Predicting conservation choice: Beyond the cost-minimization assumption

Models of residential energy conservation choice typically require the assumption that consumers make conservation choices so as to minimize life-cycle costs. Empirical evidence has shown discount rates to vary widely, which has cast doubt on the cost-minimization assumption. We propose an alternative model of conservation choice that does not assume minimized life-cycle costs, and that accounts for non-financial factors influencing choice. Perceived comfort effects, social norm influences, perceived costs and savings, and time required for installation are identified as causal influences on conservation choice, and these factors are measured using a household survey. The resulting model is shown to predict conservation choice correctly in 74.7% of the households. A new statistical technique (CART) is used to show that perceived savings have little predictive power.