The Incidental Economist

Medicare Advantage plans consume more taxpayer money per beneficiary than FFS. This is a fact.

That’s not to say MA plans can’t achieve lower costs, only that taxpayers aren’t getting the benefit if and when they do. Beneficiaries do derive some benefit from the over-payments, but only at the rate of 14 cents on the dollar (that does not mean the other 86 cents is plan profit, more here). It is also true that there are some spillover effects. FFS costs may be lower when MA enrollment is higher, but to my knowledge, this has not been quantified with data more recent than 2001. A lot has changed in the MA program since then. The size of the spillover today is uncertain, as far as I know (happy to be corrected on this).

But back to taxpayer costs. Below is a figure I’ve posted before that illustrates the relationship between MA payments and FFS costs. (Actually, it shows “benchmarks,” not payments, because plan-level payments aren’t available to me as a researcher. But we know from MedPAC that payments are only a few percentage points below benchmarks on average, and still well above FFS costs.)

In this figure, each circle is a county and its size is proportional to the county’s MA enrollment. Circles centered above the 45-degree line correspond to payments (benchmarks) above FFS costs. All the circle centers are above the line. Additional details are in a prior post.

Given the fact that MA plans are paid above FFS costs, you’d think folks in favor of harnessing the efficiency of the market to achieve taxpayer savings would be appalled. I am. This has got to stop. And that does not mean “eliminate private plans.” It means “stop overpaying them.” There are a lot of ways that could be achieved. I prefer a competitive bidding approach that includes FFS among the bidders. Details next week.

In a Kaiser Health News column today, James Capretta misunderstands my prior critique of Medicare Advantage (MA), a critique that I believe applies to Rep. Paul Ryan’s voucher plan (which his own staff has admitted to me). Capretta wrote,

Critics say Medicare Advantage plans — the private insurance options offered to beneficiaries — are inefficient and costly. But those same critics oppose vouchers for Medicare — even though that would set up a direct competition between the private plans and the traditional fee-for-service program. …

After all, if the case critics (see Austin Frakt’s August 19 KHN column) make is correct and private insurers simply can’t do the hard work of cost control as well as the government, then Medicare’s “public option” would presumably win this contest.

Kinda makes it sound like I’m one of those critics, that I think private insurers are incapable of reducing costs. Where did I write that? What I really wrote was this,

The politics of Medicare are such that Ryan’s idea, paying for care entirely through private plans, costs more. That’s not due to a market failure, but a political one. Congress likes to spend money; insurers, providers and beneficiaries like to receive it. Congress spends even more when it can satisfy those interests under the guise of a seemingly pro-market, pro-competitive program.

MA plans have a political problem. They are paid above FFS cost whether they can achieve below-FFS cost or not. The same political problem would befall Rep. Ryan’s voucher program because under his plan the HHS secretary has the discretion to set voucher levels. We can do better than that.

In a post on August 23 I followed up on this idea and described a voucher plan that does not suffer this problem. If an independent body, shielded from politics, is in control of voucher levels and other conditions described in that post are met, I would support a voucher plan. Moreover, I am in favor of a competitive bidding system (I’ve written about it in many posts going back a year or more) with a level playing field, as described in my August 24 post. I’ll describe it more fully next week.

Capretta clearly didn’t read all my writing on this topic. He just presumed I was anti-private plan. I never wrote that. I hope he’ll read my post next week on competitive bidding and be in touch if there is any uncertainty about my position.

I’m always surprised at which posts draw heat and how they do so.  I’ve been getting a lot of email about this post.  Most of it is in the form of gloating, as if I’ve “lost” because I showed that the government is rationing.

Um…  no.

First of all, please get over the idea that this is a game one side wins or loses.  This is much too serious.

Second, please get over the idea that I’m on some imagined side.  I’m not a partisan, nor do I believe “all government is good” as some of you seem to think. I will call out government for things I think are bad just as I will call out the private sector.  We are all part of a big health care system that I think is terribly flawed.  When someone says that we don’t need health care reform, they believe it’s all good.  I don’t.

Third, as I told a commenter (perhaps over-excitedly) if you think this sort of rationing isn’t happening in private insurance as well, I’ve got a bridge to sell you. Our whole system rations by ability to pay. We hear about it more from government because that information is public. Private insurance more often gets to keep its books closed. That doesn’t make it better.

