Professional Documents
Culture Documents
Ecological Economics
Analysis
a r t i c l e i n f o a b s t r a c t
Article history: Sustainable development poses a major challenge to Western societies as many of their citizens have lifestyles
Received 1 June 2015 with resource use beyond the earth's carrying capacity. Sustainability transitions will be easier to implement if
Received in revised form 2 March 2016 they also increase individuals' well-being. We investigate the relation between the ecological footprint and the
Accepted 13 March 2016
subjective well-being at the individual level, using a questionnaire carried out in Flanders (Belgium). Our results
Available online 20 April 2016
suggest that a lower footprint does not reduce well-being in the sense that we find no significant correlation. In a
Keywords:
next step, we investigate the direct impacts of the different ecological footprint components on subjective well-
Subjective well-being being. Switching to a more environmentally friendly diet and not using electricity for domestic heating create
Ecological footprint win–win situations as these actions decrease the ecological footprint while increasing reported levels of well-
Individual level being. Finally, we investigate the socio-demographic determinants of the ecological footprint and subjective
well-being to look for indirect impacts. Having a relationship and being a house owner increase subjective
well-being and decrease the ecological footprint. Better social life and health and living in a pleasant environment
increase subjective well-being with no cost in terms of ecological footprint.
© 2016 Elsevier B.V. All rights reserved.
1. Introduction some of these changes will be easier to implement if they are at the same
time beneficial for individuals' well-being. This joins in with the call of
There is widespread attention and an urgent need for a more sustain- Helliwell et al. (2013, p. 108) that “the analytical community needs to help
able development and for sustainability transitions, especially in Western us understand which policies would raise well-being in a sustainable way”.
countries. The lifestyles of citizens in these countries have significant eco- We want to contribute to this call by integrating information about the en-
logical consequences in terms of resource and energy use. WWF (2014), vironmental impact of individuals and their reported subjective well-being.
for instance, reveals that global society needs 1.5 earths to meet its current Therefore we will also examine and compare if people's demographic char-
demands on nature as measured by the ecological footprint. The Ecologi- acteristics are related to their well-being and their environmental impact.
cal footprint per capita of high-income countries is about five times higher In recent decades, the literature on subjective well-being (SWB) is
than that of low-income countries. As a result, researchers (e.g. Gardner booming and self-reported data on happiness or life satisfaction is be-
and Stern, 2002) often argue that a series of behavioural changes is need- coming increasingly available. Many studies summarize the main fac-
ed across different policy fields, such as mobility, food and consumption, tors influencing subjective well-being, amongst others Blanchflower
production processes, etc. Policymakers are aware of these challenges and Oswald, 2004; Dolan et al., 2008; Stiglitz et al., 2009; Helliwell
and are taking sustainability issues into account at various levels: at the et al., 2012; O'Donnell et al., 2014. Key factors include age, income,
global level (e.g. the Post 2015 agenda), at the European level (e.g. sus- education, work, family life, social capital, religion, community and
tainable growth as a key goal in the Europe 2020 strategy) and at the governance, mental health and physical health. Furthermore, the im-
local level (e.g. the movement towards sustainable cities or regions). portance of personality traits, adaptation and expectations are indisput-
Stiglitz et al. (2009) relate sustainability issues to the concept of subjec- able (e.g. Veenhoven, 1991; Diener, 2000; Helliwell et al., 2012).
tive well-being by stressing the challenge of maintaining current levels of We aim to compare information on subjective well-being to informa-
well-being for future generations. Many of the necessary behavioural tion on the environmental sustainability of an individual's lifestyle. A fre-
changes have an impact on the way people live and on their well-being, quently used measure for the (un)sustainability of individual behaviour
which might impede their acceptance by citizens. Policies that promote is the ecological footprint (EF) – the amount of acres of biologically pro-
ductive land that are needed for the individual's consumption and activ-
ities. The largest part of the EF is the area of productive land that is
☆ We would like to thank Ecolife vzw for calculating the ecological footprint of the
respondents in our survey.
required to absorb an individual's CO2 emissions – i.e. his carbon foot-
⁎ Corresponding author at: Sint-Pietersplein 6, 9000 Gent, Belgium. print (CF). In the literature considerable attention is devoted to the CF
E-mail address: Elsy.Verhofstadt@UGent.be (E. Verhofstadt). and its main determinants. Kerkhof et al. (2009) report diverging results
http://dx.doi.org/10.1016/j.ecolecon.2016.03.021
0921-8009/© 2016 Elsevier B.V. All rights reserved.
E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89 81
for the relation between household income and CO2 emissions: there is a we have collected that information for one and the same population.
negative relation between income and emissions in the Netherlands and As a result we can investigate whether there is a direct link between
the UK, whereas emissions increase with income in Sweden and Norway. (a) SWB and (b) the EF and its components. Moreover, in line with
Büchs and Schnepf (2013) show that besides income, also household the approach of Lenzen and Cummins (2013), we want to investigate
size and composition is highly relevant for carbon emissions. Larger if there is an indirect link: which socio-demographic determinants
households have higher emissions in general, e.g. due to heating and have an impact on both the individuals' SWB and their EF? We aim to
laundry, but lower emissions per capita. The authors also find higher identify common determinants with a positive effect on subjective
emissions for elderly or unemployed people (who spend more time at well-being and a negative effect on the footprint. Our objective is also
home), for the higher educated and for those living in rural areas to formulate policy advice based on these results.
