Beyond energy labels: estimating housing energy efficiency and energy underconsumption using administrative microdata

Abstract A reliable measure of the energy efficiency of housing is essential, both for evaluating the effectiveness of energy efficiency policies and for assessing energy poverty. Across the EU, Energy Performance Certificates (EPCs) or energy labels are commonly used for this purpose. However, these data are often outdated or incomplete, and only weakly correlated with actual energy consumption —a discrepancy known as the performance gap. As a result, EPCs are poorly suited for evaluating energetic housing quality or measuring energy poverty. We address these limitations by developing and implementing a model that estimates housing energy efficiency by combining EPC data with additional administrative resources. Our approach improves upon previous studies through richer data integration and more precise model calibration. The resulting model explains 51% of the variation in energy expenditure based on housing characteristics, compared to 40% when using EPCs alone. We demonstrate the model’s application in assessing energy poverty through the LILEQ indicator (i.e. Low Income, Low Energy Quality), showing that it correlates more strongly with survey-based consensual indicators of energy poverty (e.g. EU-SILC), than commonly used indicators based on actual expenditure (e.g. share of energy expenditure). Finally, we illustrate how the model can be used to detect energy underconsumption and monitoring housing quality over time. In summary, we present a calibrated, data-driven model of housing energy efficiency that outperforms EPCs and enables the development of higher-quality, policy-relevant measures of energy poverty and housing conditions.