عوامل بهره وری کل درون زا و تفاوت درآمد بین کشوری

Using a class of endogenous growth models that exhibit international spillovers, we show that most of the cross-country differences in output per worker are explained by barriers to the accumulation of rival factors (physical and human capital) rather than by barriers to the accumulation of knowledge. This is shown theoretically, by comparing models with exogenous and endogenous TFP, and quantitatively by using a carefully calibrated version of the model. The main finding is that barriers to the accumulation of physical and human capital explain up to 64% of income gaps relative to the US.

In his classic study of the mechanics of economic growth and development, Lucas (1988) famously asks: “Is there some action a government of India could take that would lead the Indian economy to grow like Indonesia's or Egypt's? If so, what, exactly? If not, what is it about the ‘nature of India’ that makes it so?” Seminal papers by Parente and Prescott (1994), Klenow and Rodríguez-Clare (1997), and Hall and Jones (1999) have shown that differences in total factor productivity, or TFP, are key for understanding income differences, and Prescott (1998) has called for a theory of TFP. These studies suggest that differences in physical capital per worker and human capital cannot account for income differences. Their main policy implication, which constitutes an answer to Lucas’ question, is that the elimination of barriers to accumulation of knowledge is key for economic development.

Countries with higher factor intensities X tend to exhibit higher total factor productivity A. In our database, the covariance term accounts for around 35%35% of the cross-country dispersion of output per worker. What explains this large covariance? Models with exogenous TFP, such as the Solow model, are silent about the nature of this observed relationship. However, the covariance term is crucial for understanding the ultimate causes of cross-country income differences. In fact, two landmark studies by MRW and KR (1997) arrive to opposite conclusions about the relative importance of X versus A partly because they impute the covariance term differently. Their way to assign the covariance is arbitrary because their underlying model, the neoclassical growth model, has no predictions about this covariance. In contrast, endogenous growth theory provides a natural explanation for the covariance term. If technological progress is costly, then economies with more factors abundance can undertake more R&D activities, accumulate a larger stock of knowledge, and become more efficient. Thus, growth theory suggests that the covariance term, or part of it, must be assigned to X.