(Büchs and Schnepf, 2013). Also Menz and Kühling (2011) point to the In the next section, we provide an overview of our research ques-
important role of age. Druckman and Jackson (2009) focus on the in- tions and hypotheses. Section three presents the concepts, the survey
creased recreation, leisure and household activities (such as commuting, and the data while section four contains the empirical results. In section
eating, heating, …) as important drivers of the increase in CO2-emissions. five, we conclude.
The relationship between sustainability and well-being is not com-
monly explored in the literature. When there is a link to the environ- 2. Research Question and Hypotheses
ment or to sustainability in the subjective well-being literature, the
focus is often on the impact of the state of the environment. Tempera- We first explore the direct link between an individual's ecological
ture is highly significant (positive), both for happiness (Rehdanz and footprint and his or her subjective well-being. Afterwards we link
Maddison, 2005) and for life satisfaction (Maddison and Rehdanz, SWB to the components of the footprint: food intake, energy use
2011). Many studies confirm the negative impact of air pollution (heating and electricity), paper use, car use, use of public transportation
on life satisfaction (Welsch, 2006; Rehdanz and Maddison, 2008; and travel behaviour. Finally, we investigate the indirect link between
MacKerron and Mourato, 2009; Ferreira et al., 2013; Li et al., 2014). Fi- SWB and the EF in the sense that we identify a number of common de-
nally, natural disasters such as floods also have a significant negative terminants for both concepts (see Lenzen and Cummins (2013)). If such
impact on life satisfaction (Luechinger and Raschky, 2009). Besides determinants influence the EF and SWB in opposite directions, they can
looking at the state of the environment, some subjective well-being create win–win situations in the sense that they can positively influence
studies also look at the impact of an individual's attitudes towards the SWB while at the same time reducing the EF. When win–win variables
environment on his or her reported subjective well-being. Ferrer-I- have policy relevance, we can formulate policy advice about the possi-
Carbonell and Gowdy (2007), for instance, find a negative relation bility of creating double dividends. Other variables or policies could cre-
between the concern for ozone pollution and life satisfaction and a pos- ate a lose–lose situation. In both these cases, be it win–win or lose–lose,
itive relation for concern about species extinction. In studies on the eco- policy implications are evident. We also expect to find variables that in-
logical or carbon footprint of individuals the focus lies exclusively on fluence both the EF and SWB in the same direction. They are beneficial
household characteristics and behaviour (see Druckman and Jackson, for one outcome but have an adverse effect on the other and thus entail
2009; Kerkhof et al., 2009; Büchs and Schnepf, 2013), so that a direct a trade-off for the policymaker. This might be the case for many of the
link to subjective well-being is absent in that literature. variables from the carbon footprint literature (e.g. income and car own-
To our knowledge there are three studies that, to some extent, com- ership). Finally, we will also have win-neutral (or lose-neutral) vari-
bine information on the EF and subjective well-being. Welsch and ables: determinants that have a positive (or negative) and significant
Kühling (2011) show that, for a German sample, environmentally impact on only one of both outcomes.3
friendly consumption1 raises the reported life satisfaction of individuals. Based on the SWB and EF/CF literature we have formulated a number
Their measure of environmentally friendly behaviour contains some el- of research hypotheses that can be found in Table 1. In the table we only
ements related to the EF, but it is not a comprehensive measure of the include determinants that have been studied in either (a) both research
environmental impact individuals have through their behaviour. Next, fields or (b) one of the three studies (cfr. supra: Lenzen and Cummins,
Ericson et al. (2014) indicate that empathy, compassion and non- 2013; Ericson et al., 2014; Welsch and Kühling, 2011) that combine
materialistic values contribute to both subjective well-being and to a both research fields. In our analyses we will also consider determinants
more sustainable way of life. Finally, Lenzen and Cummins (2013) inte- that appear only in the literature of one of the fields (SWB or EF/CF), yet
grate two surveys from different samples of the Australian population — we do not formulate a priori hypotheses for these variables.
the Australian Unity Wellbeing Survey and the Australian Household
Expenditure Survey — and identify common determinants for subjec- 3. Data
tive well-being (SWB) and for the carbon footprint. Their aim was to
separately extract trends from both surveys on the basis of a common Our analysis is based on self-reported information obtained from a
set of determinants. Lenzen and Cummins (2013) find a set of variables survey (N = 1286) in Flanders (the northern Dutch speaking part of
that are beneficial for well-being, but have an adverse influence for the Belgium) in 2013.4 In the sample, participants are weighted to achieve
footprint: income, car ownership/car use, education/qualifications and similarity to the (univariate) frequency distributions in the Flemish
recreation and leisure. They also find that the more educated report to population (data from the Labour Force Survey 2012) for life situation,
be more environmentally concerned, yet this does not result into acting gender and age.5 In line with the population, the weighted sample
more pro-environmentally.
