Ecological Economics 127 (2016) 80–89

Contents lists available at ScienceDirect

Ecological Economics

journal homepage: www.elsevier.com/locate/ecolecon

Analysis

Linking individuals' ecological footprint to their subjective well-being☆

E. Verhofstadt a,⁎, L. Van Ootegem a,b, B. Defloor a, B. Bleys c
a
SHERPPA, Ghent University, Belgium
b
HIVA – University of Louvain, Belgium
c
Department of General Economics, Ghent University, Belgium

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
There is widespread attention and an urgent need for a more sustain-
able development and for sustainability transitions, especially in Western
countries. The lifestyles of citizens in these countries have significant eco-
logical consequences in terms of resource and energy use. WWF (2014),
for instance, reveals that global society needs 1.5 earths to meet its current
demands on nature as measured by the ecological footprint. The Ecologi-
cal footprint per capita of high-income countries is about five times higher
than that of low-income countries. As a result, researchers (e.g. Gardner
and Stern, 2002) often argue that a series of behavioural changes is need-
ed across different policy fields, such as mobility, food and consumption,
production processes, etc. Policymakers are aware of these challenges
and are taking sustainability issues into account at various levels: at the
global level (e.g. the Post 2015 agenda), at the European level (e.g. sus-
tainable growth as a key goal in the Europe 2020 strategy) and at the
local level (e.g. the movement towards sustainable cities or regions).
Stiglitz et al. (2009) relate sustainability issues to the concept of subjec-
tive well-being by stressing the challenge of maintaining current levels of
well-being for future generations. Many of the necessary behavioural
changes have an impact on the way people live and on their well-being,
which might impede their acceptance by citizens. Policies that promote
☆ We would like to thank Ecolife vzw for calculating the ecological footprint of the
respondents in our survey.
⁎ Corresponding author at: Sint-Pietersplein 6, 9000 Gent, Belgium.
E-mail address: Elsy.Verhofstadt@UGent.be (E. Verhofstadt).
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
Helliwell et al. (2013, p. 108) that “the analytical community needs to help
us understand which policies would raise well-being in a sustainable way”.
We want to contribute to this call by integrating information about the en-
vironmental impact of individuals and their reported subjective well-being.
Therefore we will also examine and compare if people's demographic char-
acteristics are related to their well-being and their environmental impact.
In recent decades, the literature on subjective well-being (SWB) is
booming and self-reported data on happiness or life satisfaction is be-
coming increasingly available. Many studies summarize the main fac-
tors influencing subjective well-being, amongst others Blanchflower
and Oswald, 2004; Dolan et al., 2008; Stiglitz et al., 2009; Helliwell
et al., 2012; O'Donnell et al., 2014. Key factors include age, income,
education, work, family life, social capital, religion, community and
governance, mental health and physical health. Furthermore, the im-
portance of personality traits, adaptation and expectations are indisput-
able (e.g. Veenhoven, 1991; Diener, 2000; Helliwell et al., 2012).
We aim to compare information on subjective well-being to informa-
tion on the environmental sustainability of an individual's lifestyle. A fre-
quently used measure for the (un)sustainability of individual behaviour
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
required to absorb an individual's CO2 emissions – i.e. his carbon foot-
print (CF). In the literature considerable attention is devoted to the CF
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
for the relation between household income and CO2 emissions: there is a
negative relation between income and emissions in the Netherlands and
the UK, whereas emissions increase with income in Sweden and Norway.
Büchs and Schnepf (2013) show that besides income, also household
size and composition is highly relevant for carbon emissions. Larger
households have higher emissions in general, e.g. due to heating and
laundry, but lower emissions per capita. The authors also find higher
emissions for elderly or unemployed people (who spend more time at
home), for the higher educated and for those living in rural areas
(Büchs and Schnepf, 2013). Also Menz and Kühling (2011) point to the
important role of age. Druckman and Jackson (2009) focus on the in-
creased recreation, leisure and household activities (such as commuting,
eating, heating, …) as important drivers of the increase in CO2-emissions.
The relationship between sustainability and well-being is not com-
monly explored in the literature. When there is a link to the environ-
ment or to sustainability in the subjective well-being literature, the
focus is often on the impact of the state of the environment. Tempera-
ture is highly significant (positive), both for happiness (Rehdanz and
Maddison, 2005) and for life satisfaction (Maddison and Rehdanz,
2011). Many studies confirm the negative impact of air pollution
on life satisfaction (Welsch, 2006; Rehdanz and Maddison, 2008;
MacKerron and Mourato, 2009; Ferreira et al., 2013; Li et al., 2014). Fi-
nally, natural disasters such as floods also have a significant negative
impact on life satisfaction (Luechinger and Raschky, 2009). Besides
