Professor Perri of Appalachian State sent me a very interesting point about why increases in the minimum wage could increase employment– while decreasing hours of employment at the same time.
When the minimum wage rises, the cost of increasing HOURS PER WORKER increases faster than the cost of increasing THE NUMBER OF WORKERS if there are any fixed costs (hiring and training costs) per worker. Thus, in addition to the the other two effects when the minimum wage rises [which are the SCALE effect (the marginal cost of production rises, reducing profit maximizing output and the usage of all inputs), and the FIRST SUBSTITUTION EFFECT (firms use more capital and less labor)], the SECOND SUBSTITUTION EFFECT is for firms to increase the number of workers, and decrease hours per worker. This, last effect offsets (at least partially) the first two effects, making it harder to find disemployment effects of the minimum wage. Some will assert fixed cost are not large for low-skill workers, but indeed they may not be insignificant.
Thus, workers lose hours, which makes them worse off since, if they tended to prefer fewer hours, firms would have been forced to offer them given how competitive the labor market is.
The point needs some expansion. If the marginal cost of a worker were constant, it would not work. Then, the employer would want to hire just one worker, to minimize the fixed cost, and work him 1000 hours per week. What is crucial is that the initial number of hours be enough that the marginal cost of a worker is rising.
A good model is that the marginal cost of a worker is constant for the first 40 hours per week, and then rises sharply with overtime, and then rises gradually as the worker’s effectiveness falls because he gets tired. Without the minimum wage increase, the employer and workers might find it optimal to hire workers at more than 40 hours per week, because it spreads the fixed cost of training. After the wage increase, the optimal number of hours will fall.