Asymmetrical Information - Megan McArdle

01.15.13

Are Gold-Plated Health Benefits Making American Workers Worse Off?

Health care costs have eaten up our compensation growth. Is that a bad thing?

One of the tricky parts of discussing inequality is how you compute income.  Americans take a lot of their compensation in the form of tax-advantaged benefits, particularly health care costs.  Michael Tanner of Cato argues that if you look at total compensation, inequality may not be increasing at all.

Ramesh Ponnuru thinks that this is a stupid argument:  

Conservatives are right that trends in total compensation look better than trends in wages. But that’s not a reason for complacency. It’s a problem. What the numbers mean is that increases in health-care costs have depressed wage growth, and sometimes kept wages from rising at all.

If there’s a consensus among health economists about anything, it’s that employer-provided health benefits come out of wages. If health insurance were cheaper, or the marketplace were structured so that most people bought health coverage for themselves rather than getting it with their jobs, people would be paid more and raises would be higher.

If people consciously decided to spend most or all of their pay increases on health care, that would be one thing. They don’t. Current policies elaborately disguise how much health coverage costs people. Because the federal government taxes wages but not health benefits, employers provide more of the latter and less of the former than they otherwise would. Most people have no idea how much money they have forgone in wages because of those benefits. They never see the money.

Health benefits are, of course, valuable to people, and the increase in their cost over the last generation -- which those statistics on rising compensation show -- partly reflects that medicine can do more than in the past. It also reveals a lot of waste and inefficiency, however, rather than increased well- being.  

I am second to none in my hatred of the tax exclusion for employer-provided health insurance, which distorts the market (and the tax code) in all sorts of ways.  Nonetheless, I'm not as sure as Ponnuru that the compensation being funneled into health insurance is all that much less valuable than wages.

How would we know if the money was wasted?  Well, if we had evidence that employees would prefer cash to benefits.  Or if we had evidence that employer-provided health insurance was making health care cost more, without buying more benefits.  

I'm not sure how true either of these things is.  During health reform, everyone on the pro-Obamacare side seemed convinced that there were magic pots of money to be found buried under the health care system--a bunch of stuff that obviously cost too much, or shouldn't be done at all.  But they seemed to have a hard time actually locating the magic money; instead, the legislation relied mostly on blanket reimbursement cuts, and a "magic asterisk", in the form of IPAB, which was supposed to identify some more targeted cuts later.  A lot of what were supposed to be obvious cuts in worthless stuff like end-of-life care and expensive cancer treatments turned out to be painful discussions about how much we were willing to spend in order to give someone a small chance at extra months or years of life.

I don't have any obviously metaphysically correct way to answer that question.  Neither does anyone else who has so far weighed in.

So we turn to revealed preference: would people rather have cash than the option to have an expensive cancer treatment at some point in the future?  Again, not clear to me.  The best data we have about this sort of tradeoff is from unions, who explicitly make thesm during collective bargaining (or as administrators of multi-employer health plans).  I don't see them giving up gold-plated health benefits for wage gains--rather the reverse, in fact.  They frequently give up other stuff in order to protect their health benefits.  It's possible that this represents some sort of principal-agent problem, with the union negotiators selling out the membership, but I've seen no evidence that this is the case.  

Now, maybe people wouldn't choose to buy the marginal new treatments if they had to actually face the price tag--if your insurance company said, "Do you want coverage for experimental cancer coverage at $100 a month?" you might say, "No, thanks."  But even in a perfectly free market, you wouldn't actually make all the tradeoffs that explicitly, because it would take days to write the insurance contract, and there would nonetheless be wrangling over which treatments counted as "experimental". In other words, consumers would probably waste some money on stuff that they wouldn't want if they had perfect information, just as they do with cars and cell phone contracts.  So it's hard to tell how much less they we would all be spending if you gave folks the cash and let them buy insurance for themselves. I mean, I personally would have a high-deductible plan.  But my emotional intuitions about things like insurance and extended warranties are . . . atypical.  

All of which is to say that it's not clear to me that consumers are actually worse off than they would be if they'd been getting more cash, and less health insurance.  I'm not arguing that they aren't worse off, only that I think it's hard to tell.

I do think it's quite clear that they feel worse off, because there are about six layers of insulation between them and the good they are consuming.  Even people who use the expensive neonatal services or get an advanced cancer treatment don't think to themselves, "Hey, good thing I've been paying an extra $300 a year for the past six years, because this is awesome!"  For one thing, they dont' know the prices, and for another, they rarely perceive the direct tradeoff between benefits and wages, since those things are almost never negotiated.    

That is a real problem: personal, and political.  One that I unfortunately have no idea how to fix.