looking at the state of the environment, some subjective well-being
studies also look at the impact of an individual's attitudes towards the
environment on his or her reported subjective well-being. Ferrer-I-
Carbonell and Gowdy (2007), for instance, find a negative relation
between the concern for ozone pollution and life satisfaction and a pos-
itive relation for concern about species extinction. In studies on the eco-
logical or carbon footprint of individuals the focus lies exclusively on
household characteristics and behaviour (see Druckman and Jackson,
2009; Kerkhof et al., 2009; Büchs and Schnepf, 2013), so that a direct
link to subjective well-being is absent in that literature.
To our knowledge there are three studies that, to some extent, com-
bine information on the EF and subjective well-being. Welsch and
Kühling (2011) show that, for a German sample, environmentally
friendly consumption1 raises the reported life satisfaction of individuals.
Their measure of environmentally friendly behaviour contains some el-
ements related to the EF, but it is not a comprehensive measure of the
environmental impact individuals have through their behaviour. Next,
Ericson et al. (2014) indicate that empathy, compassion and non-
materialistic values contribute to both subjective well-being and to a
more sustainable way of life. Finally, Lenzen and Cummins (2013) inte-
grate two surveys from different samples of the Australian population —
the Australian Unity Wellbeing Survey and the Australian Household
Expenditure Survey — and identify common determinants for subjec-
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
set of determinants. Lenzen and Cummins (2013) find a set of variables
that are beneficial for well-being, but have an adverse influence for the
footprint: income, car ownership/car use, education/qualifications and
recreation and leisure. They also find that the more educated report to
be more environmentally concerned, yet this does not result into acting
more pro-environmentally.
Lenzen and Cummins (2013) indicate that, at present, there exist no
data that combine information on individuals' subjective well-being
and their ecological footprint.2 Our study aims to address this issue, as
1 4
Environmentally friendly consumption is measured using five aspects: organic food,
low energy light bulbs, low energy household appliances, solar thermal heating systems
and green electricity.
2
At macro level, the New Economics Foundation (nef (2012)) links the Ecological Foot-
print to well-being in the “happy planet index (HPI)” and Cloutier et al. (2014) assess the
happiness impact of sustainable practises in communities or cities. However, these studies
do not contain an analysis at individual level, including socio-economic and demographic
information.
81
we have collected that information for one and the same population.
As a result we can investigate whether there is a direct link between
(a) SWB and (b) the EF and its components. Moreover, in line with
the approach of Lenzen and Cummins (2013), we want to investigate
if there is an indirect link: which socio-demographic determinants
have an impact on both the individuals' SWB and their EF? We aim to
identify common determinants with a positive effect on subjective
well-being and a negative effect on the footprint. Our objective is also
to formulate policy advice based on these results.
In the next section, we provide an overview of our research ques-
tions and hypotheses. Section three presents the concepts, the survey
and the data while section four contains the empirical results. In section
five, we conclude.
2. Research Question and Hypotheses
We first explore the direct link between an individual's ecological
footprint and his or her subjective well-being. Afterwards we link
SWB to the components of the footprint: food intake, energy use
(heating and electricity), paper use, car use, use of public transportation
and travel behaviour. Finally, we investigate the indirect link between
SWB and the EF in the sense that we identify a number of common de-
terminants for both concepts (see Lenzen and Cummins (2013)). If such
determinants influence the EF and SWB in opposite directions, they can
create win–win situations in the sense that they can positively influence
SWB while at the same time reducing the EF. When win–win variables
have policy relevance, we can formulate policy advice about the possi-
bility of creating double dividends. Other variables or policies could cre-
ate a lose–lose situation. In both these cases, be it win–win or lose–lose,
policy implications are evident. We also expect to find variables that in-
fluence both the EF and SWB in the same direction. They are beneficial
for one outcome but have an adverse effect on the other and thus entail
a trade-off for the policymaker. This might be the case for many of the
variables from the carbon footprint literature (e.g. income and car own-
ership). Finally, we will also have win-neutral (or lose-neutral) vari-
ables: determinants that have a positive (or negative) and significant
impact on only one of both outcomes.3
Based on the SWB and EF/CF literature we have formulated a number
of research hypotheses that can be found in Table 1. In the table we only
include determinants that have been studied in either (a) both research
fields or (b) one of the three studies (cfr. supra: Lenzen and Cummins,
2013; Ericson et al., 2014; Welsch and Kühling, 2011) that combine
both research fields. In our analyses we will also consider determinants
