Who emits most? Associations between socio-economic factors and UK households' home energy, transport, indirect and total CO2 emissions

a r t i c l e i n f o Does the association between household characteristics and household CO 2 emissions differ for areas such as home energy, transport and indirect emissions? This question is policy relevant because distributional implications of mitigation policies may vary depending on the area of emissions that is targeted if specific types of households are likely to have higher emissions in some areas than in others. So far, this issue has not been examined in depth in the literature on household CO 2 emissions. Using a representative UK expenditure survey, this paper compares how household characteristics like income, household size, education, gender, worklessness and rural or urban location differ in their association with all three areas as well as total emissions. We find that these associations vary considerably across emission domains. In particular, whilst all types of emissions rise with income, low income, workless and elderly households are more likely to have high emissions from home energy than from other domains, suggesting that they may be less affected by carbon taxes on transport or total emissions. This demonstrates that fairness implications related to mitigation policies need to be examined for separate emission domains. Since households contribute substantially to the UK's total emissions – around 74% according to Baiocchi et al.'s (2010) estimate, including indirect emissions – a reduction of household emissions is essential for meeting the UK's carbon reduction targets. Additional climate change mitigation policies (in the following " mitigation policies ") will thus be needed to reduce household emissions. To examine potential fairness implications of these policies, we need to analyse the distribution of emissions across household groups. Two points are particularly relevant here: first, if factors other than income are associated with emissions, mitigation policies that put a price on emissions will have varying effects on different types of households independent of their income. Characteristics other than income thus need to be considered in distributional analysis of emissions. However, some characteristics such as income and education or income and rural/urban location are related to each other. Whilst bivariate analysis may find that each of these characteristics is associated with emissions, multivariate analysis is required to determine whether or not characteristics such as education or rural location remain associated with emissions after income is controlled for. So far, only few studies employ multivariate analysis to control for relationships between different factors, but examples …