The Health Care Cost Shifting Myth


Picture 12 There is a pervasive notion that providers of health care can make up for lower payments received from one set of payers (e.g. Medicare, Medicaid, uncompensated care) by increasing prices charged to other payers (e.g. private insurance companies). To the extent it occurs cost shifting offsets attempts to control overall health care costs through reduced fees paid by public insurers. It makes “bending the cost curve” harder.

However, it is a myth that providers can fully shift costs. That they could do so violates, in most cases, principles of economics. Moreover, empirical evidence suggests cost shifting, where it occurs, is done so a minimal level: only a small fraction of decreased payments by public payers shows up as an increase in charges to private payers. Losses associated with one payer are largely not recouped from another.

Some take price discrimination as evidence of cost shifting. However, price differentials are not necessarily the recouping of losses from one payer by overcharging another. As described in the 2001 Health Affairs paper by Richard Frank “Prescription Drug Prices: Why Do Some Pay More Than Others Do?” price discrimination can be due to unequal bargaining power across classes of purchasers. In other words, in maximizing profits, providers charge different prices to different market segments. In such cases, by definition, profits cannot be further increased by cost shifting.

It’s true that cost shifting could theoretically occur under specific conditions. One case is when a provider has monopoly power that it has not fully exploited, for instance charging private insurers less than it could. More fully exploiting its monopoly power with respect to those payers, such a provider can recoup losses. Still, there is a limit to how much of the lost revenue can be recouped. The monopoly profit-maximizing price level imposes a ceiling.

Another instance in which cost shifting could occur is in a more competitive market in which all providers have roughly the same level of undercompensated care. All competitors in such a market might choose to increase charges to private insurers by the same amount, maintaining their relative competitive positions. However, if one competitor elects to reduce costs or reduce its burden of undercompensated care, it might be able to charge private insurers less then others, thereby increasing its market share. So, cost shifting may not be a stable equilibrium.

The literature provides estimates of the extent of cost shifting in cases where it is theoretically possible. The March 2009 MedPAC Report to Congress: Medicare Payment Policy (Chapter 2A) includes a summary of such evidence. It concludes that the dominant dynamic in the market is that hospitals with strong market power have abundant financial resources. In turn they have a high cost structure (perhaps due to provision of relatively higher quality care) that causes lower or negative Medicare margins. In contrast, hospitals that are forced to run efficiently are adequately funded by Medicare payments. That is, Medicare payments are sufficient to cover costs but some hospitals run inefficiently and make it appear otherwise. Therefore, MedPAC has concluded that increased Medicare payments to hospitals would not reduce rates charged to private insurers. The primary effect would be to induce lower cost operations.

The MedPAC report cites mixed evidence from the literature on the level of cost shifting, as does the December 2008 CBO report Key Issues in Analyzing Major Health Insurance Proposals. A few studies from the 1980s found evidence of cost shifting at a rate of up to fifty cents on the dollar. However, conditions in the 1990s were less conducive to cost shifting and the rates were found to be on the order of a 0.4 to 1.7 percent increase in private payments in response to a 10 percent reduction in Medicare and Medicaid fees. In a 2005 study of geographic variation in health costs of the Federal Employees Health Benefits Program, the GAO concluded that the considerable variation it found was not due to variations in payments from other payers.

In conclusion, cost shifting is not as large and widespread a phenomenon as some would believe. Under some market conditions it is inconsistent with economic theory. And, while it can occur under other market conditions it is far from a dollar-for-dollar shift in costs. The most recent studies of the phenomenon find little evidence of cost shifting or very low levels of it. Claims that reductions in public payments for health care will necessarily show up as commensurate increases in private payments are unfounded.

Austin Frakt blogs at The Incidental Economist and is a health economist and principal investigator with the Department of Veterans Affairs’ Health Services Research and Development Service and assistant professor with the Boston University School of Public Health, Department of Health Policy and Management. The views expressed in this post are his alone and do not necessarily reflect the positions of Boston University or the Department of Veterans Affairs.

Categories: Uncategorized

Tagged as: , ,

58 replies »

  1. It seems to me that this dialog, and most other “cost shifting” dialogs, myopically overlook the full public cost of providing care to non-insured and underinsureds. The significant costs of providing health care at county hospitals and in the ER’s of non-profit and for-profit hospitals seems to be totally overlooked in these discussions.

  2. Cost shifting is alive and well in Wisconsin. Our (unscientific) research indicates that all area nursing homes use cost-shifting to the tune of 20-30%. If the law required private pay to be the same as Medicaid, Medicaid would no longer have the luxury of a select group of private citizens paying what is a Medicaid responsibility. True, Medicaid costs would rise to make up the difference, but at least *all* of the public would be paying for those public services. Medicaid might also be a little more vigilant when they no longer have a captive group to absorb some of their costs.

  3. If you had done your home work and consulted with providers your blog would be credible, but suggesting that cost shifting it is a myth, only confirms the fact you are a non factor.

  4. Unbelievable…the idea the cost-shifting is a myth. Ignorant or deliberate obsfucation: you choose.
    You want to understand the reality of cost-shifting? Look at Aetna’s miraculous turnaround from the brink of extinction. It’s a case study in how cost shifting really works.

