Using a decomposition framework and Kenyan data from 1986, 1998 and 2005, this study analyzes the factors associated with the likelihood of unemployment among the urban youth labour force and the disproportionately higher vulnerability to unemployment among female youth compared to male youth. Overall, the results indicate that household-headship, training, marital status and being male as opposed to being female are significantly correlated with the likelihood of being unemployed. Level of formal education appears less important while experience appears to be more important for female youth. After controlling for potential endogeneity of training results indicate that access to training/skills could help to diminish overall youth unemployment by about 58% and by 53% and 51% for females and males, respectively. The decomposition analysis indicates that the observed gender gaps in youth unemployment are largely explained by differences in average characteristics between female and male youth. Household-headship exerts the most positive effect in widening the differential. Over time, the combined positive effect of human capital variables declines sharply, thus narrowing the gap. Marital status increasingly limits young women from being employed, thus widening the gap. Overall, the analysis provides limited justification for employment discrimination in the youth labour market along gender lines.