Lenzen and Cummins (2013) indicate that, at present, there exist no 3
For the interpretation of the results, only variables that have a statistically significant
data that combine information on individuals' subjective well-being effect are important. We included the concept ‘win- or lose-neutral’ for those determi-
and their ecological footprint.2 Our study aims to address this issue, as nants that are significantly related to only one of the two outcomes. For policy makers the-
se determinants can be considered as ‘second best’ solutions to stimulate either ecological
behaviour or increase well-being.
1 4
Environmentally friendly consumption is measured using five aspects: organic food, We call the survey LEVO 2013 which is short for LEvensomstandigheden in Vlaande-
low energy light bulbs, low energy household appliances, solar thermal heating systems ren Onderzocht – Dutch for “Inquiry into the life circumstances in Flanders”. The LEVO sur-
and green electricity. vey has been carried out yearly since 2010 with a focus on subjective well-being measures
2
At macro level, the New Economics Foundation (nef (2012)) links the Ecological Foot- and their determinants. The 2013 edition included an additional series of questions on the
print to well-being in the “happy planet index (HPI)” and Cloutier et al. (2014) assess the Ecological Footprint.
5
happiness impact of sustainable practises in communities or cities. However, these studies Chi-Square Goodness of Fit Tests show that the univariate distributions for the weight-
do not contain an analysis at individual level, including socio-economic and demographic ed sample equal the expected ones (from the Labour Force Survey) for life situation
information. (p = 0.999), gender (p = 0.992) and age (p = 0.593).
82 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89
Table 1
Hypotheses (based on literature) regarding impact of variables on SWB and EF.
Beneficial for SWB and EF: Beneficial for SWB but harmful for EF:
▪ Living together with people: SWB literature + Büchs and Schnepf ▪ Elderly: SWB literature + Menz and Kühling (2011)
(2013); Lenzen and Cummins (2013) ▪ Higher income: Lenzen and Cummins (2013)
▪ Altruistic and conscientious: Ericson et al. (2014) ▪ Car ownership/more car use: Lenzen and Cummins (2013)
▪ Environmentally friendly consumption: Welsch and Kühling (2011) ▪ Higher education: SWB literature + Büchs and Schnepf (2013); Lenzen and Cummins (2013)
Harmful for SWB and EF: ▪ Recreation and leisure: SWB literature + Druckman and Jackson (2009); Lenzen and
Cummins (2013)
▪ Unemployed: SWB literature + Büchs and Schnepf (2013)
▪ Living in rural areas: Dolan et al. (2008) + Büchs and Schnepf (2013)
consists of 49% male and 51% female respondents, with an average age use and c7 travel behaviour. The family composition is taken into ac-
of 50 years old. The majority (55%) is working full-time or part-time, count where relevant to rescale family outcomes into individual im-
the pensioners (27%) are the second largest group. Appendix 1 offers pacts (for heating, paper use, car use, use of public transportation and
detailed descriptive statistics for the weighted sample. travel). Appendix 2 provides an overview of the different EF questions
that feed into Ecolife vzw's footprint calculator. With qjk the answer to
3.1. Concept and Measurement of Subjective Well-being question k for component j, and there are Qj questions for component
j, the footprint calculator is a function fcj relating the answers to the
Our approach to linking utility and subjective well-being is similar to questions to the calculation of the components:
Layard et al. (2008). Assume that each individual i has a level of experi-
enced utility, which depends on M observable determinants xi1 , … , xiM, c j ¼ f c j q j1 …q jQ j : ð3Þ
such as the individual's income, his social contacts or his age.
An individual i's ecological footprint EFi is calculated as the sum of
ui ¼ f i ðxi1 ; …; xiM Þ ð1Þ
his scores on each of the components cij:
self-evaluation questions were inspired by the work of Anand and van EF component Correlation with SWB
Hees (2006). They had to indicate to what extent they agreed with Food 0.018
the following questions using a scale from 0 (completely disagree) to Heating 0.059
10 (completely agree): Electricity −0.066*
Paper −0.036
− I have a good social life (e.g. having friends, being part of associa- Car use 0.029
Public transport 0.014
tions); Holidays 0.108**
− I live and spend my time in pleasant environments (e.g. house, work,
⁎⁎ p b 0.01.
physical environment); and ⁎ p b 0.05.
− I consider myself to be in good physical health.
Table 3 Table 4
Results of a stepwise OLS regression for SWB using the EF-items. OLS regression models for SWB and EF.
