I had the same reaction as Pablo Halkyard at the PSD Blog to yesterday’s article in the NYTimes on Bolivian water privatization so here is his post:

The adverse selection story is a wonderful example of McCloskey’s argument that great rhetoric persuades even when it shouldn’t.  The market for lemons is simple enough for your friends to understand but profound enough for them to be impressed at your learning, so it’s a hard story not to tell!      

The facts of the matter, however, are that adverse selection is not an important part of the market for automobiles (trucks), or of auto, life insurance or health insurance (on the latter see below).

One reason adverse selection may not be that important in practice is because buyers and sellers use testing and certification to remove the most important information asymmetries.  You can buy a decent used car, for example just get it inspected or certified.  Only if such adjustments are illegal, or in some other way not allowed, will adverse selection become important.

Second, the asymmetry may run in favor of the sellers.  Do I really know more about my own life expectancy than an insurance firm that has access to sophisticated actuarial models?  And, assuming that I do have extra information is it all that important?  After all "the race is not to the
swift, nor the battle to the strong, neither yet bread to the wise… but
time and chance happen to them all."  Or, more prosaically, the signal is near irrelevant when the noise to signal ratio is high. 

Third, propitious selection can be more important than adverse selection.  What sort of person buys a lot of life insurance?  Is it people who expect to die soon?  Or is it the sort of person who is so worried about not leaving their family in trouble that not only do they buy life insurance they also buckle their safety belt and eat healthy?  The price of life insurance falls the more you buy so evidently insurance companies believe it is the latter.

Everyone talks about adverse selection in the market for health insurance but in fact non-group policies in these markets are not relatively expensive and not hard to get.  The national average annual premium for reasonably generous coverage for a single person is just $2,268.

Sure, that’s a lot of money but the point is that it’s not a lot relative to what an employed person and their employer would pay for similar coverage in the group market.  There is no evidence for an adverse-selection death spiral in the market for health insurance.  That’s not surprising because non-group health insurance is medically underwitten (i.e. medical inspections just like car inspections).  Most people are accepted a few are not.  Only in states that require insurance companies to accept all or most buyers are rates high relative to the group market (rates in New Jersey, an outlier, are almost three times as high as the national average.)

There are problems in the health insurance market, including a lack of long term insurance, job lock and the inequity of affordability, but adverse selection is not one of them.

Thanks to Bryan Caplan, Robin Hanson, Tyler Cowen, Tim Harford, and Ray Lehmann for discussion.

Addendum: Comments are open.

Two doctoral students in Switzerland are running an online test to see whether people can spot the difference between true stock prices and a random walk.  I was surprised to guess correctly 3 times out of 3.

It’s not just GM, United Airlines and the Federal government who have made unsustainable promises to current and future retirees.  State and local governments have also been irresponsible, to the tune of perhaps a trillion dollars in unfunded liabilities.

For years, governments have been promising generous medical
benefits to millions of schoolteachers, firefighters and other
employees when they retire, yet experts say that virtually none of
these governments have kept track of the mounting price tag. The usual
practice is to budget for health care a year at a time, and to leave
the rest for the future.

Off the government balance sheets – out
of sight and out of mind – those obligations have been ballooning as
health care costs have spiraled and as the baby-boom generation has
approached retirement.

…most states and cities have set aside no money to pay for retiree
medical benefits. Instead, they use the pay-as-you-go system – paying
for former employees out of current revenue.

The plethora of discount
rates and taxes obscures the basic point in Feldstein’s argument against
capital taxation.  Here is a bare-bones version.

There are three
goods, labor, apples and oranges.  Assume that the government taxes
oranges at a higher rate than apples.  A tax on oranges is also a tax
on labor since you need labor to buy oranges and if the price of
oranges is high the value of your labor is low.

Now let’s show
that a reduction in the orange tax matched by an increase in the labor
tax to keep total tax revenues constant can make everyone better
off.  The simplest case is to assume a tax on oranges so high
that no one buys any oranges.  Orange tax revenue is therefore zero.
Now we get rid of the tax on oranges and add an equal-revenue tax on labor
(zero).  So long as the consumer cares at all about oranges he now buys
more oranges and is better off (because he now consumes a variety of
fruit and has an increased incentive to work).  The consumer will still be better off even if we replace the zero-revenue orange tax with a
small labor tax which increases government revenues.  The zero-revenue assumption makes the argument obvious but is not at all necessary for the results.

The basic point is that the tax on oranges distorts the labor-leisure choice and the apples-oranges choice.  A tax on labor distorts only the labor-leisure choice and so is preferred.

For Feldstein’s argument rename oranges as savings, apples as present consumption and labor as income.  To see a counter-argument introduce more people into the model and rename oranges as yachts.