  5. And, the data from Medpac clearly shows that doctors are not voting with their feet and leaving Medicare in droves.
    The threatened medicare cuts have also been passed off every time they come up. It is a bankrupt system refusing to take the action necessary to save it that knows when they do take needed action the providers will walk

  6. Looks like my reply was eaten, so I’ll try again. So, Medpac studies don’t fly here, but incomplete summaries of the literature from nearly two decades ago by trade journals do fly? Sorry, but I know a few of the Medpac commissioners, and I’ll trust their efforts a hell of a lot more than Entrepreneur magazine.
    Entrepreneur’s summary has one thing right. Those studies found that in the early 1980s, immediately post-PPS, there was some evidence of this type of dynamic cost-shifting (to differentiate from the static price discrimination cost-shifting that Austin began with in his note). Even those studies found that a 10% cut in Medicare prices only led to a 3-4 percent increase in private prices. And they found that even as the 1980s ended and the 1990s began, the effect became weaker and weaker–both smaller and bordering on being statistically insignificant. And all of these these studies ended their analysis with data that predated the growth of managed care.
    For dynamic cost-shifting to happen, economic theory suggests that ther have to be two conditions. First, providers must have market power compared to private insurers. Second, they must not be fully exploiting that market power to begin with.
    Most economists agree that this was a reasonably accurate description of the situation in the 1980s. But, very few think it’s true in the 1990s or 2000, as insurers consolidated, reducing comparative market power of providers, and as competition increased and reduced the ability or willingness of providers to less than fully exploit whatever market power they had left. As a result the most recent study (Zwanziger and Bamezai 2006), which analyzed data through 2001, showed that there was no significant cost shift from Medicaid and that the Medicare cost shift had shrunk to less than half of what it had been in the 1980s.
    And not a single one of these studies accurately deal with the fact that the simple price comparison is inappropriate, because it fails to account for public support of capital costs.
    The data on hospital profitability that I cited comes directly from the AHA–if anything, they are under-representing the profits from Medicare among hospitals.
    Who besides Medicare says Medicare pays a fair price? Well, I count on what people actually do, rather than what they say they might do if Medicare cuts their rates by 10 percent (hey, notice that the doctors did not say they would just raise private prices to offset the Medicare cuts on their income–seems to me that contradicts your cost shifting argument!).
    And, the data from Medpac clearly shows that doctors are not voting with their feet and leaving Medicare in droves. And, despite your polemics, that is the third option open to every doctor and hospital that you ignore.
    The costs that are often cited in these complaints about Medicare are horribly inflated fictions of what some hospitals would like to receive if they could go back to the world of cost+ reimbursement. The reality is that there are thousands of hospitals and hundreds of thousands of doctors and physician groups who have made their practices more efficient and make a profit on Medicare. The question should not be why Medicare won’t pay inefficient providers more. It should be why more providers are not working as hard as the others to become more efficient in their processes of care.

  7. http://www.entrepreneur.com/tradejournals/article/114982248_2.html
    Why is it we should dismiss all of these studies?
    Another possibility for hospital managers to compensate for reduced Medicare payments existed by which the charges were (generally) a tree reflection of costs, the percentage paid by Medicare declined, and then Medicare patients were cross-subsidized from other revenue sources, most likely private insurance (which could account for recently escalating premiums for private health insurance). There is certainly a historical precedent for this kind of cross-subsidy by private payers (cost shifting by hospitals) to compensate for underpayments by Medicare (Guterman 2000), so this was a likely scenario. Ginsburg (2002), Dobson (2002), and McGuire (2002) recently analyzed the evidence concerning the effects of reduced public payments on hospital cost shifting to private payers. Clement (1997) provides an empirical perspective on cost shifting in hospitals from the 1980s and early 1990s, before the implementation of the 1997 BBA. Zwanziger, Melnick, and Bamezai (2000) analyzed data from California for the time period 1983 through 1991 and provided clear evidence of cost shifting in hospitals from Medicare to private payers. Thus, the empirical basis establishing cost shifting in hospitals from Medicare to private payers has long been established.
    And according to a 2007 AMA survey of 8,955 physicians in the United States, 60 percent of doctors said they plan to limit the number of new Medicare patients and 40 percent of doctors said they plan to limit the number of established Medicare patients that they treat if Medicare payment rates are cut by 10 percent in 2008.

  8. GTWMA what studies are you referring to? And MedPac studies don’t fly here.
    First off even if you aren’t seeing new patients it pays 5% more to sign the annual agreement;
    “The Medicare payment amount for PAR physicians is 5% higher than the rate for non-PAR physicians.”
    Not to Mention;
    “Carriers provide toll-free claims processing lines to PAR physicians and process their claims more quickly.”
    “Limiting charges for non-PAR physicians are set at 115% of the Medicare approved amount for non-PAR physicians. However, because Medicare approved amounts for non-PAR physicians are 95% of the rates for PAR physicians, the 15% limiting charge is effectively only 9.25% above the PAR approved amounts for the services. Therefore, when considering whether to be non-PAR, physicians must determine whether their total revenues from Medicare, patient copayments and balance billing would exceed their total revenues as PAR physicians, particularly in light of collection costs, bad debts, and claims for which they do accept assignment.”
    Not like the government makes it a balanced decision, you either accept PAR or they really screw you.
    If 50% make money on Medicare that means 50% lose money, lets name names and see who is winning and losing. If physician owned and specaility hospitals are the only ones making a profit I think that proves we have more problems then good points to build on.
    Studies like this tell me we have problems;
    Results: The average Medicare dollar margins were )$712 (95% confidence interval [CI] = )$729 to )$695) for ED admissions and $22 (95% CI = )$2 to $47) for elective admissions. Medicare dollar margins for ED admissions were lower than those of elective admission for the most common DRGs. ED admission was associated with greater patient severity of illness.
    1.1 million admissions is a nice study to work from. Specality hospitals don’t have ED admissions like public and rural hospitals do.
    Who besides Medicare says Medicare pays a fair rate?

  9. Nate explain to me why half of hospitals report they make a profit on Medicare. If Medicare payments are so inadequate, explain to me why the percentage of physicians who say they accept all new Medicare patients is almost exactly the same as the percentage of physicians who say they accept all new privately insured patients. Medpac’s annual reports on payment adequacy measures show little of the massive “walking away” that you claim. 95 percent sign participation agreements, the same as in past years. 92 percent are accepting new Medicare patients, no change from prior years. The number of physicians to beneficiaries continues at the same levels. Beneficiaries report few, if any, access problems.
    And Medicare does not represent 50% of the physician market. For the typical physician, it represents just over 30% of their revenue.
    For well run practices and hospitals, Medicare payment not only covers marginal costs, it’s profitable. And the objective data on physicians and beneficiaries shows just that.