We present unstandardized coefficients, as all included variables are dummies the effects Residence
are comparable. Rural 0.014 0.631 0.106 0.001
Self-evaluations
Health 0.130 0.000 −0.011 0.761
Social life 0.124 0.000 0.066 0.095
Pleasant environment 0.159 0.000 −0.008 0.840
(reducing EF and increasing SWB11). All other variables from Table 3 re- Personality traits
Positive attitude 0.152 0.000 −0.008 0.808
duce the EF but are at the same time decreasing SWB: a limited con-
Altruistic and conscientious 0.063 0.032 −0.046 0.169
sumption of meat or fish, living in a small apartment or house, having Emotionally concerned −0.074 0.011 −0.025 0.455
or using no car and spending holidays at home or b200 km from Expectant 0.047 0.112 0.068 0.043
home.12
Attitudes towards the environment
Income is known to be an important driver of the ecological foot- Concern about environmental issues 0.040 0.172 −0.101 0.002
print (see, for instance, Csutora, 2012) so that some of the observed re- Prob N F = 0.000 Prob N F = 0.000
lations in the model may just be income effects. Therefore, we did a Adj. R2 = 0.274 Adj. R2 = 0.095
robustness check including income as a fixed control variable in the re-
gression that supports the conclusions in the paragraph above. The
same holds when also including age and gender as additional control
variables.13
X X X X
EF ¼ γ 0 þ j
γ oj o j þ j
γlj l j þ j
γtj t j þ j
γaj a j þ ε: ð7Þ
4.2. Indirect Relation: Effect of SWB-determinants on the Ecological
Footprint
Table 4 presents the standardized coefficients (with their signifi-
Now we deal with the indirect effects of SWB-determinants on the cance levels) estimated using an OLS procedure.14 The table includes
ecological footprint. There are two specifications: one for SWB and only those variables that are significant for either SWB or the EF, or for
one for the EF: both, in a stepwise OLS regression with all determinants available (see
Section 3.3).15 As a robustness check, Appendix 4B shows the table
X X X X with all determinants available.
s ¼ β0 þ βoj o j þ βlj l j þ βtj t j þ βaj a j þ ε ð6Þ
j j j j Concerning the estimation methodology, the same remark as above
applies: the coefficients in Table 4 are best interpreted as conditional
correlations. In Appendix 3 we present the correlation matrix. This
shows that many of the well-being determinants are to a certain extent
11
related (e.g. socio-economic position and age, education and income,
Electricity as heating is used by only 10% of families in Belgium in 2013 (11% in 2012)
etc.).
and more often in apartments or terraced houses (Vreg, 2013). The fraction of gas as
heating increased from 62% in 2012 to 66% in 2013 (Vreg, 2013). A possible explanation
why using electricity as heating for the house decreases well-being is that electricity is
used more frequently for heating in less recent and smaller houses. The fact that for our
sample people using electricity as heating report a lower average (7.38 vs 7.65;
14
p = 0.006) on the question ‘to what extent they live in a pleasant environment’, supports Van Praag and Ferrer-i-Carbonell (2010) have a theoretical discussion on the estima-
this hypothesis. tion methods in happiness economics and conclude: “It follows that we do not have to be
12
These aspects (except food) were not included in the analysis of Welsch and Kühling very anxious on which particular estimation method is used” (p. 15). The OECD guidelines
(2011) which was restricted to organic food, low energy light bulbs, low energy household (2013a) further support this view but also suggest to always examine the differences be-
appliances, solar thermal heating systems and green electricity. tween OLS and ordered probit. Also for our results, the conclusions are similar when using
13
When including income as a fixed control variable in the estimation of SWB in Table 3, an ordered probit model for subjective well-being. The sign of the coefficients was always
the results for the EF components are similar and income is not significant (p = 0.089). As the same. Regarding significance levels there was only a difference for the dummy ‘stu-
identifying common determinants for well-being and the footprint is the focus of the next dent’, which was below the 5% significance threshold in the OLS model (p = 0.037) but
section, we did not include all SWB determinants in this model. The extended SWB model slightly above it in the ordered probit model (p = 0.063). Moreover the skewness
(adding all significant variables from Table 4 to the stepwise model) is presented in Ap- (−0,994) and kurtosis (1,913) of the SWB variable are acceptable to justify the normality
pendix 4A. As the impact of most EF aspects on SWB is driven by specific behaviour or cir- assumption (George and Mallery, 2010; De Vaus, 2002).
15
cumstances of people, most of the direct effects disappear: only the negative impacts of The stepwise regression results are not reported. They provide similar results with the
mainly eating ‘frozen vegetables and fruit preserves’ and living in a ‘small apartment or exception of three significance levels for socio-economic position in the SWB model. In the
terraced house’ are withheld. As income was not significant in Table 4, it was not included stepwise model both ‘incapable to work’ and ‘househusband/wife’ are significant at 5%
in Appendix 4A. However, the robustness check in appendix 4B, including all possible de- level while in the presented results this is only at 10% level. On the other hand ‘being a stu-
terminants, illustrates the non-significant (p = 0.970) impact of income on SWB. dent’ was not significant at 5% level in the stepwise model.
E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89 85
Table 4 shows that being unemployed has a negative impact on SWB Table 5
but not on the EF. As a result, we consider this as lose-neutral variable Summary of the results.
