The proposal presents a systematic and complete agenda to study the gender labor market discrimination in Kenya. This agenda will allow us not only to properly detect gender discrimination but also fully understand its consequences. I seek to do this by using three different sources of information. The first source of information is a new and rich longitudinal data set containing detailed labor market histories of a representative sample of Kenya. My second source of information will be a new survey, carefully designed to measure cognitive and non-cognitive abilities at the individual level. Finally, I plan to use the information from an audit study in which I will send written applications to real job advertisements. Altogether, these sources of information will provide a unique opportunity to (i) re-analyze the previous findings in the literature and (ii) go beyond what have been done in Kenya to detect and understand the gender labour market discrimination phenomenon. With the first source of information I will correct the estimates of gender discrimination by adding controls for the labor history of the workers, marital and maternal histories, family characteristics (parents’ education, number of siblings, etc.) and school performance and quality. The second source of information will be the first attempt in Latin America to measure cognitive and non-cognitive abilities and their effects on economic outcomes. Recent outstanding theoretical and empirical literature centers its attention on these issues. Finally, with the audit study (the third component of my agenda) I will be able to follow the field experiment approach to study the presence of gender discrimination in the labor market. In this experiment I will send pairs of identical written applications differing only in the applicant’s gender. Additionally, I will explore the importance of the neighborhood effects and beauty by introducing the dimension of geographical location and including pictures in the job applications.
1.0 CHAPTER 1.
Gender and social discrimination in the labour market are the one key issue on public policies’ discussion in Kenya. This proposal contributes to fill this lack of evidence. It doesn’t matter how much has been done to study labour market discrimination, racial, ethnic or gender, the issue of detecting is still unsettled. In the usual regression analysis there are several problems of unobservable variables that clearly bias the result and, on the other hand, the experimental studies have been under discussion for not correctly measuring discrimination. In Kenya, despite the fact that average years of schooling of Kenya female workers are not statistically different from those of male workers, pure average wages of male workers are 25% higher. In fact, previous studies suggest that gender discrimination is factors in determining wages in the Kenyan labour market. Estimates of the Blinder-Oaxaca decomposition give “residual discrimination” a significant participation on the total wage gap. The evidence also shows stable and systematic differences in the returns to education and to experience by gender along the conditional wage distribution. Additionally, it has been shown that “residual discrimination” is higher for women with more education and experience. Furthermore, Kenyan female labour force participation is particularly low. This is lower for married women and in fact, the higher participation is found in separated or divorced women. This later fact may be interpreted as evidence of women preferences for non-market activities 1.2 The statement of the problem
This proposal presents a systematic and complete agenda to study the gender labour market discrimination in Kenya. This agenda will allow us not only to properly detect gender discrimination but also fully understand its consequences. I seek to do this by using three different sources of information. The first source of information is a...