Measuring Aggregate Human Capital

We construct a set of human capital indexes for the states of the United States for each Census year starting in 1940. To do so we propose a new methodology for the construction of index numbers in panel data sets. Our method is based on an optimal approach by which we choose the `best' set index numbers by minimizing the expected estimation error subject to some search constraints. Some of the empirical findings are that the stock of human capital in the United States grew twice as rapidly as the average years of schooling and that human capital inequality across states went up during the 1980s (while the dispersion of schooling actually fell). We conclude that using the average years of schooling for the empirical study of existing growth models may be misleading.