Farm Size and the Adoption of Rice Production Technologies in Taiwan

In this paper, we test a conventional predictive model of the diffusion of rice blasticides, using data from Taiwan. The model has modest predictive power, accounting for one third of the variation in adoption (R2 = .334). The impact of farm size on the model is assessed under the assumption that it reflects the impact of social structural constraints on individual farmers. Except for the relationship between total production and adoption, farm size was found to have little effect on the model. This may have occurred because of the relatively homogeneous farm population in Taiwan. For homogeneous populations, predictive models, which incorporate individual and psychological variables, appear appropriate to the extent that they refer to the adoption of low cost, unsophisticated technologies.