Energy poverty indicators: Inconsistencies, implications and where next?

Energy Poverty in all climates and economies is recognised as problematic, whether related to insufficient warmth or inadequate cooling, making its definition and the identification of which households are most at risk of central importance. Different definitions of Energy Poverty identify not only varying numbers of households at risk, but households with different characteristics, and provide an ambiguous basis for both academic studies and policy design. To confirm and illustrate these universal issues we use a large dataset from the United Kingdom to demonstrate how three commonly used indicators, including two based on official measures, can imply very different targeting policies. In particular, our rich panel dataset confirms and extends the finding that older households are more likely than their younger counterparts to be identified as energy poor when using expenditure-based indicators, but are less likely to self-identify as unable to afford adequate warmth. Similar discrepancies between Energy Poverty indicators relate to the presence of children in a household, household size, and geographical location. This methodological and historical illustration from the UK, which has recently introduced its third official definition of Energy Poverty in a decade, is relevant to any country seeking to measure and address the plight of households who struggle to achieve desired in-home temperatures. The observed contradictions emphasise the importance of obtaining direct evidence on the fundamental underlying issues encapsulated by Energy Poverty. Where in-home temperatures are the focus this means complementing the current measures with information on achieved temperatures and households' temperature preferences. •Compares perception-based and expenditure-based Energy Poverty indicators in the UK.•Identifies inherent difficulties in interpreting results from different metrics.•Little overlap between expenditure and perception-based indicators is observed.•Each indicator identifies different household types as energy poor.•In-home temperature and temperature preference data are key for future research.