Some Simple Economics of Mandated Benefits

Via MR commentator John Chilton, here is a link (JSTOR, but free pdf here) to the classic and excellent treatment by Lawrence Summers on mandated benefits.  My view is that the main case for mandated benefits is simply to note that public provision is often worse and that direct subsidy, such as a cash transfer, is not always feasible.

The case against is simple too.  Say that previously unprovided health insurance would have cost the employer 60 and would have been valued by the worker at 40.  You're imposing a tax of 20 on the employment relation.  In the short run firms will hire less labor and during a recession is an especially bad time to produce that effect. 

In the longer run, if the market is competitive, wages will fall by 20.  We're forcing relatively poor workers to consume more medical insurance, and more medical care, than they wish to, at the expense of their cash income.

Do not forget the excellent words of Ezra Klein:

Indeed, the main impact of health-care
reform on health may be that if it could contain costs, we'd have money
to spend on things that actually do make us healthier.

He means things other than health care in the narrow sense.  I don't know if Ezra opposes the mandates approach (compared to what?) but his quotation indicates a problem with it, even from the standpoint of health alone, much less considering the other pleasures of spending or saving cash.

Of course some of the people covered by the mandate would otherwise end up showing up at emergency rooms.  Treating them that way would get tacked on to my medical bills, one way or another.  With a mandate they are no longer my financial headache. 

With this new change, who's better off?  Me.  Who's worse off?  The previously uninsured poor person.

You might say: "We are covering more people, at a lower price, than we had thought possible."  That sounds like a kind of triumph.  But if you cut through to the actual analysis, your paternalism has to be a lot stronger than your egalitarianism for you to support this kind of measure.