  10. GTWMA,
    I don’t beleive Medicare covers marginal cost. I don’t beleive Medicaid pays anything close to marginal cost. If they did why do so many doctors opt out? If your argument was even close to accurate then an economist would expect every provider to particiapte in those plans. Why would any provider choose to not participate in a plan representing 50% of the market based on payment if it covered his marginal cost?
    I don’t think highly of physician business skills but find it hard to beleive that many are making the bad business decision to walk away from profitable business.

  11. Now, back to the issue of marginal costs. Seems as if you accept the idea that evidence shows that Medicare covers marginal costs. So, that gets us to the next stage of the argument–cost shifting is mostly a question of who pays the fixed costs of business. The two most important points about this are: (1) this is not a situation unique to health care. Whenever you have suppliers and demanders with market power–and both Medicare AND hospitals and physicians have significant market power–you are going to have this tussle over who pays their fair share of fixed costs. And, (2) the question of who pays fixed costs cannot be answered solely by looking at prices. Medicare, and governments in general, pay for the fixed costs of business for health care organizations in a lot of different ways, including tax advantages, grants, discounted loans and more. To truly compare whether Medicare is paying its fair share, you would need to factor into any discussion not just the comparative prices, but a fair comparison of how government helps organizations pay those fixed costs.
    As for the juvenile debate about whether economists are in ivory towers or not, here’s my view. We both have things to learn from each other, and the sooner some of us get out of elementary school name calling the better. The advantage of the tower is it gets you to look beyond the surface views (WAAAAH, Medicare isn’t paying me enough) to get to the heart of the matter (cost shifting is really a complex question abut how we pay fixed costs in our health system and requires looking at much more than price comparisons). When we get to the heart, rather than just stay on the surface, we might actually find we agree about some ways to improve health care.

  12. You are assuming, Nate, the provider income is a set figure that must be paid. It’s not. It’s determined by the most efficient payment to get the provider to provider services.
    And there is ample evidence that there are many, many physicians and hospitals who have figured out care so well, that they provide higher quality care at lower cost than the Medicare payment.
    The question isn’t whether Medicare’s payment is too low. It’s why these other providers insist on receiving their income, rather than insist on providing the rest of lower cost and higher quality service.

  13. A=provider income
    B=public payment
    C=private payment
    A must be greater then $x for providers to stay in business. If you reduce B then you must increase C, that is the simple math that people understand and the simple math that applies in the real world, outside the ivory towers and their theroies.
    We have controlled experiment ask any hospital administrator or provider asking private insurance for an increase why and they will tell you.

  14. [Re-try. Apologies for duplicate posts (if any).]
    Wow. This has been quite a discussion. I can’t possibly respond to all comments and won’t. Many have been debated back and forth quite well by others anyway.
    My piece was a summary of economic analysis of cost shifting conducted by others. The aim of those analyses was to identify the causal effect of public payment on private payers. That is a very specific and challenging exercise and different from other financial or accounting exercises one could do and others have done.
    It is possible, and likely, that different people and disciplines use the term “cost shifting” to mean different things based on different types of analyses. It is also possible, and likely, that individuals readily generalize their specific experience and market to all experiences in all markets. Cost shifting of one form or another likely exists at a high level in some markets while not existing significantly in others. One’s experience can be both valid and not reflective of the whole. (Note one could make a similar argument to refute one or another study: the data are specific and results not generalizable.)
    I’ll close with the essence of the cost shifting question as *I* see it. I do not expect, nor do I think it necessary, that all see it this way.
    A = B + C
    Does this expression mean that (1) a change in B causes one in C, or (2) a change in C causes one in B, or (3) changes in B and C are both caused by something else, or (4) is this just an accounting identity that must hold and says nothing about how B and C causally relate to one another or to anything else?
    Questions such as these are at the heart of much of economics. One’s mind can trick oneself into believing one or another of the above four options are true. If one is honest one can only say that the truth is unknown without further investigation via controlled experiment or very careful statistical work. Much debate is over a lack of understanding that the answer is otherwise fundamentally unknowable and/or a misunderstanding of the techniques statistics and economics bring to bear on such problems.

  15. Readers of these comments might be interested in an article posted by Science News “Medicare changes threaten access to radiation therapy”.
    (I am not an oncologist or associated with Science News). If anything the article supports the above post by Nate by expressing the concern physicians have that Medicare and Medicaid reimbursements will not be sufficient to meet physician office expenses and keep them in business.
    Steve Zeitzew
    an Orthopaedic Surgeon in Los Angeles

  16. I don’t think there is any question cost are shifted. That doesn’t mean there are not alternatives to reduce or eliminate cost shifting but you can’t deny that cost are shifted.
    Why cost are shifted doesn’t change the fact they are. By denying there is any cost shifting people want to skip the entire argument of sufficient public reimbursement for services rendered. For the sake of acheiving what is important;
    Fine there is no such thing as cost shifting.
    I now want to discuss the inadequate reimbursement of providers by public plans. The label of the discussion has changed but the facts haven’t. Providers could not afford to stay in business reimbursed solely at Medicaid and Medicare rates. While some providers could improve efficency and lose less if paid at Medicare no one could get by on Medicaid. Most can not cut enough to get by on Medicare. As long as those two plans continue to underpay good doctors for legitimate work the system is going to be screwed up. As long as they pay any crooked doctor for questionabale work the system will be screwed up. Fix the public plans everything else will fall in place.

  17. I would remind readers of anything written by an economist. Beware! Economists are like lawyers, they can find an economic answer for any situation that perfectly suits their clients’ desires. This article sounds just like the economic equivalent to the Dan White twinkie defense for killing San Francisco’s mayor and city councilman.

