Evaluating homeowners’ retrofit choices – Croatian case study

• Published in: Energy and buildings 2018-07, Vol.171, p.40-49
• Main Authors: Matosović, Marko, Tomšić, Željko
• Format: Article
• Language: English
• Subjects: Case studies; Data processing; Discrete choice; Energy conservation; Energy costs; Energy efficiency; Energy efficiency in buildings; Energy efficiency policy; Energy policy; Homeowners; Housing; Income; Investment; Investment decision; Power efficiency; Refurbishment; Regression analysis; Regression models; Residential areas; Retrofitting; Subsidies
• ISSN: 0378-7788; 1872-6178
• DOI: 10.1016/j.enbuild.2018.04.020

•The national energy efficiency retrofit scheme in Croatia was analyzed.•No difference in willingness-to-pay among different income classes was found.•Final investment cost determines whether the household will invest or not.•A significant degree of free-riding was found for windows replacement activities. In this paper, a model of homeowners’ energy efficiency retrofit choices is presented based on the database of a project implemented by the national energy efficiency retrofit scheme in Croatia during 2015 and 2016. The analysis was performed on the cleaned dataset of 4610 implemented projects in privately owned family houses. We have analyzed available data to answer how much money consumers are willing to pay for energy refurbishment and how this corresponds to household characteristics. We have developed linear regression models for estimating costs and energy savings of different measures and have further applied multinomial and nested discrete choice models on investment choices made by homeowners. The resulting models provide estimations of willingness to pay for energy efficiency refurbishment in private family houses in relation to income class the owners belong to. The results show that homeowners in all income classes have a similar level of willingness-to-pay for similar energy efficiency measures, with main differences being their capability to invest. With the current level of subsidies, only three of four observed measures show a significant level of free-riders. This suggests there could be a better allocation of public funds, where less money could be spent for subsidies to generate a similar amount of savings. Furthermore, the proposed model replicates existing experience and provides insight into inconsistencies which appear when the effectiveness of implemented measures is analyzed.