One of the main strategies for reducing energy demand—and thereby carbon emissions— is increasing energy efficiency. In the Netherlands, energy efficiency policy and energy reduction targets in the residential sector are based on the theoretical energy consumption of buildings, which often over- or underestimates the measured energy consumption. The difference between the actual energy consumption and the calculated energy consumption of a building is referred to as the energy performance gap (EPG). As a consequence of this discrepancy, the estimated energy savings are inaccurate and the energy-saving targets are unattainable. Narrowing the EPG is necessary to develop more accurate estimates of energy savings and realistic targets, and to improve the effectiveness of energy reduction policies and campaigns. Most studies on the EPG have focused predominantly on data from the social rental housing sector, failing to represent the national distribution of home ownership type. At the same time, homeowners and tenants have been shown to behave differently regarding energy consumption.
This thesis investigates the effect of home ownership on the actual natural gas consumption and the EPG by descriptive statistics, correlation analysis, and multiple linear regression on a representative sample of the Dutch housing stock. The multiple regression analysis controls for building and occupant characteristics that are expected to influence the actual gas consumption and the EPG, in order to measure the ceteris paribus effect of ownership type on the actual gas consumption and the EPG.
The results show that ownership type does not have a practically significant effect on actual gas consumption or the EPG, while controlling for building and household characteristics. However, without controlling for these factors, there is a moderate positive correlation between home ownership and actual gas consumption, and a weak positive correlation between home ownership and the magnitude of overpredictions. This suggests that observed differences in gas consumption or EPG between ownership groups may be explained by building and household characteristics, rather than by potential behavioral differences. Specifically, the positive correlation between home ownership and actual gas consumption can be explained by the larger floor area, type of buildings, higher income, and larger household size. The positive correlation between home ownership and the size of overpredictions is explained by type of building and larger floor area. Thus, there are no major differences in energy consumption behavior between homeowners and tenants that cause large differences in their actual gas consumption or EPG. Nevertheless, the distinction of ownership type may still be of practical use to policymakers. Targeting homeowners could be an efficient way to promote energy-saving measures in the largest and highest energy-consuming dwellings.