Sherry A. Glied, Adam Striar
S. Glied and A. Striar, Looking Under the Hood of the Cadillac Tax, The Commonwealth Fund, June 2016.
One effect of the Affordable Care Act’s “Cadillac tax” (now delayed until 2020) is to undo part of the existing federal tax preference for employer-sponsored insurance. The specific features of this tax on high-cost health plans—notably, the inclusion of tax-favored savings vehicles such as health savings accounts (HSAs) in the formula for determining who is subject to the tax—are designed primarily to maximize revenue and minimize coverage disruptions, not to reduce health spending. Thus, at least initially, these savings accounts, rather than enrollee cost-sharing or other plan features, are likely to be affected most by the tax as employers act to limit their HSA contributions. Because high earners are the ones benefiting most from tax-preferred accounts, the high-cost plan tax will probably be more progressive than prior analyses have suggested, while having only a modest impact on total health spending.
The Affordable Care Act’s high-cost plan tax (HCPT), popularly known as the “Cadillac tax,” is a 40 percent excise tax on employer plans exceeding $10,200 in premiums per year for individuals and $27,500 for families. The tax is scheduled to take effect in 2020. Employer and employee premium contributions will count against the threshold, as will most employer and (pretax) employee contributions to health savings accounts (HSAs), Archer medical savings accounts (MSAs), flexible spending accounts (FSAs), and health reimbursement accounts (HRAs).1 Guidelines for determining which coverage is taxed are partly subject to interpretation by the Internal Revenue Service (IRS); the information provided here reflects the most up-to-date guidance.2,3
Previous studies have highlighted several problems with the HCPT. Its effect is likely to vary according to regional differences in wages, input prices, and other factors that affect premiums, many of which are beyond the control of employers and employees.4 The tax will probably also disproportionately affect workers with high health care costs and those who have negotiated generous benefit packages, such as public-sector union workers.5
This brief focuses on how the HCPT is likely to affect health care spending. Initially, the tax will probably have little impact on plan offerings but is likely to affect tax-preferred savings vehicles. Our analysis of data from the National Compensation Survey and the Survey of Consumer Finances suggests that the tax will be more progressive than prior analyses have indicated and will have smaller short-term effects on health spending.
Internal Revenue Service (IRS) clarifies earlier administrative ruling by adding provision to tax code formally excluding ESI from taxable employee income.6
IRS formally recognizes “cafeteria plans,” which allow employees to choose from array of benefits and exclude value of these benefits from gross income.7
Flexible spending accounts (FSAs) become popular.8 Established by employers, they typically involve pretax salary-reduction arrangements.
Tax treatment of ESI extended to self-employed individuals.9
Health Insurance Portability and Accountability Act of 1996 (HIPAA) establishes Archer medical savings accounts (MSAs)10 for self-employed persons, employees of small firms, and workers enrolled in high-deductible health plans. Employees and employers can contribute on pretax basis.11
IRS formally acknowledges health reimbursement accounts (HRAs), established and funded solely by employers.12
Medicare Prescription Drug, Improvement, and Modernization Act of 2003 authorizes health savings accounts (HSAs), allowing employee and employer pretax contributions.13
By reducing the price of health benefits relative to other goods, the tax exclusion for health insurance encourages employers to offer insurance and employees to obtain coverage.14–18 Economic theory suggests that the tax exclusion encourages people to buy more generous plans than they otherwise would.
By increasing plan generosity, the tax exclusion may lead to higher spending through greater demand for medical care.19 The relationship between plan generosity and medical spending depends on how responsive spending is to out-of-pocket costs. Research suggests that consumers respond to higher cost-sharing by reducing their use of services, particularly for more discretionary expenditures,20 rather than by seeking less costly providers.21–24
In contrast to the effect of the general tax exclusion, favorable tax treatment of savings vehicles encourages people to choose less generous health insurance with higher coinsurance and deductibles. This is certainly the case for MSAs and HSAs, which can be used only in conjunction with high-deductible plans, and may also be true of FSAs, which do not impose such restrictions.25 At the same time, among people with similar coverage, those with savings accounts are likely to use more services than those who do not have such accounts.26
Both the tax exclusion of health insurance and the favorable tax treatment of savings vehicles are regressive in nature. High-income workers benefit the most because they have the highest tax rates and also have access to generous benefits, take up these benefits, and contribute to savings accounts.
The HCPT affects only health expenses above a threshold. There is still a strong tax benefit below this threshold, making it unlikely that employers will respond to the tax by eliminating offerings altogether. This tax benefit is maximized at the threshold ($10,200 for single coverage) (Exhibit 1).
