Sarah Nowak, Christine Eibner
S. Nowak and C. Eibner, Rethinking the Affordable Care Act’s “Cadillac Tax”: A More Equitable Way to Encourage “Chevy” Consumption, The Commonwealth Fund, December 2015.
The Affordable Care Act’s “Cadillac tax” will apply a 40 percent excise tax on total employer health insurance premiums in excess of $10,200 for single coverage and $27,500 for family coverage, starting in 2018. Employer spending on premiums is currently excluded from income and payroll taxes. Economists argue that this encourages overconsumption of health care, favors high-income workers, and reduces federal revenue. This issue brief suggests that the Cadillac tax is a “blunt instrument” for addressing these concerns because it will affect workers on a rolling timetable, does relatively little to address the regressive nature of the current exclusion, and may penalize firms and workers for cost variation that is outside their control. Replacing the current exclusion with tax credits for employer coverage that scale inversely with income might allow for regional adjustments in health care costs and eliminate aspects of the tax exclusion that favor high-income over low-income workers.
How much of the variation in health insurance premiums across states is explained by factors that can be controlled by enrollees and insurers (such as plan generosity or utilization management), and how much is the result of factors outside of enrollees’ control (such as worker demographics, regional variation in health care costs, and competition)? In this issue brief, we explore that question to understand how and when a new tax created by the Affordable Care Act, due to take effect in 2018, will affect workers’ costs. We also offer an alternative proposal.
Under current U.S. tax policy, employer health insurance spending is excluded from income and payroll taxes. Economists have criticized this exclusion on the grounds that it reduces federal tax revenue and provides a greater benefit to people with higher incomes: Since these individuals pay more in taxes, they have more to gain when an employer-provided benefit such as health insurance is excluded from tax calculations.
There is also concern that the tax exclusion could lead firms and workers to choose overly generous health insurance coverage with low out-of-pocket cost-sharing requirements, broad provider networks, and little or no incentive to use cost-effective care. Overly generous plans could increase health care utilization and contribute to health care cost inflation,1 a concern supported by empirical research showing that individuals in more generous health plans use more health care.2 When individuals move to less generous plans, they often use less needed and unneeded care.3
The Affordable Care Act’s so-called Cadillac tax is an attempt to address these concerns. It applies a 40 percent excise tax on total employer premium spending in excess of $10,200 for single coverage and $27,500 for family coverage, starting in 2018.4 Premium spending below these thresholds will continue to be excluded from taxable income. The threshold will grow at the level of the consumer price index (CPI) plus one percentage point in 2018 and 2019, and keep pace with CPI growth starting in 2020. Since health insurance premiums have historically risen faster than CPI, an increasing share of firms will likely be affected by the Cadillac tax each year. The law allows for some adjustments to the Cadillac tax thresholds to avoid penalizing firms based on the age and gender of their workers, but there is no adjustment for regional variation in health care costs, and limited ability to make adjustments based on industry.5
The Cadillac tax does little to address the regressive nature of the employer tax exclusion—high-income workers will still benefit more than low-income workers. But capping the exclusion may slow health insurance premium growth by reducing consumers’ health care utilization.
Once the Cadillac tax takes effect, the Congressional Budget Office and the Joint Committee on Taxation estimate that relatively few employers will continue to offer a taxed plan6; those that do will face higher premiums. More likely, employers will switch to plans that require higher cost-sharing for workers, plans with narrow networks, and plans that strictly manage utilization through the use of gatekeepers for specialty care and preauthorization for expensive drugs or treatments. These plan features are meant to limit the use of low-value health care services, slowing the growth of health care costs.7
How This Study Was Conducted
To conduct the analysis, we used state-level data from the Medical Expenditure Panel Survey-Insurance Component (MEPS-IC), the Centers for Medicare and Medicaid Services (CMS), the U.S. Bureau of Labor Statistics (BLS), and other sources to estimate factors that contribute to cross-state variation in health insurance premiums. In addition, we used data from the MEPS-IC, coupled with health care cost inflation trends reported by the Congressional Budget Office8 to estimate when 10 percent or 50 percent of workers with employer-sponsored insurance in each state were likely to be affected by the Cadillac tax. Additional details on our approach can be found in our separate technical appendix.
But the Cadillac tax could be a blunt instrument for promoting cost-conscious care because it does not adjust for regional differences in wages, input prices, and other factors that may affect the cost of health insurance premiums but are outside the direct control of firms and their workers.
To understand the relationship between regional variation and the tax’s effect on workers, we estimate the year in which at least 10 percent and at least 50 percent of each state’s employer-insured workforce will be affected by the tax. Workers are considered to be affected by the tax if they (or their employer) must pay the tax, or if their employer would have to alter its current benefits package to avoid the tax.
