Tuesday, August 24, 2010

A Liberal Voucher Plan for Medicare

Cross-posted from Critical Condition on National Review Online.

In my previous post, I wrote that it would be nice to see liberals come up with constructive alternatives to Paul Ryan’s Road Map for Medicare reform. Fortunately, Austin Frakt has picked up that gauntlet, with six posts over the past three days on his own reform plan.

Frakt criticizes Medicare Advantage, the most market-oriented component of Medicare, because it has higher costs on average than traditional, government-run Medicare (14 percent higher, according to the latest figures). This development surprised most experts, who believed that privately administered plans would be more efficient than traditional Medicare. So why are Medicare Advantage plans more expensive? Frakt argues:
The answer is that health care cost control is tough, technically and politically. Provider groups typically resist it. When it pertains to Medicare, beneficiaries resist it too. By adding another private-sector layer to the program–health insurers–the Advantage program invites a third source of political pressure. Rent-seeking by providers and insurers, as well as the power of the beneficiary constituency, align in their encouragement of higher Advantage payments. Congress, apparently, is willing to yield to that encouragement.
In summary, Frakt is arguing that, for insurers, it’s easier to yield, and say “ok, we’ll pay for that benefit,” and pass the costs on in the form of higher premiums, than it is to actually clamp down on costs and get yelled at by politicians. He's right to point this out. But it is also true of government-run Medicare, so it doesn’t really explain why private insurance costs more. Rent-seeking insurers only account for a small fraction of the discrepancy.

The real reason that Medicare Advantage costs more is that the private insurers who participate in it are smaller, and less subsidized, than traditional Medicare. The government plan stacks the deck in its own favor.

Because traditional Medicare controls more than 35 million patients, the government has the market power to force hospitals and doctors to accept lower prices. Without those 35 million patients, the hospitals and doctors would lose too much business. Large private insurers under Medicare Advantage might control a few million patients each, and therefore don’t have that kind of influence. In fact, for the private insurers, market power goes in the other direction: hospitals tell insurers, “we’re going to jack up our rates, and either you pay them, or we won’t accept your patients.” Hospitals can’t do that as easily to the government. (Indeed, Frakt has written eloquently about this problem.)

Another key factor is that traditional Medicare benefits from its status as a government program. 3 percent of Medicare’s administrative costs are paid for by other government departments. In addition, private insurers are subject to taxation, whereas traditional Medicare is not. Between 8 and 13 percent of Medicare expenditures—$60 to $100 billion a year—go to pay fraudulent claims, something that doesn't show up in traditional Medicare's true costs. (One of the reasons private plans cost more is that they spend more to weed out fraud.)

Frakt’s own voucher proposal contains many promising ideas for Medicare reform, but it doesn’t address the problems above. Frakt advocates a pilot program in which competitive bidding is introduced to Medicare. This is an interesting idea, and Rep. Ryan would be well advised to consider it. His other suggestions, such as establishing a minimum benefit package, are within the realm of a reasonable debate.

But no pilot program is going to succeed, when traditional, government-run Medicare enjoys such significant structural advantages over the private sector. This is why the Bush-era reforms subsidized the extra 14 percent cost of Medicare Advantage, and why the private-sector Medicare market will wither now that Obamacare has taken those subsidies away.

Frakt looks into, but ultimately dismisses the experience of the popular, market-oriented Federal Employees Health Benefits Program (FEHBP), which the Clinton-era Bipartisan Medicare Commission saw as a model for a reformed Medicare program. “FEHBP’s experience does not count,” writes Frakt. “It’s apples to oranges,” because it is “largely untested for [the Medicare] market.” This isn't a persuasive argument: FEHBP plans do include retirees, who do just fine in the program. And any meaningful reform of Medicare, by definition, is “untested.”