that appear only in the literature of one of the fields (SWB or EF/CF), yet
we do not formulate a priori hypotheses for these variables.
3. Data
Our analysis is based on self-reported information obtained from a
survey (N = 1286) in Flanders (the northern Dutch speaking part of
Belgium) in 2013.4 In the sample, participants are weighted to achieve
similarity to the (univariate) frequency distributions in the Flemish
population (data from the Labour Force Survey 2012) for life situation,
gender and age.5 In line with the population, the weighted sample
3
For the interpretation of the results, only variables that have a statistically significant
effect are important. We included the concept ‘win- or lose-neutral’ for those determi-
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.
We call the survey LEVO 2013 which is short for LEvensomstandigheden in Vlaande-
ren Onderzocht – Dutch for “Inquiry into the life circumstances in Flanders”. The LEVO sur-
vey has been carried out yearly since 2010 with a focus on subjective well-being measures
and their determinants. The 2013 edition included an additional series of questions on the
Ecological Footprint.
5
Chi-Square Goodness of Fit Tests show that the univariate distributions for the weight-
ed sample equal the expected ones (from the Labour Force Survey) for life situation
(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.
Win–win or lose–lose variables
Beneficial for SWB and EF:
▪ Living together with people: SWB literature + Büchs and Schnepf
(2013); Lenzen and Cummins (2013)
▪ Altruistic and conscientious: Ericson et al. (2014)
▪ Environmentally friendly consumption: Welsch and Kühling (2011)
Harmful for SWB and EF:
▪ Unemployed: SWB literature + Büchs and Schnepf (2013)
consists of 49% male and 51% female respondents, with an average age
of 50 years old. The majority (55%) is working full-time or part-time,
the pensioners (27%) are the second largest group. Appendix 1 offers
detailed descriptive statistics for the weighted sample.
3.1. Concept and Measurement of Subjective Well-being
Our approach to linking utility and subjective well-being is similar to
Layard et al. (2008). Assume that each individual i has a level of experi-
enced utility, which depends on M observable determinants xi1 , … , xiM,
such as the individual's income, his social contacts or his age.
ui ¼ f i ðxi1 ; …; xiM Þ
When individual i is asked to report his subjective well-being (SWB),
the individual applies a strictly increasing function to ui to arrive at an
answer si. The answers individuals provide, come with a noise term εi
as not all determinants are observed or the answers might be influenced
by the individual's mood at the moment he is answering the question.
si ¼ g i ðui Þ þ εi
In line with the OECD guidelines on measuring subjective well-being
(OECD, 2013a) we use life satisfaction scores as our measure of subjec-
tive well-being (SWB). The question “All things considered how satis-
fied are you with your life in general?“is measured on a scale from 0
‘very unsatisfied’ to 10 ‘very satisfied’. Based on an assessment of evi-
dence on validity and relevance, on understood results and on devel-
oped policy uses, O'Donnell et al. (2014) conclude that this question is
to be preferred when measuring subjective well-being. For our sample
the average life satisfaction is 7.24 (with standard deviation 1.50).
3.2. Concept and Measurement of the Ecological Footprint
To measure sustainability we use the ecological footprint (EF), as
this composite indicator has emerged as the world's primary measure
of humanity's demand on nature. The EF is expressed in global hectares
(gha) and captures “how much area of biologically productive land and
water an individual, population or activity requires to produce all the re-
sources it consumes and to absorb the waste it generates, using prevail-
ing technology and resource management practises” (GFN, 2012). The
largest part of the EF is the carbon footprint that looks at the amount
of productive land that is required to sequester carbon dioxide emis-
sions. On average the CF makes up 55% of the total footprint worldwide
(WWF, 2012). The ecological footprint also includes built-up land, fish-
ing grounds, forests, crop land and grazing land.
The EF of all respondents in our dataset is calculated using Ecolife
vzw's footprint calculator that was developed according to the WWF-
methodology (World Wildlife Fund) and draws on data related to
seven footprint components: c1 food intake, energy use (c2 heating
and c3 electricity), c4 paper use, c5 car use and c6 public transportation
Trade-off variables
Beneficial for SWB but harmful for EF:
▪ Elderly: SWB literature + Menz and Kühling (2011)
▪ Higher income: Lenzen and Cummins (2013)
▪ Car ownership/more car use: Lenzen and Cummins (2013)
▪ Higher education: SWB literature + Büchs and Schnepf (2013); Lenzen and Cummins (2013)
▪ Recreation and leisure: SWB literature + Druckman and Jackson (2009); Lenzen and
Cummins (2013)
▪ Living in rural areas: Dolan et al. (2008) + Büchs and Schnepf (2013)
use and c7 travel behaviour. The family composition is taken into ac-
count where relevant to rescale family outcomes into individual im-
pacts (for heating, paper use, car use, use of public transportation and
travel). Appendix 2 provides an overview of the different EF questions
that feed into Ecolife vzw's footprint calculator. With qjk the answer to
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
questions to the calculation of the components:


c j ¼ f c j q j1 …q jQ j : ð3Þ
An individual i's ecological footprint EFi is calculated as the sum of
ð1Þ
his scores on each of the components cij:
X7
E Fi ¼ c :
j¼1 ij
ð4Þ
For the individuals in our sample, the EF ranges between 2.87 and 17
(with an average of 6.94) global hectares. This is very much in line with
EF data for Belgium (2008). According to WWF (2012), the average per
ð2Þ capita footprint in Belgium is 7.11 gha (4.72 gha is the EU average,
2.7 gha is the world average).
3.3. Determinants of Subjective Well-being and the Ecological Footprint
In order to analyse the indirect relation between SWB and the EF, a
number of socio-demographic variables are selected from the empirical
literature on SWB and the EF, as reviewed supra in Section 1 and
Section 2 (Table 1). Some of the determinants are measured objectively
while others were collected using self-assessments.6
Individual i's score on objective indicator m is denoted oim. The fol-
lowing objective variables are taken into account:
- personal situation of the respondent (student, pensioner, unem-
ployed, person incapable of working, housewife and employed);
- income;
- house ownership;
- the educational degree obtained by the respondent;
- number of people living in the house;
- being religious;
- having a relation;
- gender;
- age and age squared7; and
- specification of residence: centre of municipality, city centre, rural or
outskirts of town.
6
Stiglitz et al. (2009) state “Objective and subjective dimensions of well-being are both
important” (p. 15)
7
Age squared is included to capture the possibility that life satisfaction is U-shaped over
the life cycle (see for example Blanchflower and Oswald, 2008).
 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89
Respondents were asked to make self-evaluations of specific aspects
of their lives (social life, living environment and physical health). These
self-evaluation questions were inspired by the work of Anand and van
Hees (2006). They had to indicate to what extent they agreed with
the following questions using a scale from 0 (completely disagree) to
10 (completely agree):
− I have a good social life (e.g. having friends, being part of associa-
tions);
− I live and spend my time in pleasant environments (e.g. house, work,
physical environment); and
− I consider myself to be in good physical health.
Individual i's score on self-evaluation question n is denoted lin.
Next, the respondents were asked to rate themselves on twelve per-
sonality traits: extravert, altruistic, conscientious-dutiful, worried, hard
to please, creative, optimistic, self-confident, emotional, progressive,
jealous and having high expectations. The answering format was a nu-
meric 7-point scale (coded from − 3 to + 3). These traits combine
both traditional personality traits (i.e. the Big Five personality traits)
and traits related to expectations and adaptation. The format of the per-
sonality questions (a 7-point scale, based on descriptors, neutral pre-
sentation) was based on Gosling et al. (2003). In order to reduce the
number of personality traits in the regression analysis and because of
their mutual correlation, we performed a principal component analysis
on the twelve single items. This resulted in four factors for which we
choose the following catch-all names:
− ‘positive attitude’ (extravert, creative, optimistic, self-confident, pro-
gressive);
− ‘altruistic and conscientious’ (altruistic, not jealous, conscientious –
dutiful);
− ‘emotionally concerned’ (worried, emotional); and
− ‘expectant’ (hard to please, high expectations).
Individual i's score on personality trait p is denoted tip.
Finally, we also include information on the individuals' attitudes to-
wards the environment: their knowledge about environmental prob-
lems and concern for environmental issues. The concern question was
relevant for the results of Ferrer-I-Carbonell and Gowdy (2007), the
knowledge question is new. Both questions were rated on a scale from
0 (very little knowledge respectively not worried at all) to 10 (very
much knowledge respectively very worried). Individual i's score on atti-
tude question r is denoted air.
4. Direct and Indirect Relations Between Footprint and Well-being
First, we examine whether there is a direct relation between the EF