The range of spending types covered by the HCPT may encourage employers to reduce costs by reducing payments to savings accounts, for several reasons. First, these plans have little intrinsic insurance value—higher-income workers (the group most likely to have these plans) can usually use other savings to pay for expenditures.
Second, these savings vehicles are quite widespread and often hold balances that are quite large relative to premiums, so employers can substantially reduce their exposure to the HCPT by adjusting the contribution limits. The most recent estimates suggest that about half of all American workers had access to an FSA. In 2015, 26 percent of firms offering benefits offered at least one HSA or HRA compatible plan, and 24 percent of covered workers were enrolled in such plans.27 The latter figure has increased by 11 percentage points since 2010.28
Employer contributions to HSAs averaged $568 for single coverage and $991 for family coverage in 2015.29 Since some of these accounts go unfunded, these figures rise to $809 and $1,412, respectively, when including only funded accounts.30 Employer contributions to HRAs followed similar patterns, with employers contributing an average of $1,079 for single coverage and $2,001 for family coverage.31 These contributions constituted a significant proportion of employer spending on employee health benefits altogether. Employer HSA contributions accounted for approximately 11 percent of total employer contributions (including premiums) to single plans and 8 percent of contributions to family plans.32 For HRA plans, these figures were 18 percent for single plans and 14 percent for family plans.33
Third, in addition to these employer contributions, employees can choose how much to deposit into their FSA or HSA (up to predefined limits established by the IRS and the employer), but employers are responsible for making HCPT payments if total costs exceed the limits.34 Thus, employers face increased exposure to unexpected tax liabilities. Surveys have indicated that employers might eliminate or limit savings accounts because of their concern about these liabilities.35,36
The availability of savings accounts may encourage people to enroll in less generous health insurance plans with lower-than-average premiums, offsetting the costs associated with savings accounts. In principle, this might mean that offering tax-favored savings accounts reduces overall payments for employer-based health care plans. In practice, however, the value of applicable coverage from these plans—which includes employee contributions—often exceeds that of traditional preferred provider organizations (PPOs) and health maintenance organizations (HMOs). In 2015, the average premium savings of $1,008 for single coverage was less than the average contribution to an HSA (including both contributors and noncontributors). By 2018, the total cost of the average HSA-compatible high-deductible health plan, including premiums and account contributions, is projected to be $1,653 more than the cost of the average HMO plan.37 The combined cost is even higher with FSAs, which are usually coupled with PPO-type plans—the most costly plan types.38,39
Reducing tax-favored contributions to FSAs and HSAs will raise tax revenues. Paradoxically, these changes may not reduce health care spending. The immediate effect will be increased out-of-pocket costs because more of these costs will have to be met with after-tax dollars, which is likely to reduce total spending. But the changes will ultimately slow participation in high-deductible plans, which is likely to increase spending, since people with higher cost-sharing plans, such as the high-deductible plans required for HSAs, use less medical care than do those with conventional plans.
Prior analyses of the HCPT have shown that it is relatively progressive.40,41 However, these analyses may understate its progressivity, at least in the early years, because information about health savings account ownership and contributions is not readily available. Ownership and funding of FSAs and HSAs are more skewed toward high-income workers than are health benefits more broadly. Studies show that high earners are much more likely to fund savings accounts and to fund them fully.42,43
One measure of the use of FSAs and HSAs is whether an employer offers the accounts to employees. Employers of highly paid employees are most likely to offer such accounts (Exhibit 2).44
Exhibit 2. Workers’ Access to Flexible Spending Accounts and Health Savings Accounts, by Wage Percentile, 2015
|Low wage||Middle wage||High wage|
|Account||<10th percentile||10th–24.99th percentile||25th–49.99th percentile||50th–74.99th percentile||75th–89.99th percentile||≥90th percentile|
Note: FSA = flexible spending account; HSA = health savings account.
Source: Bureau of Labor Statistics, National Compensation Survey, 2015.
However, many employees who are offered savings accounts do not use them or do not make significant contributions to them. Our analysis of the available data show that account ownership is skewed toward higher-income earners and that account balances are highest among the highest earners (Exhibit 3).45 High earners benefit most from making contributions to these accounts because they have both the highest tax rates and the most discretionary income available for saving.
Exhibit 3. Ownership and Average Balances of Tax-Favored Savings Accounts, 2007–2013
|Weighted average balance ($)|
|Income percentile||Households (%)||All accounts combined||Smallest account|
Note: Data are for households with at least one savings account, no children at home, and a head of household under the age of 65. All dollar amounts are in 2013 dollars. Responses used to calculate these figures may include balances in Coverdell or 529 (state-sponsored) education savings accounts, health savings accounts, and/or medical savings accounts.
