Measuring the Stock of Human Capital for Comparative Analysis

This paper summarizes the outcomes of the first phase of the OECD human capital project. In so doing, it shows the feasibility of applying the lifetime income approach to measuring human capital for comparative analysis, both across countries and over time. It also highlights the feasibility of applying the methodology to the categorical data (i.e. by 5-year or 10-year age group) that are typically available within the OECD statistics system, rather than to data by single year of age required by the original Jorgenson- Fraumeni methodology. The results in this paper indicate that the estimated value of human capital is substantially larger than that of traditional physical capital. Ratios of human capital to GDP are in a range from around eight to over ten across countries, broadly in line with those reported in a number of national studies. The distributions of human capital by age, gender, and education show that men dominate women in terms of their human capital holdings. In addition, people with higher education are better off than those with lower education, and the same is true for younger people compared to their older counterparts, although the detailed patterns vary across countries. Decomposition analysis of changes in the volume of human capital demonstrates that changes in population structure between men and women had little effect on the change of human capital per capita. While in all countries higher educational attainment contributed positively to the change of human capital per capita, this is not always sufficient to offset the negative effect of population ageing; as a result, the volume of human capital per capita appeared to have declined in some countries over the observed period. Finally, sensitivity analysis confirms that estimates of the value of human capital depend on the choice of the two key parameters, i.e. annual real income growth rate and discount rate, while within-country distribution of human capital and trends of the volume of human capital are less sensitive to these assumptions.