  18. Wow! What a discussion! Does cost shifting exist? I don’t know but someone at the Colorado Department of Insurance thinks it does and I can objectively measure it… I am supposed to answer the following question on a survey:
    32. Percentage increase in medical (non-Rx) costs due solely to medical
    cost-shifting – Please enter the year over year percentage increase in nonprescription
    drug medical costs associated with the shifting of medical
    costs onto the insured members of your health plans in Colorado.
    So hang on guys… I will figure it and let everyone now.

  19. To recap:
    Medicare pays rates roughly at the marginal cost of care overall. Hospital decides it needs to keep its margins up so demands higher payments from private insurers to achieve margins of 3%, 5% or whatever on its total book of business. The insurer accepts the higher demanded rate, or something close to it. Some say that counts as cost shifting and some say it doesn’t.
    Those who say it doesn’t point to the fact that there are other ways this scenario could have played itself out, dependent on market conditions. The hospital could have reduced it’s operating costs (some do). The hospital could refuse to accept Medicare. The insurer could have played hardball and refused the higher rate, resulting in either the hospital backing down or the insurer dropping it from the network. Those are all business decisions. Sometimes, highly constrained by market conditions like local market share and expected public/customer response.
    Austin and others say the fact that these possibilities exist means that market conditions determine the rate private insurers pay. Nate and others say that in fact private insurers often have no good business decision other than to accept higher rates than Medicare and so this is cost-shifting.
    Is this just a semantic debate about the meaning of “cost shifting?” Did anyone define what they meant by the term? Or are we disagreeing about the relevance of possible responses that wouldn’t result in higher rates paid by private plans? Nate seems to exclude them as irrelevant to assigning cost shifting on the grounds that the plans had no practical choice but to accept higher costs, while Austin seems to argue that the fact that these practical considerations are market conditions is relevant and means we aren’t dealing with cost shifting.

  20. do they treat just public enrollees or do they spend 1-2 days a week in the clinic then have their private practice on the side. Can’t really tell from the website and your primarily comment contracdicts my just.

  21. Nate asks:
    > anyone ever see a provider treating just
    > public plan enrollees? Seeing as how they
    > are an underserved market if there was really
    > money to be made why is no one targeting it?
    Well, yes.
    Five or six FQHCs on this list see primarily Medicaid and uninsured patients. The universities don’t 😉 As for making money, nobody’s starving, but they’ll never be particularly well-off.

  22. The last few comments cogently reflect how some of the prior post demonstrate that emotional symbols and ad hominem arguments can become sacrosanct to those primarily informed by ideologically.
    To Austin’s empirical point, Medpac simply points out, (among other things) that 1,138 hospitals reporting the highest Medicare losses (-20%), were actually some of the more profitable hospitals overall, with a median total margin of 4.6%, relative to a 3.4% overall margin for 964 hospitals earning a margin of 7.6% treating Medicare patients.
    If we take a step back from absolutist hyperbole, irrespective of how we label it, we can probably concur that it is fairly likely for least some portion of 964 hospitals, competition has probably sharpened quite a few pencils.

  23. Jim you might want to be aware that 50%+ of all employer provided insureds are in self funded plans, thus there really is no insurance company and thus any argument about them living off the public teet mute.
    “tax subsidies for private health insurance”
    interesting way of stating this, by the governemnt taking less of my money from me they are subsidizing me? In your view all my money really belongs to washington and only for their good graces I am allowed to keep some of it?
    There are providers that treat only private insureds, anyone ever see a provider treating just public plan enrollees? Seeing as how they are an underserved market if there was really money to be made why is no one targeting it?

  24. What about “loss shifting”? As Medicaid and Medicare reimbursements fall, in some cases they are below cost, an increasing number of physicians shift their practices to care for fewer patients in those plans. Some private practices have simply gone broke and no longer exist. Some physicians have left their Los Angeles based private practices. I don’t believe there are any orthopaedic surgeons in Los Angeles who can afford to participate in MediCal HMO’s (Medicaid). Surgeons in some specialties can restrict their practices to outpatient surgery centers and to paying patients shifting the potential losses in caring for other patients to other physicians. Physicians sometimes avoid or leave hospitals that compel them to be on-call. Some of these losses are shifted to hospitals that must pay physicians to be on-call. Urgent care centers capture some paying patients shifting losses to hospital based emergency rooms. Patients are shifted to hospitals supported by government, some in California supported with additional state funds for disproportionate care to the uninsured. HMO’s build their own hospitals and clinics largely restricted to care of patients who pay for that HMO, effectively shifting others elsewhere. Patients who cannot afford care may eventually required emergency care, shifting losses to the emergency room. These patients shift some of these losses to themselves by enduring illness, shortened lifespan, and disability along with the economic consequences. Outsourcing shifts some loss. For many years GM opened plants elsewhere in part as a response to medical costs. Production of the Camaro was shifted to Canada saving hundreds of dollars per vehicle in healthcare costs alone, shifting losses to laid off workers and the American economy. Some companies shift healthcare losses to employees by restricting some to part-time positions not eligible for healthcare benefits.
    Physicians in increasing numbers are choosing salary supported positions shifting the potential for loss to the hiring organizations. These groups may have some influence negotiating with insurance companies, but it was my experience when I was in private practice that the insurance companies universally told me to “take it or leave it”. I couldn’t find a way to further reduce my costs without an even greater loss in capacity to practice effectively.
    Steve Zeitzew,
    Orthopaedic Surgeon in Los Angeles

  25. Edward Kleinbard, the top economist for the Joint Tax Committee in Congress, says that tax subsidies for private health insurance now total more than $300 billion a year. The Milliman study using 2006 numbers, approximates operating expenses for nearly 5000 community hospitals at 600 Billion.
    Private insurance is paying out approximately 300 Billion to hospitals annually. The remaining 300 Billion is paid directly by the federal government.
    The mythical (but plausible) cost shift from public to private insurers is justification for private insurers to drive up insurance premiums and to beg for more government giveaways. The beauty of this situation, the blame for the rise in premiums is placed squarely onto the shoulders of government.
    In a nutshell, a cost shift from public to private insurers allows private insurers to milk more money fron the federal government, retain more revenue and blame the federal government for the problem.

