Schooling, Skills and Segmentation: Effects of Childhood Socioeconomic Status on Adult Earnings in Urban Kenya

Abstract

Over the past half-century, Kenya has pursued an ambitious agenda of educational expansion. Policies that formally extend and democratize access to schooling have enjoyed widespread public support. Programs like free primary education, which successive governments have attempted to implement three times, are considered harbingers of socioeconomic equity that present marginalized children with new pathways to professional success. These initiatives aim to upset socioeconomic reproduction by providing children from uneducated, low-income households with the tools necessary to succeed in a modern economy. In practice, however, these programs have strained the government’s fiscal and organizational resources. Expansion has often arrived at the cost of quality, meaning that children leave school with variable skills and academic knowledge. Kenyan industrialization has simultaneously stagnated, spawning a massive informal urban labor market characterized by low pay and low human capital barriers to entry. Discriminatory processes like nepotism and bribery frequently govern access to the high-skill, lucrative formal sector. The confluence of these educational and economic realities calls into question whether simply expanding educational access can actually mitigate socioeconomic reproduction in practice. Using data from the World Bank’s 2013 STEP Skills Measurement Program, I discovered substantial earnings gaps between urban Kenyan workers based on their parents’ level of education and their affluence at age 15, two indicators of childhood socioeconomic status. Higher levels of parental education and childhood affluence were both associated with substantial earnings premiums. I attempted to deconstruct these disparities to determine how differences in educational attainment (years of schooling) and literacy contribute to these gaps. I also investigated whether workers from more advantaged backgrounds enjoy greater access to formal labor markets than those from less advantaged backgrounds, given the same level of educational attainment and literacy. I first found that both parental education and childhood affluence were positively associated with educational attainment. I next discovered that, when controlling for educational attainment, childhood affluence remains positively correlated with literacy, though parental education does not. Third, my analysis revealed that educational attainment and literacy contribute independently to earnings in urban Kenyan labor markets. Differences in educational attainment can explain the entire effect of parental education on earnings. Differences in educational attainment and cognitive skills can both partially explain the earnings advantage associated with higher childhood affluence. Finally, I found that higher levels of parental education and childhood affluence do not increase a worker’s probability of entering the formal sector when controlling for school attainment and literacy. Even when controlling for school attainment, literacy, and labor market sector, childhood affluence remains positively correlated with earnings. Overall, I conclude that when attempting to reduce socioeconomic reproduction by expanding education, the Kenyan government must consider both the quantity and quality of schooling to maximize future earnings payoffs for students.