Friday, February 08, 2008

Will This Give Mark Kleiman the Vapors, Too?

by Ken Houghton

Mark Kleiman blew some smoke a couple of days ago, catching, among others, the normally-brilliant Battlepanda with his "I'm a pundit, I don't have to know anything" post.

More on that one later, but let's do a high-level comparison. Kleiman lets Harold Pollack attack the Gruber paper. Why? Because of this (inane and unworthy of even a Huffington Post contributor; see later post) quote:
In particular I assume that 95% of those who would not voluntarily choose to insure are forced to insure through the mandate.
So we are left to wonder whether he will now attack the Obama campaign and David Cutler in the same manner, after Cutler declared (via Brad DeLong):
Our estimates, based on studies in the literature, is that we can get 98% or 99% coverage without a mandate for adults. There may be some small pockets of people who choose not to buy it.

I'll wait patiently for Kleiman to realise that Cutler's estimate has the same basis as Gruber's, with much less empirical data to support it.

UPDATE: Let's put the best spin possible on this. Cutler assumes a 98-99% adoption rate wihtout mandates. In the discussion, he declares:
If there are free riders, Obama is open to mandates....

Richard Eskow: Would mandates be considered at that point?

David Cutler: He hasn't ruled anything out. It's a matter of priorities. The fact is, the policy differences on the mandate issue aren't that large at all. Sen. Obama believes they're an option down the road, if other approaches don't work.

Let us leave aside that what Senator Obama believes doesn't match with his advertisements or his public statements (Krugman's point here and here), which implies either Cutler isn't as deeply in the loop as he appears to be, or Obama isn't so good a politician as the world believes he is.*

Cutler is essentially saying, "Yes, mandates would increase the participation rate, but we would prefer not to use them." Gruber is saying, "Mandates should improve the participation rate—especially given that the enforcement mechanisms are already in place—so I'll make a reasonable assumption that, even with them, there will still be a small fraction of the populace that will remain uninsured.**"

But the former assumes a significantly higher adoption rate than the latter, even though both agree that a Mandate would insure more people.

*And, yes, I do fit in that world.

**Indeed, if we assume sufficiently high income and a sufficiently safe lifestyle, insurance could become an unnecessary expense. Bill Gates, for instance, could pay for all the health care he needs out of petty cash.

Labels: , ,

Depends on what Cutler means by "studies in the literature," no?

It's not as if the Obama plan works entirely on the honor system -- it includes mechanisms (default enrollment, penalties for late enrollment) -- and I do seem to recall seeing article titles in the Journal of Health Economics addressing how various mechanisms affect enrollment rates.
I've been going through the recent papers, and can't find anything that is anywhere near that optimistic.

Gruber's model—which makes what is probably an optimistic across-the-board assumption on the relative health of the newly-insured—estimates about 50% of the uninsured without a mandate, which is about a 90% total-insured level.

Getting that additional 40-45% that Cutler assumes, even with opt-out requirements and late penalties seems dubious--see Plan D.
Well, the question then is the degree of subsidy. (This, of course, is where things would be a lot simpler if market-oriented center-leftists didn't waste their time trying to put lipstick on the current pig of a system.)
Um, Ken,
If you read my post, you'll see that I actually disagree vigorously with Kleiman. I only say that his line of reasoning is the "most persuasive" because the other Obama people are assuming that everybody is just going to join the Obama plan because of its sheer fabulousness from being associated with Obama.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?