Jonathan Cohn would like you to believe that conservatives are so irrational in their hatred of Obamacare that they even despise parts of the law that should make them cheer, such as the Cadillac tax on high cost health plans. Mr. Cohn is correct in asserting that "writers like James Capretta and Robert Moffit have long called for reducing or eliminating the tax breaks for employer sponsored insurance." But there's a world of difference between how conservatives would cap or eliminate the current employer tax exclusion and the Cadillac tax, which even Mr. Cohn concedes "isn't really a tax so much as a convoluted attempt to undo an existing tax break" and which even progressive blogger Matt Yglesias describes as "slightly kludgy."
Although it does not become effective until 2018, the Cadillac tax-under which any employer or health insurer offering a plan that costs more than $10,200 for an individual and $27,500 for a family would typically pay a 40 percent excise tax on "excess benefits," i.e., the amount exceeding the threshold-addresses a very legitimate policy problem. Health benefits are treated differently than cash wages: the latter are subject to federal, state and local income taxes as well as Social Security and Medicare payroll taxes, the former are not. Consequently, taxpayers effectively are providing what amounts to a subsidy that averages 35% of the cost of employer-provided health benefits. This tax-free treatment of employer-provided health coverage is the largest single tax expenditure in the federal budget, resulting in the loss of over $300 billion in federal revenues in 2013; in addition, states and local governments will lose roughly $38 billion in revenues due to the exclusion.
Why The Tax Exclusion is Inefficient
This subsidy is both inefficient and unfair. Because it is open-ended, it encourages employees to cover as much of their medical spending as possible through their employer health plan so as to maximize the amount of spending to which this discount applies. No one uses their auto insurance to put gas in their car (or even to pay for arguably essential preventive maintenance such as oil changes) because these are routine expenses that everyone can and should budget. But because of the tax exclusion 156 million Americans can and do use their health insurance plans to cover the equivalent routine maintenance expenses for their bodies.
But if gas and oil are necessities for your care, what's the big deal if auto insurance pays for them? After all, if you're going to pay for them anyway, who cares whether it's Joe Driver or his auto insurance company? Three adverse consequences flow from using a third party payer to pay this kind of expense:
- First, consumers may become less price-sensitive knowing that all or nearly all of any higher price they pay for oil changes will be borne by everyone else in the insurance pool. Steven Brill's recent Time expose and a more recent New York Times piece on the high cost of colonoscopies should settle any questions about whether this phenomenon is widespread in American medicine. In general, Americans have the highest medical prices on the planet.
- Second, consumers may undertake preventive activities more frequently than they would otherwise (changing oil every 1,000 miles instead of every 3,000). Case in point: about one-quarter of Medicare patients undergo colonoscopies more often than clinically recommended.
- Clearly, some of this wasteful spending can be avoided by erecting rules and monitoring to preclude this, but these in turn lead to the third unavoidable consequence: higher administrative costs. When someone else pays the bill, the payer always will need to undertake at least some form of monitoring activity to ensure that a) the service was needed/allowable; b) the service was actually provided to the customer (the most common forms of Medicare fraud are durable medical equipment never provided and services never performed); and c) that the price did not exceed some specified "reasonable" level. Otherwise that payer may be subject to massive fraud or excessive payouts that would make them uncompetitive compared to payers that were more careful about who got paid what amount. Thus, even if consumers remained prudent shoppers (even though it is not in their interest to do so when someone else is paying most of the tab) and somehow are cajoled into using precisely the identical amount of preventive care that they would if they paid for such care on their own, these administrative costs-even if they amounted to only 5 cents on the dollar- would make buying the service through a third party payer more expensive than if the identical bill had been paid directly by the consumer.
All three reasons explain why we do not observe auto insurance policies that cover the costs of fill-ups and oil changes, nor homeowners policies that cover the cost of mowing the grass. It's more sensible and less expensive to let consumers handle such expenses on their own. In reality, there's a wide range of tastes regarding how often to change one's oil (or mow one's grass), what grade of oil to use etc. We could easily squander a lot of resources needlessly arguing with third party payers over whether a given such expense should be covered.
Economists have measured how much excess spending is induced by the tax exclusion. As I noted in a post a few weeks ago, getting rid of the tax exclusion for employer-based health coverage entirely would lower group premiums by 45%! That's a rough measure of how much waste is encouraged by our current ill-conceived system of subsidized health insurance coverage. Instead of a system where every dollar of extra employer-provided health insurance is subsidized by Uncle Sam, we'd be far better off with a system that provided fixed-dollar contributions to coverage. Under such a system, Americans-among the best shoppers on the planet-would have an incentive to seek out coverage whose premium was close to the amount of the subsidy provided since every dollar of premium above that amount would be paid by the family instead of partially subsidized by taxpayers. This is not an ivory tower fantasy. In essence, it is the very system that has worked extremely well for over 50 years for federal employees and their dependents (not to mention members of Congress).
