To what extent can disparities in compositional and structural factors account for the gender gap in unemployment in the urban areas of Kenya?

In recent years, there have been sharp changes in the Kenyan labour market. Most notably, Kenya has experienced a remarkable increase in female labour force participation in its urban areas over the period 1986 to 1998. The sharp increase in female LFPR has not been matched by an increase in their employment rate and consequently unemployment amongst women remains a pressing problem. In contrast, male unemployment rates are substantially lower and have not increased significantly over time. This paper uses data from two time periods, 1986 and 1998, to identify the factors that influence the likelihood of being unemployed and to examine why women are more vulnerable to unemployment than men are. Using a decomposition framework, the paper establishes whether the gender gap in unemployment is driven by differences in observable characteristics between women and men (a composition effect) or differences in the returns to these characteristics in the labour market (structural effect/discrimination). The analysis shows that the overall likelihood of being unemployed is heavily influenced by sex, marital status, household-headship and human capital characteristics such as experience and level of education. The decomposition estimates display that for both periods, gender gaps in unemployment are overwhelmingly, about 81 to 84 per cent, attributed to the composition effect.