Economic growth in the long run

We present new data on real output per worker, schooling per worker, human capital per worker, real physical capital per worker for 168 countries. The output data represent all available data from Maddison. The physical capital data represent all available data from Mitchell. One major contribution is a new data set of human capital per worker, the foundation of which comes mostly from Mitchell. We provide original estimates of schooling per worker & per young worker. Using standard Mincer human capital construction, we find that variation in inputs can explain about 46% of the variation in long run living standards using standard covariance accounting. With capital intensity covariance accounting, variation in inputs explains about 25% of the variation in long run living standards. With intergenerational human capital accumulation with or without spillovers, we find that variation in inputs can explain at least half of the long run variation in living standards.