  26. Economists, policymakers, non-economist researchers, and people in the trenches of provider reimbursement have been arguing the cost shifting issue for years.
    There is an old joke about economists that they observe something in practice and wonder if it can work in theory. Hospital managers in my experience always say they need to charge higher prices to private to cover costs not paid by public payers. That said, it’s in the interest of hospital and other providers to say one thing (to lobby for higher Medicare and Medicaid reimbursements) and do another (maximize their revenue stream from whatever sources).
    My own personal take on the issue is that the “principles of economics” suggest that we should be very skeptical of claims that “cost-shifting” is occurring, because its often a case of mislabeling something else. One big hole in the excuse that cost-shifting is inevitable is the assumption that costs are fixed, so they have to be shifted.
    The discussion above notes that a good portion of the relative price differentials paid by different payers is not cost-shifting per se, but price-discrimination. This is not a cost-shifting phenomenon about covering some fixed level of cost, but a demand-side phenomenon about the relative market power of various payers.
    I think that Dr. Frakt’s basic point is that simply raising public reimbursement rates will not relieve private payers of paying higher relative prices for health services. As MedPAC and CBO studies would support, increases in public payment rates would increase revenues to providers, but while these increased revenues might be used to support patient care, they won’t (by themselves) result in greater efficiencies. But, the appearance of “cost-shifting” might diminish, only because the increased revenue flow might allow for less-intensive negotiations with payers outside the public sector. However, in a competitive market, the incentives are for the bargaining parties to exploit every advantage they can, and providers will still be seeking to extract higher payments from whatever source they can, whether lobbying for higher Medicare and Medicaid payment, or negotiating for higher private pay rates.

  27. Eagerly awaiting Austin’s research on actual costs to hospitals for provision of specific services. & no fair suggesting that code assignment=identification of costs to provide actual services.
    The confounding myth is that costs for specific services can be adequately identified, not that cost shifting can be discerned. Is there a hospital that scrupulously breaks down costs to specific services rendered, adequate to providing conclusive proof for or against Austin’s assertion? They generally know their total costs for gauze & slippers, but simply don’t capture costs at the product level.
    As a result, all service costs are “shifted” in some fashion; usually in more than one fashion. The semantic static overwhelms any attempt at making an objective case for or against.
    in this particular debate the academios have had their propeller beanies handed to them.

  28. GTWMA writes:
    > These studies all show that Medicare clearly pays
    > the marginal costs of treating its beneficiaries,
    > which is what any payer must do.
    Well, yes, this is the minimum any payer must do over the long run. In the aggregate though, payers must pay at least the long-run average cost of treating all beneficiaries. Since Medicare is the 900# gorilla, they pay something like the long run marginal cost of treating their beneficiaries. The difference between marginal and average is made up the the other half of the payers. The commercial insurers call this “cost shifting”. Economists call it something like “price taking under duopolistic monopsony facing a monopoly seller”.
    The term “cost shifting” is propaganda. We don’t call it “cost shifting” when a ticket bought today for a flight leaving at noon is priced higher than a ticket for the same flight bought a month ago, or when flying standby costs less than buying a ticket more than a few minutes before the flight leaves. It is simply a consequence of differences in market power among buyers and sellers. Of course (as Nate will point out) we do not have airline fare insurance. But the presence or absence of insurance is a market condition also.
    Hospital economics are sort of like airline economics: very high fixed costs, relatively low variable costs. It does not cost much more to fly a full airplane than to fly a half-full airplane on any given day. True also of hospitals. You mgiht spend less on soap, meals, even drugs, but 65% of your cost is labor, 15% or so plant and equipment. If you start laying people off for a half-day or a day without paying them, your workforce will leave, so you’re going to spend 80% no matter what.
    I have thought for a long time that Medicare paying more will not result in commercial insurers paying less. It is more likely to result in higher executuve salaries or better shareholder dividends, and maybe better indigent care.

  29. As PN points out, if Medicare hired a real administrator(s) that knew how to process claims and manage risk they could pay providers a fair price per unit and still reduce total cost. Medicare has shown numerous times that cutting rates doesn’t save money, it penalizes the honest and good providers more then the corrupt.

  30. Milliman’s study can be dismissed for two reasons: 1) It only measured the difference in payments and did not establish any mechanism or proof that a cost shift has occurred and 2) It is clearly biased.
    If it were biased but actually did show what you wished it showed, then it would have to be considered.
    If a parts supplier was being paid $1 per widget by GM and $1.25 per widget by Ford, would you say that the supplier was “cost shifting” from GM to Ford? That would be ridiculous.

  31. Nice analogy Peter on the whack-a-mole. I eagerly await an answer to your question though I suspect it may not be forthcoming.
    If Miiliman’s study can be dismissed simply because it serves the insurance industry’s interest could not the same be said of almost every post by a physician or a certified CCHIT vendor that graces these pages?

  32. As the former president of a managed medical network, I paid providers substantially higher fees than Medicare and more than most insurers. Yet we reduced medical costs by more than 40 percent on the average. How does your model explain those results? (Over 100,000 providers)
    A common confusion in cost analysis is to equivocate on costs and fees. Actually, fees make up part of the medical costs, utilization makes up the other component. It is relatively simple to try to solve the medical cost problem by pushing down on fees but like a medical wack-a-mole, the costs increase somewhere else – utilization. The case in point – Medicare.

  33. “Under the original Medicare fee-for-service model, we lose $464 per patient each year.”
    Then get your costs in line.

