Claim: Instead of helping low-wage workers, an increase in the minimum wage will cause them to lose their jobs.
Certainly that’s the prediction from the classical model of the labor market. Everyone’s being paid precisely their marginal product (their value added to their firm’s production at the margin) and to increase their pay by mandate is to render them unaffordable. In fact, the prediction from the simple model is that even a few pennies increase in the minimum wage would lead to extensive job losses.
And yet…we’ve had dozens of federal minimum wage increases and now have 19 states with minimums above the federal level. Surely if the classical model were correct, we’d have clearly seen its negative impacts by now.
Which makes it, like all good economic questions, an empirical one. Thankfully, there’s been extensive research on the question which is very usefully reviewed here by economist John Schmitt of CEPR. Pay particular attention to the results from all those natural experiments created by all the variation among the states.
I’ve often said the true elasticity of job loss with respect to a minimum wage increase hovers about zero. I’ve seen good studies finding small negatives and good ones finding small positives. John presents this very cool graph from a “meta-analysis”—a study of a bunch of studies—showing precisely that result. There are outliers on both sides of zero, but the strong clumping around zero provides a useful summary of decades of research on this question.
These results don’t mean that no worker ever loses her job when the minimum wage goes up. But they do mean that the vast majority get a raise and that anyone blithely citing the classical model, as Rep Boehner did the other day (“When you raise the price of employment, guess what happens? You get less of it”), is speaking from prejudice, not from the evidence.