Many of you are using this example to say that government should get out of health care.  They are rationing.  They are the ones denying you care.  They should get out.

Really?  We’re talking about the permanently disabled here.  These people are no longer gambles; they are absolutely, positively going to cost a lot of money.  What free market private insurance solutions do you propose for them?  Which private insurance companies are going to compete for their business?  No one wants to insure a guarranteed loss.  It has to be government.

Do you have another solution?

Every once in a while, when people let their guard down, the truth slips free.  Most of the time they are able to cage it, to hold it prisoner.  But not always.

Kaiser Health News had a good piece up on Friday about the waiting period for people with full disabilities who qualify for Medicare:

Under federal rules, most people with disabilities who are younger than 65 aren’t eligible for Medicare until more than two years after they qualify for Social Security disability income. A coalition of more than 65 organizations led by the Medicare Rights Center has been pushing Congress to do away with the waiting period. But the effort has stalled because of the high cost to the federal government – an estimated $113 billion over 10 years, according to the Congressional Budget Office. That takes into account a $32 billion reduction in federal spending on Medicaid, the state-federal program for the poor and the disabled. Many people with disabilities go on Medicaid while they wait to become eligible for Medicare.

Some will tell you that the two year waiting period is to make sure people are really disabled before they are let into Medicare.  It’s to catch the people who are otherwise committing fraud.  Unfortunately, these hypothetical freeloaders aren’t nearly as common as real life situations like these:

After Russ Hillard developed Huntington’s disease, a devastating neurological disorder, he lost his $35,000-a-year job as a welder and, with it, his health insurance.

His wife, who was working part time, had insurance, but it didn’t come close to covering the medical bills for the incurable disease, which causes uncontrolled movements, emotional problems and the loss of cognitive abilities. Eventually, Hillard qualified for Medicare, which covers disabled people under 65 after a two-year waiting period. But the coverage didn’t kick in until after the family went deeply into debt and had to take out a $20,000 loan on their home in Methuen, N.H. The delay was a “cocktail for disaster,” says Hillard’s daughter, Laura Quinn, who is 29.

So why don’t we do away with this waiting period?  Why don’t we let people who are disabled get the care they need right now?  We’re not talking about the worried well here.  We’re talking about people with permanent life-altering disabilies.

KHN asked Joseph Antos, who is a health care policy scholar at the American Enterprise Institute.  Here’s what he said:

“Across the board eliminating the two years just doesn’t seem practical… This really is a money issue.”

I give Mr. Antos credit.  No hemming and hawing.  He just said it.  It’s the money.

Remember that the next time someone tries to scare you away from universal health care systems in other countries with specters of rationing.  They will tell you that in this country, we don’t ration.  They will tell you that in this country no one waits for care care because they are too old or because they are deemed not worth it.  But they are wrong.  Mr. Antos just told you so.

In this country we make people wait for care all the time.  We have waiting lists because “it’s a money issue”.  We ration by ability to pay.

And that may be the most indefensible kind of rationing of all.

Reader Ryan Stevens encouraged me to take a closer look at what the ACA legislation actually says about ACOs. Fortunately, the job is made dramatically easier by HealthReformGPS. In two separate posts the amazingly thorough staff of that blog cover the nitty-gritty on ACO provisions as they pertain to Medicare and Medicaid.

I encourage interested readers to check out the posts themselves. I’m not going to attempt to summarize them because, frankly, it’s too boring. However, in the very lengthy section on open questions, the following part jumped out at me:

How will federal antitrust enforcers view the establishment of ACOs? Will ACOs be insulated from potential antitrust claims to the extent that the ACO providers collectively negotiate payments with private third-party payers outside of Medicare? Will ACO certification include a determination that ACO are “clinically integrated” and thus fall within the federal antitrust exception? Will the federal government create an express safe-harbor from antitrust scrutiny for ACO activities under certain conditions?

I’d love to know the answers to those questions.

Reihan Salam injected the notion of a “level playing field” between FFS Medicare and Medicare Advantage in any competitive bidding payment system. About that, Bob Coulam, coauthor of Competitive Pricing for All Medicare Health Plans, e-mailed the following.