I hope Frakt takes more time to learn about the FEHBP. He makes a disappointing comment, in reference to FEHBP disquisitions by the Commonwealth Fund, Heritage Foundation, and Wikipedia:
Ugh! This is no way to learn about a program. Can I have some analysis from a source that’s broadly viewed as unbiased please? How about CRS? (I searched, not finding something easily.) How about GAO? (I also looked.) How about the academic literature? (Yep, I checked.) I’m sure I am missing some good documents. Do you know of any?
Does Frakt really believe that government and academic research is “unbiased,” whereas think-tank research is not? Research, whatever its origin, is research. Either the data is accurate and the methodology sound, or not. Research from Harvard can be horrible, and research from conservative and liberal think tanks can be thorough.

At any rate, as has been pointed out elsewhere, Frakt (and anyone else interested in the subject) would benefit from reading Walton Francis’ 300-page book on Medicare and the FEHBP program, Putting Medicare Consumers in Charge: Lessons from the FEHBP. It’s a through and fair-minded examination of the program. Francis also wrote the Heritage report that Frakt considers biased, but that is manifestly unfair to Francis, who is one of the nation’s leading experts on FEHBP.

At the end of the day, this entire argument misses the point. The emphasis in Medicare reform, on both sides, has been on who the insurer should be: the government, or a private entity? But the reason that Medicare keeps costing more is because it heavily subsidizes the excessive consumption of health care, whoever does the actual insuring. As John Goodman writes in a perceptive post:
The fundamental problem in health care is not that we are using too much of one payment mechanism and too little of another. The problem is that the person who benefits from the service is not the same as the person who pays the bill.
If health care is nearly free to retirees, don’t be surprised that they use a lot of it. If you want to bring costs down, and encourage retirees to spend their health care dollars wisely, they need to have an actual stake in doing so. That stake comes from sharing more of the costs.


  1. Wow Avik! You make some great points. But I'm very distracted by your strong attack of Frakt. He seems to upset you tremendously. Why? He doesn't attack you this way. On the contrary, he seems to be very constructive and not personal at all. Doesn't he deserve the same in return?

  2. Hi Anonymous,

    Does it come across like an attack? That is certainly not how it is intended. Austin and I agree on more than we disagree. I spend a great deal of the piece saluting Austin for providing a constructive proposal, and acknowledging the ways in which Paul Ryan (and others) should take his points into account.

    I agree with much of his proposal. The main things I disagree with (and they are not minor things) are his diagnosis of the underperformance of Medicare Advantage, and his view that the work of people at think tanks is hopelessly biased vs. the work of people in academia and government.

    Austin knows that I make a genuine effort to represent his views fairly. If he feels I haven't done so in this case, I am more than happy to correct the piece accordingly.

  3. " Does he believe it is also ludicrous that people can shop meaningfully for cars, computers, and cell phones?"

    I certainly don't. Those are easy to shop for. I think the better analogy would be credit cards. Many lawyers with degrees in finance have written about how difficult it is to understand credit card fine print. I think the same holds true for health insurance. It is foten difficult to figure out what they really do or do not cover. My MBA office manager fields a lot of calls from patients and my staff, MDs with a few PhD/MDs thrown in. Are you better off with a higher deductible or co-pay. Should you want ER visits covered or primary care visits. Then, remember that we are dealing with an older group, one that is often not tech savvy enough to run to the net to look for info.

    More than half of my patients with artificial valves do not know what valve has been replaced (I have been keeping track.) Most people just arent that medically sophisticated.


  4. Separately, most states are dominated by one or two insurance companies. What percentage of the market are you thinking you need to control to get lower prices?


  5. One thing that is not highlighted in this discourse is that Medicare Advantage (and all the private Medicare predecessors) are voluntary, and that healthier than average persons have signed up historically, leaving the sickest in traditional Medicare. So, not only has private Medicare cost more, but has done so while caring for healthier patients. Even when Medicare HMO premiums were set at 95% of Avg adjusted per capita cost it still cost more than traditional Medicare would have for those patients because those signing up for private plans were healthier. A real test of competitive bidding (assuming the elusive level playing field were found) would be for all persons in a geographic area to be included in the bid, but of course it is politically impossible to force beneficiaries into private plans. Therefore, selection effect/cherry picking will remain an overriding issue in any imaginable private Medicare plan.

