[Peter Suderman] Ezra Klein thinks markets don't work efficiently for health insurance because patients lack the ability to accurately assess doctor performance:

I spent some of the morning running almost random searches to try and remember the name of my doctor. Eventually, I got him. And a bunch of reviews of him. Most of the reviews, to be sure, were good. The consensus appears to be that he's sort of mean to you, but quite thorough. But that's all the reviewers were able to evaluate: Was he courteous? And did he do lots of stuff?

Those are not particularly useful metrics on which to evaluate doctor quality. Etiquette is important, but neither here nor there to diagnosing appendicitis. And "doing stuff" can actually be bad, and needlessly costly. The question is whether the doctor is doing the right stuff. But patients simply don't have the requisite tools to evaluate that question. They can't review doctors like they can review books, or movies, because impressions of the experience aren't terribly relevant, or insofar as they are relevant, they're a second-order concern. There have been some efforts to get insurers to rate doctors, but their incentive, of course, is to rate doctors highly when they're being stingy, and helping the insurer make profits.

Here's where I wonder if some sort of independent rating agency not beholden to either doctors or insurers would work. And either way, would a largely or fully public health care system, like Ezra wants, solve the problem of lack of patient knowledge? I'm not convinced. Patients seeking information about doctors would still have to turn somewhere, and that would likely mean health care metrics provided by the government, which would still be subject to various outside incentives (keeping costs down to lower the taxpayer burden, fiddling with the stats to show better performance). We've all seen The Wire and heard about "juking the stats." Public institutions aren't exactly perfectly honest when it comes to self-assessment.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.