This NR blog reflects a rather constant complaint of those who favor private options in Medicare.  They claim that the terms available for plans can’t be level, in that the federal government is incapable of administering a level playing field and instead will favor its own FFS plan in a competitive bidding system. Their arguments would have the effect of shielding private health plans from competitive pricing.

Bob Berenson, in Medicare Disadvantaged and the Search for the Elusive “Level Playing Field,” notes the importance of having a consensus on what constitutes a level playing field–and an agreement that there should be a level playing field–both of which have been lacking.  He argues that MedPAC’s method of leveling [setting payments to FFS costs]–which is now law [in some approximation] and will in due course take effect–is too crude.  His proposal is to use administrative data to adjust FFS payment levels to reflect plan costs and other concerns.  As my co-authors and I said in our recent book, Berenson doesn’t construe a “level playing field” in nearly so simple a way as does MedPAC, and he urges a conversation on what, politically, we actually want to level.

Agreed! There should be more discussion of this, with the aim of a pilot or demo.* I would not welcome ripping out Medicare as we know it to replace it with something that is largely untested for that market. For this reason, FEHBP’s experience does not count. It’s apples to oranges.

Also, keep in mind how this discussion started, with my observations on Rep. Ryan’s Medicare voucher plan, which does not call for competitive bidding. Nor does it call for the establishment of standardized plans upon which to focus bids. Thus, already all participants in this discussion have gone well beyond what Rep. Ryan has proposed to something that has a far better chance of meeting the goal of efficient use of taxpayer funds. It seems, then, there is broad agreement that Rep. Ryan’s formulation could use some improvement.

*Later: In their AEI paper, Coulam, Feldman, and Dowd make a strong case that a competitive pricing demonstration is not necessary–that Medicare should move straight away to a competitive bidding system that includes FFS among the bidders. Since I respect the authors (at one time or another I’ve worked with each of them), I take their argument seriously. Perhaps a demo of the type of system they describe is overly cautious. I would retain my preference for a demo for any idea that deviates considerably from what the authors describe and simulate. Rep. Ryan’s proposal would be one such idea. The only other reason to consider a demo (or a pilot) is politics. Skeptics may not accept that competitive bidding can save money. But they may accept testing the idea with broad implementation contingent on proven results.

Reihan Salam wonders,

The conditions Austin outlines for the kind of Medicare voucher system he would endorse are worth reading. I am, as regular readers know, a great believer in experimentation. I wonder, however, if Austin is proposing a true level playing field behind Medicare Advantage and Medicare FFS.

It matters a great deal what this “true level playing field” is. If Reihan points me to something specific, I’ll look. In the meantime, it is worth remembering that the key is to reduce Medicare costs below trend (well below). If a level playing field achieves that and all the other conditions I outlined are satisfied, I’d be very happy.

Also, note that I wanted an IPAB-like entity to govern how plan bidding would work. To the extent the playing field is leveled and remains so, it would be done by them. Finally, the whole arrangement has to be proven effective in a pilot. It’s about results, not theory.

As I recall, getting past that first step–a bidding demo or pilot–has been a pretty tough sell. Why don’t we work on that? A Medicare Competitive Pricing Demonstration anyone?

The following is from a 2007 paper by John Cawely and Andrew Whitford that appeared in the Journal of Health Politics, Policy and Law.

The Federal Employees Health Benefit Plan is administered by the OPM and offers health insurance coverage to federal employees and their dependents; about 9 million people are enrolled in 350 different plans. A broad difference between FEHB and Medicare is that the former considers applications only from comprehensive, prepaid medical plans (HMOs, PPOs, and point-of-service plans). These health benefits carriers must offer a complete line of medical services, including doctor’s office visits, hospitalization, emergency care, prescription drug coverage, and treatment of mental conditions and substance abuse. The FEHB does not contract with companies offering only services such as dental/vision plans, prescription drug plans, supplemental insurance, and disability insurance. It also does not write contracts with FFS carriers.

