Human capital and convergence : theory, estimation and applications
In growth theory, convergence analysis tries to answer three fundamental questions "Are poor countries catching up with richer ones How quickly And what are the determinants of this process" This thesis deals with issues that are relevant to all these questions. It begins by setting out the key theoretical contributions to the analysis of the role of human capital in growth and convergence. Secondly, attention is turned to the way that convergence is estimated from data. The econometric techniques used in the convergence literature usually assume that shocks are uncorrelated across countries. We claim that this is unlikely for most data sets and investigate the use of an estimator so far ignored, namely the annual panel estimator where shocks are allowed to be correlated. Our analysis indicates that this estimator is more efficient than conventional ones for plausible values of cross-country error correlation. The study then turns to the analysis of the third question. Although differences in human capital endowments and rates of investment have long been recognised as crucial elements for explaining observed GDP gaps, nevertheless, human capital proxies are rarely significant in growth regressions. In this study some possible solutions to this puzzle are explored. We estimate aggregate returns to education in Italy and Spain, and compare our results with the predictions of competing theoretical frameworks. In general, our empirical analysis identifies a positive role for human capital, and stresses the relevance of theoretical models in which human capital has a fundamental but indirect role in the catching up process. The final part of the thesis proposes a new methodology designed to estimate technology levels and to test whether part of observed convergence is due to technology convergence. The results seem to confirm the existence of technology catch-up among regions.