  34. “by estimates, I mean the results of statistical analysis.”
    Would those be the same statistical analysis used to project the cost of Medicare since its inception? Haven’t those been way off every time they did them? How did those Part D statistical analysis work out? Don’t states use your statistical analysis to project Medicaid cost? Can you point me to an example of where statistical analysis was actually close to being right?
    I suppose Austin knows better then this guy to what his own data means?
    “The Everett Clinic, which cares for more than 28,000 Medicare patients, just completed an independent review of our Medicare finances. The results are sobering.
    Under the original Medicare fee-for-service model, we lose $464 per patient each year. In 2009, the Everett Clinic will lose approximately $11.7 million treating Medicare patients. With anticipated growth in the Medicare population, it is estimated that the loss will grow to $70 million in the next five years.”
    “To compensate for losses on Medicare, physicians are forced to increase the fees charged to other patients.”
    But you have the catch all excuse that anyone who disagrees with you is protecting their stake in the game, thus everyone who actually does this for a living is lieing and some disconnected economist with statistical analysis is correct. I can’t imagine why government healthcare plans always end up broke…

  35. Cost shifting is too complicated. Merely jack up the volume, do more scopies and bopsies, and buy a computerized record keeper that spits out pages of gibberish to justify a level 5 visit code.

  36. the Milliman study “quantifies” the cost shift but it doesn’t show how the cost shift happens. Also, it was bought and paid for by the Commercial insurance industry and was not peer reviewed. How is that relevant as anything but propaganda?
    I’d say the same to you: If there is only one hospital in town how does any of your argument apply? Why is it cost-shifting and not just negotiating with the only payers that negotiate?

  37. Hospitals don’t merely shift cost between payers, they shift costs between different services. Any hospital that specializes in trauma or burn care is going to see a high volatility and high average in costs per claim, but the reimbursement rates won’t vary much, and they probably won’t make a very large profit margin off of those claims.
    That doesn’t mean they are any more or less efficient, they just don’t have the luxury of counting on a predictable and steady stream of revenue. Rural and suburban hospitals don’t have to account for nearly as many possible services, and thus don’t have to shift costs around to make up for lost revenue. They can then start offering more high margin services, like child birth and elective surgeries, that help to subsidize Medicare and Medicaid underpayments.
    Different categories of services are disproportionately used by different populations, and providers will definitely charge more for services where they know they will be reimbursed. This is the essence of cost shifting.
    If cost shifting weren’t a problem, then why would we have so many providers and clinics who refuse to accept any new Medicare or Medicaid patients, if at all?
    Hospitals have to operate just like any other business: recruit the best care givers and make a healthy profit to invest in better equipment, better facilities and better efficiency. If we used the M/M model, everything would be covered at cost, and there would be no ability to invest in improvements.

  38. Tim,
    The reality is those type of contracts DO exist in the market, stupid or not. A hospital who is the only game in town and has 55% goverment business – simply changes their charge master 2-3 times a year to make up for some of the difference. These type fo contracts apply to all kind of carriers and payors inlcuding BCBS. In addition to the absurdity of the charge levels, the place of service differences in a cost of a procedure are dramatic. An MRI at an outpatient hospital can be 3-4 times more expensive (in charges) than one at a free standing facility…..Yes it is stupid but t is also reality that we have to address.

  39. Nate, by estimates, I mean the results of statistical analysis. There are several reasons to rely more on that than on the data you cite. First, it’s by independent researchers, rather than those with a stake in proving cost shifting. Second, it relies on data from many different markets over many years. Third, it uses more sophisticated statistical models to prevent bias.
    So, yes, there are many reasons why those results are less valid. And there’s PLENTY of evidence that reducing public payments spurs efficiency–basically that’s the entire history of DRGs. If you doubt it, let’s go back to cost-plus reimbursement and see what happens.

  40. Austin’s study of other peoples work 3 or 4 times removed is real. The first hand experience of people like myself and Scott is not.
    Austin’s work has validity becuase he is an economist, Milliman’s does not becuase they are a benefits actuary who does this stuff all day.
    You say your self they are estimates, I’m not using estimates I am using real actual contracts and the real actual claims I process every day. But those are less valid then Austin’s estimates?
    You don’t force providers to achieve efficency by reducing their public reimbursement. Those are two completly seperate issues. No one would disagree providers need to get more efficient. Cutting reimbursements doesn’t accomplish that, it just results in greater cost to private insurance. Something that someone actually working in the trenches every day would see. Private insurance has no negiotating power over the majority of hospitals. I see PPO contracts of 0% to 5%, they aren’t anything more then prompt pay discounts, but when it is the only hospital for 100 miles there is not much you can do.

  41. Nate, we’re not talking about theories. We are talking about real world estimates made by real world economists using real world data. Austin summarizes it well. These studies all show that Medicare clearly pays the marginal costs of treating its beneficiaries, which is what any payer must do. There are PLENTY of providers who treat Medicare beneficiaries for less than what Medicare pays AND provide quality that far exceeds the typical provider. The question isn’t why Medicare isn’t paying its fair share. It’s why more providers can’t achieve the level of efficiency and quality that the best Medicare providers can.

  42. becuase the phrase cost shifting means taking expenses one plan should be paying and making another plan pay for them. Medicare and Medicaid do not reimburse enough to cover the services they receive. Private plans are forced to make up the difference. You can’t just redefine the word.
    How can you claim medicare doesn’t have anything to do with it. If you removed all the private plans the hospitals would be broke and close becuase their cost was not covered. You don’t seem to understand the PR involved in cutting a hospital from your network. If your members think it is your fault they will leave for another network. If they blame the hospital they might actually change hospitals.
    This is real world activity here, not the pretty little models economist draw up where everything functions like they want it to. When hospitals are cutting service and receiving state bail outs the public beleives they need higher reimbursements. When HCA is spending lavishly and buying everyone out people think they are reimbursed to much. This public perception has everything to do with the PPOs ability to negiotate a contract.
    Read up on any of the big hospital/PPO battles of the past 10 years, it all comes down to public perception.

