payments reduced the use of both effective and less effective care, as defined by a team of physicians. Differences in out-of-pocket requirements had no effect on most participants’ health, but among the poorest and sickest participants, those who faced no requirements of that kind were healthier by some measures than those who did.

Effects on Employers and Workers. By increasing the tax liability of people enrolled in high-cost employment-based plans more than the excise tax will, the alternatives considered here would probably increase the financial burden on some people with substantial health problems. In particular, some employers and workers would avoid the increased tax liability by shifting to plans with lower premiums and requirements for more out-of-pocket payments, which would increase costs the most for people who used the most services.

In general, workers with higher income face higher income tax rates and are more likely to enroll in plans with high premiums. Therefore, limiting the exclusion from income taxation, as the third alternative does, would reduce that benefit more for people with higher income. The two alternatives that limit the exclusion not only for income taxation but also for payroll taxation would still increase tax liabilities more for higher-income people, on average, because they tend to enroll in plans with higher premiums.

Under all three alternatives, employees of firms that had a less healthy workforce or that operated in an area with above-average health care costs would be more likely to see their tax liability increase. In higher-cost areas, those increases in people’s tax liability might exert pressure on health care providers and insurers to reduce prices or decrease unnecessary care.

Although these alternatives would reduce total spending on health care, they would increase after-tax premiums for some people enrolled in employment-based insurance, particularly those whose premiums were above the limits imposed by each alternative and who therefore would newly be paying taxes on that portion of their premiums. In addition, because all three alternatives would impose a limit on the exclusion that was lower than the excise tax thresholds that exist under current law, employers would have a heightened incentive to keep premiums low, which could cause them to refrain from hiring older workers (who tend to spend more on health care and to raise average premiums) or to reduce the compensation of older workers. That effect would be particularly likely among employers with fewer employees over whom to spread risks.

Effects on the Number of Uninsured People. The tax increases in these alternatives would lead fewer employers to offer health insurance, thus increasing the number of uninsured workers. Most people whose employers stopped offering coverage would buy it in the nongroup market, either in the health insurance marketplaces or elsewhere. The federal subsidies available to low-income people through the marketplaces would give many of those people an affordable alternative to the employment-based coverage that they had lost, and the penalty for lacking insurance would give many high-income people an incentive to buy insurance even without a subsidy. Nevertheless, some workers whose employers stopped offering health insurance would forgo coverage, CBO and JCT anticipate.