1 Introduction
Adam Smith stated that “consumption is the sole end and purpose of all production” (Smith
1904) thus putting consumption firmly in the field of industrial ecology. In this chapter we specifically focus on the carbon emissions caused by household consumption, as these have been estimated to be accountable for around 72 % of carbon emissions on a global basis (Hertwich and Peters
2009; Wilson et al.
2013). Thus the study of households
1 and how the environmental impacts for which they are responsible may be reduced is key to achieving a low carbon future.
Accordingly, in this chapter, we first examine the accounting perspective required to scrutinise the carbon emissions for which household consumption is responsible (Sect.
2). Section
3 reviews the evidence concerning the determinants of household carbon missions. In Sect.
4, we introduce the ‘rebound effect’ – a phenomenon that confounds attempts to reduce carbon footprints, making reducing emissions more of an uphill task than often acknowledged. In the final section (Sect.
5), we broaden the focus to look at households in the wider context of systems of production and consumption, and possibilities of win-win solutions that offer potential to reduce carbon while at the same time enhancing well-being.
Household
consumption is a wide ranging topic and inevitably there are many limitations to this chapter. One of these is that we focus here on consumption by Western households. A second is that we do not review the prolific literature on the driving forces behind household consumption or the ways that household consumption may be reduced.
2 And a third limitation is our focus on ‘carbon’ emissions. However carbon is defined (see Sect.
2), use of carbon emissions as a single indicator can lead to policies that, while beneficial in terms of reducing global warming, may lead to unexpected and unintended detrimental consequences in terms of other environmental impacts. For example, Benders et al. (
2012) analysed five environmental impact categories: global warming potential, acidification, eutrophication, summer smog and land use. Combined analysis of the five impact categories found that food has the largest environmental impact,
3 whereas analysis of greenhouse gas emissions
alone indicated that housing has the largest impact. Nevertheless, climate change caused by anthropogenic carbon emissions is currently accepted as the most urgent environmental threat (IPCC
2014) and is thus considered a useful indicator for the focus of this chapter.
2 Consumption Accounting and Carbon Footprinting
In this Section, we first set out the importance of the type of accounting framework used for exploring household carbon footprints, and explain how this is different from the default framework normally applied by governments in assessing their emissions and in international treaties. The framework is best introduced by posing the question: to what extent should Western consumers take responsibility for the things they buy? If, say, a UK consumer purchases a TV manufactured in China
, which nation should take responsibility for the emissions incurred during its manufacture? This dilemma illustrates two different accounting approaches that must be untangled as we strive to devise strategies for a more sustainable future. According to accounting by the production perspective
, China should take responsibility as the emissions arose on Chinese territory. This is the approach used in the Kyoto Protocol and is the most commonly used accounting approach (Bows and Barrett
2010; Wiedmann
2009). An alternative is the consumption perspective. According to this perspective, the UK should take responsibility, as export to the UK was the driving force motivating production, and a UK consumer is the primary beneficiary of the final product (Druckman et al.
2008; Peters and Hertwich
2008a; Peters et al.
2011; Lenzen
2008; Lenzen et al.
2007; Jackson et al.
2006).
The accounting perspective used is particularly important because accounting according to the production perspective
shows that many Western economies are successfully reducing their carbon emissions. However, when the consumption perspective
is used for accounting, not only are carbon emissions often found to be higher than compared to the production accounts, but they also tend to exhibit a rising trend (CCC
2013; Baiocchi and Minx
2010; Ahmad and Wyckoff
2003; Peters and Hertwich
2006a; Baiocchi et al.
2010). The reason for the differences shown between the two accounting perspectives is the quantity of carbon emissions embedded in trade, which is the subject of Wiedman’s chapter (Chap.
8) in this book. An example of the importance of the carbon embedded in trade is given by Li and Hewitt (
2008) who found that, through trade with China
, the UK reduced its production based carbon dioxide emissions by approximately 11 % in 2004, compared with a non-trade scenario in which the same type and volume of goods are produced in the UK.
While many studies explore the carbon emissions embedded in trade,
4 some studies focus specifically on the role of imported goods and services and their associated emissions in the carbon footprints of households
5 (Hertwich and Peters
2009; Lenzen et al.
2006; Munksgaard et al.
2005; Nijdam et al.
2005; Peters and Hertwich
2006b). Peters and Hertwich (
2006a) put forward a general rule that countries with a high proportion of imports and relatively clean electricity generation are likely to have a significant proportion of their household carbon emissions attributed to imports. This means that, due to the supply chain emissions embedded in imported goods, households drive emissions in other countries as well as in their own country. For example, Weber and Mathews (
2008) found that nearly 30 % of the carbon dioxide emitted to meet household demand in the US occurred outside the borders of the US.
