Description
Switzerland boasts arguably the highest density of green properties in the world. In 2008,
more than 15 percent of total new construction received the Swiss energy building label Minergie. The
spatial distribution of these green buildings, however, is highly heterogeneous. In some regions, more
than half of the new dwellings are built according to the Swiss green building standard. In others, this
share is still negligible. The purpose of this paper is to identify the determinants of the distribution of
green housing.
Journal of Financial Economic Policy
What drives “green housing” construction? Evidence from Switzerland
Marco Salvi J uerg Syz
Article information:
To cite this document:
Marco Salvi J uerg Syz, (2011),"What drives “green housing” construction? Evidence from Switzerland",
J ournal of Financial Economic Policy, Vol. 3 Iss 1 pp. 86 - 102
Permanent link to this document:http://dx.doi.org/10.1108/17576381111116777
Downloaded on: 24 January 2016, At: 21:40 (PT)
References: this document contains references to 20 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 800 times since 2011*
Users who downloaded this article also downloaded:
Steve Swidler, (2011),"Homeownership: yesterday, today and tomorrow", J ournal of Financial Economic
Policy, Vol. 3 Iss 1 pp. 5-11http://dx.doi.org/10.1108/17576381111116786
Beverley Searle, (2011),"Recession and housing wealth", J ournal of Financial Economic Policy, Vol. 3 Iss 1
pp. 33-48http://dx.doi.org/10.1108/17576381111116867
Dag Einar Sommervoll, Gavin Wood, (2011),"Home equity insurance", J ournal of Financial Economic
Policy, Vol. 3 Iss 1 pp. 66-85http://dx.doi.org/10.1108/17576381111116768
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
What drives “green housing”
construction? Evidence from
Switzerland
Marco Salvi
Zu¨ rcher Kantonalbank and Department of Architecture,
Swiss Federal Institute of Technology Zu¨ rich, Zu¨ rich, Switzerland, and
Juerg Syz
Diener Syz Real Estate, Zollikon and Shanghai, and Universita¨ t Zu¨ rich,
Zu¨ rich, Switzerland
Abstract
Purpose – Switzerland boasts arguably the highest density of green properties in the world. In 2008,
more than 15 percent of total new construction received the Swiss energy building label Minergie. The
spatial distribution of these green buildings, however, is highly heterogeneous. In some regions, more
than half of the new dwellings are built according to the Swiss green building standard. In others, this
share is still negligible. The purpose of this paper is to identify the determinants of the distribution of
green housing.
Design/methodology/approach – For 2,571 Swiss municipalities, the author computes the green
building share of new residential buildings. Data are collected for several variables measuring
demographic, geographic, social, cultural, and political aspects that – according to the authors’
hypothesis – may in?uence green building activity. Count regression is used to estimate the impact of
these variables on the demand for green buildings.
Findings – It is found that differences in income levels and cultural af?liation between Swiss
municipalities account for the largest part of the variation in green building activity. The impact of
homeowners’ stance on environmentalism is highly signi?cant but less important. Government
subsidies do not seem to trigger additional green housing activity.
Originality/value – The paper presents one of the ?rst empirical analyses regarding the
determinants of green building activity. Thanks to a comprehensive dataset, the authors are able to
investigate the impact of potential drivers of “green housing” construction activity. The regional
variation in governmental incentives is analysed and delivers valuable insight for policymakers
interested in spurring the development of green buildings.
Keywords Environmental regulations, Residential homes, Switzerland
Paper type Research paper
1. Introduction
The Swiss property market is an ideal playground to examine the determinants of the
demand for green properties. Indeed, Switzerland has one of the highest densities of
energy-ef?cient buildings in the world (Salvi et al., 2010). By mid-2010, more than 16,000
new and retro?tted buildings had received the Swiss green building label “Minergie”.
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The authors are indebted to Andrea Horeha´jova´, Julie Neeser, and Andreas Bro¨hl for helpful
comments and research assistance. The authors would also like to thank Steven Swidler,
Erika Meins, and Philippe Thalmann for their precious help and encouragement. The comments
of two anonymous reviewers are gratefully acknowledged.