Assuming that premium inflation exceeds CPI growth, the share of workers affected by the tax will grow over time and vary across states because of regional differences in premium levels. Because premiums for most employer-sponsored plans are below the Cadillac tax thresholds, only a small share of workers is likely to be affected initially (e.g., before 2020), but that percentage grows over time. We do not assess the types of avoidance strategies that workers and firms might take, or the effect of this tax avoidance on enrollees’ health care consumption and health outcomes. In addition, we do not quantify whether or which firms and workers are likely to pay or avoid the tax.
Most of what influences state-by-state variation in employer premiums is outside the control of businesses and their workers.9 This includes demographics, the share of people working in the health care sector, industry, and regional differences in health care costs. Exhibit 1 shows that these factors account for more than half the variation in premiums. “Plan generosity,” which encompasses plan deductibles and whether the plan is an HMO (an indicator of a relatively restrictive network and relatively strong care utilization management practices) explains a comparatively small share of total variance in premiums across states.
Exhibit 2 shows the projected year in each state when at least 10 percent of workers with employer-sponsored health plans will be affected by the Cadillac tax. In two states, Alaska and New Jersey, 10 percent of workers with employer-sponsored insurance could be affected by the tax as early as 2018. Fifteen states, including California, Illinois, New York, Connecticut, and New Hampshire, could also have as many as 10 percent of workers affected in the near future (between 2020 and 2022). But most states will not have 10 or more percent of workers affected until after 2023.
Exhibit 3 shows similar differences in the amount of time it will take for at least 50 percent of a state’s workers to be affected by the tax. Again, Alaska is at the front of the pack, along with New Hampshire and Connecticut, with 50 percent of workers affected before 2032. California, Texas, Wisconsin, and several states in the Midwest and Northeast also have more workers affected sooner. However, it will be almost 20 years before a majority of workers in most states are affected by the tax.
Exhibits 2 and 3 show that workers with employer-sponsored insurance will be affected by the Cadillac tax much earlier in some states than in others. Such regional variation could make it difficult in the relatively near term for low-income workers in some states and regions to afford needed care if, for example, they are faced with the choice of paying the tax or moving to a plan that requires extremely high cost-sharing.10
One alternative approach would be to restructure the current employer-tax exclusion to more closely parallel the tax credit approach used in the Affordable Care Act’s marketplaces. Marketplace tax credits are based on the price of the second-lowest cost silver plan available in an individual’s community, and vary in proportion to the enrollees’ income. Individuals with incomes between 100 percent and 400 percent of the federal poverty level are eligible for tax credits if they do not have an affordable offer of coverage from another source. Eligible enrollees must contribute a certain percentage of their income toward health insurance; these percentage contributions increase as income goes up. Once enrollees meet the required contribution, they receive a tax credit that covers the remaining premium cost, up to the price of the second-lowest-cost silver plan available in the enrollees’ community. The silver plan has a 70 percent actuarial value, meaning that—on average—the plan pays for 70 percent of an enrollee’s health care expenditure.
In contrast, the average actuarial value for family coverage through an employer plan is 83 percent, and—in the year the Affordable Care Act was signed—nearly 25 percent of enrollees in employer coverage had plans with actuarial values above 90 percent.11 While marketplace enrollees may purchase a more expensive plan if they wish, they must pay out-of-pocket for the remaining premium costs, over and above the cost of the second-lowest-cost silver plan. Because the silver plan provides less generous benefits than a typical employer plan, the marketplace approach encourages “Chevy” rather than “Cadillac” health care consumption.
The Affordable Care Act’s marketplace tax credits provide a model for subsidizing employer-sponsored coverage that is progressive and less prone to regional variation than the employer tax exclusion. Eliminating the current tax exclusion for employer insurance and offering employees a tax credit when they enroll in employer coverage could allow for adjustments based on the regional cost of health care. Tax credit amounts could scale with income, and—like marketplace tax credits—they could be capped at the cost of a 70 percent actuarial value plan. Like the Cadillac tax, the employer tax credit would be applied to the total plan premium—this would reduce costs for both workers and businesses. Businesses might in turn pass these savings back to workers in the form of higher wages.12 Such an approach could insure more people than the ACA at no additional cost to the federal government.13
Such a large-scale change, however, could disrupt the employer insurance market and cause some firms to drop coverage. A compromise solution could retain the current Cadillac tax but make adjustments based on regional variation in price levels, and potentially other factors, such as industry. Policymakers could also consider a “safe harbor” approach, in which the Cadillac tax would not be applied if a worker could not obtain a minimum-generosity plan at a premium below the Cadillac tax limit. This approach could be helpful for firms and workers in high-cost areas, who may have trouble finding even “bare bones” plans at prices below the Cadillac tax threshold if health care cost growth continues to outpace CPI growth.
While few firms are likely to face this type of situation in the short run, it is possible over time. For example, some older, small-group workers in Alaska already face employer premiums that are approaching the Cadillac tax threshold, even for bronze plans. The bronze plan covers on average 60 percent of an enrollees’ health expenditure, the minimum level of coverage necessary to meet the Affordable Care Act’s individual mandate requirements. While the Cadillac tax threshold will eventually be adjusted to account for employees’ ages (details of the adjustment have not yet been finalized), it is conceivable in some high-cost areas that even minimally generous employer plans will eventually be subject to the Cadillac tax.
