Publications: ISAAA Briefs No. 14 - 1999
José Benjamin Falck-Zepeda, Greg Traxler, and Robert G. Nelson Department of Agricultural Economics and Rural Sociology Auburn University
Summary Introduction Background The Empirical Model Results Discussion Summary and Conclusions References Appendix
Summary We used an alternative data set from Plexus Marketing Research, Inc. and Timber Mill Research, Inc. to reestimate rent creation and distribution from the adoption of Bt cotton in 1996. This alternative data set allowed us to compare the sensitivity of the estimates for 1996 presented in a previous paper. Results from both estimations indicate that farmers gain between 43% and 59% of all rents created from the introduction and adoption of Bt cotton. In contrast, the innovators (Delta and Pine Land, and Monsanto) gain between 47% and 26% of all rents in 1996. Preliminary results from the estimation of rent creation and distribution for 1998 indicate that farmers and innovators share almost equally the rents created by adopting Bt cotton. Farmers gain 43% whereas the innovators gain 47% of total rents. Regionally, there were winners and losers from the adoption of Bt cotton in 1998. Regions with low adoption rates, such as California and Missouri, lost because farmers suffered a price reduction of cotton lint without having the benefits of the technology. We performed a sensitivity analysis to evaluate results by reducing the yield and/or cost change assumptions in half. In the worst-case scenario, where yield increases and cost reductions were reduced by 50%, farmers still captured 21% of the total rents, whereas the innovators gained 74% of total rents. Results for the three-year analysis have been fairly consistent. Farmers share the rents created by the technology almost equally with innovators, even when a monopolistic structure for the input market is assumed. Improving the reliability of these results requires modeling rents using an ex ante framework where risk is considered in the analysis. The naive estimations presented here and in our other papers need to be formalized into a model that introduces the characteristics discussed earlier. There is a need to validate results from surveys whose intention is to measure cost and yield benefits or losses due to technology. This is the only way that we can assure that measurements of the benefits and losses of the technology are measured accurately. Finally, there is an urgent need to quantify the environmental externalities, particularly to quantify the benefits of decreased pesticide releases into the environment. These benefits may be significant and are not currently included in the framework presented here. List of Tables
|
|||||||||||||||||||||||||||||||||