JFEP
3,1
86
Journal of Financial Economic Policy
Vol. 3 No. 1, 2011
pp. 86-102
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381111116777
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In 2008, roughly 15 percent of newly constructed buildings successfully completed the
Minergie certi?cation process. This paper draws on data collected from the Swiss
market to investigate the question of “who builds green houses and why”. Previous
research has shown that homeowners do value the expected future cost savings
generated by investments in energy-ef?cient buildings. Several authors have
documented the incentive effects of higher energy prices on the demand for
energy-ef?cient technologies, see, e.g. Hausman, 1979; Beresteanu and Li, 2008; Linn and
Klier, 2009. However, this paper argues that more moderate utility bills alone do not
explain the demand for green properties.
We observe that the spatial distribution of green buildings in Switzerland is highly
heterogeneous. In some cities, more than half of the new constructions are built
according to the Swiss green building standard. In other regions, this share is negligible,
suggesting that there are more subtle drivers of the demand for green buildings than
energy cost savings. To detect these drivers, we investigate the determinants of green
housing activity that lead to regional clusters in Switzerland. Our approach is closely
related to Kahn and Vaughn (2009) who study clusters of Leadership in Energy and
Environmental Design (LEED) registered buildings and hybrid cars in the USA.
However, in their paper, these authors analyze just 10,000 registered LEED buildings
(765 of themcerti?ed) scattered across the USA. Inthis paper, we explore a market where
the density of certi?ed green buildings is by two orders of magnitude higher. Moreover,
thanks to its decentralized political structures and to the intensive use of direct
democratic instruments at the federal, cantonal, and municipal level, Switzerland offers
an ideal situation for studying the impact of environmentalism and government
subsidies on green building activity. We use a unique dataset, including all newly built,
Minergie labeled residential buildings in Switzerland. We relate the green housing
density in Swiss municipalities to corresponding demographic, geographic, social,
cultural, and political attributes. We include a measure of environmentalism based on
voting data as well as government subsidies offered at the level of the 26 Swiss cantons.
We ?nd that – among all investigated variables – differences in income across
municipalities account for the largest part of the explained variation in green building
activity. Linguistic af?liation, as a proxy for cultural norms, turns out to have a strong
impact on the regional distribution of Minergie residential buildings as well. The
in?uence of political af?liation, as measured by voting data, is statistically signi?cant
but less important. Government subsidies for green buildings do not appear to have any
positive impact on the clustering of green properties.
The paper is organized as follows. In Section 2, we describe the characteristics of the
Swiss green building standard Minergie and its rapid propagation. We also document
the large regional differences of green building activity. In Section 3, we develop six
hypotheses for potential drivers leading to the observed regional clusters of green
housing activity. We present the regression results with regard to the correlates of
green housing adoption in Section 4. We draw our conclusions in Section 5.
2. Green buildings in Switzerland
2.1 The Swiss green building standard “Minergie”
Minergie is the leading eco-label for energy-ef?cient buildings in Switzerland[1].
Anon-pro?t private association, Minergie is supported by its members, which include the
federal government, the cantons, schools, companies, individuals, andvarious associations.
“Green housing”
construction
87
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Minergie offers both eco-labeling and eco-certi?cation. Third parties, usually a cantonal
authority, certify Minergie buildings. There are three levels of Minergie building
certi?cations. The basic Minergie certi?cation is used broadly for new and retro?tted
buildings. To attain the standard, the building must achieve a reduction of at least
25 percent in general energy consumption in comparison to the average conventional
building. In addition to this requirement, fossil-fuel consumption has to be less than half of
that of the average conventional building. Minergie-Pis a stricter certi?cation that requires
very low energy consumption and is especially demanding with regard to heating energy
demand. This standard broadly corresponds to the German Passivhaus Standard. Finally,
Minergie-ECO involves an additional certi?cation that veri?es the use of
environmental-friendly building materials. In this paper, we do not differentiate between
the sub-labels as the basic Minergie label covers 95 percent of the certi?ed buildings[2].
