Education, Inequality and Earnings in Ghana

Abstract

This thesis uses the fifth and sixth rounds of the Ghana Living Standards Survey (GLSS) conducted in 2005-2006 and 2012-2013, respectively, to examine the themes of education, inequality and earnings in Ghana. Three key objectives are pursued. The first objective is to assess how educational attainment, measured by the average years of schooling, relates to inequality in education. The second objective is to examine how education influences inequality in earnings through an assessment of the heterogeneous returns to education. The final objective is to provide an analysis of the effect of educational mismatch (over- and under education) on earnings and the return to education. Consistent with these objectives, this thesis is organised into three empirical chapters. The first chapter examines the nature of the relationship between educational attainment in the form of average years of schooling and the various measures of inequality in educational attainment based on the educational Kuznets hypothesis. By employing a pseudo-panel estimation approach, the study finds evidence of a non-linear relationship between the mean years of schooling and educational inequality with a threshold occurring around nine years of education. The results suggest that advances in educational attainment may not necessarily be associated with a reduction in educational inequality in Ghana, but may increase it depending on where the increase occurs. That is, while education expansion from the lowest percentiles of the educational attainment distribution up to the median is found to be associated with declines in educational inequality, an expansion in the attainment of persons in the higher percentiles is associated with a rise in educational inequality. Also, increases in the proportion of people with secondary, post-secondary and tertiary education were found to be associated with rising educational inequality in absolute terms over the period covered by the analysis. Public policy should, therefore, be focused on increasing educational attainment particularly at the basic education level given the benefits of increasing human capital and making its distribution more equal, as rising inequality in educational attainment has potentially adverse consequences for economic growth and income inequality. The second chapter assesses the heterogeneous returns to education across individual income groups within the human capital framework, given the implications of such patterns for the inequality-reducing role of education. The analysis is conducted within the operationalised sectors of employment: the public, formal private, self-employment and agriculture. Recent studies on Ghana that have assessed whether there exist variations from the mean returns to schooling have resulted in mixed outcomes. While some studies have found variations from the mean to exist with the returns to education being highest at the top decile of the earnings distribution and declining down the distribution, others have found the returns to education to be highest at the bottom decile of the earnings distribution and decreasing as one moves up the earnings distribution. Quantile regression is used to examine whether there is individual heterogeneity in the returns to schooling. The results reveal that the impacts of education on earnings are not equal across the conditional earnings distribution in the pooled employment sectors or the public and formal private sectors of employment. However, such patterns are not observed for self-employment. The returns to education are highest in the bottom decile of the earnings distribution and lowest in the top decile for the public sector, indicative of the fact that within the sector, education has a more equalising effect on earnings. For the formal private sector, returns are highest in top earnings decile and significantly lower in the bottom earnings decile, giving credence to the view that education may widen earnings inequality within the sector. For the pooled employment sectors, although education appears to potentially widen the earnings differences between workers with low-quality skills and those with high-quality skills in the GLSS 5 (2005-06); however, it tends to bridge the earnings gap between the least advantaged and the most advantaged in the GLSS 6 (2012-13). This change in the effect of education along the conditional earnings distribution over the period of analysis may be accounted for by a marked deficiency in the relative demand for highly skilled workers over the period rather than a significant increase in the demand for low skilled workers in the Ghanaian economy. Moreover, the observed uneven effects of education along the conditional earnings profile strengthens the view that public policy must seek to improve upon the quality of education, especially among the least advantaged. The final chapter provides empirical evidence on the effect of educational mismatch on earnings and the returns to education in Ghana. The study utilises a data-based criterion, being the modal measure, to determine the educational requirements of jobs in the two-digit classification for the GLSS 5 (2005-06) and the three-digit classification for the GLSS 6 (2012 13) of occupations. The analyses are conducted based on the economic sectors of employment (i.e. agriculture, industry and service), the employment status (i.e. wage and self-employment) and the institutional sectors of employment (i.e. public, formal private and informal sectors). The results suggest that over- and under-education exists in Ghana’s labour markets. While in the GLSS 5 (2005-06), 34 percent of workers are classified as over-educated and 9 percent under-educated, in the GLSS 6 (2012-13), 42 percent and 12 percent of workers are classified as over-and under-educated, respectively. The incidence of over- and under- education also varies depending on the sector of employment. Furthermore, given other characteristics, over educated workers tend to earn more while under-educated individuals tend to earn less than University their co-workers who possess the job-specific level of education. Nonetheless, the over educated (under-educated) earn less (more) than workers with the same level of education and engage in jobs for which they possess the job-specific level of schooling. The occurrence of penalties associated with being over-educated may be indicative of a waste of resources perhaps due to government subsidization. Government policy can, however, be directed at enhancing skills utilization in enterprises. Some best practices include the adoption of occupational forecasting models to anticipate future skills needs and supplies and the setting up of sectoral and occupational skills councils to advise on policy.