• How much cost shifting should we expect today?

    Yesterday I illustrated a theory that predicted some degree of cost shifting if hospitals are not profit maximizers. It also suggested that hospitals cut costs instead of just raising private prices in response to shortfalls in public payments. The question everyone should want the answer to is, for a dollar reduction in public payments, what proportion is shifted to private payers and what proportion leads to cost cutting?

    My best answer, based on a thorough review of the literature, is that about 20 cents of each dollar shortfall in public payments is shifted to private payers, the rest is absorbed in cost cutting. This is quite a bit different from what you’ll hear from the insurance or hospital industries. They want us to believe that 50 to 100 cents on the dollar is cost shifted. Relative to those empirically unsubstantiated claims, a belief that cost shifting is pure fiction (0 cents on the dollar) is closer to the truth!

    In my view, the two most convincing and careful studies of cost shifting are those by Cutler and Wu. Other papers use somewhat less rigorous methods or are focused on a time period or geographic region that make them less applicable to today or inappropriate for national extrapolation. (A review of all recent papers on cost shifting is here. It includes more detail than provided below on the papers by Cutler and Wu.)

    In his paper, Cutler asks to what extent do lower Medicare payments lead to lower costs (reduced services and lower quality) and to what extent is the cost level maintained and the burden of covering them shifted to the private sector? His answer depends in part on the nature of the private market, which varied considerably over the two time periods he examined—1985-1990 and 1990-1995. (I described the differences in the health care market between these two time periods in prior posts (here and here).)

    Cutler found that over the 1980-1985 period hospitals shifted costs dollar-for-dollar. Over 1990-1995 Cutler found no evidence of cost shifting. His interpretation is unambiguous. In the late 1980s, Medicare payment cuts were financed by shifting costs to the private sector. With the rise of managed care in the early 1990s, cost shifting was no longer feasible due to the price sensitivity and market power of managed care plans. Thus, in the early 1990s cost cutting was the dominant response to lower Medicare payments.

    Wu has provided what is, perhaps, the most careful study of the cost shifting hypothesis (summarized in this prior post). With a long difference model using 1996 and 2000 Medicare hospital cost report data, she examines the effect on private prices of reductions in Medicare payments to hospitals as a result of the Balanced Budget Act of 1997. Moreover, she considers the heterogeneity of that effect across private-public payer mix, levels of hospital competition and share of hospitals in the market with for-profit status. Of all studies I reviewed, Wu’s provides the strongest mitigation against and test of the potential confounding endogeneity of Medicare payment, thereby providing the most plausible estimate of its causal effect on private prices.

    Wu found that, on average hospitals shifted 21 cents of each Medicare dollar lost to private payers. The degree of cost shifting varies by hospital bargaining power: a one standard deviation increase in such power increases the cost shifting rate to 33 cents on the dollar. There is no statistically significant evidence of heterogeneity in cost shifting by for-profit, teaching, or public hospital status. Nor does it vary by HMO market penetration or change of it. A smaller degree of cost shifting occurs in markets with a higher share of discharges from for-profit hospitals.

    Among the papers that use methods sophisticated enough to estimate a causal effect of Medicare payment levels on private prices, Wu’s is based on the most recent data. We have no more recent, credible estimates of the cost shift rate. Until we do, we must go with Wu’s figure as our best estimate.

    Given these findings, what can be said about the likelihood of cost shifting in the future? Relative to the period in which cost shifting is most likely to have occurred at a relatively high rate (approximately 1987-1992), there is now greater price competition in the commercial market due to network-based contracting. On the other hand, relative to the period in which there was likely no cost shifting (when tightly managed care dominated, the mid-1990s), price competition is weaker because consumers no longer accept networks with the same level of restrictions as existed then.

    To the extent that hospitals still possess some unexploited market power, perhaps some cost shifting is possible, but it is unlikely to be dollar-for-dollar as suggested by industry-funded reports (yes, I’ve reviewed those too: here, here, here). The most credible estimate in the recent era (since the demise of tightly managed care) is that provided by Wu, which found that 21 cents of each dollar in Medicare hospital payment reduction was shifted to private payers. Thus, over time the markets have changed making cost shifting far less likely.

    Does that mean we’re through with arguing over cost shifting? Don’t bet on it. But, when you hear claims or implications of inevitable, large-scale shifting, be skeptical. The vast majority of shortfalls in public payments are absorbed by cost cutting. If you believe in evidence, you’ve got to believe in that.

    This concludes my planned posts on hospital cost shifting. If you really want to see more, check out the papers I referenced in Monday’s post. They cover the entire subject, from a heath economics point of view.

    If you’ve followed all my recent cost shifting posts, here’s a question for you: would you expect more or less cost shifting in the pharmaceutical industry as compared to the hospital industry? I think theory provides good guidance, if you can argue your way toward which theory applies. Next week I’ll summarize some empirical results in this area.

    • My guess is that pharmaceutical companies do maximize their profits, so there should be little or no cost shifting.

    • I have now read all your cost-shifting posts and your 2 working papers.
      Very interesting and convincing.
      Here’s a question though.
      If we take Wu’s conclusions of $0.20 cents cost-shifted to private payers, that leaves $0.80 of each dollar in Medicare decrease to be absorbed through cost cutting.
      Are there any studies which look at how hospitals actually cut costs? Some portion could be from increased efficiency (reduced waste) but there is also the real danger of decreasing quality—either through offering fewer desired services or by reducing the absolute quality of services which are delivered. Neither of these latter two options are cause to celebrate.

      • @Beth – Some of the studies I reviewed look at cost cutting (Cutler’s, for example). Also, Wu has a new NBER working paper on cost cutting. I’ll be reading and posting on it. However, I have a sense that you’re after the health effects of cost cutting. I do not think I’ve seen anything on that. It’s probably hard to study.

        Bottom line: if 80 cents go to cost cutting then something is being lost. Is the loss bad for health? I’m sure some of it is and some of it is not, and it may not be the same things for everyone, given heterogeneous needs for and responses to care.

    • Cutler’s article is next on my list to read after Morrisey’s book. I look forward to your review of Wu’s next paper.

      Of course the bottom line is health, though I think there are other measures of quality which matter as well which would be a sig. loss even without adverse measurable effects on health–such as the “caring” side of medicine (the time doctors, nurses, transporters, etc have to spend with patients and their families), wait times, and even the aesthetics of the hospital experience which can effect a patient’s sense of comfort or anxiety.
      I agree these would be hard to measure, but that does not make them unimportant. (I don’t mean to imply that you think they are.)

      One other question–In your reading, is there anything which would lead you to think that these same arguments about cost shifting/cost cutting and Medicare are any different when applied to the problem of uncompensated care for the uninsured? Since the uninsured as a group pay very close the same percentage as Medicaid, my guess is that the arguments should apply.
      I did get the 3 articles you mentioned (Families USA, Hadley and Kessler) which address this issue. They are now in my stack.

      Thank you kindly for your reply.

      • @Beth – I have not examined the case of the uninsured in any depth. I don’t see how the issues could be much different though. The issues of hospital-plan negotiation and underlying costs all seem the same to me.

        It’s a cliche that insuring the uninsured will relieve the rest of us of the cost shift from them. I challenge anyone to identify the reductions in premiums and taxes we’re expected to enjoy.

    • From what you have been writing on this blog and your cost-shifting articles, I expected you to have this answer–but just though I’d check to be sure.
      Thank you.