The list of the certi?ed buildings is publicly available on the Minergie web site. Real estate
agents routinely advertise the presence of the label as a part of the sale process.
As with other green building labels, the implicit assumption of Minergie is that the
energy consumption of a dwelling is a function of its building standard. Energy
consumption estimates are basedonthe characteristics of the materials applied and used
to assess whether a new or retro?tted building quali?es for the Minergie label. As of
today (2010), the requirements of the basic Minegie standard described above set a limit
of 38 kWh per square meter of ?oor area and year. This corresponds roughly to the lower
bound of the energy rating “B” of the European Energy Performance of Buildings
Directive (EPBD)[3]. The use of active ventilation is mandatory to obtain certi?cation.
Since its launch in 1998, the Minergie label has been quite successful. In principle,
properties of all types – be it of?ce buildings or residential housing – can be certi?ed if
they meet the label’s criteria. In practice, however, private homeowners are at the
forefront of green building activity in Switzerland, as most green properties belong to
residential owner occupiers and private owners of residential multi-family buildings.
As of August 2009, 11,555 or 91 percent of all certi?ed buildings are residential
buildings, whereof 68 percent are single family and 32 percent multi-family homes.
Of the 9 percent of non-residential units, about 70 percent are owned by the public sector.
Schools, sports facilities or of?ce buildings make up the larger part of this category.
Because of the predominant share of residential buildings, we focus our analysis on this
segment[4]. Table I summarizes the distribution of certi?ed buildings by property type.
The number of new Minergie buildings tripled between 2004 and 2009. While at
the beginning of this period, only 5 percent of the new buildings received the label, this
proportionincreasedto15percent in2008. Figure 1showsthe number of Minergie-certi?ed
Number of Minergie-certi?ed buildings Percentage of total
Single-family homes 7,810 62
Multi-family homes 3,745 29
Others, whereof 1,101 9
Administration of?ces 432
Schools 308
Sport facilities 73
Total 12,656 100
Source: See Data Appendix
Table I.
Minergie-certi?ed
buildings in Switzerland
as of mid-2009
JFEP
3,1
88
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
new residential buildings and their share of total new residential buildings since 1998.
However, newconstructionrepresents onlya small part of the total built stock. Hence, only
about 1 percent of the existing Swiss buildings have been certi?ed so far. Nevertheless, to
the best of our knowledge, the rate of green buildings is higher in Switzerland than in
comparable countries. Indeed, Minergie’s penetration rate in Switzerland is roughly
280 times higher than LEED’s rate in the USA, where it represents the most widely used
green building label[5]. In England, the number of residential buildings in the
energy-ef?ciency rating bands “A” and “B” represented only 0.3 percent of the housing
stock in 2008 (UK Department for Communities and Local Government, 2010).
2.2 The spatial distribution of green housing
The geographical distribution of Minergie buildings in Switzerland is highly clustered.
Most Minergie houses are located in the northern and northeastern part of the country,
as well as in the cities of Bern and Geneva. To compare green building activity between
municipalities, we divide the number of Minergie-certi?ed new buildings by the
number of total residential buildings constructed between 1998 and 2008. Figure 2
shows a map of the spatial distribution of this share. The city of Zurich stands out with
a share of approximately 20 percent, followed by regions in the agglomeration of
Zurich. Alpine touristic resorts such as Davos and Zermatt also stand out for their high
share of energy-ef?cient buildings. In these resorts, roughly one new building in ten
has received the Minergie label. At the other end of the scale, the cantons of Ticino and
Jura and the area around the lake of Geneva exhibit very low green housing activity.
There, the share of new green buildings accounted for less than 2 percent.
On ?rst inspection, the share of Minergie-certi?ed houses seems to mirror the
linguistic regions in Switzerland. In the German-speaking part, every ?fth new
residential building completed in 2008 obtained the Minergie certi?cation, while in
French-speaking Romandy and in the Italian-speaking region only one in 12,
respectively, one in 14 did. On the other hand, French-speaking Geneva tops the list of
green building activity among Swiss cities, as shown in Table II. The heterogeneous
distribution of green building activity raises the question of the drivers of the demand
for green buildings. We address this question in the remainder of the paper.