Accounting according to the consumption perspective
is commonly known as ‘footprinting’: this is the approach adopted in this chapter, and in particular the chapter is concerned with carbon footprinting. However, the definition of what is included in a carbon footprint is contentious, as shown in Table
9.1. In this chapter we take a relaxed approach to what we mean by ‘carbon’ and include reviews of studies that range from assessing carbon dioxide emissions only to those that take a more comprehensive greenhouse gas approach. The main difference in results is that emissions due to food make up a larger portion of the carbon footprint of a household when the analysis is extended to a basket of greenhouse gases. What we are more stringent about in this chapter is that we take a whole supply chain, life cycle approach to assess the carbon emissions caused by households.
Table 9.1
Recent definitions of a carbon footprint
“The carbon footprint is a measure of the exclusive total amount of carbon dioxide emissions that is directly and indirectly caused by an activity or is accumulated over the life stages of a product” (Wiedmann and Minx 2007: 4) |
“A carbon footprint is equal to the greenhouse gas emissions generated by a person, organization or product” (Johnson 2008: 1569) |
“A measure of the total amount of CO 2 and CH 4 emissions of a defined population, system or activity considering all relevant sources, sinks and storage within the spatial and temporary boundary of the population, system or activity of interest. Calculated as CO2e using the relevant 100-year global warming (GWP100)” (Wright et al. 2011: 69) |
“Climate footprint: A measure of the total amount of CO 2, CH 4, nitrous oxide, hydrofluorocarbons, perfluorocarbons and sulfur hexafluoride emissions of a defined population, system or activity considering all relevant sources, sinks and storage within the spatial and temporal boundary of the population, system or activity of interest. Calculated as CO 2 equivalents using the relevant 100-year global warming potential” (Williams et al. 2012: 56) |
“A measure of the amount of carbon dioxide released into the atmosphere by a single endeavour or by a company, household, or individual through day-to-day activities over a given period” (Collins English Dictionary 2012) |
There are two basic categories of a household carbon footprint. First are direct emissions that arise due to direct energy use in the home (such as gas for space and water heating, and electricity for lighting and powering appliances and gadgets) and due to burning personal transportation fuels (petrol and diesel). Second is ‘embedded’ emissions, such as those that arise during our example of the manufacture of a TV made in China
. Embedded emissions along supply chains (arising domestically and abroad) account for the majority (around 60–70 %) of the carbon footprints of Western households (Druckman and Jackson
2010; Dey et al.
2003; Bin and Dowlatabadi
2005; Baiocchi et al.
2010).
6
Estimates of household carbon footprints are generally derived from expenditure data. Carbon emissions arising from expenditure on transportation fuels and energy use in the home are relatively easily estimated from information on prices, the carbon content of fuels and information from each country’s Environmental Accounts. Estimation of carbon emissions embedded in other expenditures is harder and requires information on the technologies used to manufacture all products and services purchased, wherever in the world this may occur. This is generally done using Environmentally Extended Input-Output Analysis (EE-IOA) (Hertwich
2011; Munksgaard et al.
2005; Baiocchi et al.
2010; Weber and Matthews
2008; Weber and Perrels
2000; Lenzen et al.
2004; Wiedmann
2009). EE-IOA is a top-down methodology that combines information on the structure of the economy with environmental data (see Miller and Blair (
2009)). There are some notable exceptions to this methodology. The first is hybrid analysis which combines process-based, bottom-up Life Cycle Assessment (LCA) with top down EE-IOA (Benders et al.
2012). Another exception is the work by Girod and de Haan (
2009,
2010) who use a bottom-up LCA methodology only, based on physical functional units such as kg of food, person kilometres and living square meters.
4 The Rebound Effect
The discussions above lead to many suggestions concerning ways that the carbon
footprints of households may be reduced, but this is beyond the scope of this chapter, as explained in Sect.
1. However, a systemic issue that works against many measures suggested for reducing emissions (such as installing loft insulation and travelling less) is the ‘rebound effect’. The rebound effect, in relation to households, can be explained as follows (Sorrell
2007; Maxwell et al.
2011): When an action is carried out that is intended to save energy, it will often result in saving money also. However, a household always uses its income in some way or other. For example, when purchasing a car, suppose the purchaser decides to buy one that is more fuel efficient than the average car on the market. Knowing his normal mileage, he can calculate the fuel saved, and hence by how much he will expect to reduce his
carbon footprint. However, as less fuel is now used for his normal journeys, less money is spent on this fuel. This freed up money might be spent on driving further, which will result in more carbon emissions. This is called the direct rebound effect. Alternatively, the money saved might be spent on something entirely different from motor vehicle fuel, such as taking a vacation. This will also give rise to more emissions, and this is known as the indirect rebound effect. Alternatively, he might decide, rather than to spend the money, to save it and therefore he puts it on deposit in a bank. The bank, however, then invests the money, and this investment, in turn, gives rise to carbon emissions. This is another example of the indirect rebound effect.