(and not as lose–lose variable as hypothesised in Table 1). Being a pen- Win–win or lose–lose variables Trade-off variables
sioner is also a lose-neutral variable. Because pensioners live alone more Beneficial for SWB and EF: Beneficial for SWB but harmful for EF:
often (the average number of people living in the house of a pensioner is
▪ Consumption of seasonal and ▪ Satisfying social life (only significant
1.8 compared to 3.1 for the complete sample) they have a higher EF. On
fresh products at 10% level)
the other hand, the SWB of pensioners is not different from that of ▪ Having a relation Beneficial for EF but harmful for SWB:
working people. ▪ House ownership
▪ A limited consumption of meat
‘Having a relation’ and being a ‘House owner’ are both win–win var- Harmful for SWB and EF:
of fish
iables that positively influence SWB while reducing the EF. Having a re- ▪ Using electricity as heating ▪ Living in a small apartment or
lation’ increases SWB and it also implies (mostly) that one is not living for the house house
alone such that the EF reduces. House-owners have a lower EF because ▪ Having or using no car
they pay more attention to insulation and more often use green energy ▪ Spending holidays at home or
b200 km from home
(probably also because the government gave subsidies to households Win- neutral variables Lose- neutral variables
installing solar panels). We do not find the expected trade-off character Beneficial for EF and neutral for SWB: Harmful for EF and neutral for SWB:
of income: while respondents with a higher income level do have
▪ Concern about environmental ▪ Living in rural areas
a higher EF, the impact of income on SWB is not significant in the issues ▪ Pensioners
regression model. Consequently, income is found to be a lose-neutral Beneficial for SWB and neutral for EF: ▪ Income
variable.16 The educational dummies have no significant relation with Harmful for SWB and neutral for EF:
▪ A positive attitude
either of the outcome variables. ▪ Altruistic and conscientious ▪ Unemployed
Living in rural areas is a lose-neutral variable as it increases the EF ▪ Living in a pleasant environment ▪ Emotionally concerned
and has no influence on SWB. A potential explanation for the EF effect ▪ Good health
is that people from rural areas live more often in a large (semi-) de-
tached house (39.3% for people from rural areas compared to 28% for
the complete sample). They also use their car more often (86.5% of
those living rural use their car daily compared to 76.5% for those living
elsewhere).
Being healthier and living in a more pleasant environment increase priority when designing policies to influence human behaviour towards
subjective well-being without influencing the EF. Consequently we con- sustainability. The following policy implications can be formulated
sider these variables as win-neutral. If ‘social life’ is interpreted as proxy based on our empirical findings:
for recreation and leisure, we expected a trade-off character (see hy-
pothesis in Table 1) however for the EF social life is only significant at ▪ increase the consumption of local and seasonal products and de-
10% level. crease consumption of frozen vegetables and fruit preserves;
Contrary to our hypothesis, being ‘altruistic and conscientious’ is ▪ stimulate co-housing (this is the general idea behind ‘having a part-
only beneficial for SWB (no impact on the EF). In general the personality ner’ and also it is why pensioners have a higher EF);
traits are more relevant for SWB than for the EF. ▪ avoid electricity for the heating of houses. (see fn 10 for comments
The variable ‘concern about environmental issues’ impacts negative- on this result); and
ly on the EF without influencing SWB, so that this variable is found ▪ make homeownership more appealing (or adapt energy efficiency
to be win-neutral. The variable ‘knowledge about environmental norms for the rental market).
problems’ has no significant relation with either of the outcome
variables.
Our analysis also identified a number of trade-off variables that
5. Discussion and Policy Implications reduce the EF, yet they do so at the cost of a lower SWB. Examples
include having no car or not using a car, staying at home or nearby
Identifying variables that simultaneously increase and individual's during holidays and living in a small house or an apartment. A crucial
level of subjective well-being and reduce his or her ecological footprint question for these variables is how to convince people to take into
is a research question with high relevance for policy making. Given the account the long run, when most decisions are taken with a short-
urgent need to promote more sustainable lifestyles for the sake of future term mind set. “Policymakers, citizens and organizations need informa-
generations (see, for instance, Stiglitz et al., 2009) policy-makers should tion about what sustains well-being over time, to help guide decision-
(at least) focus on ways to reduce our environmental impact that are making in the present” (OECD, 2013b, p. 175). The proposals formulated
not detrimental for our current level of well-being. The aim of our above can be regarded as a first step, although one can question if
paper was to combine information on reported subjective well-being such policies will be sufficient to reconcile subjective well-being and
with information on the ecological footprint at the individual level. sustainability. Also, other views on well-being (e.g. taking a point of
Combining information in this way for one and the same sample has view that goes ‘beyond’ subjective or individual well-being) could
not yet been done to date (Lenzen and Cummins, 2013). lead to further insights on the relation between the individual and the
Table 5 summarizes the results of our empirical analysis. Four vari- planet.
ables are either win–win or lose–lose. If possible, these should get As noted above, a limitation of our approach has to do with the
quantitative methodology: an OLS strategy does not allow to establish
a causal link between the independent and the dependent variables.