Figure 1.
Number of Minergie
certi?ed new residential
buildings and their share
of total new residential
buildings, 1998 to 2008
3,000
2,500
2,000
1,500
1,000
500
0% 0
1999 2000 2001
Number of Minergie new homes
Share of Minergie new homes of total new homes
2002 2003 2004 2005 2006 2007 2008
2%
4%
6%
8%
10%
12%
14%
16%
18% “Green housing”
construction
89
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. The drivers of green building activity
3.1 Hypothesis development
Even though at least one Minergie building is present in more than half of the Swiss
municipalities, the share of green buildings varies widely across the municipalities. We
develop six hypotheses to explain this heterogeneity. Our list of likely correlates of green
housingactivityincludes demographic, geographic, social, cultural, andpolitical aspects[6].
3.1.1 Income. If green buildings are superior goods, their demand will be strongly
positively related to income[7]. We test the hypothesis that Minergie buildings are
more likely in richer municipalities:
H1. Green building demand increases with income.
3.1.2 Age. Some researchers have argued that the willingness to pay for environmentally
friendly goods declines with age (Hersch and Viscusi, 2005). We test this hypothesis by
including the municipalities’ age distribution in our regressions:
Rank City
Percentage of
all new buildings
Number of Minergie-certi?ed
new buildings
1 Geneva 34.5 39
2 Zurich 33.2 249
3 Bern 19.8 26
4 Winterthur 15.1 97
5 Lucerne 14.7 22
6 Basel 8.5 9
7 St Gall 8.4 18
8 Lugano 3.1 10
9 Lausanne 1.2 4
Source: See Data Appendix
Table II.
Recent green construction
activity in Switzerland’s
largest cities (2004-2008)
Figure 2.
Minergie’s share of new
residential buildings in
Swiss regions, 1998-2008
1998-2008
0% - 2%
2% - 4%
4% - 6%
6% - 9%
9% - 12%
12% - 15%
15% - 20%
JFEP
3,1
90
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H2. Green building demand is negatively related to age.
3.1.3 Cultural norms. At a more general level, cultural differences may be important in
addressing environmental problems (Milton, 1996). This may also in?uence green
building activity. As a multilingual country, Switzerland has natural cultural
boundaries within its national borders. We investigate whether the different linguistic
regions vary in their af?nity towards green housing:
H3. Green building demand varies with linguistic af?liation.
3.1.4 Geography. The energy consumption of a building depends crucially on heating
demand, which is a function of the difference between internal and external temperature
(MacKay, 2008). Outside temperature may thus affect the demand for energy-ef?cient
buildings. Unfortunately, average local temperature and heating degrees data is only
available for the limited number of communities in which a meteorological station is
located. In Switzerland, however, differences in temperature are strongly correlated with
altitude, which can be easily obtained for every community[8]. We thus include the
average height above sea level as a proxy for this demand driver:
H4. Green building demand is positively related to altitude.
3.1.5 Government subsides. We are interested in measuring the impact played by
governmental subsides on regional green building activity. The Swiss cantons have a
wide discretion when ?xing the amount of the subsidies to be granted to energy-ef?cient
buildings. As a result, payments for Minergie-certi?ed newbuildings vary considerably
fromcanton to canton. In 2008, the canton of Bern paid subsidies totaling CHF2.2 million
while 11 out of the 26 Swiss cantons, including the canton of Zurich, did not make any
subsidy payments. Among the cantons supporting Minergie new buildings, payments
ranged from CHF 3,700 per building in canton Ticino to CHF 31,340 in the canton of
Valais. We use the average subsidy payment per new Minergie building in each canton
to investigate the effectiveness of governmental programs:
H5. Green building demand is positively related to government subsidies.
3.1.6 Environmental activity. Environmentalists tend to be more likely to purchase
green products and may be willing to pay more for environmentally friendly products
(Kotchen and Moore, 2007, 2008; Kahn and Vaughn, 2009). We suppose that
municipalities with a large share of the population supportive of green ideas are more
likely to have a higher share of green buildings. We measure environmentalismbased on
revealed preference political data:
H6. Green building demand is positively related to the degree of
environmentalism in the municipality.