Why The Tax Exclusion is Unfair
But inefficiency is only half the problem with the employer tax exclusion. The other problem relates to fairness, as shown in the figure below from my book. As currently structured, the tax exclusion provides the biggest subsidies-both in terms of dollar amounts and the percentage of premium that is subsidized-to those in the highest income brackets. Thus, Bill Gates receives a far higher subsidy to pay for his health coverage than does a Microsoft janitor. As well, the subsidy is restricted to employer-sponsored health coverage. Workers at businesses that don't offer insurance, the unemployed (and, to a lesser extent, the self-employed) all have to pay for policies with after-tax dollars.
This is why experts from across the ideological spectrum-not just conservatives-favor gradually eliminating the exclusion by starting with a limit (i.e., "cap") on the dollar amount of the exclusion, with most proposals suggesting that this limit be set at the 75th percentile of premiums (i.e., 25% of employees are currently enrolled in plans whose premiums would exceed this limit).
What's Wrong with the Cadillac Tax
The 40% excise tax on high-cost health plans works to some extent like a cap on the exclusion, i.e., above a certain dollar threshold, there is a very strong incentive to avoid plans with higher premiums. Indeed, to the extent that employers respond by simply swapping such plans for equivalent coverage at a lower cost, the effects on workers would be identical. That is, to workers, it will not matter whether the motivation for an employer swapping coverage came in the form of an excise tax or a cap on the exclusion. In either case, workers will end up with a less expensive plan.
But, of course, if employers could provide equivalent coverage for a lower premium, then in a competitive labor market they presumably already would have done so without the inducement of an excise tax. So the reality is, the path to a lower premium will take the form of either restrictions on coverage (such as managed care) or higher cost sharing. Indeed, most experts, including the Congressional Budget Office, expect employers to respond by the use of higher deductibles and copayments and/or termination of employer contributions to health and flexible spending accounts. This will increase out-of-pocket burdens on workers, though a portion of this added cost likely would be offset by increases in (taxable) cash compensation. But such consequences will be the same regardless of whether the Cadillac tax or a cap on the tax exclusion provides the motivating force to employers to move in this direction.
However, since the tax is expected to raise $80 billion between 2013-2023 (even though it doesn't even start until 2018), it is clear that Congress was not expecting (or perhaps even intending) for employers with high-cost plans to dodge the tax entirely. And this is important since the distributional effects of a Cadillac tax versus gradually eliminating the tax exclusion are substantially different. Consider a low wage worker earning $16,000 and a management worker earning $100,000 whose employer provides a health plan that would qualify it for the 40% tax. (Remember that the 40% applies only to the difference between the actual cost of the plan and whatever dollar threshold applies to that plan-not to the entire premium cost of the plan.) Since the low wage worker has an income too low to pay federal income taxes, the tax exclusion effectively provides a subsidy equal to 15.3% of plan premiums, i.e., the amount of payroll taxes that are avoided because of the exclusion. If single, the higher-paid worker would be in a 28% federal tax bracket, and in my state would pay an additional 7.75% in state income taxes. All told, every dollar of premiums for such a worker would implicitly be subsidized by more than 50% (i.e., 15.3%+28%+7.75%= 51.05%)-effectively a 50 percent off coupon courtesy of U.S. taxpayers. If we're giving out tax-financed coupons to encourage workers to buy health coverage, shouldn't the biggest ones go to families with the lowest incomes?
Moreover, remember that the Cadillac tax does nothing about the underlying tax exclusion despite its inefficiency and unfairness. So the high income worker in this example who is in a $12,000 employer plan for single coverage still is going to get the 51% discount coupon for every penny of those premiums. The excise tax merely requires that worker to essentially give back 40% of the $1,800 difference between the cost of the plan and the $10,200 threshold. Put a different way, the excise tax simply reduces the subsidy from 51% to only 11%. This individual still will be receiving a $198 subsidy for what the government has declared is a gold-plated health benefits plan! But now consider the low wage worker. That worker still would get their 15.3% discount via the tax exclusion on the entire $12,000 premium, but would have to pay the identical 40% penalty on the $1,800 in "excess" premiums above the threshold. Whereas the high-paid worker simply will have the size of their discount coupon reduced by about 80% (from 51% to only 11%), the low wage worker will see his or her discount coupon converted into an actual 25% penalty. In short, the Cadillac tax will repeatedly create the following bizarre situation across employers with gold-plated health plans: two workers receive identical health benefits, yet in one case, the higher-paid worker will get a 10% subsidy from Uncle Sam while another lower-paid worker will pay a penalty of 25%. How fair is that?
