Households in the south-west of England are some of the most carbon intensive in Europe, a new study shows.

The paper, published this month in the journal Environmental Research Letters, is the first to break down the embodied greenhouse gas emissions from household consumption across the EU.

Its data, mapped below, shows the emissions associated with the consumption of products by an average citizen across 177 regions in 27 EU countries (bar Croatia, the most recent addition to the EU).

The paper argues that such consumption-based emissions accounting is important to consider alongside the territorial, production-based emissions typically required to be reported by governments, since they highlight the links between local consumption and its global environmental consequences.

The breakdown into regional data could also empower regions to implement more focussed mitigation strategies, the paper argues.

The results show that people living in Eastern Europe, such as in Romania, Bulgaria and Hungary, have some of the lowest carbon footprints in the EU, with the per capita average in some regions just a third of that in several British regions and in Luxembourg.

Carbon Brief explores the trends and drivers behind the data.

Emissions across different EU regions

The map below shows the average per capita emissions across the 177 EU regions. The data includes direct household emissions, such as tailpipe emissions from vehicles, and indirect emissions, such as those created in the supply chains of purchases, which are then counted as being embedded in that product.

Per capita household carbon footprint across 177 EU regions

Per capita household carbon footprint across 177 EU regions in tonnes of CO2 equivalent per capita (tCO2e/cap), encompassing both direct and embodied emissions. Note that only national averages are shown for Sweden and the Netherlands. Source: Ivanova 2017.

The map visualises the large disparities between the consumption-based emissions of different European regions, often even within the same country.

Italy, Spain, Greece and the UK all displayed large differences in the overall footprint across their regions, while other countries such as Denmark, France and the Czech Republic were found to be more uniform. Overall, the difference between regional footprints within the same country varied from just 0.6 tonnes of CO2 equivalent per capita (tCO2e/cap) to over ten times this value.

As the map above shows the UK stands out as the only country besides Luxembourg to top the highest bracket of emissions in some regions (the south west).

When the EU is considered as a whole, the authors found the top 10% of people with highest household emissions are responsible for around 15% of EU’s household emissions, while the bottom 10% emit only 5%.

Emission drivers

On average across all countries, transport emissions were the largest source of emissions, accounting for around 30% of the carbon footprint of households, mostly due to the use of private vehicles. Transport emissions were found to be highest in France, Greece and most of all Luxembourg, where they accounted for 83% of per capita emissions. This is largely due to Luxembourg’s so-called “tank tourism” effect: the country’s lower fuel price compared to its neighbours, which leads to people crossing the border to fill up. (Although this has subsided significantly since the study’s data was collected in 2007.)

Food was also a significant source of emissions, making up around 17% of EU household emissions (though this varied from 11% to 32% across regions). The capital region of Denmark had the highest food-related emissions with 3.9 tCO2e/cap, around 27% of its total footprint.

Consumption based emissions across the EU 27 by sector. Values do not add to 100% due to rounding. Source: Ivanova 2017. Chart by Carbon Brief using Highcharts.

The study found big differences in the contribution of animal products to food emissions. For example, they were responsible for 33% to 38% of food emissions across regions in Belgium and Denmark, but 79% of food emissions in Slovenia (which also had the highest animal- based emissions per capita in absolute terms).

Clothing contributed only around 4% of household emissions on average, with some UK regions, such as London and Northern Ireland, showing some of the highest emissions due to clothing.

Emissions from services accounted for an overall 14% (though this ranged from 7% to 41% depending on the region), with Spain the country with the highest relative importance of services, largely associated with its hotel and restaurant services.

Around 10% of household emissions were associated with other manufactured products, such as furniture, while 22% was associated with housing emissions such as fuel and electricity impacts.

Total household footprint

It is worth noting, however, that emissions per person can vary substantially from the total emissions of a region. For instance, people living in cities tend to have lower transport footprint per capita, even though the transport footprint overall may be very high.

The second map here shows the total embodied household carbon footprint that a region is responsible for in megatons CO2-equivalents (MtCO2e). Unlike the per capita map above, the data presented here is a function of a region’s total population as well as its per person consumption.

Total household carbon footprint across 177 EU regions

Total household carbon footprint across 177 EU regions in tonnes of CO2 equivalent, encompassing both direct and embodied emissions. Note that only national averages are shown for Sweden and the Netherlands. Source: Ivanova 2017.

As for the per capita map, this also clearly shows large differences in emissions between regions. The study found that some regions are responsible for a large share of a country’s emissions: the North Rhine-Westphalia and Bavaria in Germany, for example, together accounted for around 40% of Germany’s embodied emissions.

More monitoring

Diana Ivanova, the study’s lead author and PhD candidate in industrial ecology at the Norwegian University of Science and Technology, says the overview of regional differences given in the study are important as sustainability is not something to be considered solely on the national and international level.

She tells Carbon Brief:

“These interregional differences within countries have implications for the regions’ impact mitigation potential and, as some would argue, for their responsibility to contribute to national mitigation goals.”