A second important difference is that the FEHB lists broad types of benefits that contracting plans may cover — not the mandatory set of benefits and services required in MA. Beneficiaries may reduce the package of services they receive and thereby reduce their exposure to premiums and cost sharing. OPM exercises discretion in negotiating with plans over benefits packages; these negotiations also impact beneficiaries, because they can potentially reallocate costs between the program and subscribers over time. There are two broad types of plans, nationally rated and locally rated, which differ in how premiums are set to vary across local markets. Plans that participate are selected by fairly basic criteria: meeting a set of qualifications and offering at least a minimum benefit package. OPM negotiates benefits and rates annually, but it does not solicit competitive bids. However, competition by multiple plans in the same area may reduce costs to beneficiaries (Feldman, Thorpe, and Gray 2002). In FEHB, the federal government pays 75 percent of the costs of the plan up to a prespecified national ceiling.

The following citations from the paper may be of use.

Caplan, Craig F., and Lisa A. Foley. 2000. Structuring Health Care Benefits: A Comparison of Medicare and the FEHBP. American Association of Retired Persons (AARP) Policy Institute Research Report No. 2000-05, May.

Feldman, Roger, Kenneth E. Thorpe, and Bradley Gray. 2002. Policy Watch: The Federal Employees Health Benefits Plan. Journal of Economic Perspectives 16 (2): 207 – 217.

I don’t know enough as I’d like about the Federal Employees Health Benefits Program (FEHBP). In particular, I’d like to learn more about how it is actually governed and run, in contrast to Medicare Advantage, and what the results are for costs.

I’ve just spent a half hour poking around and this is what I’ve found so far:

A questionable Wikipedia entry. Read it. It is very pro-FEHBP and anti-Medicare. It is misleading about Medicare in at least one place, saying there is no bargaining with providers by original Medicare. That’s true but ignores Medicare Advantage, which do bargain with providers (excluding PFFS plans) . I’m not saying the entry is factually wrong (I admit I don’t know enough to judge). I’m just saying it seems very one-sided.

Some info from Commonwealth. Here are a few bits from a longer Commonwealth paper on FEHBP:

Since 1969, FEHBP has experienced slightly lower premium growth per enrollee than private health insurance overall (10.6% vs. 11.0%) but higher growth than Medicare (8.9%). …  In the last four years, Medicare has outperformed FEHBP by a far greater margin, with premiums growing at only about one-third the FEHBP rate. …

[T]he savings FEHBP achieves through competition have not been greater than savings from Medicare’s system. …

Although FEHBP has operated with little government intervention, its spending is only about one-tenth that of the Medicare program. It seems unlikely that government policymakers will remain immune from plan and provider pressure if the stakes are raised to Medicare’s level.

Based on data comparisons over 28 years, the FEHBP ties Medicare in cost control without regard to benefit changes over time. Taking into consideration these benefit adjustments, FEHBP costs, as with private-sector insurance generally, have increased less than Medicare costs over most or all of the life of the Medicare program.

I’m sure I am missing some good documents on this topic. Do you know of any?

Later: This post has been edited to reflect corrections suggested in the comments.

I’ve narrowly critiqued Representative Paul Ryan’s Medicare voucher plan on the grounds that it is too similar to an expansion of the Medicare Advantage (MA) program, which itself is more costly per beneficiary than traditional fee-for-service (FFS) Medicare. The basis for my understanding of Rep. Ryan’s plan is his own staff, with whom I’ve spoken. The basis for my understanding of MA is my own career studying and publishing on it. So, I’m quite confident in my ability to accurately compare the two and draw policy conclusions.

However, I like to be constructive, and there is a Medicare voucher program I would support, but it must include at least all the elements I list below. Rep. Ryan’s does not. I would, by the way, also support expansion of FFS Medicare under certain conditions as well. How can I simultaneously support annihilating FFS Medicare and expanding it? It’s easy. I support whatever demonstrably works. Read the rest and you’ll see what I mean.

I would support a Medicare voucher proposal that included at least the following provisions:

