A few readers have written me or asked in the comments why I am not so crazy about HSAs.  From the past, read here and here, and here, or here is an index of previous MR posts on the topic; in any case my take is relatively straightforward:

1. I favor tax-free savings (albeit with some fiscal qualifications), so you can make a case for HSAs on this ground, noting that we do already have other tax-free savings vehicles.

2. HSAs take one market segment — usually a relatively wealthy and health care-satisfied segment — and introduce one marginal improvement of incentives.  This doesn’t seem to help much in terms of lowering aggregate costs.

3. HSAs introduce greater care into any single medical expenditure by creating a direct private opportunity cost for the spender.  I am less sure it will limit medical expenditures in general; that depends on how people frame withdrawals, once funds are committed to an HSA account, and to what extent they use HSAs for what would have been cash payments anyway.

4. As Paul Krugman says, "too much health insurance" is not the fundamental problem in the health care market.  (Unlike Krugman, I don’t see single-payer plans as the solution; I see the incentives of producers, combined with the fear and unreasonableness of buyers, as the key problem on the cost side.)

5. Re Bryan Caplan on Singapore, HSAs might work much better in another setting, noting that the other features of Singapore also might account the difference in performance in health care systems. 

6. Given #1 and #2, it is easy for me to believe that HSAs bring net social benefit.  It is much harder for me to see HSAs as "the one health care idea we would promote if we had one shot at health care reform."  The main beneficiaries are the healthy and the wealthy, and, while I am all for helping those people, surely that is odd, no?

7. I will profess my agnosticism on many health care policy issues, but one of the better plans is Jason Furman’s and/or spending more on medical R&D and some public health programs and lots of cost-lowering deregulation while in the meantime getting expenditures and costs under control.  I also recommend Arnold Kling’s work.

Last week Congress cut benefits to Medicare recipients and liberal pundits applauded.  Indeed, Paul Krugman said this was "Kennedy’s Big Day" and "the first major health care victory that Democrats have won in a long time."  Of course, Krugman and the others who applauded this "victory" didn’t say that they were cutting Medicare benefits – even though that is exactly what they were doing – instead they framed the victory as one over privatization and waste.  Here’s the story.

Medicare beneficiaries can enroll in Medicare’s fee for service plan or they can choose Medicare Advantage joining, for example, an HMO.  In the latter case, Medicare pays the HMO a rate per enrollee and the HMO competes to obtain enrollees by offering them a package of benefits and premiums. 

Now what you will be told about Medicare Advantage is that it is more expensive than traditional Medicare.  Thus the CommonWealth Fund says:

Private Medicare Advantage (MA) plans were paid an
average 12.4% more per enrollee in 2005 compared with what the same
enrollees would have cost in the traditional Medicare fee-for-service
program…

That much is true.  But why are MA programs more expensive?  The answer, which one gets by innuendo and implication, is that Medicare Advantage programs are wasteful and the extra money is being pocketed by corporations.

The CommonWealth Fund says:

"…eliminating extra payments to private plans could save Medicare a projected $30 billion over five years." (italics added)

Paul Krugman says:

the fastest-growing type of Medicare Advantage plan, private
fee-for-service, costs taxpayers 17 percent more per beneficiary than
Medicare without the middleman.  (italics added).

Robert Waldmann is least careful and in a comment on Tyler’s article on means testing says

Cowen doubts that expanding the public share of health insurance would
reduce costs. We have a test case medicare vs medicare advantage
accounts. They cost, on average 12% more per patient…

Thus the message is that traditional Medicare is cheaper because it eliminates the middleman, doesn’t involve private corporations, and is more efficient at lowering costs.  None of this is true.

I’ll give you the full story in a minute but let me first point to one clue that something is amiss.  According to all of the above "enrollment in these plans has been growing rapidly" (Krugman).  Now why would so many Medicare beneficiaries opt out of low-cost, efficient Medicare and into high-cost, inefficient MA plans?   

While you puzzle over the clue let’s cover the necessary background.  Here is how the MA program pays a private provider (quoting the CBO).

Private plans that want to participate in the Medicare Advantage program must submit bids indicating the per capita payment for which they are willing to provide Medicare’s Part A (Hospital Insurance) and Part B (Supplementary Medical Insurance) benefits–and to take on the financial risk of doing so.

The government compares those bids with county level benchmarks that are determined in advance through statutory rules. The benchmarks are the maximum payment the government will make for enrollees in private plans; in most cases the plans’ bids (and the resulting payments) are lower than the benchmarks….

If a plan’s bid is less than the benchmark, Medicare pays the plan its bid plus 75 percent of the amount by which the benchmark exceeds the bid.

