Does education signal ability in Ghana? An analysis comparing wage earners with the self-employed

When education signals underlying abilities, income returns to education may not reflect true increases in productivity. This may be a particular concern in developing countries, where education is often prescribed as a major path to escaping poverty. Unfortunately, because education signalling occurs when underlying worker characteristics are difficult to observe, it is extremely difficult to obtain estimates of education signalling. This study uses Spence's (2002) model of signalling to develop a testable hypothesis: in the presence of education signalling, wage earners will see a higher return to education than the self-employed doing similar work. Using the 2005-2006 round of the Ghana Living Standards Survey, the study finds that returns to education are consistently higher for those in the wage-earning sector, which supports an education signalling hypothesis. Signalling appears to be more prevalent at higher levels of education and in large, unionised, and professional firms.