Source: Federal Reserve Board, Survey of Consumer Finances, 2007, 2010, and 2013.
As noted above, employers are initially likely to respond to the HCPT by limiting employer and employee contributions to FSAs and HSAs. This change will have the greatest effect on high-income employees, who are most likely to be contributing substantially to these accounts and who face the greatest gap between pre- and post-tax dollars.
The data suggest that in addition to the primary effect of reducing contributions to savings accounts, the HCPT will have secondary effects on employee plan choices. With reduced contributions, employees will favor plans with lower premium and out-of-pocket costs, since they will have to pay for these costs with after-tax dollars.46 This preference, in turn, will encourage employers to offer plans with similar attributes.
In general, employers have two options for reducing the cost of employee health coverage: increasing cost-sharing or offering plans with more restrictive provider networks. Since the HCPT makes higher cost-sharing less attractive to employees, it will probably lead to somewhat greater participation in plans with narrow networks than previously projected.
The HCPT was designed to maximize revenues and minimize the possibility that employers will drop health insurance coverage. By taxing only the portion of plan costs that exceeds a threshold, the HCPT ensures that most employer-sponsored benefits will remain intact. However, the HCPT limits the tax benefits of health savings accounts, which may reduce the appeal of high-deductible plans. While the HCPT does not erase the distortionary effects of the favorable tax treatment of health insurance, it is a more progressive tax. Finally, employers may respond to the HCPT by offering plans with narrow networks rather than increased cost-sharing.
About This Study
This analysis uses data from the 2015 National Compensation Survey (NCS) conducted by the Bureau of Labor Statistics (BLS) and the 2007, 2010, and 2013 Survey of Consumer Finances (SCF) from the Board of Governors of the Federal Reserve (FRB). The NCS component uses descriptive statistics to examine differences in access to FSAs and HSAs according to income groupings. The BLS supplies these income groupings, which are calculated on the basis of average hourly rates from the March 2015 Employer Costs for Employee Compensation survey conducted by the BLS.
The SCF component combines the 2007, 2010, and 2013 surveys for a more robust sample. All dollar figures are indexed to the 2013 Consumer Price Index for All Urban Consumers. This component uses descriptive statistics to examine differences in HSA ownership and account balances according to income category. This survey combines HSA ownership and account balance statistics with those for Coverdell and 529 (state-sponsored) education accounts. To isolate HSAs, we excluded households with one or more children present and households in which the head of the household was over the age of 64. Using the weights supplied with the survey, we calculated two different measures of account balances across income categories for households with at least one account: the weighted mean of combined balances across all household accounts and the weighted mean balance of the lowest-value account. Weights were used to adjust for sampling error by approximating an equal-probability sample.
Savings Account Characteristics for Each Income Percentile, 2013
Note: Data are for households with at least one savings account, no children at home, and a head of household under the age of 65.
1 On-site medical clinics offering more than de minimis care, retiree coverage, multiemployer plans, and executive physical programs are also subject to the tax. Tax-favored contributions to vision and dental plans, long-term care coverage, and hospital indemnity coverage are excluded.
2 K. Levin, “Section 4980i—Excise Tax on High Cost Employer-Sponsored Health Coverage,” in Notice 2015-16 (Internal Revenue Service, 2015).
3 K. Levin, “Section 4980i—Excise Tax on High Cost Employer-Sponsored Health Coverage,” in Notice 2015-52 (Internal Revenue Service, 2015).
4 S. Nowak and C. Eibner, Rethinking the Affordable Care Act’s “Cadillac Tax”: A More Equitable Way to Encourage “Chevy” Consumption (The Commonwealth Fund, Dec. 2015).
5 B. Herring and L. K. Lentz, “What Can We Expect from the ‘Cadillac Tax’ in 2018 and Beyond?” Inquiry, Winter 2011–12 48(4):322–37.
6 M. A. Thomasson, “The Importance of Group Coverage: How Tax Policy Shaped U.S. Health Insurance,” American Economic Review, Sept. 2003 93(4):1373–84.
7 L. E. Irish, “Cafeteria Plans in Transition,” William & Mary Annual Tax Conference, 1984.
8 J. Mulvey, Health Care Flexible Spending Accounts (Congressional Research Service, June 13, 2012).
9 “Self-Employed Health Insurance Deduction,” in “Knowledge Base” article TaxSlayer Pro.
10 Presentation by Thomas Wilder on behalf of America’s Health Insurance Plans, National Committee on Vital and Health Statistics: United States Department of Health and Human Services, 2015.
11 C. Rapaport, Tax-Advantaged Accounts for Health Care Expenses: Side-by-Side Comparison, 2013 (Congressional Research Service, Nov. 8, 2013), pp. 1–15.