The safe harbor approach would prevent businesses and their workers from facing the Cadillac tax if a minimally generous plan could not be obtained below the Cadillac tax limit. Regulators could define a minimum-generosity plan based on the price of a 60 percent or 70 percent actuarial value plan. Under this approach the effective Cadillac tax limit would become the minimum of either the local bronze (or silver) premium or the Cadillac tax amount specified under current law.
The Cadillac tax as currently designed is likely to create inequalities based on geographic differences in costs and other factors outside of workers’ control, without addressing the regressive aspects of the current employer tax exclusion. Alternative approaches to taxing generous health plans, including providing income-base tax credits to workers who enroll in employer-sponsored plans, might be equally effective at containing health care cost growth, and would be more equitable.
1 M. S. Feldstein, “The Welfare Loss of Excess Health Insurance,” Journal of Political Economy, March–April 1973 81(2 Pt. 1):251–80; H. A. Chernick, M. R. Holmer, and D. H. Weinberg, “Tax Policy Toward Health Insurance and the Demand for Medical Services,” Journal of Health
Economics, March 1987 6(1):1–25; and R. Feldman and B. Dowd, “A New Estimate of The Welfare Loss of Excess Health Insurance,” American Economic Review, March 1991 81(1):297–301.
2 W. G. Manning, J. P. Newhouse, N. Duan et al., “Health Insurance and the Demand for Medical Care: Evidence from a Randomized Experiment,” American Economic Review, June 1987 77(3):251–77; and K. Baicker, and D. Goldman, “Patient Cost-Sharing and Healthcare Spending Growth,” Journal of Economic Perspectives, Spring 2011 25(2):47–68.
3 K. Baicker, and D. Goldman, “Patient Cost-Sharing and Healthcare Spending Growth,” Journal of Economic Perspectives, Spring 2011 25(2):47–68; and M. Beewkes Buntin, A. M. Haviland, R. McDevitt et al., “Health Care Spending and Preventive Care in High-Deductible and Consumer-Directed Health Plans,” American Journal of Managed Care, March 2011 17(3):222–30.
4 The Cadillac tax applies to employer-sponsored insurance plans only, and does not affect individual market insurance.
5 The Cadillac tax will be adjusted to account for higher premium costs among high-risk professions, including law enforcement, fire protection, first-responders (e.g., paramedics), longshoremen, and individuals engaged in the construction, mining, agriculture (not including food processing), forestry, and fishing industries.
6 Congressional Budget Office, Options for Reducing the Deficit: 2014 to 2023 (Washington, D.C.: CBO, Nov. 13, 2013), https://www.cbo.gov/content/options-reducing-deficit-2014-2023.
7 An additional concern, not addressed in this issue brief, is that higher cost-sharing could discourage both low-value and high-value care, ultimately leading to worse health outcomes. For a discussion of this issue, see K. Baicker, and D. Goldman, “Patient Cost-Sharing and Healthcare Spending Growth,” Journal of Economic Perspectives, Spring 2011 25(2):47–68.
8 Congressional Budget Office, The Budget and Economic Outlook: 2015 to 2025 (Washington, D.C.: CBO, Jan. 26, 2015), https://www.cbo.gov/publication/49892.
9 Potentially, health care costs may be affected by plan generosity at the regional level, for example if regions with more generous health care benefits have higher utilization, which in turn drives up prices. We have attempted to control for these issues in our model. For example, our health care cost index is net of cost differences that are driven by patient utilization decisions, such as whether to see a low-cost nurse practitioner rather than a higher-cost primary care physician.
10 Technically, the tax is levied on insurance companies. However, the CBO anticipates that most of the tax would be passed on to employers in the form of higher premiums. In turn, these higher premiums would likely be passed on from employers to their workers, either in the form of larger employee premium contributions or (over time) reduced wages. For additional explanation, see K. Baicker and A. Chandra, “Myths and Misconceptions About U.S. Health Insurance,” Health Affairs, Nov.–Dec. 2008 27(6):w533–w543.
11 J. R. Gabel, R. Lore, R. D. McDevitt et al., “More Than Half of Individual Health Plans Offer Coverage That Falls Short of What Can Be Sold Through Exchanges as of 2014,” Health Affairs, June 2012 31(6)1339–48.
12 K. Baicker and A. Chandra, “Myths and Misconceptions About U.S. Health Insurance,” Health Affairs, Nov.–Dec. 2008 27(6):w533–w543.
13 E. A. Saltzman, C. Eibner, and A. C. Enthoven, “Improving the Affordable Care Act: An Assessment of Policy Options for Providing Subsidies,” Health Affairs, Dec. 2015 34(12):2095–2103.