Interpretation and Assessment of Benefit-Cost Findings

Abstract

Agricultural production, input use, and productivity have been evolving over time, with substantially different patterns among U.S. states; so, too, has the pattern of spending on agricultural research and extension by the federal and state governments. In our econometric models linking these patterns of R&D spending and agricultural productivity, we have imposed a great deal of structure on the research lag and state-to-state spillover relationships. So long as they are appropriate and do not lead to estimation bias, these restrictions are helpful in reducing the number of free parameters to be estimated and improving the precision with which they are estimated. The resulting estimates indicate strong linkages between research and extension spending and productivity, and high payoffs to past investments. They also signal slower future productivity growth, especially if the past slowdown in the rate of growth in spending on farm-productivity-oriented research and extension spending will be sustained into the future.