and SWB. Afterwards, we focus on their indirect relation by identifying
common determinants that have a similar or a diverging impact on
well-being and on the footprint.
4.1. Direct Relation Between Well-being and the Ecological Footprint
(and Components)
Our results show no significant correlation between the EF and SWB
(r = −0.012; p = 0.719). So having a lower footprint is not associated
with reporting a higher level of well-being. In a more positive view, one
can also interpret this as that having a lower footprint does not reduce
one's level of subjective well-being. Table 2 shows the correlation be-
tween SWB and the different components of the EF. For two compo-
nents of the EF there is a small but significant correlation: being more
wasteful in electricity consumption correlates negatively with life satis-
faction (this effect disappears in the multivariate analysis of the next
83
Table 2
Pearson correlation of SWB with the different components of the EF.
EF component Correlation with SWB
Food 0.018
Heating 0.059
Electricity −0.066*
Paper −0.036
Car use 0.029
Public transport 0.014
Holidays 0.108**
⁎⁎ p b 0.01.
⁎ p b 0.05.
section) while travelling further for holidays (and thus being less sus-
tainable) is positively correlated with life satisfaction.8
To investigate the direct relation between SWB and the EF compo-
nents, we assume a linear relationship for expressions (1) and (2)
above, which is standard in the literature. The expression we estimate
has the form (dropping the individual subscripts for ease of notation):
X
s ¼ α0 þ j
α qj q j þ ε: ð5Þ
Table 3 presents the results of a stepwise OLS regression for subjective
well-being using the answers to the EF-questions as explanatory vari-
ables. We use the original responses to the EF-questions and not the EF
components (expressed in gha) as it is the behaviour as such that is rel-
evant for SWB. The original EF-questions are coded into dummies (see
Appendix 2 for the reference categories), so that each of the qj in
Eq. (5) is a dummy and each parameter αqj measures the impact on
SWB. We opt for a stepwise OLS regression (variables are included
when p-value b 0.05), as the number of dummy-variables to capture all
items is very large and only a limited number of them are significant.9
We acknowledge that, to establish a causal relationship between the in-
dependent variables and the dependent variable, an instrumental
variable (IV) strategy or panel data are more appropriate. However, it is
not possible to perform a panel data approach, as we dispose only of
one observation per individual. For the IV approach, the independent var-
iables need to be instrumented. In the present context it is difficult to
conceive of an instrument influencing the components but not SWB. As
a consequence, our coefficients are best interpreted as conditional
correlations.
Electricity use, type of paper used and the use of public transport are
components of the EF that have no impact on SWB (as they do not ap-
pear in the final stepwise regression model presented in Table 3). The
main item that is good for both SWB and the EF is the consumption of
seasonal products and fresh products (consuming frozen vegetables
and fruit preserves has a negative impact compared to the reference
of consuming fresh products). So in line with Welsch and Kühling
(2011) we find that switching towards a more environmentally friendly
diet can have a positive effect on subjective well-being. Also not using
electricity as heating for the house10 creates a win–win situation
8
A Spearman correlation (as robustness check) also pointed to a significant (p = 0.046)
positive correlation (0.066) with car use.
9
A (regular) basic model (non-stepwise) provides similar results as the effects present-
ed in Table 3. Only the significance level for spending holidays b200 km from home is
above the 5% threshold (p = 0.073) and an additional significant effect appears regarding
the use of mainly recycled paper (p = 0.012) having a negative relation with satisfaction.
10
The impact of using electricity as heating for the house on the EF is almost double to
that of the impact of oil fuel (the reference category) or gas.
84 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89