13 J. O’Sullivan, H. Chaikind, S. Tilson et al., Overview of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (Congressional Research Service, 2004), p. 28.
14 M. Smart and M. Stabile, Tax Credits and the Use of Medical Care (National Bureau of Economic Research, July 2003).
15 M. V. Pauly, “Taxation, Health Insurance, and Market Failure in the Medical Economy,” Journal of Economic Literature, June 1986 24(2):629–75.
16 M. Stabile, “The Role of Tax Subsidies in the Market for Health Insurance,” International Tax and Public Finance, Jan. 2002 9(1):33–50.
17 A. B. Royalty, “Tax Preferences for Fringe Benefits and the Health Insurance Offered by Employers,” Journal of Public Economics, Feb. 2000 75(2):209–27.
18 J. Gruber and J. Poterba, “Tax Incentives and the Decision to Purchase Health Insurance: Evidence from the Self-Employed,” Quarterly Journal of Economics, Aug. 1994 109(3):701–33.
19 M. V. Pauly, “Taxation, Health Insurance, and Market Failure in the Medical Economy,” Journal of Economic Literature, June 1986 24(2):629–75.
21 N. Sood, Z. Wagner, P. J. Huckfeldt et al., “Price Shopping in Consumer-Directed Health Plans,” Forum for Health Economics and Policy, March 2013 16(1):1–19.
22 N. Lurie, C. J. Kamberg, R. H. Brook et al., “How Free Care Improved Vision in the Health Insurance Experiment,” American Journal of Public Health, May 1989 79(5):640–42.
23 W. G. Manning, H. L. Bailit, B. Benjamin et al., “The Demand for Dental Care: Evidence from a Randomized Trial in Health Insurance,” Journal of the American Dental Association, June 1985 110(6):895–902.
24 A. M. Haviland, M. D. Eisenberg, A. Mehrotra et al., Do “Consumer-Directed” Health Plans Bend the Cost Curve over Time? NBER Working Paper No. 21031 (National Bureau of Economic Research, March 2015).
25 W. Jack, A. Levinson, and S. Rahardja, “Employee Cost-Sharing and the Welfare Effects of Flexible Spending Accounts,” Journal of Public Economics, Dec. 2006 90(12):2285–301.
26 D. K. Remler and S. A. Glied, “How Much More Cost Sharing Will Health Savings Accounts Bring?” Health Affairs, July–Aug. 2006 25(4):1070–78.
27 G. Claxton, M. Rae, M. Long et al., Employer Health Benefits: 2015 Annual Survey (Henry J. Kaiser Family Foundation/Health Research and Educational Trust, Sept. 2015).
34 T. Berkley, HSA Account Contributions Through Payroll Face Risk of Early Elimination If Not Exempted from “Cadillac Tax” (HSA Consulting Services, LLC, May 26, 2015).
36 B. Faler, “‘Cadillac Tax’ Could Wreck Popular Medical Accounts,” Politico, Aug. 31, 2015.
37 B. Herring and L. K. Lentz, “What Can We Expect from the ‘Cadillac Tax’ in 2018 and Beyond?” Inquiry, Winter 2011–12 48(4):322–37.
38 G. Claxton, M. Rae, M. Long et al., Employer Health Benefits: 2015 Annual Survey (Henry J. Kaiser Family Foundation/Health Research and Educational Trust, Sept. 2015).
39 G. Claxton and L. Levitt, How Many Employers Could Be Affected by the Cadillac Plan Tax? (Henry J. Kaiser Family Foundation, Aug. 2015).
40 L. J. Blumberg, J. Holahan, and G. B. Mermin, The ACA’s “Cadillac” Tax Versus a Cap on the Tax Exclusion of Employer-Based Health Benefits: Is This a Battle Worth Fighting? (Urban Institute, Oct. 2015).
41 G. B. Mermin and E. Toder, Distributional Impact of Repealing the Excise Tax on High-Cost Health Plans (Tax Policy Center, July 2015).
42 U.S. Government Accountability Office, Health Savings Accounts: Participation Increased and Was More Common Among Individuals with Higher Incomes (GAO, April 2008).
43 L. A. Helmchen, D. W. Brown, I. Z. Lurie et al., “Health Savings Accounts: Growth Concentrated Among High-Income Households and Large Employers,” Health Affairs, Sept. 2015 34(9):1594–98.
44 Bureau of Labor Statistics, “National Compensation Survey—Benefits” (BLS, U.S. Department of Labor, 2010–2015).
45 We limited our sample to families without children at home in order to exclude families most likely to have education savings accounts.
46 Reducing contributions will probably also encourage employees to participate in dental, vision, and long-term care insurance plans, which remain sheltered from tax.