Cost shifting, you may recall, is the idea that providers can make up for shortfalls in Medicare or Medicaid revenue by increasing the prices charged to private payers. I've been following the literature on it for years, summarizing that on hospital cost shifting through 2011 in a publication in the Milbank Quarterly and updating that summary through the summer of this year with one in Health Services Review. Since that publication, a few more papers on cost shifting have come out. This post summarizes their findings.
As I wrote in an October post on this blog, there is very little evidence that hospitals have been able to shift costs in recent years. The work by Vivian Ho and colleagues (PDF), published in Health Management, Policy and Innovation is consistent with this conclusion. In an analysis of Texas hospitals over 2000-2007, they found "no evidence that reductions in Medicare prices or Medicaid prices per uninsured are associated with increases in private pay prices."
In Health Affairs, Chapin White and Tracy Yee examined the effect of Medicare payment cuts on number of staffed beds and Medicare inpatient volume in 10 states over the period 1995-2009. They found that price cuts lead to fewer staffed beds and to Medicare inpatient volume cuts. Though not itself a direct test of the cost shifting hypothesis, these findings are consistent with prior work that found that hospitals cut costs and decrease private pay rates* in response to Medicare shortfalls.
White and Yee estimated that a 10% decrease in Medicare price led to 4.6% fewer Medicare discharges and a 6.3% reduction in staffed beds. At a national level,
a 10 percent Medicare price cut was associated with 100,000–900,000 fewer discharges among the elderly, and 400,000– 7,000,000 fewer hospital days (the ranges represent plus or minus two standard errors). The simulated reduction in Medicare inpatient hospital spending was about $14.4 billion from the direct price effect (that is, when we held the volume constant) plus another $1.5–$11.9 billion from the reduction in volume.
In an NBER working paper, Jeffrey Clemens and Joshua Gottlieb consider cost shifting by physician organizations. As far as I know, there has never been a study of cost shifting that found that physicians engage in the tactic. Clemens' and Gottlieb's study is no different in this regard. They found that private rates follow those of Medicare, a spillover effect that is the opposite of cost shifting.
We assess Medicare's influence on fees for physicians' services, finding its in influence to be substantial. A $1 change in Medicare's relative, cross-service payments leads to a $1.30 change in private payments. When Medicare mistakenly pays generously for low-value services, the private sector follows its lead. Medicare similarly moves the level of private payments when it alters fees on an across-the-board basis. Medicare thus in influences both the relative valuation of, and aggregate expenditures on, physicians' services.
The cost shifting hypothesis has been consistently rejected by every publication (seven of them, by my count) since 2011. Though cost shifting is not impossible, under the right circumstances, I think we can very confidently say from the evidence that it has not been possible in recent years.
* Private pay rates would be expected to fall along with Medicare's if a hospital attempted to decrease Medicare volume in exchange for increasing private pay volume. To attract more of the latter, it would have to charge private payers less, not more as cost shifting would suggest.