مجموع تفاوت بهره وری عامل: فن آوری مناسب در مقابل بهره وری

Recent development and growth accounting studies have established that total factor productivity is an important source of cross-country differences in income levels and growth rates. This paper makes two contributions. First, it examines the sensitivity of the development accounting results to the Cobb–Douglas specification of the production function. Second, within the Cobb–Douglas framework, it weighs evidence of the two alternative explanations of total factor productivity differences: the inefficiency view and the appropriate technology view. To accomplish these tasks, the world production frontier is estimated using a nonparametric deterministic approach known as data envelopment analysis. I find that the fraction of income differences explained by physical and human capital increases from 32% to 55% when departing from the Cobb–Douglas assumption. There is also evidence consistent with the appropriate technology view; countries with an inadequate mix of inputs are unable to access the most productive technologies. Moreover, the world technology frontier appears to be shifting out faster at input combinations close to that of the R&D leader. However, inefficiency appears to be the main explanation for low incomes throughout the world; it explains 43% of output variation in 1995, and its importance has increased over time.

Recent empirical literature on economic growth has investigated the proximate causes of the enormous differences in per-capita incomes across countries. In most cases, the motivating question is the role of measured factors of production, such as physical and human capital, relative to that of unobserved total factor productivity (TFP). Most researchers assume that output is given by the Cobb–Douglas production function, y=Akαh1-αy=Akαh1-α, and decompose cross-country variation in the levels of output per worker (y)(y) into parts attributed to the variation in factors (kk and hh) and TFP (A)(A). The results usually show that differences in incomes are largely a consequence of differences in TFP (Klenow and Rodriguez-Clare, 1997, Hall and Jones , 1999, Easterly and Levine, 2001 and Caselli, 2005). Klenow and Rodriguez-Clare report that about 40% of income differences can be explained by human and physical capital, while the remaining 60% is due to TFP. When applied to growth rates, this decomposition is the cross-sectional version of familiar growth accounting (Solow, 1957), and the results are similar to those for income levels. Easterly and Levine (2001) report, among what they call “new stylized facts of growth,” that around half of the average per-capita output growth and 90% of the cross-country variation in growth rates are explained by TFP. This leads them to conclude that “the residual (TFP) rather than factor accumulation accounts for most of the income and growth differences across nations.”1

Recent empirical growth literature has established, through development and growth accounting studies, the importance of total factor productivity for understanding differences in income levels and the variation in growth rates across countries. This paper attempts to use a new empirical approach to shed light on two issues. First, I examined how sensitive the findings of the development accounting literature are to the assumption of Cobb–Douglas production function. Second, within the Cobb–Douglas framework, I looked for evidence of the two alternative explanations of total factor productivity differences: The inefficiency view and the appropriate technology view.