Editor’s Note: This post by Len Nichols of the New America Foundation appears in tandem with yesterday’s publication on the Health Affairs Web site of an article by Jack Hadley of George Mason University, John Holahan of the Urban Institute, and coauthors, which estimates the cost of covering uninsured Americans. Additional bloggers on the Hadley-Holahan article include Henry Aaron of the Brookings Institution and Tom Miller of AEI.
Jack Hadley and colleagues’ Health Affairs Web Exclusive makes a number of important points. The fundamental one for the upcoming debate on health care reform is this: the amount of net new health spending that is required to cover the uninsured is about 5% of current national health spending, or slightly less than 1% of gross domestic product (GDP). If the economic cost of the premature death and prolonged morbidity of the uninsured is estimated to be as much as $204 billion in 2006, which is equal to or greater than the cost of insuring them, then covering the uninsured makes economic sense, over and above any moral considerations.
At the same time, credible estimates suggest we currently spend as much 5% of GDP on health care that adds no clinical value. Therefore, we can probably finance much of the cost of a coverage expansion with system savings in the long run, if we implement smart delivery system reform at the same time. Hadley and colleagues have helped document and strengthen the case for comprehensive health reform, and I applaud that.
My quibble with the paper is over its treatment of cost shifting and private insurance premiums. Some of these details may be clarified in a longer paper by the same authors on the Kaiser Family Foundation Web site, which was not available when I wrote this post earlier this month. That said, the argument in this paper about hospital cost shifting seems to overemphasize some and underemphasize other rather important points.
Misleading allocation of public funds
First, the authors count Veterans Affairs (VA) and Indian Health Service (IHS) federal subsidies as public moneys that are used to finance care for the uninsured, thus reducing the potential amount that might need to be shifted to private payers. VA plus IHS spending is nontrivial. According to their Exhibit 3, VA and IHS spending is $7 billion in 2008 dollars. This is almost as much as the estimated $8.6 billion in Medicaid disproportionate-share hospital (DSH) payments, the current program most directly focused on financing care for the uninsured.
The problem with counting VA and IHS moneys against what must be shifted to private payers to cover the garden-variety uninsured is that VA hospitals and IHS facilities by and large do not treat the garden-variety uninsured. While it is true that the Current Population Survey (CPS) classifies people who only have VA or IHS coverage as uninsured, the number of people who are enrolled in either of these programs alone according to the Medical Expenditure Panel Survey (MEPS), the basis of the authors’ estimates for gross uncompensated care, is vanishingly small. Thus, the paper reduces estimated uncompensated care (computed from actual MEPS sample participants) by an amount equal to the independent estimate of the large public stream of funds. However, these funds do not necessarily defray the costs of caring for the uninsured for the vast majority of community hospitals that actually treat them. This surely leads to an underestimate of the costs shifted to private payers to finance care for the uninsured.
Second, the authors leap from the fact that hospital markups don’t rise in tandem with the percentage of the population that is uninsured to the conclusion that uninsured costs are unimportant to hospitals. As the data in their Exhibit 4 show, hospital markups vary considerably as a result of a variety of factors. The authors acknowledge that this somewhat cyclical trajectory is mostly driven by the waxing and waning of relative bargaining power between health plans and hospitals and by fluctuations in public payer rates.
In the past 20 years, both rates bargained by private insurers and public payer rates have experienced greater and more frequent fluctuations than have uninsured rates, which have crept slowly upward at a relatively constant rate. The fact that anticipated costs of uncompensated care are built into prices negotiated with private health plans (and, implicitly, private premiums) affects the level of prices, not their rate of growth, because the aggregate rate of growth of the national rate of uninsurance has been too small to drive price growth. Estimates of the cost shift have tried to capture the level of prices and the percentage of the “hidden tax.”
Trend analysis does not refute the very real phenomenon of charging private payers a markup to finance anticipated care for the uninsured, when market conditions permit, which sometimes they do and sometimes they do not. When market conditions do not permit a markup to finance uncompensated care, hospitals scream for public payment relief, which they typically get after a while (as they did after the first wave of BBA payment cuts went into effect in the late 1990s).
However, in a given locale, if the uninsurance rate increases rapidly as a result of deteriorating economic conditions or Medicaid eligibility cutbacks, local hospitals that treat the uninsured have no choice but to raise private rates if they can. Therefore, specific trends in uninsurance rates can matter in the short run. This is hard to prove in the absence of localized data. However, too many people involved in these transactions believe that hospitals raise private rates as a result of spikes in the number of uninsured for me to discount it based on heroically constructed but flawed aggregate data.
To conclude, I think that this paper makes a major contribution to our understanding of how much additional spending will be required to cover the uninsured in the short run. On the other hand, it is not as strong in its decomposition of exactly how current care for the uninsured is implicitly financed. That relatively minor weakness should not detract from the larger irrefutable point: our nation can realistically afford to cover all Americans.