Lewin Group's Estimates of Increased Coverage Under McCain Health Program Based on Bogus Assumptions

The estimable Len Burman of the Tax Policy Center clears up a few points of confusion on McCain's health care plan. Apparently, the main reason the Lewin Group estimated that there would be a huge increase in coverage under the plan is that they assumed McCain would spend 5 times what he has said he would to subsidize the uninsured through "GAP" programs. This is completely bizarre. Of course it is often the case that analysts trying to forecasting a policy's effects need to make assumptions to fill in blank spots on policy proposals. But in this case, the Lewin Group appears to have decided that the McCain campaign must be lying about how much would be budgeted for these subsidies, because if, as the campaign has said, the budget would only be $7-10 billion a year, McCain's plan would make only a tiny dent in reducing the number of uninsured.

It's unclear to me whether the Lewin Group made this assumption because they were trying to produce an analysis more favorable to McCain or because the McCain campaign manipulated them into this position. I can't find the $7-10 billion figure in the Lewin Group report, even though it's been widely reported in the press. It's impossible to believe that the Lewin Group analysts weren't aware that McCain's campaign has said that $7-10 billion will be budgeted for the GAP program. Why would they possibly leave that out of their 182 page analysis of the candidates' plans?

I must conclude that contrary to what I wrote earlier, the Lewin Group is not a trustworthy source. Here's Len's full post:

When TPC analyzed Senator McCain’s proposal to replace the income tax exclusion for employer-sponsored health insurance with flat refundable tax credits of $2,500 for single coverage and $5,000 for family coverage, we found only modest net effects on coverage. Our model predicted that more than 21 million people would gain insurance coverage in the individual nongroup market by 2013 while 16 million would lose employer-based coverage. Despite a $1.3 trillion price tag over the next decade, the proposal would yield only modest and temporary gains.

A couple of factors drove that result. One was the $7-10 billion per year that Senator McCain’s campaign says it would spend on its Guaranteed Access Plan (GAP). That’s a fraction of the $100 billion annual cost we estimated for covering those with serious health problems who’d otherwise lack insurance. Ignoring the campaign’s statements about its own plan, John Sheils of the Lewin Group assumed that the government would provide $470 billion in subsidies over ten years ($47 billion per year) for the GAP, half of it financed by a new assessment on insurance premiums. By Sheils’s estimate, 5.8 million people would gain coverage under that plan. Unlike Sheils, we judged the funding proposed to be inadequate and the plan’s details too nebulous, so we did not model the GAP’s effect on either cost or coverage.

Sheils also concluded that Senator McCain’s proposals to limit health care costs would be effective, something my colleagues in the Urban Institute’s health policy center doubt. Moreover, Sheils appears to assume that firms are less sensitive to changes in the price of health insurance than TPC does, which means that fewer firms drop coverage. All told, Sheils’s more optimistic scenario produces much more coverage. Lewin estimates that the number of uninsured would fall by 21 million people in 2013.

There is one way that McCain’s plan might really boost coverage: if just about anything could be labeled as “insurance.” With no minimum standards, insurers could design products that cost less than the tax credit amounts, even for people with serious pre-existing conditions. For example, they might sell a single policy that covers the first $2,000 of medical expenses for a $2,500 premium. This sounds like a bad deal, but if the entire tab is paid by the federal government and it is all a sick person can find, it’s better than nothing. State regulators couldn’t block such policies because Senator McCain’s plan would allow insurers to market products across state lines, meaning that shady insurers could just set up shop in a state with no regulation.

Of course, token insurance that doesn’t cover major costs would be cold comfort to those with expensive health problems, but it would make the statistics on coverage look better. And if you think that insurers would not offer such products or that consumers would not use their tax credits to purchase them, consider the experience with the short-lived health EITC, a small, but poorly designed subsidy intended to help low-income families acquire health insurance coverage for their children. Unscrupulous insurers sold nearly worthless policies—often using fraudulent methods—to credit recipients according to a 1993 Congressional investigation (summarized by CBPP).

And, despite its limited value, the coverage gains would carry a heft price tag. If all uninsured people bought non-group coverage qualifying for the tax credits, the cost of the plan could almost double, from our estimated $1.3 trillion to $2.5 trillion.
All that said, it is remarkable that a conservative Republican is proposing more than a trillion dollars in refundable tax credits for health insurance. Those seeking a bipartisan compromise that could significantly expand health insurance coverage might take heart from that. But then again, Senator McCain also insists that his plan, in fact, has no budgetary cost over ten years, suggesting that he might not really be serious about the tax credits.