Another type of rebound effect that commonly arises is the ‘embodied’ rebound effect, and this is better illustrated through an example of loft insulation. In this example a person who installs loft insulation can calculate how much energy (and hence carbon emissions) will be saved through reduced fuel use. However, energy is used in the manufacture of the loft insulation, and, following the consumption accounting principle discussed in Sect.
1, carbon emissions from this energy use are the purchaser of the insulation material’s responsibility. Hence these emissions offset the expected savings, and this is known as the embodied rebound effect.
If a measure is expected to achieve a reduction of 100 kgCO2e then a rebound effect of 30 % implies that only 70 kgCO2e was saved, and a rebound effect of 100 % implies that no carbon was saved. A rebound effect greater than 100 % means that the measure resulted in more, not less, emissions, and, from this view, it would have been better not to have done the action at all. This is known as ‘backfire’.
Until relatively recently, although the rebound effect was a well-known phenomenon, there were few studies that had estimated to what extent it is a problem with respect to households. In the last few years, however, studies have been carried out to explore it focusing on various different countries. These include Lenzen and Dey (
2002) and Murray (
2013) for Australia; Alfredsson (
2004) and Brännlund et al. (
2007) for Sweden; Mizobuchi (
2008) for Japan; Kratena and Wuger (
2010) for Austria; and Thomas and Azevedo (
2013) for US and Druckman et al. (
2011a) and Chitnis et al. (
2013,
2014) for the UK. These studies generally consider a variety of measures such as abatement actions (for example, reducing the amount of food wasted, reducing household room temperature thermostat settings and replacing short car journeys by walking or cycling) and energy efficiency measures (for example, installation of cavity wall insulation, loft insulation, condensing boiler, water tank insulation, energy efficient lighting and purchase of an efficient car). Chitnis et al. (
2014), who estimated the rebound effect in terms of GHG emissions
, found rebound to be around 0–32 % for measures affecting domestic energy use and around 25–65 % for measures affecting vehicle fuel. The possibility of backfire was found for measures that reduce food waste, with estimates being around 66–106 % (Chitnis et al.
2014). In general, rebound was found to be larger for lower income groups (with some exceptions) as they have a higher proportion of expenditure on direct energy (as discussed in Sect.
3.1) and this expenditure has relatively high income elasticities (Chitnis et al.
2014).
The conclusion from this rebound effect work is not that encouragement to carry out the abatement and energy efficiency actions should be abandoned: indeed, for all except food waste under certain conditions, considerable carbon emissions can be saved through these means and therefore it is imperative that such actions should be supported. However, it is vital that governments take into account the rebound effect when estimating reductions in carbon emissions that can be achieved, else they stand in danger of systemically missing their carbon reduction targets.
Nevertheless, efforts should be made to minimize the rebound effect wherever possible. The best way to do this is to encourage a wholesale shift in expenditure patterns towards low carbon goods and services. The rebound effect studies also highlight the importance of investment decisions, and Druckman et al. (
2011a) show that in order to achieve zero rebound, the money saved through the abatement or efficiency actions should be invested in carbon neutral or reducing investments.
This chapter has explored the drivers and components of household
carbon footprints. Evidence shows that ‘hair-shirt’ policies, particularly within the realm of recreation and leisure, are unlikely to gain enough traction to achieve the widespread changes needed (Soper
2008). The ‘holy grail’ is thus to devise low carbon lifestyles that achieve maximum happiness. However, economic growth (the policy goal of most governments
11) aims to increase incomes. But it is generally found that as incomes increase, carbon footprints are likely to increase while well-being levels off (Lenzen and Cummins
2013; Jackson
2009). This raises the question: which policies enhance well-being, or at least do not reduce well-being, while being environmentally beneficial? Such activities represent win-win opportunities for encouraging activities which give rise to relatively low quantities of carbon emissions while at the same time enhancing well-being and happiness.
Reviews of the literature reveal that social activities such as conversing with friends and family, making love, reading and carrying out hobbies are low carbon activities that generally make people happy (Csikszentmihalyi
2006; Holmberg et al.
2012; Kahneman et al.
2004; Caprariello and Reis
2012; Nassen and Larsson
2015). For many of the activities that generally enhance happiness, the carbon emissions depend on how they are carried out. For example, being close to nature and physical activities such as walking, exercising and sport can be relatively low carbon if carried out without the use of personal transportation. Csikszentmihalyi (
2006) talks about how goal-orientated activities can induce high levels of happiness. His theory is that when a person is carrying out an activity that is all-encompassing, in that the activity requires total concentration and focus (in other words, the person is “in the flow”) then a high state of happiness can be achieved. Examples of this include playing a musical instrument or singing in a choir, both of which can be done in relatively low carbon ways, but one of Csikszentmihalyi’s examples is the state of flow achieved during downhill skiing, and, depending on where one lives, this can be a very high carbon activity. Gatersleben et al. (
2008) investigated how volunteering can yield high levels of happiness and, again, this may be carried out in high or low carbon ways. Shopping is an example of an activity that generally brings happiness, but is, arguably, rarely a particularly low carbon activity.
