Krugman writes:

I see that a number of commenters failed to click on my link in this post. Um, it’s true that the Times article did not include data on benefits as opposed to wages. But other studies have — notably the big EPI study (pdf). The Times results match what those studies say about wages — and those studies find that the lack of evidence for overpayment remains once you take benefits into account.

Has Krugman referenced Andrew Biggs on the EPI study? Indeed, how many of those who’ve cited the EPI study as evidence asked basic questions about whether retiree health benefits are worth zero dollars and if it makes sense to count pension benefits as currently funded rather than as promised, i.e., as contractually obligated? Also, is it sound to assume that all state workers would immediately go to work for the largest business enterprises in the state, e.g., “correcting for firm size”? Given that Microsoft and Boeing and GM can only absorb so many workers, my guess is that the answer is no. The EPI staff is smart enough to understand that these are legitimate questions.

Public sector workers are not, on average, grossly overpaid compared with the private sector — period. You can fiddle at the edges of this conclusion, but it’s just not possible to conclude, based on any honest assessment of the data, that schoolteachers are the new welfare queens.

I find this hilarious. In a similar vein, the median consumer of medical insurance doesn’t generate extremely high costs. Rather, it is a few outliers at the right end of the bell curve. This is why we ought to approach some patients differently from others. That is, we don’t want rigid rules that force us to compensate very productive workers the same way we compensate not-so-productive workers.