Training, Productivity and Wages: Direct Evidence from a Temporary Help Agency

Firms frequently provide general skill training to workers at the firm's cost. Theories proposed that labor market frictions entails wage compression, larger productivity gain than wage growth to skill acquisition, and motivates a firm to offer opportunities for skill acquisition, but few studies directly test the hypothesis. We use unusually rich data from a temporary help service firm that records both workers' wages and their productivity as measured by the fees charged to client firms. We first document that the firm provides upfront training, and show that both workers' tenure and the initial fee charged to clients are positively related to the length of training, but the initial wage paid to workers is not. We then demonstrate that the fees charged to clients grow faster over workers' tenure than the wages paid to workers. Finally, we find that about one-quarter of the fee growth is associated with client quality upgrading, but that workers receive none of this growth. Each of these results are consistent with wage compression that skills acquired through training and learning-by-doing increases productivity more than wages.