Yesterday I gave you the research case for a minimum wage increase. The starting point for the research case against a minimum wage increase should be David Neumark and William Wascher's book Minimum Wages which makes a full-bore case that minimum wage rules are perverse. That starts with findings of short-term disemployment effects, but continues with the argument that the impact is particularly severe on the youngest unskilled workers who suffer lasting damage by failing to get inside the world of formal paid labor quickly. They also argue that minimum wages are poorly targeted policy. Many of the workers who earn more as a result of the minimum wage are not poor, and many poor families receive little help from the minimum wage.
This is precisely the line of research that Arindrajit Dube and his co-authors are critiquing in their "new minimum wage" research that I referenced yesterday. Naturally Neumark & Wascher have (with J.M. Ian Salas) a critique of their critique of their work. Something worth noting is that in principle policymakers could actually design minimum wage policies that are meant to shed light on the issue. We could, for example, randomly assign counties different minimum wages on a spectrum of $5 an hour to $15 an hour and see what happens.
Meanwhile, though this cuts against the general charts-and-NBER-abstracts tendencies of the wonk blogosphere, the clearest case against a minimum wage hike has (in my view) more to do with freedom than with research. You've got a guy who wants to give someone $8 to do something that'll take an hour and another guy who wants $8 and is happy to do the thing in exchange for the money. Now Barack Obama's going to fine them for agreeing to trade $8 for the work? Seems perverse. In the real world, obviously, the perversity of this is greatly mitigated by the existence of formal exemptions and weak enforcement. If you pay a neighbor's son $10 to mow your lawn and it takes him 70 minutes, you're going to be able to get away with it even in a world of a $9 minimum wage. Which is probably as it should be.