Click here for a guide to following the health care reform story online.
An emerging narrative of the White House health reform negotiations (reportedly hitting warp speed because Democrats fret they may lose Ted Kennedy's Senate seat in Tuesday's special election in Massachusetts) is that Big Labor used its powerful muscle to emasculate the excise tax on high-end "Cadillac" health insurance policies. That's true in the superficial sense that yes, the Cadillac tax now poses a smaller threat to union members, and yes, AFL-CIO President Richard Trumka very publicly opposed the Cadillac tax and has suggested labor support for Democrats might prove tepid in 2010 should Obama take working people for granted. But this version of the facts is belied by the pretty obvious truth that in 2010 labor has almost no clout. Perhaps you've noticed that the unions utterly failed to achieve what was by far their No. 1 priority in health care reform—creation of a "public option" government health insurance program. The Cadillac tax got shrunk not because labor is strong but because the tax itself is almost laughably weak. Trumka was blowing on milkweed.
If I were to credit anyone with removing the chassis from the Cadillac tax, it wouldn't be the AFL-CIO. It would be Health Affairs, a peer-reviewed journal on health policy whose current issue contains a quietly devastating paper titled "Taxing Cadillac Health Plans May Produce Chevy Results."
The Cadillac tax is premised on the idea that overly generous benefits drive up the cost of health insurance. This would surely be true if health insurance policies with overly generous benefits actually existed in any meaningful number. (Raise your hand if you've got one.) The Health Affairs study, written by four researchers from the National Opinion Research Center, took the novel approach of examining health insurance plans not in some economist's hypothetical universe but here on Planet Earth. It found that a health plan's relative generosity (as measured by actuarial value) accounted for a grand total of 3.7 percent of the variation in the cost of family coverage. Add in the type of plan (health maintenance organization, preferred provider organization, etc.) and you've accounted for a mere 6.1 percent of the cost variation.
The real determinant of a health plan's cost isn't actuarial value. It's demographics—the average beneficiary's age, occupation, geographic location, and so on. Different cohorts face dramatically different likelihoods of requiring doctor or hospital care. As Joseph White and Timothy Jost observed on the Health Affairs blog, the health reform bill "defines a Cadillac price, not a Cadillac plan. (Fortune magazine columnist Allan Sloan has written much the same, as has Alec MacGillis of the Washington Post.)
Last week I asked Jonathan Gruber, an MIT health economist who is strongly in favor of the Cadillac tax—and who, I subsequently learned, has a $297,600 contract with the Health and Human Services Department to run the numbers on various health reform proposals—whether he'd written anything in response to the Health Affairs study. His reply (via e-mail): "The point is that you don't need to respond to the Health Affairs study."
But one way to look at the changes to the Cadillac tax is largely as a response to the Health Affairs study. Is the tax aimed at inhibiting a cost driver that doesn't really exist? Then no great harm in nosing up the threshold from $23,000 to $24,000, and further still if medical inflation exceeds current projections. Might it undo existing collective bargaining agreements? Exempt 'em for five years. Worried that employers will drop dental and vision coverage? Exempt it. Employers and workers need transitional relief? Give it to 'em! Once you recognize that the Cadillac tax is built on a faulty theory about medical inflation—that the tax's value is entirely symbolic—then there's almost no carve-out you can't justify.
The Washington Post's Ezra Klein argued that the excise tax was "virtually unchanged" after the White House came to terms with Trumka and Co. But according to the Wall Street Journal's Laura Meckler and Naftali Bendavid, the combined changes knock an estimated $59 billion out of the $149 billion that the Cadillac tax was expected to raise over 10 years. This is trivial only in the sense that the Cadillac tax was always a lousy revenue raiser compared to its House counterpart, a 5.4 percent surtax on family incomes above $1 million, which was estimated to raise $460 billion over the same time period. (For reasons I can't fathom, the millionaire surtax is thought unsellable in the Senate, whereas the Senate bill's 0.90 percent Medicare surtax on family incomes above $250,000, which may now—as it should be—get extended to investment income, is not.) The Cadillac tax rate, 40 percent, is set well above the top income-tax rate (35 percent), giving the lie to proponents' argument that the change merely undoes the existing tax exclusion at the high end. As a consequence, it's assumed insurers will in nearly all cases lower the value of health plans rather than pay the tax. Most revenues will therefore come indirectly, as employers divert dollars previously spent on tax-free health insurance into taxable income. That's the theory, anyway.
The principal respect in which the Cadillac tax continues to retain its bite is in its annual cost-of-living adjustment, set at the general inflation rate plus 1 percent. Everybody agrees that the medical inflation rate will be well above that, with the result that over time the Cadillac tax will shrink benefits, increase co-payments, or raise taxes on those of us with Chevy health plans. By that I don't mean the handful of Rust Belt workers still holding union cards or even the much larger group of unionized public employees. I mean everybody. The indiscriminate nature of this cost-saving scheme will likely make a great many voters very, very angry, and they will likely be heard in Washington. When that happens, expect Congress to step in and either water down the Cadillac tax further or toss it out altogether. Neither would be any great loss.
E-mail Timothy Noah at firstname.lastname@example.org.