In short, leaving aside that it flagrantly violates the President's firm pledge against raising taxes on those with incomes below $250,000 a year, the Cadillac tax (first proposed by liberal Senator John Kerry) is completely indefensible from the standpoint of fairness. But now perhaps you can begin to understand the political logic behind this compromise proposal. Labor unions were fiercely opposed to capping/phasing out the tax exclusion since their lavish health benefits would be most susceptible to being limited by such a cap. Therefore, they got policymakers to swap an excise tax that appeared to achieve the same purpose. The typical union worker is closer to the high-paid worker than to the low-paid worker in my example, so swapping the tax exclusion for the Cadillac tax effectively meant retaining a small residual subsidy for such workers even if they remained in very lavish union-sponsored health plans, as opposed to losing their subsidy altogether. In contrast, low-wage workers not only lose a tax break, they actually start paying an additional tax to help bankroll Obamacare. Under this administration, it pays to be well-connected!
The Real Purpose of the Cadillac Tax: To Raise Revenue
But here's the dirty little secret of the Cadillac tax. It may encourage some cost-saving plan-swapping, but in the long run, by design, it is intended to raise revenue in the same shockingly regressive manner just described. Between 2018 and 2023, the CBO projects that revenue from this "tiny" tax will more than quadruple; in contrast, revenue from the penalty payments by those who are uninsured will rise only 20% during the same period. How can this possibly be? It's quite simple: the dollar thresholds used to determine which plans are subject to this tax will be indexed using the Consumer Price Index +1 percentage point in 2018 and 2019 and CPI only thereafter. However, the actual rate of increase in premiums for employer-sponsored coverage over the past decade-despite the dramatic slowdown in health spending in recent years-has averaged more than 15% a year above inflation! Thus, just as tens of millions of Americans inadvertently became subject to the Alternative Minimum Tax due to its lack of indexing-even though the AMT originally targeted only a very tiny fraction of taxpayers-over 100 million Americans vey quickly will find themselves in what the government labels "Cadillac" health plans. In fact, Bradley Herring, a health economist at Johns Hopkins Bloomberg School of Public Health, estimates that as many as 75 percent of plans could be affected by the tax just in the next decade. But unlike a cap on the exclusion-in which the low-paid worker in my example would pay extra taxes equal to 15.3% of the amount above the cap and the high-wage worker would pay added taxes of 51% on the identical amount-the Cadillac tax would subject both workers to the same massive tax rate despite their hugely disparate incomes (not to mention hugely disparate tax-financed discounts on their health coverage). In short, both the Cadillac tax and a cap on the tax exclusion would raise federal revenue, but the former would do so in an unconscionably regressive fashion.
Jonathan Cohn concedes this adverse impact, but explains "We also know that other parts of Obamacare, like tax credits for purchasing insurance and guarantees of coverage for people with pre-existing conditions, will help the sick and the poor far more than the Cadillac tax will hurt them." He goes on to claim "unlike liberals, conservatives don't seem particularly troubled by the implications for the chronically ill. Either that, or conservatives do a remarkably good job of disguising their anxiety." Yet every conservative reform plan I have seen that tackles the exclusion replaces the upside-down subsidy with some form of tax credit that either is neutral across income groups (i.e., every taxpayer who purchases health insurance being qualified for the identical dollar amount of tax credits regardless of income) or which provides subsidies that vary by income (with the highest subsidies going to the lowest income households) and/or health status (with the highest subsidies again going to those who need them the most, i.e., the sick). Mr. Cohn is far too knowledgeable about health policy matters not to be aware of this. Readers can draw their own conclusions about what would have motivated him to pen such a factually false claim.
Similarly, Matt Yglesias is dead wrong in describing the Cadillac tax as "a bit of a silly workaround, but functionally it works the same way as a phase-out." Whereas conservatives favor a solution that would truly replace the inefficient and unfair tax exclusion with a far more sensible subsidy structure, Obamacare never gets rid of the tax exclusion and instead layers on top of it a highly regressive tax that is used to subsidize coverage of other small firm workers in the exchanges-not the workers who are paying the tax. In other words, rather than fix a rather obvious wrong, Obamacare adds another wrong on top of it. Surely two wrongs don't make a right. The most sensible approach would be to repeal the Cadillac tax in favor of a more sensible reform of the tax exclusion. And when all the evidence is taken into account I truly cannot understand why any progressive would disagree.
