The different intensities of household consumption across European expenditure deciles can be attributed to a combination of two plausible causes: first, the composition of consumption baskets could systematically differ according to the level of household expenditure [@tobben_regional_2017]. Second, the energy and carbon intensity within individual final consumption categories could systematically differ according to the level of household expenditure. This second aspect is usually not captured in individual country footprint studies due to limitations of EE-MRIO models (homogeneous product assumption).
The reason we are able to capture some of the differences in energy and carbon intensity across European expenditure groups is the strong spatial clustering of household incomes in Europe (spatial autocorrelation). Therefore, national households are very unevenly distributed across European expenditure deciles resulting in different intensities that reflect efficiency differences in energy supply as well as final use (figure 1).
The reason we are able to capture some of the differences in energy and carbon intensity across European expenditure groups is the strong spatial clustering of household incomes in Europe (spatial autocorrelation). Therefore, national households are very unevenly distributed across European expenditure deciles resulting in different intensities that reflect efficiency differences in energy supply as well as final use (Figure 1).
The final consumption categories (housing, mobility, food, goods, and services) contributed very differently to the environmental footprint of European households in 2015 (Figure 3). On average, housing and mobility were the two largest categories, accounting for about two thirds of both the energy and carbon footprints. In addition, the sectoral footprint variation across the expenditure deciles was also highly variable (Figure 3). For housing there was very little difference between deciles in both the energy and the carbon footprint. The bottom four deciles even had higher carbon footprints from housing than most top deciles, which can be explained by the extreme differences in intensity shown in Figure 2. Mobility was the most unequal category, with footprints in the top decile `r mobility_co2eq_10_10` times higher than the bottom decile, corroborating findings in [@ivanova_unequal_2020] and [@oswald_large_2020]. Goods was the second most unequal final consumption category (90:10 ratios of `r goods_energy_10_10` for energy, `r goods_co2eq_10_10` for carbon), similar to services (90:10 ratios of `r services_energy_10_10` for energy and `r services_co2eq_10_10` for carbon) and then food (90:10 ratios of `r food_energy_10_10` for both footprints).
The final consumption categories (housing, mobility, food, goods, and services) contributed very differently to the environmental footprint of European households in 2015 (Figure 3). On average, housing and mobility were the two largest categories, accounting for about two thirds of both the energy and carbon footprints. In addition, the sectoral footprint variation across the expenditure deciles was also highly variable (Figure 3). For housing there was very little difference between deciles in both the energy and the carbon footprint. The bottom four deciles even had higher carbon footprints from housing than most top deciles, which can be explained by the extreme differences in intensity shown in Figure 2. Mobility was the most unequal category, with footprints in the top decile `r mobility_co2eq_10_10` times higher than the bottom decile, corroborating findings in [@ivanova_unequal_2020] and [@oswald_large_2020]. Goods was the second most unequal final consumption category (90:10 ratios of `r goods_energy_10_10` for energy, `r goods_co2eq_10_10` for carbon), similar to services (90:10 ratios of `r services_energy_10_10` for energy and `r services_co2eq_10_10` for carbon) and then food (90:10 ratios of around `r food_energy_10_10` for both footprints).
```{r figure3, out.width="100%", fig.cap="**Figure 3: Household energy and carbon footprints by final consumption category and European expenditure decile in 2015, further broken down by source and location. 'Direct' (direct energy use and carbon emissions from households) plus 'Domestic' (energy use and carbon emissions along the domestic national supply chain) make up that part of the household footprint coming from within national borders, while 'Europe' is the part from other countries within the sample (plus Italy and Luxembourg), and 'non-Europe' from all other countries.**"}
The geographical source of the energy and carbon footprints also varies by consumption category (Figure 3). The housing footprint was almost entirely domestic, with `r housing_co2eq_direct`% and `r housing_energy_direct`% respectively coming from direct household energy use and carbon emissions from heating and cooling, and the rest embedded primarily along the domestic supply chain. The mobility footprint, on the other hand, was around one fourth non-European. The majority of the mobility footprint, above 60%, came from vehicle fuel, either directly from households, or indirectly, i.e. embedded along the supply chain. The goods footprint was mostly non-European, while services and food were both around one third non-European. These results suggest that proposed future carbon border-adjustment mechanisms [@european_commission_communication_2019] will especially impact the mobility and goods footprints of the higher deciles, and to a lesser extent the food and services footprints.
The geographical source of the energy and carbon footprints also varies by consumption category (Figure 3). The housing footprint was almost entirely domestic, with `r housing_energy_direct`% and `r housing_co2eq_direct`% respectively coming from direct household energy use and carbon emissions from heating and cooling, and the rest embedded primarily along the domestic supply chain. The mobility footprint, on the other hand, was around one fourth non-European. The majority of the mobility footprint, above 60%, came from vehicle fuel, either directly from households, or indirectly, i.e. embedded along the supply chain. The goods footprint was mostly non-European, while services and food were both around one third non-European. These results suggest that proposed future carbon border-adjustment mechanisms [@european_commission_communication_2019] will especially impact the mobility and goods footprints of the higher deciles, and to a lesser extent the food and services footprints.
