The effect of label information on farmers' pesticide choice

Pesticide labels provide farmers with information about potential health and environmental consequences associated with their use. This study analyzed the effect of label information on farmers' herbicide choices. Two approaches were used as theoretical framework for the analysis: a hedonic model using the household production theory and a discrete choice random utility model. The empirical estimation of the models was based on a sample of U.S. soybean farmers. Characteristics of the herbicides used as explanatory variables included health and environmental characteristics displayed on herbicides labels and efficiency measures calculated using relevant studies from the agronomic literature. Models estimated using information available on the labels were compared with a model estimated with variables obtained from MSDS's. This comparison was done to assess the assumption held in previous studies that farmers have a detailed and complete understanding of all the scientific measures used to evaluate the human and environmental risk of pesticide use (e.g., LD50 values). Finally, results from the estimated models were used to analyze the consistency of herbicide risk indices among alternative model specifications.;The results suggest that health and environmental statements displayed on pesticide labels (which generally reflect higher level of risk) reduce significantly the probability of selection, although, they seem to be less important than the herbicide production characteristics and cost. It was estimated that farmers are willing to pay ;Finally, the statistical results suggest that farmers' understanding of the human safety and environmental characteristics of herbicides is more in line with the information displayed on the labels than with the information contained in the MSDS's. This result has important implications for estimation and use of pesticide risk indices. In fact, it was shown that risk indices estimates are very sensitive to model specifications based on different assumptions regarding farmers' knowledge of herbicides' human and environmental characteristics.