Autores: Eldon V. Ball, Carlos San Juan, Camilo Ulloa. Ewepa 2011 - XII European Workshop on Efficiency and Productivity Analysis - Verona - Italy
Tuesday 30 October 2012, by Carlos San Juan
Efficiency in Agriculture III
Several recent studies of the agricultural sector provide evidence of convergence of total factor productivity (TFP) across states. McCunn and Huffman (2000) found evidence of "catching-up" in levels of TFP (i.e., — convergence), although they rejected the hypothesis of declining cross-sectional dispersion (i.e., —convergence). Ball, Hallahan, and Nehring (2004) also found evidence of —convergence after controlling for differences in relative capital intensities.
The speed of convergence and whether it is transitory or permanent in nature play an important role in characterizing regional disparities in income and, hence, have important implications for the design of agricultural policy.
According to Barro and Sala-i-Martin (1992), there is —convergence if states with lower levels of productivity tend to grow faster than the technology leaders, and —convergence if the dispersion of their relative TFP levels tends to decrease over time. Thus, —convergence is a necessary but not a sufficient condition for sigma-convergence. An important implication of this result is that income inequality across states or regions may persist due to shocks (e.g., cyclical fluctuations in economic activity) that tend to increase dispersion.
This paper explores the relationship between the business cycle and convergence in levels of agricultural productivity. Two alternative explanations have been proposed in the literature to explain why convergence patterns may be related to the business cycle. The first is based on the pro-cyclical nature of the innovation process and the time lags between technological innovations and diffusion processes. According to this argument, productivity leaders tend to innovate more during periods of expansion in response to positive demand shocks. However, due to the existence of informational barriers, productivity followers, who tend to learn by imitation, postpone the adoption of innovations until economic downturns. The second explanation is based on the relation between competition and productivity. Productivity followers have more incentive to reduce their costs during downturns when negative demand shocks increase the probability that these firms will exit the industry.
Overall, these arguments point to faster rates of convergence during contractions in economic activity and to slower rates of convergence, or even divergence, during periods of expansion. Despite these arguments, few researchers have estimated the impact of the business cycle on productivity convergence.
An exception is provided by Escribano and Stucchi (2008). Using firm level data for the Spanish manufacturing sector, the authors test the catch-up hypothesis across different phases of the business cycle. They find strong evidence in support of the innovation-imitation hypothesis. Firms tend to diverge during periods of expansion and to converge during recessions, a result of both time lags in the diffusion of technical information and the pro-cyclical nature of innovation.
In this paper, we closely follow the methodology of Escribano and Stucchi (2008). First, we test the catchup hypothesis using a model specification that ignores the business cycle (i.e., the benchmark model). Then we investigate the possible impacts of the business cycle on the convergence process.
We find strong evidence of "catching-up" across the business cycle. Moreover, the speed of convergence was greater during periods of contraction in economic activity.