Table 3 Table 4
Results of a stepwise OLS regression for SWB using the EF-items. OLS regression models for SWB and EF.

EF component Satisfaction SWB EF

Unst. Coef. Sig. St. coef. Sig. St. coef. Sig.

Constant 7.516 0.000 Socio-economic position

Food Mainly local and seasonal products 0.222 0.020 Student 0.070 0.037 −0.012 0.755
Mainly frozen vegetables and fruit preserves −0.862 0.000 Pensioner 0.054 0.106 0.213 0.000
Consumption of meat and fish: maximum −0.297 0.028 Unemployed −0.066 0.031 0.015 0.669
3 times a week Incapable to work −0.052 0.088 0.000 0.991
Heating Small apartment or terraced house −0.598 0.002 Househusband/wife −0.060 0.064 0.001 0.973
Electricity as fuel for heating −0.295 0.013
Car use No car (or never use car) −0.511 0.020 Socio-economic characteristics
Holidays Spending holidays: at home −0.479 0.000 Being religious (yes) 0.090 0.002 −0.014 0.682
Spending holidays: b200 km from home −0.320 0.024 Having a relation (yes) 0.171 0.000 −0.150 0.000
Prob N F = 0.000 Personal income (in thousands €) 0.020 0.597 0.166 0.000
Adj. R2 = 0.079 House owner (yes) 0.097 0.001 −0.077 0.020

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.

an average (semi-) detached house

Appendix 1. Descriptive Statistics LEVO 2013

well insulated and you are sparingly with heating

Mean/proportion SD Range

Life satisfaction 7.24 1.50 0–10

Ecological footprint 6.94 1.85 2.87–17.00

Socio-economic position oil fuel

Working 55.2%
Student 5.6%
Pensioner 26.8%
Unemployed 4.0%
Incapable to work 3.6%
Househusband/wife 4.7%

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

Attitudes towards the environment

in Europa, by car
Knowledge about environmental issues 5.06 2.02 0–10
Concern about environmental issues 5.89 2.09 0–10
Appendix 3. Correlation matrix (correlations that differ significantly from zero (p b 0.05) are indicated in bold).

(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

E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89

Age2 (7) 1.00 0.07 −0.08 0.51 −0.24 0.02 0.17 0.24 0.04 −0.03 −0.07 −0.11 0.04 0.23 −0.14 −0.19 −0.11 −0.01 −0.13 −0.03
Gender (8) 1.00 −0.10 0.09 −0.09 −0.31 0.00 0.11 0.05 0.01 −0.02 0.00 0.38 0.07 −0.10 −0.01 −0.04 0.02 −0.20 −0.01
Relation (9) 1.00 −0.13 0.14 0.18 0.14 −0.04 −0.06 0.07 0.04 0.07 −0.01 0.06 0.12 0.07 0.07 0.10 0.11 0.10
Lower educated 1.00 −0.48 −0.20 0.00 0.17 0.04 0.04 −0.09 −0.08 0.06 0.11 −0.18 −0.11 −0.10 −0.06 −0.26 −0.16
(10)
Higher educated 1.00 0.35 0.02 −0.13 −0.05 −0.05 0.11 0.03 −0.11 −0.01 0.20 0.14 0.08 0.02 0.24 0.17
(11)
Income (12) 1.00 0.11 −0.01 −0.08 0.02 0.06 0.10 −0.16 0.09 0.22 0.08 −0.02 0.01 0.17 0.11
House owner (13) 1.00 0.05 −0.02 0.05 0.09 −0.05 −0.01 0.14 0.06 0.05 0.03 0.08 0.01 0.03
Religious (14) 1.00 0.03 0.04 −0.05 −0.07 0.06 0.15 −0.05 −0.05 −0.03 −0.01 −0.09 0.00
Centre of 1.00 −0.46 −0.36 0.03 0.02 0.02 −0.09 0.03 0.04 0.03 −0.04 −0.01
municipality (15)
Rural (16) 1.00 −0.37 −0.06 −0.02 0.02 0.00 0.02 0.02 0.09 −0.05 −0.05
Outskirts of town 1.00 0.03 0.01 0.01 0.07 0.01 −0.06 −0.07 0.06 0.06
(17)
Positive attitude 1.00 −0.01 0.01 0.00 0.21 0.28 0.25 0.10 0.07
(18)
Emotionally 1.00 0.01 −0.04 −0.14 −0.10 −0.02 −0.13 0.05
concerned (19)
Altruistic and 1.00 −0.02 0.03 0.02 0.14 −0.01 0.11
conscientious
(20)
Expectant (21) 1.00 0.08 −0.02 −0.01 0.15 0.10
Subjective health 1.00 0.34 0.37 0.09 0.08
(22)
A good social life 1.00 0.53 0.07 0.06
(23)
A pleasant 1.00 0.03 0.16
environment (24)
Knowledge 1.00 0.51
environment (25)
Concern 1.00
environmental
(26)