  6. Hi Don, that is an excellent point and one I should have raised in the piece -- thanks for bringing it up.

    Steve: the problem here is that they are competing against traditional Medicare. If they try to save money by cutting benefits or increasing cost-sharing, beneficiaries will go back to traditional Medicare. The phenomenon you're referring to is in the employer-sponsored market, where some of the same issues can arise if employers are insensitive to the costs of their health plans. That is starting to change.

    On the issue of sophisticated consumerism: how many people know how a car works? Or understand the technical features of cell phones? Almost none. They buy a specific product for a lot of subjective and objective reasons. Efficient markets are driven by the behavior of a minority of highly sophisticated consumers who drive incremental buying patterns.

  7. The so-called"cost shift" is another factor lurking in our current system.

    Hospitals and physicians can accommodate payments from Medicare/Medicaid that do not cover their costs, in part because they have been able to charge their insured patients more. (Not the uninsured of course -- whose costs hospitals and physicians also charge to insured patients).

    This tactic shifts costs away from Medicare/Medicaid and onto private payers. It makes private insurance more expensive than it would otherwise be and makes Medicare look better than it is.

    Milliman recently released a study that measures cost shifting at 11% of private insurance premiums. That is, government-imposed lower reimbursements drive private insurance costs higher, in fact requiring 11% higher premiums.

    This is not a recent phenomenon. Michael Crichton estimated the value of the cost shift at 14% in 1970, and research has produced similar conclusions during the intervening years.

    When private insurance goes away, hospitals and physicians will not be able to deal with government-imposed reimbursements by charging their privately-insured patients more.

    Will hospitals and physicians simply shrug and accept lesser incomes? Doubtful - if nothing is done to help them manage their costs. Has anything been done to help them manage their costs? Despite political rhetoric, there is no evidence of it. And that means the outcome of future "negotiations" will be very different with the likelihood that Medicare and Medicaid premiums will spike upward.

    I am now eligible for Medicare. "Original" Medicare is clearly a lousy insurance plan, it costs over $900 per month per person, and you have to buy supplemental insurance to protect yourself. The end of cost shifting means Medicare is gonna get a lot more expensive. I can scarcely contain my enthusiasm.

  8. Avik,

    As a reader of your blog and Frakt's I may be able to help. First your post includes a photo of him. That is a strong signal that it is about him as a person, not about his ideas. I grant that you discuss his ideas, but some of your phrasings seem designed to be personally insulting (I'll get to one of them below). My advice, ditch the photo.

    Suggesting that someone you're debating with go read a book (in this case Francis') is the height of snark. Frakt did request more reading on this subject, seeking something that was generally regarded as unbiased (what's wrong with that?). The polite way to make a suggestion would be right on his blog, not in a snide comment in your own post.

    Lastly, Frakt made the astute point that standardization increases competition. Product differentiation is source of market power and higher prices. That's true in any market, but particularly so in a concentrated one like health insurance. Part of harnessing the power of the market is also to recognize its limitations and establish conditions to minimize the distortion they would impose.

  9. Avik
    Having read above, I am a bit surprised. I don’t have the time to parse all your points (unfair, sorry), but I think you are a bit over the top in your criticism. Austin, while an academic, is hardly “ivory tower” in the pejorative sense in which I think you mean it. Looking at his evidence-based posts in total over a long period, I always get the sense data drives him, and when he states his opinions, he says so. Frankly, I thought is conclusions on FEHBP were on target, as I did my own independent searching out of self-interest (not reading any books on subject, I have job after all), and had difficulties finding objective summary reviews as well(although they are probably out there). In terms of think tanks, I think you are right, nothing is bias free, but please—left or right, these shops are the greater of the evils as compared to peer reviewed, “independent” studies or overviews. You wont change my mind on that count.

