• Meme-busting: Hospitals shift costs

    This is an ongoing series on health care system “memes” that continue to permeate our debate, even when evidence shows them to be false. The introductory post contains links to all entries.

    Judging from a recent comment, among other things, it seems that it is still taken for granted that hospitals shift costs, or make up for shortfalls in public payments by charging private insurers more.

    I’ve written so much on cost shifting that it is hard to think of another way to make the point that hospitals are limited in their ability to do this. For a list of many of my posts on this subject, some of which link to my paper on it, see the hospital cost shifting FAQ entry.

    Here’s a new approach to the subject, based on the exceptional reporting of Christopher Flavelle of Bloomberg News. Let’s take a look at what some of the hospital and insurance industry groups have claimed. Flavelle’s briefing paper has some terrific quotes.

    From the AHA:

    A spokeswoman for the American Hospital Association, Marie Watteau, wrote in an email March 10 that “underpayment by Medicare and Medicaid is one of the many factors that go into the rates negotiated by hospitals.” Payment rates by the two programs “can be viewed as a cost of doing business that competing providers all must pass along,” Watteau wrote. “It is analogous to when gas prices go up, airlines all must raise their prices.”

    From AHIP:

    A spokesman for America’s Health Insurance Plans, Robert Zirkelbach, asked by Bloomberg Government on April 28 whether hospitals charged private payers more in response to cuts by Medicare and Medicaid, responded by e-mail, “Yes, definitely, and that is what the data show.”

    Both associations reference a 2008 paper by Miliman in support of their cost shifting claims. However,

    Rather than demonstrating cost-shifting, the 2008 Milliman paper referenced by both associations only measures the difference in what Medicare, Medicaid and private insurers pay hospitals for the same services, a phenomenon called price discrimination. One of the authors of the paper, John Pickering, said in a phone interview March 10 that despite the use of the term “cost-shifting” throughout the document, the study was “more a look at price differentials that exist,” and wasn’t intended to examine whether those differentials were caused by low Medicare and Medicaid payments.In fact, Pickering said he wasn’t aware of any evidence whatsoever demonstrating that hospitals react to Medicare cuts by raising costs for private insurers. “I haven’t seen analytics that argue the cause and effect,” he said.

    The analytics I’ve seen (from careful studies published in the health economics journals) show that, with the possible exception of the late 1980s, if and when hospitals shifted costs, it was at a rate much lower than AHIP or the AHA suggest. In many circumstances hospitals cannot shift costs at all. They lack the untapped market power to do so. In others, they can shift costs, but at about 20 cents on the dollar; private hospital prices go up by 20 cents for each dollar public prices decline. The remaining 80 cents — the vast majority of shortfalls — are made up by cost cutting not cost shifting.

    Since I’ve backed all this up before, I’ll refer you to prior posts for the details.

    • Austin thanks again for all of your clarification here. In my prior post I see that it could be taken that I implied (perhaps Paul Levy even does) that there is a 100% cost shift, where clearly your review of the data and studies show $0.20 on the dollar. In fact only working for a private insurer it’s impossible to know, for instance, how much of the hospital’s cost is being shifted.

      But $0.20 on the dollar is still significant, wouldn’t you say? If Medicare reimbursements do not track national health spend as a whole then the shortfall in public spending trends compared with the overall spend must be picked up in some form, either by cost cutting, shifting to out-of-pocket expenses, or to private insurers (which is really out-of-pocket expenses again for insureds).

      But again thanks for you work and clarification here.

    • I think the If and When is the key concept around the 20% figure. Take into consideration that administrative and overall overhead costs are generally high and 20% is not that high of a figure. However, I’d say this: Are we shifting costs or are we shifting high marginal revenue sources (i.e contributions to profit). While, perhaps one of the papers analyzes that idea, its very hard to get that data in a fashion to determine what the case really is on that. The other aspect is that you might ask yourself why monsopnistic or the oligopoly analogy buyers have eager vendors in most market situation even when lower prices are given. The answer is that the market volume justifies the business model. In other words, medicare is a pretty good buyer for most vendors and so a good business model. Thoughts?

    • BTW I meant “monopsonistic” ie. single buyer markets or the few buyer equivalent. Typing on webscreens can be difficult.

    • Dr. Frakt,

      I need help understanding something. Take a look at the MedPAC June 2010 Databook.


      Look at Charts 7-12 to 7-16

      We all know Medicare undercuts in payments. So if Medicare margins are negative, but throwing in private payors hospitals make a black margin. As well, look at chart 7-22. It is a graph of hospital markup charges.

      So if hospitals are profitable because of private payors and markup is increasing, does this not translate into cost shifting or am I defining cost shifting incorrectly?

      • No, that’s not a proper definition. Can I encourage you to follow the links in the FAQ on this? Look for “price discrimination”.

        Ask yourself if you can tell causation from the figures you cite. How do you know high private payment doesn’t cause low Medicare margins? I’ve covered published work that includes evidence consistent with just that hypothesis. It can’t be discounted so easily.

        • Ah, I followed through on your links.

          So are you arguing that 21% increase is not large and should not be a concern?