87
88 E. Verhofstadt et al. / Ecological Economics 127 (2016) 80–89

Appendix 4. Robustness Checks

A. Stepwise model for SWB from Table 3 including all significant variables from Table 4.

Unst. coeff. St. coeff. Sig.

B Std. error Beta

(Constant) 3.081 0.358 0.000

Food: mainly frozen vegetables and fruit preserves −0.785 0.225 −0.100 0.001
Heating: small apartment or terraced house −0.363 0.178 −0.063 0.042
Car use: seldom 0.271 0.123 0.064 0.027

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.

B Std. error Beta B Std. error Beta

(Constant) 2.933 0.535 0.000 8.036 0.751 0.000

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.

B Std. error Beta B Std. error Beta

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

Attitudes towards the environment

Knowledge environmental problems −0.012 0.026 −0.016 0.646 0.044 0.037 0.048 0.230
Concern about environmental issues 0.023 0.025 0.032 0.354 −0.113 0.035 −0.126 0.001
Prob N F = 0.000 Prob N F = 0.000
Adj. R2 = 0.274 Adj. R2 = 0.097

References
Anand, P., van Hees, M., 2006. Capabilities and achievements: an empirical study. J. Socio-
Econ. 35 (2), 268–284.
Blanchflower, D., Oswald, A., 2004. Well-being over time in Britain and the United States.
J. Public Econ. 88 (7/8), 1359–1386.
Blanchflower, D., Oswald, A., 2008. Is well-being U-shaped over the life cycle? Soc. Sci.
Med. 66, 1733–1749.
Büchs, M., Schnepf, S., 2013. Who emits most? Associations between socio-economic fac-
tors and UK households' home energy, transport, indirect and total CO2 emissions.
Ecol. Econ. 90, 114–123.
Cloutier, S., Jambeck, J., Scott, N., 2014. The sustainable Neighborhoods for Happiness
Index (SNHI): a metric for assessing a community's sustainability and potential influ-
ence on happiness. Ecol. Indic. 40, 147–152.
Csutora, M., 2012. One more awareness gap? The behaviour–impact gap problem.
J. Comsum. Policy 35, 145–163.
De Vaus, D., 2002. Analyzing Social Science Data: 50 Key Problems in Data Analysis. Sage,
London.
Diener, E., 2000. Subjective well-being: the science of happiness and a proposal for a na-
tional index. Am. Psychol. 55, 34–43.
Dolan, P., Peasgood, T., White, M., 2008. Do we really know what makes us happy? A re-
view of the economic literature on the factors associated with subjective well-being.
J. Econ. Psychol. 29 (1), 94–122.
Druckman, A., Jackson, T., 2009. The carbon footprint of UK households 1990–2004: A
socio-economically disaggregated, quasi-multi-regional input–output model. Ecol.
Econ. 68, 2066–2077.
Ericson, T., Kjønstad, B.G., Barstad, A., 2014. Mindfulness and sustainability. Ecol. Econ.
104, 73–79.
Ferreira, S., Akay, A., Brereton, F., Cunado, J., Martinsson, P., Moro, M., Ningal, T.F., 2013.
Life satisfaction and air quality in Europe. Ecol. Econ. 88, 1–10.
Ferrer-I-Carbonell, A., Gowdy, J.M., 2007. Environmental degradation and happiness. Ecol.
Econ. 60, 509–516.
Gardner, G., Stern, P., 2002. Environmental Problems and Human Behavior. Pearson Cus-
tom Publishing, Boston, MA.
George, D., Mallery, P., 2010. SPSS for Windows Step by Step: A Simple Guide and Refer-
ence, 17.0 Update. Pearson, Boston.
GFN (Global Footprint Network), 2012. www.footprintnetwork.org/en/index.php/GFN/
page/glossary (accessed 14/07/2014).
