مدل سازی عوامل موثر بر رشد TFP

We investigate the determinants of TFP growth of Italian manufacturing firms. Using stochastic frontier techniques, we consider three approaches for taking into account the influence of external factors, i.e., the determinants or drivers of growth. First, in our novel approach external factors may influence the technological progress, that is the shift of the frontier. To model this possible unexplored effect, we extend the standard time trend model to make it a function of the external factors. Then, following more standard approaches, we model external factors as either influencing the distance from the frontier, i.e., inefficiency, or the shape of the technology. Using a sample of manufacturing firms in 1998–2003, we find that technological investments and spillovers, human capital and regional banking inefficiency all have a significant effect on TFP growth.

The influence of external (or exogenous, environmental) factors in stochastic frontier models has been modeled with two alternative approaches. One assumes that the external factors influence the shape or structure of the technology, i.e., how conventional inputs are converted to outputs, while the other assumes that they directly influence the degree of technical efficiency, i.e., the efficiency with which inputs are converted into outputs (see, e.g., Coelli et al., 1999 or Kumbhakar and Lovell, 2000). In the literature on productivity measurement, however, no contribution explicitly considers the impact of environmental factors on the technological change, i.e., on the shift of the technological possibilities over time.

In this study we combine growth accounting with efficient frontier techniques to empirically investigate the determinants of output growth using data on Italian manufacturing firms. By applying stochastic frontier techniques, we introduce some methodological improvements to the existing empirical literature by modeling the effects of external factors on technological progress. While some of the external variables often used in this kind of studies might suffer from endogeneity bias, those we are mostly interested in (e.g., R&D spillovers, infrastructures, and regional bank inefficiency) are defined at a more aggregate level and thus do not suffer from these problems. Our results show that technology spillovers, technology investments, human capital and regional bank inefficiency are significant and economically relevant. Employing our specific dataset we fail to reject the model where external variables affect technological catch-up, i.e., efficiency or distance from the frontier, from which technological progress emerges as being quite weak and slightly decreasing over time.