Reporting on research about "Cadillac" healthcare plans

I just read an interesting article in the latest issue of the journal "Health Affairs", entitled "Taxing Cadillac Health Plans May Produce Chevy Results." The authors, J. Gabel, J. Pickreign, R. McDevitt and T. Briggs (all experienced researchers in the field of healthcare) do a statistical analysis of the premiums paid for various healthcare plans versus various factors such as value of benefits ("Cadillac versus Chevy"), type of plan (HMO, PPO, POS, high deductible), type of industry supplying the plan, and local cost of health insurance.

As written into the senate version of the proposed healthcare act, a plan is deemed a "Cadillac" plan solely on the basis of its premiums. The latest version sets the cut-off at $8500 for an individual plan and $23,000 for a family plan. This sounds like a lot of money but, as Bob Herbert of the the NY Times points out:

"Within three years of its implementation, according to the Congressional Budget Office, the tax would apply to nearly 20 percent of all workers with employer-provided health coverage in the country, affecting some 31 million people. Within six years, according to Congress’s Joint Committee on Taxation, the tax would reach a fifth of all households earning between $50,000 and $75,000 annually. Those families can hardly be considered very wealthy."

NOTE: The reason that these numbers are going up so dramatically is that the cut-off for taxation is not really indexed properly. The current indexing plans seem to call for cost of living + 1%. This is far less than the actual percentage rise in the cost of health care, which by many estimates is at least twice the rate of increase in other costs of living.

We've all heard about plans that pay for vacations at spas and full body scans and healthclubs. Is that what goes into Cadillac plans?

Not according to the statistics in the Health Affairs study. In fact, less that 4% of the premium cost of a plan is correlated with the actual benefits of the plan. Only about 6% of the cost is correlated with either benefits or type of plan (say HMO or POS). Most of the correlation is with type of company and local medical costs -- features over which the payer (individual or group) has no control. Most of the cost is actually uncorrelated, meaning no one knows why it is what it is.

Thus, Obama and the Senate have bought into a policy -- taxing health insurance solely on the basis of its cost -- that may save money, but most likely at the expense of individuals or providers who have little choice in the size of their premiums. Thus, these policy holders will either pay a steep tax (40% on amounts over the cut-off), or will be forced to give up coverage which was, all along, just Chevy, not Cadillac. Such people may be unfortunate enough to work in a high-risk industry, be middle-aged or senior, or simply live in an area with expensive medical costs. Their actual coverage, on the basis of statistical analysis, is unlikely to cover spas or body scans or any other high-end services.

This is yet another example of the curiously elliptical way that politicians often attack problems: "If I change A, it will likely change B which may change C which ought to change D, the thing I really want to affect." This Rube Goldberg kind of scheme is ridiculous. If the goal is to discourage truly luxurious and unnecessary coverage, then why not tax luxurious and unnecessary coverage? If the goal is to subsidize lower income people, then why not tax upper income people? If the goal is to control expenditures, then why not set strict limits on billing, based on actual local costs of medical services? A party or president who actually cared or thought acutely about issues would take these approaches.

But no. Obama and the Senate are pushing an ill-considered and unfair policy. They risk antagonizing one of the few segments of American society that has supported them relatively consistantly: the unions. Morever, it should be pointed out, most union members don't have "Cadillac" plans, and most people who have them are not union members. I would love to know who -- which advisors -- are pushing this policy, and what their agenda is. Obama may be a smart politician and lawyer, but does he understand statistics?