An estimation of the impact of technology on agricultural output in the Southeast

The objectives of this study were to estimate an agricultural production function for the thirteen southeastern states and to measure the timepath of the effect of technology by estimating a distributed lag function. The procedure involved estimating a Cobb-Douglas type production function using ordinary least squares regression. The impact and timepath of public research and extension expenditures, which was used as a proxy variable for technology, was estimated using a general polynomial distributed lag technique. The study used time series and cross sectional data from secondary sources. State data for research and extension expenditures covered the period from 1930 through 1968 while the other variables were measured from 1949 through 1968. The general polynomial distributed lag technique did not restrict the shape of the lag distribution. Lags were varied from 0 to 20 years. Selection of the "best" lag was based on; (1) R2, (2) the signs, magnitudes, and significance of structural parameter estimates, and (3) the reasonableness of the estimated lag distribution. Results indicated that based on the selection criteria a lag of nine years was the optimum lag with not much difference among the lags of seven to ten years. It was hypothesized that the lag distribution would have weights that increased to a maximum and then decreased; however, the results indicated a distribution with equal weights for all years. Results indicated that returns to investments in technology were quite large. Even if half of the influence is attributed to private research and extension activities, there was a return of nearly 300% to public investments in research and extension.