Consequently, our results are best interpreted as conditional correla-
tions. To the extent that these correlations also mean causality, our re-
sults can be used to design win–win policies. An analysis that
16
Note that the correlation matrix in Appendix 3 shows the expected bivariate correla- establishes this causality (e.g. by using instrument variables or panel
tions for income (beneficial for SWB but increasing the EF). The difference with the results
in Table 4 can be explained by an indirect positive effect on SWB of more income through
data) could improve the strength of our arguments. It can also be an in-
the socio-economic position, house ownership, better health and lower emotional teresting topic for future research to check the direction of the effect: is
concerns. it SWB that influences ecological behaviour, or does an ecological way of
86 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89
life lead to a higher SWB (or is there a third factor influencing both). This Appendix 2. Survey Questions for the Ecological Footprint
could also be done by performing in-depth interviews with people to in-
vestigate the motives for specific behaviours. (The reference categories for the aspects used for the model in
Another topic for further research is the challenge of developing a Section 4.1 – Table 3 – are indicated in bold).
more comprehensive analysis: we should also consider the long-run ef-
fects and relationships between sustainability and well-being. Using in-
dividual data (for the calculation of the EF) has as drawback that it
ignores that broader perspective. For example ‘having more children’
reduces the individual EF as the ecological burden per capita is
mainly fresh products
lower for a larger household, although from a long-term perspective
ever more people on this planet is not sustainable. Very often, it is
difficult for the individual to realize that his personal behaviour
(and corresponding EF) contributes to a globally increasing EF that
can provoke severe problems as e.g. extreme climate events. For in-
stance global warming will very likely have a negative impact on well- 4-5 times a week
being through reduced health and feelings of safety and security
(Sekulova, 2013). This makes that the relation (link or trade-off) in gen-
eral between well-being and sustainability is much more complex than
suggested here when (only) considering the link between individual
SWB and the EF.
Mean/proportion SD Range
Educational level
you do not specifically pay attention to your electricity consumption
Lower educated 28.4%
Higher secondary education 35.1%
Higher educated (bachelor or master) 36.5%
Socio-economic characteristics
Age 50 18.06 18–93
Gender (% females) 50.9%
mainly white paper, advertising, magazines and / or newspapers
Being religious (% yes) 75.4%
Having a relation (% yes) 79.2%
Personal income (in thousands €) 1.52 0.87 0–3.25
House owner (% yes) 83.6%
Number of people living in the house 3.11 1.39 1–8
Residence
daily for less than 50km
Centre of municipality 30.6%
City centre 13.0%
Rural 33.0%
Outskirts of town 23.3%
Self-evaluations
Health 7.38 1.63 0–10
Social life 7.41 1.43 0–10
daily for less than 50km
Pleasant environment 7.60 1.24 2–10
Personality traits
Positive attitude −0.04 1.00 −3.29–3.03
Altruistic and conscientious −0.004 1.00 −3.67–2.59
Emotionally concerned 0.06 1.01 −3.82–2.73
Expectant 0.003 1.02 −2.92–3.74
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26)
SWB 0.04 0.03 −0.15 −0.14 −0.06 0.03 0.02 −0.02 0.23 −0.06 0.10 0.11 0.18 0.07 −0.02 0.03 −0.02 0.25 −0.14 0.16 0.09 0.31 0.31 0.36 0.05 0.11
EF −0.05 0.18 −0.02 −0.03 −0.09 0.11 0.14 −0.07 −0.16 0.09 −0.01 0.09 −0.07 0.02 −0.08 0.09 −0.03 −0.03 −0.07 −0.02 0.03 −0.02 0.02 −0.02 −0.02 −0.12
Student (1) 1.00 −0.14 −0.05 −0.05 −0.05 −0.40 −0.32 −0.01 −0.18 −0.14 −0.05 −0.35 −0.04 −0.10 −0.06 0.03 0.05 0.01 −0.01 −0.13 0.03 0.07 0.09 0.02 0.01 −0.04
Pensioner (2) 1.00 −0.12 −0.11 −0.13 0.77 0.84 0.07 −0.17 0.50 −0.26 −0.16 0.08 0.20 0.05 −0.06 −0.06 −0.07 0.04 0.16 −0.18 −0.14 −0.04 0.02 −0.14 −0.06
Unemployed (3) 1.00 −0.04 −0.05 −0.14 −0.13 0.00 −0.06 −0.04 −0.03 −0.20 −0.12 −0.12 0.04 −0.03 −0.02 0.01 −0.01 −0.08 −0.08 −0.02 −0.01 −0.11 0.01 −0.04
Incapable (4) 1.00 −0.04 −0.06 −0.08 0.01 −0.02 −0.02 −0.03 −0.08 −0.08 0.00 −0.02 −0.01 −0.01 −0.03 0.08 −0.01 −0.03 −0.26 −0.03 −0.10 0.06 0.02
Housewife (5) 1.00 −0.03 −0.06 0.19 0.09 0.00 −0.07 −0.30 0.05 0.05 −0.02 0.02 −0.01 −0.01 0.13 −0.04 −0.02 0.03 −0.01 0.02 0.00 −0.02
Age (6) 1.00 0.98 0.08 −0.01 0.48 −0.21 0.10 0.20 0.25 0.04 −0.02 −0.06 −0.11 0.04 0.26 −0.13 −0.19 −0.13 −0.02 −0.11 0.00
