Investment Behavior and Energy Conservation

Binary logit and bivariate probit models were used to investigate the investment behavior of farmers relative to two energy-conserving assets, heat-recovery systems and precoolers. The bivariate probit procedure was useful in correcting for self-selectivity bias. Holdout samples and cross-validation procedures were used to develop true model statistics. Farm size, educational level of the operator, and the type of milking system in use were the important factors influencing investment behavior. As the economic costs of electrical-energy gener-ation rise and the environmental costs become more evident, energy conservation becomes a clear al-ternative to building more nuclear or sulphur-emit-ting plants. The first step to conservation is development of appropriate technologies. How-ever, after such technologies are developed, adop-tion may be very slow when the technology is embodied in durable capital items such as struc-tures or machines. An understanding of the in-vestment behavior of those who could be expected to use energy-conserving equipment or structures would be of value in predicting or attempting to modify rates of adoption and, thus, energy use. Attempts to quantify the behavioral relationships connected with investment at the firm (farm) level have generally focused on large field crop and grain storage equipment. Using a probit model, Hill and Kau found farm size, farm type, tenancy, operator age, and specific corn crop variables significant in determining investment in grain dryers in Illinois. Similarly, farm size, tenancy, corn production, and corn use variables were important determinants of grain bin investment in a tobit study by Dixon, Hill, and Saffell. A more recent multivariate analysis of tractor and combine investments found soil type, value of machinery inventory, operator age, and education to influence machinery investment decision making (Johnson, Brown, and O'Grady). In a simulated investment environment, Gustafson, Barry, and Sonka found that structural character. The authors are professor and former research support specialists, re