US productivity in agriculture and R&D

This paper examines the impact of R&D on multifactor productivity in the U.S. agricultural sector over the 1910-1990 period. We use the Bennet-Bowley indicator to measure agricultural productivity based on a multiple output-multiple input technology. We demonstrate the relationship between the price dependent Bennet-Bowley indicator and the Luenberger productivity indicator which is constructed from directional distance functions without requiring price information. These performance measures are dual to the profit function which arguably makes them especially useful in the agricultural setting. We employ time-series techniques to investigate the effect of R&D on the pattern of productivity growth. We find that we cannot reject the presence of a cointegrating relationship between the two series and that productivity growth in the U.S. agriculture responds positively to R&D expenditure with a lag of between four and ten periods.