³ÉÈË´óƬ

It is commonly thought that firms use aspects of compensation schemes, such as wage levels and promises of future raises, to reduce worker turnover, attempting to increase productivity through intertemporal relationships (such as human capital investments and dynamic incentive schemes) with their workers. Using data on Swedish white collar workers, I estimate a dynamic labor supply model to measure the empirical relevance of the belief that compensation schemes have large impacts on worker turnover. The model allows each worker to choose between staying at his current firm and moving to a particular new one, according to the compensation schemes offered by all firms. Workers consider the expected value of the wage profiles offered by each firm in addition to moving costs. The results show that the sensitivity of worker turnover to compensation schemes is low and it is not cost-effective for firms to reduce turnover by increasing future raises. Older workers rarely change employers because they face high movement costs and do not receive large wage increases upon moving. However, the turnover behavior of younger workers is more sensitive to future raises and the wages of young workers change more upon moving. These findings are consistent with a story where firms compete for younger workers much more than for older workers.