The role of the accumulation of human capital to per capita income growth has been sharply debated among economists and policy makers. One open question of this debate is how to measure human capital. The standard approach is to use the average years of education of the labour force or the school enrolment rates as proxies for the stock of human capital. However, formal schooling achievement does not fully capture all the human capital stock. In fact, other forms of human capital accumulation are unmeasured. Also, it is assumed that the productivity differentials among workers with different levels of schooling are proportional to their years of education. In order to solve these problems, we develop the Mulligan and Sala-i-Martin’s measure of human capital based, on labour income. This measure has some nice properties: is consistent with variable elasticities of substitution across types of workers, and does not impose all workers with the same amount of education to have the same amount of skill. It is also allowed for changes in the relative productivities over time and across different economies. We compute the index at the firm level and, finally, and we compare the evolution of our index with the evolution of average years of education for the Portuguese regions, highlighting the shortcomings of the latter measure of human capital. Keywords: Human capital, wages, income based measures JEL Classification: E24, J24, I21
