Tests of the labor turnover theory of wage determination in Tanzania

Labor turnover theory is one of several competing explanations of wage determination in urban labor markets of less developed countries. The theory is particularly relevant to Africa, where the flow of migrants from the cities to the rural areas is substantial. One of the applications of labor turnover theory has been to study the welfare implications of various government policies to alleviate unemployment in African cities. Empirical tests of the theory have been reported for one country--Zambia. Those tests were confined to only one prediction of the theory. In this paper, we test several predictions of labor turnover theory using data from the National Urban Mobility, Employment, and Income Survey of Tanzania (NUMEIST).We find that higher earnings in the urban areas and less attractive opportunities in the rural areas weaken the intention of a migrant to remain in the urban areas. Our results indicate that migrants with less attractive opportunities in the rural areas are more likely to be in skilled occupations in the urban areas. There is some evidence that these migrants receive higher lifetime earnings in the urban areas. Their earnings also tend to be more responsive to tenure on the job. These findings are consistent with the labor turnover theory of wage determination. An employer could also use nepotism to attract migrants with less attractive opportunities in the rural areas. However, our results are clearly inconsistent with nepotism toward these migrants.We review and integrate four previous contributions to labor turnover theory that are of special interest to development economists. Theoretical foundations are established for the behavior of workers in labor turnover theory. We also derive one new prediction from the theory. Using dichotomous probit models, we examine the determinants of the intention of a migrant to return to the rural areas and the inclination of an employer to select a migrant for a position in a skilled occupation. Using human capital earnings functions, we examine the patterns of lifetime earnings and the shapes of earnings profiles among migrants. We also test for nepotism toward a particular group of migrants.