    I would also add that while MA plans might be at a disadvantage in some ways, you portray them as a bit too overly angelic for my tastes. No gamesmanship, shenanigans, crying wolf, etc., etc., from their end? I doubt it. I cant say how far the table is tilted between Feds and marketplace or in what direction for that matter, but boy, innocents they are probably not. No lobbying, rent seeking, cherry picking? There are many sides to that debate, and Austin has reviewed that literature over time—I think quite fairly.

    Anyway, you stake out some reasonable territory, but given the positive tone of your past back and forth’s, above seems a bit too doctrinaire to me—perhaps the same thing you are accusing Austin of. My two cents, and again, I acknowledge my post lacks details, citations, but just don’t have the time—it was just your approach/tone that prompted my response and bothered me. I just should add I have never spoken to or met Austin; I just find his posts interesting and informative.


    PS—I agree about the photo and thought the same thing. Nasty, unnecessary and lacks civility.

  10. I disagree with the view that posting a photo is nasty. It has been my custom to post a photo of a writer whose work I am commenting on -- I've done it with Paul Krugman, Ezra Klein, Malcolm Gladwell, Atul Gawande, etc. Posting those photos was not considered controversial at the time.

    And I want to reiterate what I wrote in my first comment -- the intent of the post was to praise Austin's contribution, while pointing out some areas of disagreement. I regret that I did so in a way that was interpreted differently than how I intended.

    I make a real effort to represent others' views fairly, as those who have interacted with me know. I will continue to do so.

  11. Avik,

    Coming back to see how it turned out ...

    Does it really matter what your custom is? If a post seems rude to some then that's a valid opinion, whether prior ones follow the same mold or not.

    Anyway, it appears as if you cleaned this one up. But most readers have read the original and, to the extent they trust your opinion, probably came away with a skewed view of Frakt's take on things.

    It's not so much about this post but about how you handle your commentary in the future. Clearly you know how to write in a more considerate manner (as you revised this one accordingly). I recommend you continue to do so, as it will reflect well on you.

    Treating others as you would wish to be treated is the Golden Rule, is it not?

  12. "Steve: the problem here is that they are competing against traditional Medicare. If they try to save money by cutting benefits or increasing cost-sharing, beneficiaries will go back to traditional Medicare."

    Why can't they offer the same benefits as traditional Medicare? Shouldnt they have the advantage of a more efficient private enterprise model? Do they effectively need monopsony power to get lower prices from providers?

    It has been a while since I looked, but IIRC, states with many insurance companies and states with just one or two dominant companies have little difference in costs.


  13. "IIRC, states with many insurance companies and states with just one or two dominant companies have little difference in costs."

    One might well wonder if the costs aren't coming from the insurance companies, then.

    It's true there's no longer that much difference in overall medical costs from state to state, nor in the annual rates of increase in medical costs.

    It's also true that insurance companies are most effective competing on administration - and least effective competing on medical cost.

    For group insurance (and roughly 90% of insured people are in groups) the admin cost is usually 15% of the total premium or less. The other 85% of the cost pays for medical care. Is that 85% subject to competition? The evidence is slim. In fact, the public demands more of this cost be subsidized. More subsidization reduces incentives for hospitals or medical professions to find ways to reduce their costs.

    I think it's a mistake to believe that the cost of insurance can be controlled without first controlling the cost of medical care. I think it's a mistake to believe that medical costs can be controlled by arbitrary reductions to providers' revenues. I think subsidizing insurance is running thru Hell in gasoline pants.

    I would be somewhat encouraged if there were more reason to believe that the health care reform legislation were really about helping hospitals and physicians to reduce their costs of delivering medical care. Instead the focus seems clearly on subsidizing insurance, and increasing tax and regulatory burdens to pay for it. I don't see it reducing medical costs - it's more likely to increase them.

  14. Hi Philip,

    First off, thanks for your constructive criticism.

    I edited the piece because its intent was to praise Austin's contribution, and that was getting lost in all of the objections to my tone. I strive to have a civil dialogue with everyone, Austin especially, and if things I say unintentionally detract from that goal, I will certainly take note.

