The other day, when criticizing an incredibly facile minimum wage analysis in the Arizona Republic, I had meant to observe that since minimum wage jobs are such a tiny (1.5% if include jobs that work for tips) portion of the workforce, one should look at more targeted metrics to assess the effect of minimum wage hikes, such as teen employment.
Kevin Erdmann has such an analysis. He observes, "Is there any other issue where the data conforms so strongly to basic economic intuition, and yet is widely written off as a coincidence?"
Note that there is still some danger, as I wrote before, in measuring employment effects from the implementation date. Businesses plan ahead an many job losses may be occurring between the announcement and the implementation date. I know we have made all the job cuts we plan to make in response to California minimum wage increases six months ahead of the actual date the wage takes effect.
Update: The charts are obviously far from a smoking gun. That is the nature of economic analysis. In complex and chaotic multi-multi-variable systems, controlled studies are almost impossible and direct correlations are hard to find, and even when found may be coincidence. As an employer who hires a lot of summer seasonal employees in parks, I would obviously be a natural employer of teens. But I no longer do so, and it is important to understand that wages are only a part of the equation. Another major issue is one of liability. Increasingly, the legal system makes the employer liable for any action of their employees, no matter how boneheaded or how much the action is against all policy and training. I have enough trouble with employees that have years of good work history -- I am not really excited about taking a chance on an unproven 17-year-old.