  43. This discussion just shows how broken the system is and how futile the present attempts at “reform” will be. Re-arranging insurance does not reduce costs. In a government run and financed system with universal budgets there would be no wiggle room to cost shift. Prices would be controlled as they are in other countries.

  44. I’d say the same to you: If there is only one hospital in town how does any of your argument apply? Why is it cost-shifting and not just negotiating with the only payers that negotiate?
    Medicare is going to pay what Medicare is going to pay. The point of the argument is that the hospitals and payers are negotiating rates in the same way all purchasers and suppliers negotiate: by taking into account each agent’s position in the market and deciding who needs who more. What Medicare pays has nothing to do with it.
    Are you really saying that a hospital with a monopoly in a market or some other leverage (like Partners in Boston) would just ignore their clout in a world where Medicare paid 110% of costs? Why does Medicare underpayment suddenly turn hospitals into savvy negotiators and why would that savvy disappear once Medicare paid more? It doesn’t make sense. Just because the justification they’ve given you is Medicare payment cuts doesn’t mean it’s the truth.
    Only a foolish hospital executive would lower the price they charge you because someone else is paying them more than they used to.

  45. Interesting breakdown and a great start to understanding the effects of cost shifting.
    I wonder, however if you were to break the data down a little deeper and look at the ‘well financed’ hospitals vs the ‘efficient’ hospitals you would find different cost/pricing behavior. Another layer of interest would be to understand the patient population. Clearly a hospital with a small amount of care/caid billing would have less need to shift costs as compared to a hospital with a large amount of government billing.
    The dollars have to go somewhere. Unless a hospital’s revenue decreases a reduction in gov payments would have to result in a cost shift, correct?

  46. Scott,
    If I was a hospital with a “percent of billed charges” contract, I would increase my charge level as high as I could get away with, whether Medicare existed or not.
    That’s a stupid contract.

  47. One of the most egregious examples of cost shifting is forcing private plans to cover over age dependents to age 28-30 if they are disabled. Ohio and a number of other states have done this to reduce their Medicaid cost. Why a small group of citizens should be singled out to pick up the cost of these individuals that should be born by society as a whole. The fact they do this shows the hypocrisy of the single payor proponents. Both Medicare and Medicaid already dump risk they promised to cover onto other plans, this is somehow a failure of the private plans that can’t stop it?

  48. I review self insured Plan data for a living, analyzing health and pharmacy claims data for large employers. In my audits I routinely find “billed charges” from a hospital that are 1,000%+ above what Medicare reimburses for the same service. Many insurance companies only have a percentage off of billed charges contract with hospitals so the inflated “billed charges” definitely are part of the cost shifts to the private companies. I can tell you they don’t like it when I request a copy of their purchase order that would justify why an outpatient surgery has a $72,000 billed charge for surgical rods and is 1,500% above Medicare.

  49. “but it doesn’t show how the cost shift happens.”
    That is very simple, contract comes up for renewal, hospital says I want 20% increase, PPO signs contract, cost shift effected. I could use a bunch of big words and stretch it into 4-5 paragraphs but it really is that simple. The other way it increases is billed charges.
    Peer reviewed by who? Because a group of liberal academics get together and all agree on something that somehow makes it less partisan then a highly regarded actuarial firm releasing a study? Academics are meaningless, what they think and say has no effect on real world outcomes. Austin’s theorizing doesn’t help me when I sit down to redo contracts. Do you think a hospital administrator is going to care his rate increase request is contrary to Austin’s theory of economic pricing? I’ll try that argument next time and let you know how well it works.
    When a hospital is bleeding red ink and getting bailed out by the state its hard to argue their price demands are greedy and excessive. If a hospital is profitable and paying fat salaries the chance of getting the public to agree the hospital can afford to charge less is increased. The problem here is the disconnect between the way things actually work and how academics like Austin perceive them. If a hospital is charging me the health plan excessively the only recourse I have is to drop them from the plan, and even that only marginally works because many times they are the only hospital in town. My next level of recourse is draconian, I bar members from receiving any treatment but emergency care or design the plan so there are sever penalties for being treated there that go beyond the usual out of pocket caps. Either way I the insurance company or PPO is the bad guy for limiting employees’ choice. It becomes a PR battle, I have to link the hospital’s payment demands to the high cost of insurance and excessive spending by the hospital. They plead poverty and saying without the increase they will have to cut service and close their doors. There is nothing to guarantee if Medicare and Medicaid reimbursed fair rates it would all be passed on to private insurers, as long as they don’t pay fair rates we are sure additional cost will be though. No where in reality does any of Austin’s theories apply.
    The comparison is not between charges of different hospitals, it is a study of what a specific hospital charges different payors. Why should tax payors be charged 120-200% more by a tax supported hospital for the same care someone in a public plan gets? Why are public facilities allowed to accept different rates to start with? It is an undisputed fact most hospitals could not survive if all care was reimbursed at Medicaid and Medicare rates.
    Again hospital market share doesn’t have anything to do with cost. That is a basic fundamental aspect of our markets. As long as you continue to base your theories and arguments on this faulty assumption everything derived from them will be wrong. If there is only one hospital in town how does any of your argument apply?

  50. Those of us involved with employee benefits with gray hair have a rather more expansive view of what defines a cost shift than merely considering a direct correlation between medicare and private insurer balance of payments.
    For example, we recall when TEFRA and DEFRA made private insurers primary for employees formerly eligible as Medicare primary.That would be a cost shift right?
    And then there was COBRA. You can take the coverage with you and lucky you Mr Employer you get the claims charged against your loss experience or bank account if you are self insured, for former employees no less. That would be a cost shift right?
    Then there was ARRA, an estimated $65B cost shift when 65% of the premium was now subsidized by employers via a payroll tax credit. Since neither the additional claims cost for self insured plans or the additional premium charged at renewal for insured plans when COBRA elections increase 800% as they have in the manufacturing sector and claims accelerate are deductible from payroll taxes, that would be a cost shift, no?
    Speaking only for myself the points in the post above may be accurate. I am not an economist, nor did I stay at a Holiday Inn Express last night. I certainly do not dispute them My point is only that there are myriad examples of governments cost shifting which have contributed to out of control medical trend for several decades and contributed to the current predicament we face with regard to health care affordability. So when you see us brokers crapping on about cost shifting these are some specific examples.

