Graduate Labour Market Analysis in Malaysia

This thesis studies several aspects related to graduate employment in Malaysia. The first chapter examines graduate's transition from education to work which includes the analysis of the first destination choice after study and occupational types using multinomial logit model. Within each occupational category, we use Fairlie's non-linear decomposition technique to compute the differences in the participation rate between gender and ethnic groups. Women and Malay's under-representation in superior occupational types are largely due to their choice of less attractive courses that are associated with low market demand. The second chapter analyzes the wage differentials between the public-private sectors, gender and ethnic groups. The earning equation is adjusted to account for the sample selection bias due to the participation rate. We use the Oaxaca-Blinder decomposition technique to compute the wage differences between the groups. The difference in the sectoral wage is non-significant, and there is no evidence of wage differential in the public sector. Gender and ethnic pay gap only occur in the private sector where male and Chinese consistently earn higher. The third chapter explores another dimensions of graduate's transition from education to work. First, we found that higher ability graduates are more inclined to migrate in order to maximize their employment prospects as well as compensating for their superior human capital. Indeed, graduate's migration results in higher earning but not necessarily reduce education-job mismatch. Second, graduates who possess better characteristics took longer to obtain their first job but they ended up with superior occupational types and higher earning. Yet, social attributes such as being a male, a Chinese, or originating from an urban state increases the probability for faster transition from education to work. Third, using a pseudo-panel data and controlling for cohort heterogeneity shows that the remaining variables that significantly affect earning are family income and locality.