I ran across this
fine post from John while looking for something else.
There is no question that it is tough to find econometric evidence of capital-labor substitution in response to minimum wages. Some interpret that to mean that it does not occur, others that it occurs with long and variable lags that make it tough for our usual design-based strategies to sort out. I tend to favor the latter, based in part on my experience of going to China, where labor is very inexpensive and where, as a result, you are often out-numbered by the staff when you go to a store. Sometimes having more variation in the independent variable helps to clarify the sign and magnitude of the coefficient.
And, really, what we should be having a policy discussion about is not the minimum wage but about how to reduce the number of workers whose marginal product is of such a low value that the minimum wage can easily exceed it.