This discussion has highlighted some win-win approaches to reducing carbon emissions while increasing well-being, and these should be key components of strategies for moving towards a more sustainable future. But before closing this chapter it is worth taking stock and standing back to take a whole-systems approach.
A whole-systems approach requires looking at systems of production and consumption in which households play a central role. The economy is circular in nature: in simplistic terms, households earn wages from firms, and firms produce goods and services to sell to the households. Thus producers are consumers, and consumers are producers. Linking this understanding with the earlier discussion in which it was shown that one of the main determinants of a household’s carbon footprint is income, and also that households spend or invest all their income, raises another possible win-win situation: that of working-hours reduction.
Reducing the average number of hours worked per week can have both a scale effect and a compositional effect (Gough
2013). Hypothetically, due to the scale effect of fewer hours at work, workers’ incomes would be reduced, and thus expenditures and consumption would also be expected to be reduced. With each person working less, there is the possibility of increasing the number of people employed and thus reducing inequalities. High levels of inequality are associated with low levels of well-being (Wilkinson and Pickett
2009), and, furthermore, meaningful work is a generally found to be a positive factor in increasing well-being (Diener and Seligman
2004). Hence sharing the work may yield multiple benefits (Hayden
1999).
The compositional effect can be explained as follows: with lower incomes but less time at work, people’s use of time outside work would be expected to change, as would the composition of their expenditure baskets. For example, rather than buying ready-meals, people may be more inclined to cook from raw ingredients. Now such changes in time and expenditure budgets might result in higher or lower carbon emissions. For example, with less time pressure, people might walk and cycle for short journeys rather than drive. On the other hand, some people may drive further and more often to visit friends. But if we look back to the graph in Fig.
9.1, we see that there is good evidence that lower incomes will, in general, result in lower carbon footprints
.
Reducing the working week has been shown to enhance the work-life balance (Nassen and Larsson
2015; Kasser and Sheldon
2009; Eurofund
2013). For example, Hayden (
1999) records how French employees reported overall improved quality of life when their working week was reduced to 35 h. In another investigation 400 Swedish employees who had their worktime reduced to 6 h per day for 18 months reported improved life satisfaction, health and a more equal gender-balance on time spent on housework (Bildt (2007) cited in Nassen and Larsson (
2015)).
The suggestion of reducing working hours must be taken with an important warning concerning low income groups. Currently many low paid workers are struggling to meet their weekly household expenses (MacInnes et al.
2014; The Living Wage Commission
2014), and therefore any initiative to reduce the working week must be accompanied by special measures to protect them. If these are put in place, then work-time reduction offers a promising way to reduce unemployment by sharing the work, leading to reduced inequalities, while at the same time offering high prospects of increasing well-being and reducing environmental burdens (Hayden and Shandra
2009; Victor
2008; Jackson
2009; Coote et al.
2010; Knight et al.
2013; Pullinger
2014; Rosnick and Weisbrot
2007).
In conclusion, this chapter has reviewed the main determinants of Western household carbon footprints. What is clear from this body of work is that, seen from a consumption perspective, the majority of carbon impacts arise from transportation, food and housing. The need to improve systems of provision of food, energy and transportation and renovate or rebuild inefficient housing stock is therefore indisputable. However, where possible these measures should be supplemented by other approaches. For instance, through further analysis it is evident that recreation and leisure leads to the single highest proportion of household carbon emissions. Opportunities should therefore be sought for low carbon leisure activities which also enhance wellbeing. Such activities might include for instance spending time with friends and family in and around the home, or engaging in physical recreation in the local community.
One inescapable finding from this body of work is that income is one of the principal drivers of carbon emissions, with carbon footprints increasing with increasing incomes. Incomes also appear to drive the rebound effect. These understandings led us to a wider, whole-systems approach in which we view households as an integral part of the system of production and consumption. Policies on work-time reduction, with appropriate measures to safeguard low income households, can offer additional win-win opportunities that, to some extent, overcome this stumbling block. Ultimately, however, income growth is driven by economic structure. Approaches which tackle the structural implications of economic growth are also essential to a meaningful understanding of the potential to reduce carbon footprints. In summary, industrial ecology, with its wide ranging systems approach as shown in this chapter, has a great deal to contribute to the quest to devise strategies to move towards lower carbon, fulfilling lifestyles.