    Having said that, I think what we're running into in large part are key differences in the orientations of liberal vs. conservative health care people.

    The view that academics and government officials are unbiased, whereas think-tankers (or industry analysts) are biased, is a widely held view in left-of-center circles. It is reflected when people like President Obama say "our health care bill has taken in all of the cost-control ideas from all of the top experts." But as I wrote in the piece, this is not an empirically rigorous position.

    Jonathan Gruber is about as biased as it gets -- he has a personal stake in ensuring that PPACA and RomneyCare are seen as good things, because he was their intellectual architect. Whether he works at MIT or the Commonwealth Fund doesn't have any impact on his bias. His bias is his.

    Remember also that, in academic health policy circles, 90+% of faculty members are liberal to left-wing. So the peer review process has a lot of group-think. Group-think is inherently unrigorous, because it fails to take into account alternative explanations for complex phenomena.

    To summarily dismiss the work of a careful and rigorous thinker like Walton Francis, simply because he wrote a paper for the Heritage Foundation, is (1) prejudicial and unfair to Francis' intellectual integrity; (2) empirically unrigorous.

    Then, on the issue of consumerism: the view that individual consumers are not sophisticated enough to manage their own affairs is a widely, if not universally held, position for health policy people who are left-of-center. There are a couple of problems with this point of view: (1) it betrays a faulty understanding of how markets work (i.e., markets are efficient because a tiny minority of people who are highly sophisticated consumers drive incremental purchasing); (2) it is condescending towards those who value their individual liberty (since the natural logic of saying that people are unsophisticated consumers is that the government needs to step in to protect people from their ignorance).

    Because most of you who are left-of-center take the above views for granted, I think it's especially jarring when someone objects to them. That doesn't mean the objections will cease.

  15. "There are a couple of problems with this point of view: (1) it betrays a faulty understanding of how markets work (i.e., markets are efficient because a tiny minority of people who are highly sophisticated consumers drive incremental purchasing); (2) it is condescending towards those who value their individual liberty (since the natural logic of saying that people are unsophisticated consumers is that the government needs to step in to protect people from their ignorance)."

    This is a POV held by most of the docs I know. In my very conservative area, where people believe every word uttered by every talk show host, even my anarcho-capitalist fellow physicians know what it is like to dal with real live patients. When it comes time to make major medical decisions, most are vulnerable. Most do not understand that much about medical care. Most are pretty fuzzy on their own health history. It's just the way people are. When I took the CEO of a local Fortune 500 company back for his CABG, (MIT engineer if IIRC), he was terrified, and a crappy medical historian to boot.

    When it comes to the details of health insurance policies, I have been observing the problems for years as the president of our corporation. Few people know the details of what is in our health plan. My wife, first in her medical school class, has trouble figuring out exactly what is covered and not covered. Having sat on multiple hospital committees, I have had the opportunity to ask the HR folks what gives them the most trouble. Health care. People are always screwing up on what is and is not covered by their insurance.

    Do you know what is covered in your own plan? Without asking your wife?


  16. Hi Steve,

    How regularly do you read your credit card agreements? Do you know how an internal combustion engine works? How about the mechanisms behind air traffic control, or aerodynamics? I could go on.

    Contra Ken Arrow, asymmetrical information is not unique to medicine; indeed, it is nearly ubiquitous. Indeed, nearly every transaction involves asymmetrical information (i.e., the buyer knows more about what he is selling than the seller does). We rely on experts to guide us through lots of decisions in lots of different areas of the economy, and those areas of the economy have managed to function without government nationalization or micromanagement.

    The main reason why people don't know what's covered by their health plans, btw, is that they haven't chosen them for themselves. If your employer (or the govt) hands you a plan, with no decision-making on your part, you're far more likely to be ignorant of its features. If you have an individual market for health insurance, a la Switzerland, people are more aware of what they're buying (because they actually buy it for themselves).

  17. Sorry -- in that last comment I meant to say "the seller knows more about what he is selling than the buyer does."