At the WSJ, John Cogan, Glenn Hubbard, and Daniel Kessler weigh in:
[T]he administration and Congress have asserted that people with private insurance pay for care for the uninsured through “cost shifting”—higher prices charged by doctors and hospitals to recover losses from uncompensated care. […]
Our review of the research has found that there is no credible evidence of a cost shift of any substantial consequence, either within state boundaries or across state lines. […]
A study conducted by George Mason University Prof. Jack Hadley and John Holahan, Teresa Coughlin and Dawn Miller of the Urban Institute, and published in the journal Health Affairs in 2008, found that so-called cost shifting raises private health insurance premiums by a negligible amount. The study’s authors conclude: “Private insurance premiums are at most 1.7 percent higher because of the shifting of the costs of the uninsured to private insurance.” For the typical insurance plan, this amounts to approximately $80 per year.
The Health Affairs study is supported by another recent peer- reviewed study that focused exclusively on physicians. That 2007 study, authored by Massachusetts Institute of Technology economists Jonathan Gruber and David Rodriguez and published in the Journal of Health Economics, found no evidence that doctors charged insured patients higher fees to cover the cost of caring for the uninsured.
The authors go on to say that the large cost shifting estimates come from a Families USA report. Still, the body of evidence is very thin. This is all consistent with what I wrote in my forthcoming Milbank Quarterly paper on hospital cost shifting. The focus of that paper is on the cost shift due to shortfalls in Medicare and Medicaid payments, not the uninsured. But I did write this:
Families USA (2005) estimated that private insurance premiums were about 10% higher in 2005 due to the use of health services by the uninsured. Kessler (2007) and Hadley et al. (2008) both find less than a 2% effect.
So, I’m inclined to agree with the Cogan, Hubbard, and Kessler about the quote above, but not fully about this:
If anything, the likely impact of the law will be to increase, not decrease, cost shifting. According to the Congressional Budget Office, around half of the people who are expected to become newly insured under the new law will be enrolled in Medicaid. But Medicaid payments to doctors and hospitals are so low that the program creates a cost shift of its own. In fact, a long line of academic research shows that low rates of Medicaid reimbursement translate into higher prices for the privately insured.
There are two issues here. First, one has to consider where the new Medicaid enrollees come from. As I wrote in my paper, if they come from the uninsured population, then Medicaid payments may be an increase in compensation for care. However, it is likely that a sizable fraction will come from the privately insured population, a crowd-out effect about which you’ll see more about on this blog soon. Since Medicaid reimbursements are much lower than those from private insurers, that could play into cost shifting. Cogan, Hubbard, and Kessler have a point there.
The second issue is, how much cost shifting is there from Medicaid? The bulk of the studies I reviewed in my paper focused on Medicare. Those that focused on Medicaid found even lower cost shifting rates from that program. Perhaps you recall that the most plausible Medicare cost shift rate was 21 cents on the dollar (a dollar in Medicare payment shortfall causes a 21 cent increase in private hospital prices). Medicaid estimates are much lower, but they have not been examined closely with recent data. (There is no evidence of cost shifting for physician services or pharmaceuticals. Hospitals are where it’s at, if anywhere.)
Even if we assume the 21% Medicare cost shift rate applies to Medicaid too, that does not mean a 21% increase in private premiums. It’s a 21% increase in hospital payment rates. Premiums cover more than hospital care and the volume of hospital use has to be factored into the calculation to deduce the ultimate effect on premiums. No doubt there is an upward effect on premiums due to shortfalls in public payments, but it isn’t huge. The vast majority of lower public payments are translated into cost cutting, not cost shifting. That’s not necessarily a good thing either, which is just another illustration of how complicated the health system, and health policy, are.
