No question has perhaps attracted as much attention in the economics literature as "Why are some countries richer than others?" Much of the current work traces back to Solow's classic work. Solow's seminal paper suggested that differences in the rates at which capital is accumulated could account for differences in output per capita. Subsequently, human capital disparities have been given a central role in the analysis of growth and development. However, the best recent work on the topic reaches the opposite conclusion. Many argue that most of the cross-country differences in output per worker are not driven by differences in human capital (or physical capital); rather they are due to differences in a residual, total factor productivity (TFP). This project revisits the development problem. The key difference between this work and previous analyses lies in the measurement of human capital. The standard approach - inspired by the work of Mincer - takes estimates of the rate of return to schooling as building blocks to directly measure a country's stock of human capital. One problem with this procedure is that it is not well suited to handle cross-country differences in the quality of human capital. Following the pioneering work of Becker and Ben-Porath, this project models human capital acquisition as part of a standard utility maximization problem. This set up is flexible enough so that individuals can choose the length of the schooling period -which is used as a measure of the quantity of human capital- and the amount of human capital per year of schooling and post-schooling training, which is a measure of quality. This project uses evidence on schooling and age-earnings profile to determine the parameters of the human capital production function. The project then computes stocks of human capital as the output of this technology, evaluated at the individually optimal choice of inputs given the equilibrium prices.
The key finding is that when the extended neoclassical model is calibrated to match U.S. data, the model requires small differences in TFP to account simultaneously for the differences in levels of schooling and income differences between the top and the bottom deciles of the world income distribution. These quantitative results are at odds with the modern view of the role of TFP in explaining underdevelopment. The conventional wisdom is that cross-country differences in human capital are small and that consequently differences in TFP are large. Hence policies that achieve small changes in TFP cannot have large effects on output per capita. Moreover, using the Mincer approach that takes schooling as exogenous, those models effectively give up on trying to understand the impact of TFP on human capital accumulation. This project finds that, the elasticity of output per worker with respect to TFP is approximately 8. Thus, a mere 10% permanent increase in relative TFP, is predicted to increase output per worker in the long run by 80%. The model suggests that there are huge payoffs to understanding what explains productivity differences.
Broader Impacts: The research results inform development policy by addressing the following question - if human capital is important for development, as the preliminary results suggest, what policies are best suited to increase the stock of human capital? There are two related questions - what should be the size of governmental intervention as a function of the level of development? Second, how should the available resources be allocated to subsidies among early childhood education, schooling and job training? These questions are addressed in an environment wherein the key friction is the inability of parents to borrow against the future income of their children.