3.2 Measuring environmentalism
Following Kahn (2007) and Kahn and Vaughn (2009), we construct two indicators of
environmentalism at the municipal level. The basic rationale is that people “vote with
their feet” to ?nd the community that provides their optimal bundle of environmental
public goods and taxes (Banzhaf and Walsh, 2008). In Switzerland, voters have the
additional opportunity to decide directly on the bundle of environmental public goods.
Through ballot initiatives and referenda at the federal, cantonal, and municipal level,
“Green housing”
construction
91
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
they can propose or refuse particular changes in the legislation. At the municipal level,
voters are routinely asked to vote on the ?nancing of local infrastructure projects such as
schools or new of?ce buildings for the administration. The outcome of the voting on
these initiatives can be informative about the voters’ stance on environmental issues.
We base our ?rst indicator, the “greenindex”, onthe results of ?ve federal initiatives on
environmental issues, listed in Table III. All ?ve federal initiatives were rejected, most of
themby a wide margin[9]. To construct our ?rst aggregate measure of environmentalism,
we run a factor analysis on the voting results of the federal initiatives. The correlation of
the pro-environmental vote across the ?ve initiatives is high. The bivariate correlation
coef?cients range between 0.25 and 0.75. The 2008 federal initiative “Right of appeal of
NGOs” (Verbandsbeschwerdeinitiative) has the lowest correlation with other initiatives.
In the factor analysis, it receives the lowest factor loading (0.177)[10]. The factor loadings
of the other initiatives range between 0.23 and 0.32.
By this account, the list of the “greenest” communities in Switzerland closely
matches the list of the main cities with Zurich, Geneva, and Basel among the 5 percent
of the Swiss municipalities with the highest green index values. Of the ten largest
Swiss cities, only Lugano, situated in the Italian-speaking canton Ticino, does not
appear in the 10 percent of municipalities with the highest green index score.
The second measure of local environmentalism is based on the results of the 2007
election for the Swiss National Council, Switzerland’s lower house of parliament[11]. We
count the percentage of votes cast at the municipal level for the global positioning system
(GPS) and the 2004 founded Green Liberal Party (GLP). Environmental issues and the
promotion of renewable energies are at the core of both parties’ platforms. Accordingly,
the GPS endorsed all pro-environment initiatives listed in Table III. Both the GPS and the
GLP recommended the rejection of the “Right of appeal of NGOs” – initiative. On social
issues, the GPS is in general allied with left wing parties, whereas the GLP is positioned at
the center of the political spectrum. In the 2007 elections, the GLP won 1.4 percent of the
popular vote nationwide and three out of 200 seats. The GPS won 9.6 percent of the votes
and 20 seats. Unsurprisingly, green parties are strongest in the main urban areas. Six out
Ballot title Year Main aim Yes votes (%)
Participation
rate (%)
“Cut traf?c by half” 2000 Reduction of road traf?c by half over
a ten-year period
21.3 42.3
“Solar cent” 2000 Introduction of a tax of CHF 0.005 per
kWh on non-renewable energy
sources. Half of the tax ear-marked for
solar power uses
31.3 44.7
“Steering tax on non-
renewable energy”
2000 Introduction of a Pigou tax of CHF
0.02/kWh on non-renewable energy
sources
44.5 44.9
“Electricity without
nuclear power”
2003 Gradual abandonment of nuclear
energy
32.7 49.7
“Right of appeal for
NGOs”
2008 Curtailing of the rights of
environmental organizations to appeal
construction projects
33.0 47.4
Source: See Data Appendix
Table III.
Federal initiatives on
general environmental
issues in the period
2000-2008
JFEP
3,1
92
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
of the tenlargest Swiss cities belongto the 5 percent communities withthe highest share of
green parties votes. Again, Lugano stands out among this group with a share of only 4.6
percent of green votes, lower than the median at 8.5 percent. The Spearman rank-order
correlation between the two measures of environmentalism is 0.36.
The green parties did not run for election in the smaller, mostly rural constituencies.