  51. Nate, the Milliman study “quantifies” the cost shift but it doesn’t show how the cost shift happens. Also, it was bought and paid for by the Commercial insurance industry and was not peer reviewed. How is that relevant as anything but propaganda?
    Everyone knows that Commercial insurers pay more than Medicare and Medicaid, but how exactly does their study demonstrate that an explicit cost shift has taken place? How do we know that Commercial insurers would be able to pay less if Medicare paid more? Are you saying that if Medicare decided to pay at 110% of average costs, commercial insurers would also pay 110% of average costs? Wouldn’t hospitals just negotiate the same 120% of costs that they always have? What would prevent them from doing so?
    Is there evidence that hospitals with a higher case mix of Medicare and Medicaid charge Commercial insurers more than hospitals with a lower Caid/Care case mix? If not, why not? A cost shift mechanism would imply those hospitals would have more costs to shift and shift more to Commercial insurers.
    But I bet that’s not what you find. I bet hospitals that see a lot of Medicaid traffic have pretty poor negotiating power with Commercial insurers and thus charge them lower rates than hospitals that see mostly Commercial patients. Do you have any evidence to the contrary?

  52. Austin how can you even broach this subject and not mention the two studies from Will Fox and John Pickering of Milliman? It is far more relative then any study you cited, more current them most, and done by an organization far more respected. Except for the fact you don’t like their findings any debate on this subject has to start there.
    “However, it is a myth that providers can fully shift costs.”
    Interesting choice of words in that I have never seen a single person make that claim. No one says Medicare cuts $1 private pays $1 more. The argument being made is Medicare cuts $1 2-3 years later when contracts are up private pays $0.70 more. Was this poor word choice or did you mean to distort the argument in this way?
    “That they could do so violates, in most cases, principles of economics.”
    This is the problem with academics that don’t work in the real word. In your ivory tower principles matter, in the real world the piece of paper my contract is written on matters and it never heard of your principals let alone abide by them.
    “price discrimination can be due to unequal bargaining power across classes of purchasers.”
    You don’t seem to take into account when dealing with Medicare and Medicaid there is no such thing as bargaining power, rates are dictated to the hospital. Thus by definition there is no measure of bargaining power between public and private plans as there is nothing to measure.
    “It’s true that cost shifting could theoretically occur”
    When hundreds of hospital administrators, most PPOs and carriers, and studies like Milliman quantify it; aren’t we pass theoretically? My family owned a PPO for 10 years in NV, I know for a fact when public plans cut reimbursement our rates were raised. In other years we would get inflationary increases.
    “if one competitor elects to reduce costs or reduce its burden of undercompensated care, it might be able to charge private insurers less then others, thereby increasing its market share. So, cost shifting may not be a stable equilibrium.”
    You don’t seem to understand how private insurance works. These theories you are trying to apply don’t fit the insurance market. You seem to think the price at which a hospital gets reimbursed correlates into the volume of business they receive. The patient, who is almost completely insulated from hospital cost decides which hospital they go to, price is not a factor. A hospital with a terrible reputation could give away care and still not increase market share. Further most communities are served by only one hospital meaning there is zero price competition. On top of the single hospital communities the rest of the country is usually served by two competing systems that have acquired all the hospitals in the region. In very few places do you get a choice of 3-4 hospitals.
    If like you and MedPac claim there is no cost shifting why did Congress deem it necessary to repeal the portion of the original Medicare law requiring fair reimbursement to providers? MedPac is appointed by politicians and relies on the continuance of Medicare for its existence, if they came out and admitted Medicare and Medicaid drove up the cost of private insurance 15% there would be very real consequences.
    “The most recent studies of the phenomenon find little evidence of cost shifting or very low levels of it.”
    Ignoring the largest study on the matter and not done by someone advocating for reform.
    Economic theories are as bad at controlling cost as Washington Reform has been.

  53. Austin: I would submit that cost-shifting success depends on the extent of a provider’s negotiating leverage and the tolerance level of employer-purchasers for network disruption. Large, heavily branded provider systems in markets with low tolerance for network disruption are quite successful in obtaining dollar-for-dollar offset of Medicare and Medicaid losses. Unfortunately, commercial utilization is increasingly concentrating in such systems.

  54. @spike – You’ve summarized the theoretical argument well.
    @turntostoneblog – You don’t buy the theory. That’s fine. Yet as I pointed out in the piece there are cases where even theory suggests cost shifting can occur. And it does occur at a low level, as I acknowledge.
    However, dollar-for-dollar cost shifting or anything near it just doesn’t seem to happen much. Based on the evidence it seems that a better mental model is that there is little or no cost shifting. Can you point to a peer-reviewed body of recent work that suggests otherwise?

  55. It is well known that Medicare routinely pays signficantly less than cost to all providers, but especially to hospitals. Medicare gets away with this because, at the end of the day, they control a significant percentage of the market — about 50% in terms of dollars — and have the unilateral ability to set PRICES. Hospitals can, and do, make up some percentage of this underpayment through higher charges to the commercially insured and even higher prices to the uninsured who can pay. The economic arguments against this practice have limited merit in an environment where demnand is inelastic with respect to price.

  56. Thank you. I’ve been saying this for years, but I guess I’m not an actual economist so it’s not as impressive when I say it, hah. But having seen provider/contractor negotiations up close, Medicare lowering or raising their payments never came into the debate. A hospital’s clout and monopoly power came into those discussions way more frequently. If hospitals could “shift costs” at will, then why stop at only 20% higher than Medicare, why not go to 40%? Because hospitals can’t just cost shift at will. Except for Partners in Boston, apparently.