Sunday, February 17, 2013

Econ 101 and the Minimum Wage

Charlie Clarke

    Lots of work done on the minimum wage over the last 20 years, since this paper by Card and Krueger have argued that at low levels the minimum wage doesn't lead to increased unemployment.  I thought a lot of this work was based on publication bias -- the desire of the profession to publish interesting and counter-intuitive results--until I geeked out on it a little last year and the results are much more robust than I thought.  This guest post on Mike Konczal's blog provides a nice review (and much more info than anything I've seen in the recent blogosphere debates).

     Interestingly, this view has become so mainstream that it is Econ 101.  In fact, the event leading me down a nerd hole of research was over hearing Jonathon Gruber's Principles of Micro MIT opencourseware course.  He taught a Monopsony version of the minimum wage market, where low income employers (McDonald's, Wal-Mart...) have some market power and thus a minimum wage may increase employment.  Then at the end, summarized the balance of the evidence as pointing to the monopsony model.  I couldn't believe it, but he was right.  So, at least at MIT, possibly the best place to get an undergraduate education in economics, the mainstream view has shifted quite a bit.

    That said, I'm quite skeptical at the current liberal enthusiasm for the President's plan to raise the minimum wage.  None of this research focused on minimum wages when the economy was suffering from a shortfall of aggregate demand.  Just as, fiscal multiplier research may have little bearing on fiscal multipliers during a liquidity trap (self-imposed or otherwise), the distortion of a minimum wage may be quite a bit more severe under current circumstances.  


  1. I'm glad you wrote this because I was just about to write a much worse, very similar post.

    This thought is half formed, but do you think more market oriented democrats still flock to this idea as a second best alternative to a social welfare system that doesn't draw sharp distinctions between employment and unemployment? Like: "Since the government is going to give you a lot less money if you get a job, and since the government is going to make you look for a job, we want to make sure anyone who does get a job is well compensated. To the extent this increases unemployment, those people can access UI and job training programs."

  2. I think there is something to that, but now that we have an Earned Income Tax Credit it's harder to justify the minimum wage. Even Milton Friedman thought it was unfortunate that some people when employed would make very low wages, he just thought the minimum wage was a bad way to deal with it.

    Krugman appeals to an argument (by Lee and Saez) that employers respond to the EITC by lowering wages and somewhat mitigating the benefit. Thus, perhaps a best policy uses both the minimum wage and an EITC. It's an interesting theoretical argument, but it's a long way from indicating that our current minimum wage is too low. I am not at all confident that is the case.

  3. Well my preferred policy is to just give poor people wads of free money, with the wad decreasing in size as their income rises. So it all seems sort of weird and excessively complicated to me.