Institute for Development Studies, University of Nairobi

Regression analysis is used to explain rural-urban migration flows in Kenya as a function of economic incentives and the distribution of potential ethnic contacts linking urban and rural areas. These ethnic linkages not only have high explanatory power but also improve the performances of the other economic variables. The results indicate that migration is highly sensitive to the level of urban formal-sector wages but is not directed toward centres experiencing faster growths in formal-sector employment. It appears that when the urban labour- markets are expanding very slowly, ethnic contacts rather than urban jobs are a better indicator of the probability of obtaining urban employment.