87
88 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89
A. Stepwise model for SWB from Table 3 including all significant variables from Table 4.
Socio-economic position
Student 0.451 0.190 0.070 0.018
Unemployed −0.540 0.218 −0.072 0.013
Incapable to work −0.540 0.236 −0.068 0.022
Househusband/wife −0.540 0.205 −0.076 0.009
Socio-economic characteristics
Being religious (yes) 0.271 0.101 0.078 0.007
Having a relation (yes) 0.655 0.110 0.179 0.000
House owner (yes) 0.291 0.121 0.073 0.017
Self-evaluations
Health 0.095 0.031 0.100 0.002
Social life 0.143 0.036 0.139 0.000
Pleasant environment 0.197 0.043 0.164 0.000
Personality traits
Positive attitude 0.229 0.045 0.151 0.000
Altruistic and conscientious 0.097 0.044 0.066 0.026
Emotionally concerned −0.116 0.043 −0.078 0.007
Prob N F = 0.000
Adj. R2 = 0.286
B. OLS regression models for SWB and EF including all possible determinants.
SWB EF
Unst. coeff. St. coeff. Sig. Unst. coeff. St. coeff. Sig.
Socio-economic position
Student 0.449 0.238 0.070 0.059 −0.430 0.335 −0.054 0.199
Pensioner 0.126 0.216 0.037 0.560 0.959 0.303 0.226 0.002
Unemployed −0.507 0.230 −0.069 0.028 0.100 0.324 0.011 0.758
Incapable to work −0.415 0.246 −0.052 0.092 −0.018 0.342 −0.002 0.958
Househusband/wife −0.498 0.238 −0.069 0.036 0.048 0.330 0.005 0.885
Socio-economic characteristics
Age −0.005 0.017 −0.060 0.768 −0.038 0.024 −0.362 0.118
Age2 0.000 0.000 0.090 0.678 0.000 0.000 0.306 0.215
Gender (woman) 0.101 0.099 0.034 0.307 −0.079 0.138 −0.021 0.570
Being religious (yes) 0.282 0.102 0.082 0.006 −0.033 0.144 −0.008 0.821
Having a relation (yes) 0.668 0.114 0.183 0.000 −0.632 0.160 −0.140 0.000
Lower educated −0.017 0.125 −0.005 0.892 0.125 0.175 0.030 0.474
Higher educated 0.118 0.106 0.038 0.267 0.090 0.148 0.024 0.546
Personal income (in thousands €) 0.003 0.073 0.002 0.970 0.337 0.102 0.160 0.001
House owner (yes) 0.423 0.122 0.105 0.001 −0.310 0.171 −0.063 0.070
Residence
Centre of municipality −0.171 0.140 −0.053 0.221 −0.135 0.194 −0.034 0.486
Rural −0.117 0.140 −0.037 0.400 0.397 0.194 0.101 0.041
Outskirts of town −0.219 0.148 −0.062 0.139 0.100 0.205 0.023 0.626
Self-evaluations
Health 0.123 0.031 0.131 0.000 −0.023 0.044 −0.019 0.607
Social life 0.121 0.036 0.118 0.001 0.068 0.051 0.054 0.178
Pleasant environment 0.188 0.044 0.155 0.000 −0.006 0.061 −0.004 0.926
Personality traits
E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89 89
(continued)
SWB EF
Unst. coeff. St. coeff. Sig. Unst. coeff. St. coeff. Sig.
Positive attitude 0.236 0.046 0.156 0.000 −0.024 0.064 −0.013 0.705
Altruistic and conscientious 0.103 0.045 0.070 0.021 −0.050 0.062 −0.027 0.421
Emotionally concerned −0.142 0.047 −0.094 0.002 −0.025 0.065 −0.014 0.700
Expectant 0.084 0.044 0.057 0.057 0.111 0.062 0.061 0.072
References Layard, R., Mayraz, S., Nickell, S., 2008. The Marginal Utility of Income. J. Public Econ. 92,
1846–1857.
Anand, P., van Hees, M., 2006. Capabilities and achievements: an empirical study. J. Socio- Lenzen, M., Cummins, R.A., 2013. Happiness versus the environment—a case study of
Econ. 35 (2), 268–284. Australian lifestyles. Challenges 4, 56–74.
Blanchflower, D., Oswald, A., 2004. Well-being over time in Britain and the United States. Li, Z., Folmer, H., Xue, J., 2014. To what extent does air pollution affect happiness? The
J. Public Econ. 88 (7/8), 1359–1386. case of the Jinchuan mining area, China. Ecol. Econ. 99, 88–99.
Blanchflower, D., Oswald, A., 2008. Is well-being U-shaped over the life cycle? Soc. Sci. Luechinger, S., Raschky, P.A., 2009. Valuing flood disasters using the life satisfaction ap-
Med. 66, 1733–1749. proach. J. Public Econ. 93, 620–633.
Büchs, M., Schnepf, S., 2013. Who emits most? Associations between socio-economic fac- MacKerron, G., Mourato, S., 2009. Life satisfaction and air quality in London. Ecol. Econ. 68,
tors and UK households' home energy, transport, indirect and total CO2 emissions. 1441–1453.
Ecol. Econ. 90, 114–123. Maddison, D., Rehdanz, K., 2011. The impact of climate on life satisfaction. Ecol. Econ. 70,
Cloutier, S., Jambeck, J., Scott, N., 2014. The sustainable Neighborhoods for Happiness 2437–2445.
