Asymmetrical Information - Megan McArdle


Why Not Make Social Security Benefits Even More Generous

People haven't saved enough for retirement. Should the government make it up to them?

Josh Barro notes an absolutely true fact: most people don't save enough for retirement.

Therefore, says Josh, we should raise Social Security benefits to make up the shortfall:

Private saving for retirement is woeful. This typical near-retirement household has just $42,000 in retirement accounts and $18,300 in other financial assets. For most Americans, Social Security isn't augmenting private saving; private saving is (just barely) augmenting Social Security.

And as both home equity and stocks were battered over the last few years, retirement insecurity worsened. Munnell and her colleagues estimate that as of 2010, 53 percent of American households were on track to be more than 10 percent below the amount of assets they would need at age 65 to maintain their standard of living in retirement, up from 44 percent in 2007.

No entity is better positioned to fix this problem than the federal government. Employers won’t do it: They have been dropping their defined-benefit pension plans for good reasons. And with Americans working for an increasing number of employers over the course of a career, such plans become ever more inappropriate.

Individuals won’t do it: Tax advantages of retirement-saving accounts don't seem to induce people to save enough on their own. And when people do use individual retirement accounts and 401(k) accounts, they're often hit by high fees and bad investment choices.

State and local governments won’t do it, either: They're already under severe fiscal pressure, particularly due to rising health costs. Unlike the federal government, their obligation to approximately balance their budgets annually makes them unsuited to providing retirement security. Part of the way a government can shoulder the economic risks of retirement is by making deficit-financed payments when the economy is weak. For this reason, states and localities are right to work on reining in the cost of their employee pension plans, rather than expanding them.

Unlike every other player, the federal government is positioned to help. 

He suggests paying for the sweeteners with either a new tax, or by cutting Medicare benefits.  And the latter has a certain pleasing economic logic to it: why not give people cash instead of an in-kind benefit?

Well, there is a reason, with health care; in fact, there are a few of them, all of come back to the observation that health care is simply fundamentally unlike other in-kind goods that the government provides. Unlike the cost of food, or housing, or other in-kind benefits that we give to people, the value of health care varies wildly between individuals.  Some people, even the majority, might be better off with cash.  But the people who would be worse off would be really worse off.   

At this point my critics will be tempted to rejoinder that I myself wrote an article pointing out the surprising ambiguity of the data linking health insurance to mortality improvements.  All too true.  But people don't think this way.  You can explain until you are blue in the face that statistically, they won't be better off with balloon angioplasty or back surgery.  But for practical effect, you'd be better off trying to explain the fundamentals of tensor calculus.  At least they won't get angry and accuse you of heartlessly trying to kill them in order to save a few dollars on tensor fields.

Subjectively, people who want an expensive operation and can't get one are going to be incredibly upset, an upset that you will not soothe by pointing out that they got an extra $274 a month in their Social Security check.  Those people will dial up their old chum at the Senator's office and demand to know why the Medicare actuaries want them to die.  Then your health care cost controls will probably go away.

It is relatively easy to cut the budget of a public park, or give everyone $5 less of food stamps every month.  But it is very hard to cut health care benefits, in part because you can't just give everyone a slightly cheaper doctor.  You have to go around deciding which procedures will be reimbursed and at what rates, something which sounds very easy to everyone until they actually try to do it.  It is all very well to read a study which says that back surgery is mostly useless, and decide that we shouldn't cover it.  But it will be harder to stand up to the American Society of Back Pain Sufferers and the various other groups which will swarm you if you announce that back surgery will no longer be reimbursed.  

But withstanding political pressure is going to be the least of your problems.  Medical studies are very hard to do well: it is difficult to get enough participants, and to ensure that they are actually complying with your program.  Small sample sizes and strong publication bias towards positive results make it hard to get a definitive answer as to whether a given procedure should be done or not.  This can be further complicated by sloppy-to-inept statistics.

Even if you have a good study, you may find that the world of your study and the actual world in which doctors are practicing medicine are not quite the same.  For instance, I once remonstrated with a back surgeon over the poor evidence that back surgery does any good.  His exasperated reply was that he knew about those studies showing that physical therapy was better, that in fact all the back surgeons know about those studies . . . but that the back surgeons also knew something which had apparently eluded every public health researcher in America, namely that outside of small, selective studies, it is hard to get patients to actually do the physical therapy that everyone agrees is usually better than back surgery.

Apologies for getting off onto a bit of a rant here.  But "we'll just spend less on health care" has become the Laffer Curve of 21st century wonkery: a sort of parlor magic trick that is supposed to deliver free benefits without any hard choices.  Every time someone is asked how they will fund some marvelous new benefit they're proposing, they gaily reply that we ought to be spending less on health care.  It is entirely true that we oughta.  But no one--literally, no country, as far as I am aware--has come up with a sustainable way to actually spend less on health care; the best anyone has done so far is to keep health care costs from growing quite so fast.  So it seems rather unlikely that we're going to find a way to transfer, say, 2% of GDP from health care to Social Security.  And if we do manage to slow the growth of Medicare costs, the savings will have to be spent not on fun new benefits, but on less-fun items like not having a 50% budget deficit.  

As for doing it with taxes, Social Security is expected to stabilize at around 6% of GDP.  What sort of boost are we talking about?  Enough to make up for the fact that our national savings rate has fallen from about 10% of income in 1980 to 2.5% now?  That would be a huge increase, not a small one. It would have quite noticeable effects on the economy, and on the living standards of the younger, poorer workers that we are asking to transfer money to older, wealthier ones.  One of those noticeable impacts would be a further decline in labor force participation by those in their early sixties.  Another might well be a further decline in the private savings rate.  

Of course, we could give everyone smaller increases.  And I imagine they could find uses for the money.  But the same could be said for millions of non-elderly people who are stretched thin.  I'm not sure I see a strong case here for taking even more money from young people and transferring it to people who decided not to save enough.