The EU itself has encouraged environmental policies to be developed at the local level.  But the authors say a lack of region-level studies across countries could be hindering broader policy conclusions, as well as more targeted local policies. Ivanova tells Carbon Brief:

“Through regionalised emission inventory, local authorities have the power to monitor the impact and success of policies targeting local consumption practices, e.g. the introduction of flexible work schemes that reduce commuting or the construction of bike lanes. Of course, to enable this we would need a continuous monitoring of consumption-based environmental impacts.”

Break it down

To probe this further, the authors also examined the factors driving the differences seen in regional emissions.

They found that socio-economic factors, such as income, household size, education, dwelling size and basic need consumption, typically explain between 15% and 70% of the variation seen across and within countries. This is in line with previous research which has shown that geographic and infrastructural effects have a limited impact on emission differences at the regional level compared to socio-economic factors.

Income was found to be the biggest factor in emissions disparities, alone explaining 29% of disparities between regional household carbon footprint. This is likely driven by higher purchasing power and its associated environmental impacts.

In particular, income was a driving force behind variations in transport emissions (which accounts for around 30% of EU household emissions), as well as in areas such as clothing, services and manufactured products. This was especially the case in countries with higher income inequalities, such as the UK, Spain and Italy. However, in some other areas, such as food, emissions were not found to differ significantly with income.

Household size also had a significant impact on emissions, with an increase in the number of people living in one household leading to a drop in their per capita emissions, such as those from fuel and waste treatment. Finnish regions, which on average had the smallest household sizes of 1.5 persons, tended to show the highest housing footprint.

Other major factors driving differences in regional carbon footprints were local climate, carbon intensity of the electricity mix, urban-rural typology, expenditure patterns, and level of education (where a rise in education led to increased emissions).

What the study did

In order to calculate the consumption-based emissions for different regions, the authors combined regional consumption surveys with a database on the emissions from consuming different products in different countries.

National survey data can be unreliable in a number of ways. People tend to under-report some purchases, such as tobacco and alcohol consumption, and irregular purchases, while data collection may differ significantly between countries. However, the authors tried to account for this by comparing and aligning the data to national consumption figures. Note also that the data was collected between 2007 and 2013, depending on the country.

The emissions data came from the EXIOBASE database, which provides the national carbon intensities of 200 product sectors broken down by their origin and final destination. It encompasses the environmental intensity of moving products across various global supply chains, specifics on product trade between countries, as well as the direct emissions as households use these products.

Ivanova tells Carbon Brief:

“We can essentially trace back all the environmental impacts that are embodied in, say, the consumption of cheese by an average Austrian. What kind of inputs were needed to produce the cheese and where they come from, e.g. milk, labour, energy, transport.”

EXIOBASE provides data on 200 products, 43 countries and 5 rest-of-the-world regions. However, it is worth noting that the EXIOBASE database only provides an overview of the global economy in 2007 and, thus, will be outdated in some respects. In addition, its data does not provide within-country detail: if the products being consumed in one region (or by one income level) are less emissions intensive than elsewhere, this information is not picked up, with the differences between regions as shown on the maps driven purely by differences in consumption levels.

Consumption or production?

The EU has committed to cutting its territorial greenhouse gas emissions to 20% below 1990 levels by 2020, while the UK’s Climate Change Act binds it to an 80% drop by 2050.

However, some researchers argue these production-based targets don’t fully account for the international emissions countries are responsible for, as products may accumulate significant environmental impacts along global supply chains. Others are concerned using territorial targets means emissions could be reduced locally by exporting production elsewhere.

Ivanova argues both types of accounting need to be considered. She says:

“Imagine if we only focused on territorial emissions how susceptible we would be at simply moving the environmental problems associated with our consumption somewhere else, instead of actually tackling them. For example, if we started importing cars instead of producing them domestically, there may be a drop in territorial emissions, but the consumption emissions may stay the same (or even increase, depending on the production efficiency).”

An EU Commission Eurostat study released earlier this month put the average consumption-based CO2 emissions of the 28 member states at 7.2 tCO2e/cap in 2014, slightly lower than the production-based emissions of 7.3 tCO2e/cap.

Production and consumption perspective breakdown of the EU-28 in 2014 in tonnes CO2 per person. Source: Eurostat 2017.

Production and consumption perspective breakdown of the EU-28 in 2014 in tonnes CO2 per person. Source: Eurostat 2017.

However, it is worth noting that the UK is an unusually large net importer of emissions. Government statistics indicate the UK was responsible for 1050 Mt CO2e in 2013 using consumption emissions accounting versus 567 Mt CO2e for territorial emissions the same year. (Note that unlike the Eurostat data above which only looks at CO2 emissions, these are comprised of the six main greenhouse gases under the Kyoto Protocol.)

In fact, research by the Committee on Climate Change in 2013 found that the UK is one of the largest net importer of emissions in the world, second only to Japan, with the third largest per capita consumption emissions.

Therefore, some researchers have argued that the UK’s reduction targets should be stricter in order to take into account of its position as a net importer of emissions.

Whether or not country carbon footprinting ever gains traction with regards to official targets, there remains an argument that tracking the embodied emissions of products, regions and countries can give helpful insights into how best to tackle emissions.