Some commenters both here and on the NYT site complained that the work I cited in my “who-benefits-from-a-minimum-wage-increase” piece did not build in any job loss effects. What they do not seem to realize is that accounting for job loss effects in this context would only bolster the case that those who would benefit from the proposed increase are older and more well-educated than in the past.
In CBO’s oft-cited minimum wage study, based on extensive literature, the agency applied an elasticity of -0.1 for teens (a 10% increase in the minimum wage would reduced teen employment by 1%), and one-third that amount for adults. So accounting for job-loss impacts would raise the average age, schooling, parental status, etc. of affected workers.
I well remember one day many, many years ago in econometrics class when our quite prickly professor put some formula on the board and asked “what’s wrong with this estimator?” My hand shot up as I responded, “it’s biased!”
To which he replied, “Anyone can see that, Mr. Bernstein. The question is which way does the bias go.” Since then, I’ve never claimed bias without an adjective (upward, downward, or indeterminate). I recommend others do the same.