Evidence on the Minimum Wage

Brad Plumer at wonkblog presents a great summary of this paper by Schmitt at CEPR on why the classical result on the adverse effect on unemployment by raising (or establishing) a minimum wage may not be true. Some of these reasons seem somewhat intuitive, but one in particular is giving me pause:

5) Employers can respond by becoming more efficient. If minimum-wage workers suddenly cost a bit more, perhaps businesses will react by trying to squeeze more productivity out of them. Schmitt notes that there’s some evidence that this happened in fast-food chains in Georgia and Alabama. Managers started requiring better attendance and asking their employees to take on extra duties in response to a minimum-wage hike.

My main problem with this point is that it only works with the assumption that managers/firm owners are not already operating at the highest efficiency level. If it is true that employees could be more productive than they are under the current (or unchanged) wage regime, I would assume that a manager would allow this only to capitalize on some non-income related aspect of retaining workers, such as an abstraction of an efficiency wage. In this way, managers may currently be allowing workers to work at a less than optimally productive rate to incentivize longer stints of employment, saving the cost of training or hiring a new worker.

This would make some sense only if the demand for such low-skill workers were sufficiently high or working in a competitive market. However, if we are to simultaneously take stock in the monopsony model of these types of positions, this certainly isn't the case.

At the heart of the matter is the notion of whether managers and those firms hiring minimum wage employees are actually maximizing profits currently. The IZA paper with the link in Plumer's quote above provides some interesting evidence that such adjustment channels are being used, but I think it would be worthwhile to test profit maximizing behavior in various minimum wage regimes. Ozimek propounds a randomized minimum wage to take effect in various states for about an 8 year period, after which time the policy would be reassessed and allowed (or disallowed, depending on the results of the natural experiment) to continue. Recognizing the seeming harshness of such a policy, he says:

It may seem callous to put the fates of workers in the hands of random chance, but this is the most surefire way to make sure the minimum wage isn’t causing unemployment.

It's odd to me that Ozimek ignores the most obvious alternative: actual laboratory testing. After all, experimental macroeconomics is not unheard of.