Gosling, S.D., Rentfrow, P.J., Swann, W.B., 2003. A very brief measure of the Big-Five per-
sonality domains. J. Res. Pers. 37 (6), 504–528.
Helliwell, J., Layard, R., Sachs, J., 2012. World Happiness Report. The Earth Institute. Co-
lumbia University, New York (NY).
Helliwell, J., Layard, R., Sachs, J., 2013. World Happiness Report 2013. The Earth Institute.
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.
Layard, R., Mayraz, S., Nickell, S., 2008. The Marginal Utility of Income. J. Public Econ. 92,
1846–1857.
Lenzen, M., Cummins, R.A., 2013. Happiness versus the environment—a case study of
Australian lifestyles. Challenges 4, 56–74.
Li, Z., Folmer, H., Xue, J., 2014. To what extent does air pollution affect happiness? The
case of the Jinchuan mining area, China. Ecol. Econ. 99, 88–99.
Luechinger, S., Raschky, P.A., 2009. Valuing flood disasters using the life satisfaction ap-
proach. J. Public Econ. 93, 620–633.
MacKerron, G., Mourato, S., 2009. Life satisfaction and air quality in London. Ecol. Econ. 68,
1441–1453.
Maddison, D., Rehdanz, K., 2011. The impact of climate on life satisfaction. Ecol. Econ. 70,
2437–2445.
Menz, T., Kühling, J., 2011. Population aging and environmental quality in OECD coun-
tries: evidence from sulfur dioxide emissions data. Popul. Environ. 33 (1), 55–79.
nef, 2012. Happy Planet Index: 2012 report—a global index of sustainable well-being,
2012. Available online http://www.neweconomics.org/publications/happy-planet-
index-2012-report.
O'Donnell, G., Deaton, A., Durand, M., Halpern, D., Layard, R., 2014. Wellbeing and Policy —
Full Report. Legatum Institute.
OECD, 2013a. OECD Guidelines on Measuring Subjective Well-being. OECD Publishing.
OECD, 2013b. How's Life?: Measuring Well-being. OECD Publishing.
Rehdanz, K., Maddison, D., 2005. Climate and happiness. Ecol. Econ. 52, 111–125.
Rehdanz, K., Maddison, D., 2008. Local environmental quality and life-satisfaction in
Germany. Ecol. Econ. 64, 787–797.
Sekulova, F., 2013. On the Economics of Happiness and Climate Change (Doctoral thesis)
Universitat Autonoma de Barcelona.
Stiglitz, J.E., Sen, A., Fitoussi, J.-P., 2009. Report by the Commission on the Measurement of
Economic Performance and Social Progress. Commission on the Measurement of Eco-
nomic Performance and Social Progress.
Van Praag, B.M.S., Ferrer-i-Carbonell, A., 2010. Happiness economics: a new road to mea-
suring and comparing happiness. Found. Trends Microeconomics 6 (1), 1–97.
Veenhoven, R., 1991. Is happiness relative? Soc. Indic. Res. 24 (1), 1–34.
Vreg, 2013. Resultaten enquête particulieren 2013. Gedrag & ervaringen van
huishoudelijke afnemers op de vrijgemaakte Vlaamse energiemarkt. RAPP-2013-16.
Welsch, H., 2006. Environment and happiness: valuation of air pollution using life satis-
faction data. Ecol. Econ. 58, 801–813.
Welsch, H., Kühling, J., 2011. Are pro-environmental consumption choices utility-
maximizing? Evidence from subjective well-being data. Ecol. Econ. 72, 75–87.
WWF, 2012. Living planet. Report 2012. Biodiversity, biocapacity and better choices.
Available online http://wwf.panda.org/about_our_earth/all_publications/living_
planet_report/.
WWF, 2014. Living Planet Report 2014: Species and Spaces, People and Places. WWF In-
ternational, Gland, Switzerland.