• Begins with a multi-year MA competitive bidding pilot. If the pilot reveals that MA plans competing on price can reduce risk-adjusted per beneficiary costs to a level below that of FFS with no degradation in health outcomes then I’m interested in a larger competitive bidding based voucher program, otherwise I am not. The pilot should include enough geographic diversity to reveal variations across market types. It is conceivable that vouchers work well in some markets and not others, which itself suggests that a FFS option might need to exist, at least where MA can’t beat it. (Implementation of MA bidding is, itself, politically challenging. Rep. Ryan’s plan calls for MA competitive bidding during a transition period. But the ultimate voucher program under his plan would not include competitive bidding.)
• Isolates the voucher program from political meddling. Under MA, Congress establishes plan payment rules. I’ve already explained why that’s a problem. Under Rep. Ryan’s plan the HHS secretary would have the discretion to adjust the size of vouchers. That’s not any better. My view is that the details of how plan subsidies (voucher levels) are set should be controlled by an independent board, like the Independent Payment Advisory Board (IPAB) that will be established under the ACA. Until politics are more fully removed from the setting of subsidy levels I will not support further expansions of MA. Also, as mentioned, voucher levels should be based on competitive bids. So, the IPAB-like entity would just control the bidding rules and procedures.
• Includes income-related subsidies. Low income individuals should not face the same premium and cost sharing burden as those who are better off. (It is my understanding that Rep. Ryan’s plan has some accommodation for income, as does the current MA program.)
• Establishes a minimum benefit standard or a set of standardized plans. The whole point of a market-based system is to harness the power of consumer choice. But consumers can’t send meaningful signals if the market has an incomprehensible structure. One of the conditions for a competitive market is fully informed participants. The notion that seniors–or anyone–can meaningfully shop in a market with an unlimited number of plans that vary in all possible ways is ludicrous. (There is already evidence that beneficiaries don’t optimally select among the scores of Part D plans available now and that reducing the number of available plans would increase welfare.) The Medicare supplement (Medigap) market is a good model of competition within standardization. Making products more similar encourages competition. Allowing them to vary along a small number of dimensions helps consumers make sensible comparisons consistent with individual preference. Isn’t that the point? (Rep. Ryan’s proposal doesn’t call for the establishment of standardized plans.)
• A fall-back mechanism for highly concentrated markets or ones with poor quality. It is possible that too few firms will participate in some markets, raising premiums well above competitive levels or reducing quality. Some markets could be dominated by a small number of hospitals, driving hospital service prices and/or volume up, which also translate into higher premiums. The plan has to have some way of dealing with such cases. One idea is a fall-back federal option, triggered by some measure of market concentration or beneficiary harm from insufficient competition or low quality. (Rep. Ryan’s plan is silent on this issue, as far as I know.)
• Access to utilization data for research. MA plans do not provide the same level of utilization data that is available for FFS Medicare. This limits our ability to understand how private plans work, what they provide, and what the consequences are for health outcomes. If taxpayers are funding them, or successor voucher-based plans, all utilization data paid for with such funds should be available for research. This would allow qualified research organizations to independently verify the merits of or discover the problems with a market-based insurance system for Medicare beneficiaries. If we’re to move away from FFS Medicare and lose the research data associated with it, I absolutely insist on a greater level of access to private plan data than exists today. This is a serious issue that gets very little attention, despite my efforts.

That’s a minimum set of requirements. I may think of more in the future, as will others with whom I may agree. Also, for me to take any voucher plan seriously it has to be clearly politically viable. I think it’s a pretty tough sell to Medicare beneficiaries. Though MA plans are available everywhere, about three-quarters of beneficiaries opt for FFS Medicare. Thus, an overwhelming majority are voting with their feet for the cheaper option. We taxpayers should be happy about that! Maybe we should just expand FFS Medicare, getting rid of MA altogether? The evidence on cost and beneficiary satisfaction supports it.

Of course FFS Medicare has a cost problem too, though it’s less severe than MA. (Yes, MA plans offer more benefits, but taxpayers still pay for them and beneficiaries don’t value them at anything close to their cost. That’s not a good argument for expanding the program.)

My preference for FFS will strengthen if the provisions of ACA designed to lower FFS costs actually work (ACOs, bundled payments, IPAB, etc.). So, this is all about evidence. I say, do an MA competitive bidding pilot, implement the ACOs, payment bundling, and the like. Let’s see what works. Today, FFS is winning, but it will still bankrupt the country. That’s a Pyrrhic victory in which I have little interest.

PS. One could make an argument that the current arrangement, with MA and FFS Medicare coexisting, is optimal. Maybe the argument would hinge on countervailing forces of rent seeking or on the fact that MA offers models of innovation while FFS Medicare serves to mitigate market power in concentrated insurance and hospital industries. The current arrangement also appears to be a political equilibrium. I’m not making this argument here because I haven’t fully fixed it in my mind. Maybe another time.