What Happened to U.S. Health Care Costs?

Tyler Cowen points to an intriguing graph:

. . . and asks "what happened in 1980?"  If I had to guess, I'd say "some sort of data discontinuity", but other than that, I don't know.

I do, however, have a plausible candidate for what happened in 1986, when the thing starts rising briskly over a period of years: the Reagan tax simplification.  It got rid of a lot of the ability to expense lifestyle enhancements for workers, but left an important one intact: the deduction for health care.

You can argue that the employee health care deduction basically explains the entire cost differential between American systems and the others.  Take a look at this graph (you'll have seen some variant of it many times during the health care debate, but I got this particular one from a presentation by Scott Atlas at Hoover last month):

The blue lines show per capita spending; the purple, spending as a percentage of GDP. Now look at that same image, hamfistedly altered by yours truly to show what the graph would look like if the approximately 35% tax subsidy we give to employer health care benefits were withdrawn:

Suddenly, the US isn't really an outlier.  We'd still be spending more--but we're richer, and in general, rich countries spend more of their GDP on health care.

Is this dispositive?  Hardly.  But it's suggestive.  And there's some literature to back up effects that are huge, if not quite 35%, like Jonathan Gruber's recent paper.

Naturally, this puts me in mind of Ezra Klein's recent post:

The problem is that Medicare can't control costs too much better than private insurers or, as you see from the article above, doctors will simply abandon Medicare. In a world where there's only Medicare and Medicare decides to control costs, doctors can either take the pay cut or stop being doctors. And as we see from other countries, lots of people want to be doctors, even if being a doctor doesn't make you particularly wealthy. But in a world where Medicare is just one of many payers and Medicare decides to control costs, doctors can simply stop taking Medicare patients and a lot of legislators will lose their jobs.

Given how hard it's proven to tackle the tax subsidy for employer-sponsored insurance, I'm not seeing much of a change any time soon.  In fact, a recent post by Austin Frakt suggests that we're expanding it:

Are you getting this? Let me make it clear. The PPACA may make it possible for workers to get the same tax break for purchasing health insurance on the individual market (via an exchange or otherwise) as they would if they bought their employer-sponsored plan (if they're offered one). If this is the case, it removes one huge incentive for maintaining employer-sponsored coverage. With respect to taxation, it levels the playing field between the group and non-group (individual) markets.

There are some economic efficiency arguments for doing this, but it's not going to help us control costs.  Perhaps quite the opposite.