Hat tip to Loren Adler @LorenAdler for pointing me to the Bipartisan Policy Center's recent recommendation: "Replace the Cadillac tax on high-cost health insurance plans with a limit on the income-tax exclusion for ESI at the dollar amount equivalent to the 80th percentile of single and family ESI premiums in 2015 (age- and gender-adjusted)." Pages 72-77 discuss the Cadillac tax in detail. Worth noting is one other devastating aspect of the Cadillac tax that I didn't address: "Although it aims to fix many of the same problems that our proposal would address,the Cadillac tax would severely distort employer choices. It is not deductible as a businessexpense for insurers' corporate income taxes, and so would necessitate a premium increaseof almost six times the amount of the initial tax for the insurance company to maintain thesame after-tax profits. Therefore, in practice, the Cadillac tax goes well beyond changingincentives for employers and employees; it will make health insurance coverage above the threshold prohibitively expensive. Furthermore, the Cadillac tax is less progressive than our proposed cap."
 A higher threshold ($11,500 single/$29,450 family) is used for early retirees and high-risk professions such as construction workers, miners, firefighters and longshoremen.
 35% is conservative as it was measured in 2006 but it at least is a comprehensive measure of the total amount of the subsidy taking into account income and payroll taxes at all levels of government.
 According the Congressional Budget Office, there were 156 million Americans with employer-sponsored health coverage in 2013. Virtually all would qualify for some subsidy through the tax exclusion; there are additional self-employed who qualify for a tax deduction to help subsidize their purchase of non-group coverage, but such individuals would not be affected by changes to the tax exclusion.
 This is an estimate by Dr. James Godwin, a geriatrician at the University of Texas reported in an eye-opening New York Times report on the high cost of colonoscopies.
 According to Lewin Group, pure claims administration costs (as opposed to general overhead) add 4.8% to the average cost of paid health insurance claims for private health insurers and 10.9% to paid claims in the non-group market (see Appendix Figure 1). Note that Sherlock (2009) argues that both the absolute size of both figures may be too high in light of advances in claims-processing technology as well as exaggerations in the extent of economies of scale in such claims processing. But keep in mind also that claims management activities such as pre-authorization and utilization review do not count as claims processing expenditures. Thus 5 cents per dollar of claim expense may actually be a reasonable estimate of the price of running expenses through a third party. This is nearly equivalent to doubling the sales tax in the median U.S. state (median = 6.0% in 2013).
 The exact details are more complicated, but as a rough approximation, federal employees receive a fixed dollar amount equal to 72% of the average premium charged by the many plans available to them. If they choose a plan that is less expensive than this amount, they keep most of the savings; conversely if they select a more expensive plan, they pay nearly all of the added premium cost out of pocket. Walt Francis explains the system in greater detail in his book Putting Medicare Consumers in Charge: Lessons from the FEHBP.
 The self-employed are permitted a deduction for health insurance costs but this tax break applies only to income taxes, not payroll taxes. Thus, a self-employed person with $50,000 in income ends up with a lower tax subsidy than his/her counterpart working for an employer offering health coverage.
 For example, the Bowles-Simpson Commission proposed that a reduction of the income tax exclusion for employer-provided health insurance be part of a fail-safe plan that would automatically be triggered if legislation were unable to satisfy specified targets for federal funds. In particular, the commission offered an illustrative proposal that, when fully phased in, would cap the tax exclusion at the 75th percentile of premium levels in 2014, with no indexing of the cap through 2018 and a gradual phase-out of the tax exclusion by 2038. Similarly, the Debt Reduction Task Force chaired by Alice Rivlin and Pete Domenici proposed a plan to cap and phase out the tax exclusion. The task force proposal would cap the tax exclusion at the 75th percentile of premiums (including dental and vision coverage) in 2018 and eliminate new contributions to health savings accounts, another tax-free vehicle for health care spending. The cap would be phased out over a 10-year period. The center-left Urban Institute recently analyzed the impact of a cap set at the 75th percentile of the sum of premiums and other medical benefits, and indexed, or allowed to grow over time, by a five-year average of the rate of GDP growth.
 This includes both the 7.65% deducted from the employee's paycheck as well as the employer contribution of 7.65% that is largely invisible to the employee. Even though it appears that the employer bears this latter cost, most economists (and the CBO whenever it examines the distributional effects of changes in tax policy) believe that the tax ultimately is borne by the worker in the form of lower wages. That is, absent the payroll tax requirement, the employer in a competitive market would pay that 7.65% as added cash compensation.
 According to the annual Employer Health Benefits Survey sponsored by the Kaiser Family Foundation, premiums for single coverage grew 18.2% a year ($3,083 in 2002 to $5,085 by 2012) while premiums for family coverage grew even faster at 19.7% a year ($8003 in 2002 to $15,745 by 2012). According to the Bureau of Labor Statistics, the general CPI rose 27.6% during this same period, an annualized inflation rate of only 2.5%.
 Even after Obamacare is fully implemented in 2018, the CBO projects there will be 165 million Americans with employer-sponsored coverage. Thus my 100 million figure could be quite conservative, as 75% x 165 million = 124 million.