68

This paper examines the determinants of household migration from rural to urban areas in Cameroon, with special emphasis on the role of the income gap. The study uses micro-economic data of ECAM II for two consecutive years. Given that migrants are not a random section of the population, self-selection is also taken into account in order to estimate the gains from migration. The migration equation is then corrected for selectivity bias using the Heckman procedure (1979). The estimation of the probit model confirms the multidimensional aspect of the phenomenon of migration. The income differential influences households’ decision to migrate, which is consistent with the Todaro hypothesis. However, it is not the dominant factor. Factors such as access to loans, land and housing, in addition to non-agricultural activities have a more powerful effect. Some household heads may therefore migrate even though their potential income is low. Moreover, the positive impact of education on income is more evident for migrant workers. This explains why household heads with the highest level of education have a higher propensity to migrate.