Index (SNHI): a metric for assessing a community's sustainability and potential influ- Menz, T., Kühling, J., 2011. Population aging and environmental quality in OECD coun-
ence on happiness. Ecol. Indic. 40, 147–152. tries: evidence from sulfur dioxide emissions data. Popul. Environ. 33 (1), 55–79.
Csutora, M., 2012. One more awareness gap? The behaviour–impact gap problem. nef, 2012. Happy Planet Index: 2012 report—a global index of sustainable well-being,
J. Comsum. Policy 35, 145–163. 2012. Available online http://www.neweconomics.org/publications/happy-planet-
De Vaus, D., 2002. Analyzing Social Science Data: 50 Key Problems in Data Analysis. Sage, index-2012-report.
London. O'Donnell, G., Deaton, A., Durand, M., Halpern, D., Layard, R., 2014. Wellbeing and Policy —
Diener, E., 2000. Subjective well-being: the science of happiness and a proposal for a na- Full Report. Legatum Institute.
tional index. Am. Psychol. 55, 34–43. OECD, 2013a. OECD Guidelines on Measuring Subjective Well-being. OECD Publishing.
Dolan, P., Peasgood, T., White, M., 2008. Do we really know what makes us happy? A re- OECD, 2013b. How's Life?: Measuring Well-being. OECD Publishing.
view of the economic literature on the factors associated with subjective well-being. Rehdanz, K., Maddison, D., 2005. Climate and happiness. Ecol. Econ. 52, 111–125.
J. Econ. Psychol. 29 (1), 94–122. Rehdanz, K., Maddison, D., 2008. Local environmental quality and life-satisfaction in
Druckman, A., Jackson, T., 2009. The carbon footprint of UK households 1990–2004: A Germany. Ecol. Econ. 64, 787–797.
socio-economically disaggregated, quasi-multi-regional input–output model. Ecol. Sekulova, F., 2013. On the Economics of Happiness and Climate Change (Doctoral thesis)
Econ. 68, 2066–2077. Universitat Autonoma de Barcelona.
Ericson, T., Kjønstad, B.G., Barstad, A., 2014. Mindfulness and sustainability. Ecol. Econ. Stiglitz, J.E., Sen, A., Fitoussi, J.-P., 2009. Report by the Commission on the Measurement of
104, 73–79. Economic Performance and Social Progress. Commission on the Measurement of Eco-
Ferreira, S., Akay, A., Brereton, F., Cunado, J., Martinsson, P., Moro, M., Ningal, T.F., 2013. nomic Performance and Social Progress.
Life satisfaction and air quality in Europe. Ecol. Econ. 88, 1–10. Van Praag, B.M.S., Ferrer-i-Carbonell, A., 2010. Happiness economics: a new road to mea-
Ferrer-I-Carbonell, A., Gowdy, J.M., 2007. Environmental degradation and happiness. Ecol. suring and comparing happiness. Found. Trends Microeconomics 6 (1), 1–97.
Econ. 60, 509–516. Veenhoven, R., 1991. Is happiness relative? Soc. Indic. Res. 24 (1), 1–34.
Gardner, G., Stern, P., 2002. Environmental Problems and Human Behavior. Pearson Cus- Vreg, 2013. Resultaten enquête particulieren 2013. Gedrag & ervaringen van
tom Publishing, Boston, MA. huishoudelijke afnemers op de vrijgemaakte Vlaamse energiemarkt. RAPP-2013-16.
George, D., Mallery, P., 2010. SPSS for Windows Step by Step: A Simple Guide and Refer- Welsch, H., 2006. Environment and happiness: valuation of air pollution using life satis-
ence, 17.0 Update. Pearson, Boston. faction data. Ecol. Econ. 58, 801–813.
GFN (Global Footprint Network), 2012. www.footprintnetwork.org/en/index.php/GFN/ Welsch, H., Kühling, J., 2011. Are pro-environmental consumption choices utility-
page/glossary (accessed 14/07/2014). maximizing? Evidence from subjective well-being data. Ecol. Econ. 72, 75–87.
Gosling, S.D., Rentfrow, P.J., Swann, W.B., 2003. A very brief measure of the Big-Five per- WWF, 2012. Living planet. Report 2012. Biodiversity, biocapacity and better choices.
sonality domains. J. Res. Pers. 37 (6), 504–528. Available online http://wwf.panda.org/about_our_earth/all_publications/living_
Helliwell, J., Layard, R., Sachs, J., 2012. World Happiness Report. The Earth Institute. Co- planet_report/.
lumbia University, New York (NY). WWF, 2014. Living Planet Report 2014: Species and Spaces, People and Places. WWF In-
Helliwell, J., Layard, R., Sachs, J., 2013. World Happiness Report 2013. The Earth Institute. ternational, Gland, Switzerland.
Columbia University, New York (NY).
Kerkhof, A., Benders, R., Moll, H., 2009. Determinants of variation in household CO2 emis-
sions between and within countries. Energ Policy 37, 1509–1517.