So far you might think that Krugman et al. have a point.  If the benchmarks are set too high and Medicare pays the plan its bid plus 75% of the amount by which the benchmark exceeds the bid then the plans could bid their costs and get extra payments.  Now, I hope that many of you are thinking, What about competition?  Good thinking!  Indeed, if that was all there was to it, competition would push the bids below costs.  But in fact to resolve our puzzle we need not rely on competition and economic theory because here is the kicker (quoting the CBO again, italics added):

If a plan’s bid is less than the benchmark, Medicare pays the plan its bid plus 75 percent of the amount by which the benchmark exceeds the bid. Such a plan must return that 75 percent to beneficiaries as additional benefits or as a rebate of their Part B or Part D premiums.

Now the solution to our puzzle becomes clear.  Why do beneficiaries choose MA plans? 

…because such plans provide additional benefits beyond those available within traditional Medicare, including coverage for services not covered by FFS Medicare (for instance, dental services) and cash rebates of premiums or reduced cost-sharing.

In fact, the CBO estimates that the vast bulk of the increased payments to private providers flow to enrollees who get better benefits and lower payments.  Indeed, in the case of HMOs enrollees benefit twice – first because the benchmarks are higher and second because, contra Krugman et al., the HMOs actually have lower costs than traditional Medicare!  Thus the CBO writes:

In contrast, payments to HMOs averaged 10 percent above FFS costs…On average, HMOs offered extra benefits and rebates equal to 13 percent of FFS costs; those additional benefits and rebates reflected the difference between the benchmark (which averaged 10 percent above FFS costs) and the plans’ bids (which averaged 3 percent below FFS costs).

That could be written more clearly but what they are saying is that Medicare pays HMOs 10 percent more than they would pay for an enrollee in traditional Medicare but the HMOs offer the enrollee 13 percent more worth of extra benefits and rebates.  In other words, the HMOs pass on to the enrollee all of Medicare’s "extra payments" plus some.  (Note that this is exactly what one would expect in a competitive market.)

Now, I am not saying that higher Medicare payments are a good idea. But I dislike the fact that politicians are being lauded for fighting "wasteful privatization" when what they are really doing is cutting medical benefits for the elderly.   

Let’s first quote Mark Thoma’s response to my column; it is indirectly a good summary of what I argue:

I believe the political argument that giving everyone a stake in the
program helps to preserve it has more validity than Tyler does, market
failures (some of which hit all income groups) probably play a larger
role in my thinking about government responses to the health care
problem than in his, and I have more confidence than Tyler that a
universal care system has the potential to lower costs.

And now here’s me:

…the idea of cutting some government transfers provokes protest in
some quarters. One major criticism is that programs for the poor alone
will not be well financed because poor people do not have much political
power. Thus, this idea goes, we should try to make transfer programs as
comprehensive as possible, so that every voter has a stake in the
program and will support more spending.

But even if this argument
holds true now, it may not be very persuasive when Medicare costs start
to push taxation levels above 50 percent. A more modest program, more
directly aimed at those who need it, might prove more sustainable in
the longer run.

Americans have supported the growth of many
programs aimed mainly at the poor. Both Medicaid and the Earned Income
Tax Credit have grown rapidly in size since their inception. The idea
of helping the poor and not having the government take over entire
economic sectors was the original motive behind welfare programs, in
any case.

Furthermore, the argument for comprehensive and
universal transfer programs does not meet the ideal of democratic
transparency. If taking care of the poor is the real value in welfare
programs, those programs should be sold as such to the electorate. We
shouldn’€™t give wealthier people benefits just to €œtrick€ them, for
selfish reasons, into voting for greater benefits for everyone, the
poor included.

Here is another point:

Advocates of health care reform tend to be long on ideas for expanding
care and access, but short on practical solutions for cost control. The
argument is often made that single-payer health care systems in Canada
or Europe are cheaper than health care in the United States. But
Medicare is already a single-payer plan, yet its costs are
unsustainable.

Note that I am calling for higher benefits for the poor and lower benefits for higher-income groups.  That’s not a popular stance, not even with egalitarians.  In fact I view the contemporary left as oddly ill-prepared on the health care issue.  Electorally speaking, the issue is fully 100 percent in their court (and they are used to pressing it aggressively), until of course they get their way and have to "meet payroll," so to speak.  One attitude is to cite Europe and think that the production possibilities frontier can expand under better management of the U.S. system, even as you cover an extra 40 million people.  Another attitude is to face the notion of trade-offs. 

Here is the full column.  (By the way, I think that HSAs are ineffective as health care reform and that the so-called "right" is floundering on
this issue, just to get in my equal opportunity smack on the blog.)

Addendum: You can make a good argument that (some) public health programs are the best health care investment of all; I just didn’t have enough space in the column to cover that issue.

Second addendum: Greg Mankiw didn’t read so closely.  It’s not "an income tax surcharge on sick, old people."  It’s a reallocation of benefits toward people of greater need.  Is any benefit less than infinity an "income tax surcharge"?

Third addendum: Here is Paul Krugman on the topic.