Given the high level of job mobility in the United States, one might think that obtaining a low-paying job would have only temporary consequences. However, using longitudinal data, I find that state dependence in wages is large and persistent. If two comparable individuals start jobs that pay a different wage, about 60% of the wage differential is still present four years later. Moreover, about 50% of the wage differential is still present for workers who have switched employers during that period. The results indicate that the jobs acquired by individuals have long- term effects on their future careers. I also examine the mechanisms that lead to state dependence. In a stigma model, prospective employers use wages as a signal of ability. Thus, getting a poor job can lead the market to believe that an individual has low ability. In the learning-by-doing model, workers who get high-paying jobs also attain greater opportunities to acquire human capital. The evidence suggests that both stigma and learning by doing models contribute to state dependence in wages.
