Application of cumulative prospect theory in understanding energy retrofit decision

Abstract

Retrofitting residential buildings can help mitigate the effects of climate change. Cognitive biases are systematic deviations from rationality in decision making and can lead to inaction, delay, and uncertain decisions. Understanding the cognitive biases involved in residential renovation decisions and developing interventions to overcome them can help increase residential renovation rates. Despite their importance, few studies have examined the impact of cognitive biases on energy retrofits. The question addressed in this study is: “Can accounting for cognitive biases improve the prediction of homeowners’ actual investment decisions, and how can the outcomes be used to recommend potential behavioural interventions?”. Expected Utility Theory (EUT) and Cumulative Prospect Theory (CPT) are compared to evaluate which model(s) more accurately describes actual decision-making behaviour regarding energy retrofits. The EUT assumes a rational decision maker. The CPT is a quantitative model that assumes a decision-maker operating under risk and uncertainty and subject to the cognitive biases of reference dependence, loss aversion, decreasing sensitivity, and probability weighting. The influences of cognitive biases on energy retrofit decisions can be quantified if the relative performance of CPT versus EUT is more accurate. The data for these analyses come from housing surveys conducted in the Netherlands in 2012 and 2018, which also collected data on energy modules. 2,784 and 2,878 homeowners were surveyed, respectively. The model is validated by estimating the coefficients of EUT and CPT and identifying the similarities and differences between the results of the two datasets. Before estimating the parameters, four household clusters are identified using grey relational analysis and the K-Means cluster. For the first time, the EUT and CPT parameters are estimated for four clusters and two energy retrofits, double glazing and insulation, using a genetic algorithm because the equations are nonlinear. The results confirm that CPT provides a better description of the actual decision behaviour than EUT using the two previously established initial values of Layard et al. (2008) and Häckel et al. (2017) as well as the parameters estimated by the genetic algorithm. In the latter case, CPT correctly predicts the decisions of 86% of homeowners to renovate their homes to be energy efficient or not. EUT, on the other hand, overestimates the number of decisions to renovate: it incorrectly predicts retrofit for 52% of homeowners who did not renovate for energy efficiency reasons. Using the estimated parameters of CPT, the cognitive biases of reference dependence, loss aversion, diminishing sensitivity, and probability weighting can be clearly seen for different target groups. The groups with the highest average incomes and house values had the highest loss risk aversion parameters. These households invested more in installing insulation and double glazing.