Hence, this direct measure of environmentalism is not available for nine out of
26 cantons. Data on federal initiatives, however, are available for all municipalities.
We test both measures of environmentalism in the following regression analysis.
4. Empirical results
4.1 Model selection
We test the six hypotheses stated in Section 3 with data available at the municipal level.
We run count regressions on the number of Minergie-labeled properties built between
1998 and 2008 in each municipality. We take into account the non-negative
integer-valued aspect of the dependent variable. Speci?cally, we assume that the
conditional expectation of y
i
, the number of Minergie buildings in municipality i, is:
Eð y
i
jx
i
; t
i
Þ ¼ m
i
t
i
¼ expðx
0
i
b þ 1
i
Þ; ð1Þ
where x
i
is the vector of covariates and t
i
¼ exp 1
i
ð Þ is an heterogeneity factor
independent of x
i
. Taking the exponential ensures that the mean parameter is
non-negative; adding t
i
allows for unobserved heterogeneity between municipalities that
is not fully accounted for by the covariates[12]. It can be shown (Winkelmann, 2000) that
the distribution of y
i
conditional on x
i
and t
i
is Poisson distributed with:
gð y
i
Þ ¼ PðY
i
¼ y
i
jx
i
; t
i
Þ ¼
e
m
i
m
i
y
i
!
; i ¼ 0; 1; 2; . . . : ð2Þ
The heterogeneity factor, t
i
, can be integrated out of this conditional distribution under
the assumption that it is gamma distributed. This solution is called the negative binomial
model. It is more general than the Poisson regression. Its use is widespread because,
unlike the standard Poisson model, the conditional variance can exceed the conditional
mean. As such, it can accommodate over-dispersion resulting fromneglected unobserved
heterogeneity.
In contrast to standard logit regression, the use of count regression allows us the take
directly into account the fact that in 40.5 percent of all municipalities no Minergie house
has been built between 1998 and 2008. Zero-in?ated count models provide a simple way
of modeling so-called excess zeros (Winkelmann, 2000, p. 109). We thus explicitly model
the production of zero counts by specifying a Bernoulli trial that has g(y
i
) as outcome
with probability w
i
, or zero otherwise. This gives rise to a zero-in?ated negative binomial
(ZINB) model, where the probability of a non-zero event depends on the characteristics
of a municipality z
i
, speci?ed as:
w
i
¼ F
i
¼ Fðz
0
i
gÞ; ð3Þ
where the link function F is a logistic function. The standard estimator for the negative
binomial model is the maximum likelihood estimator. Estimates of the coef?cient
vectors b and g are found by minimization of the corresponding log-likelihood function
(Winkelmann, 2000). We next present estimates for various speci?cation of the ZINB
model. We discuss the estimation results for simpler models, such as the Poisson model,
the negative binomial model, and the zero-in?ated Poisson model.
“Green housing”
construction
93
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.2 Regression results
In the base model, the number of Minergie houses is regressed on several covariates
related to the hypotheses developed in Section 3. The covariates include the share of
residents in four age classes, the average altitude in the municipality, the majority
language spoken in the municipality, two indicators of environmentalism, the share of
residents in each of three income brackets, the amount of subsidies for Minergie
buildings and the total number of new buildings in the municipality. Again, we refer to
the Data Appendix for the data sources and the exact de?nition of the variables.
TableIVpresentssummarystatistics. Between1998and2008anaverageof 3.7Minergie
buildings and about 70 new buildings were completed in each municipality[13]. We notice
that there is not much variation in the population distribution by age. The median Swiss
municipality is situatedat an altitude of 613 meters above sea level. In the 25 percent richest
communities, at least 42 percent of the tax income payers are in the highest income bracket.
The ?rst column of Table V presents our base case estimation results for the vector
of parameters b in equation (1). In this speci?cation, we use the ZINB model and the
green index as indicator for environmentalism[14]. The coef?cients for the green index,
the income variables, and the language af?liation are highly signi?cant. To illustrate
the economic effect of the estimated coef?cients, we report the change in the expected
number of Minergie buildings per municipality that is associated with a given change
of a covariate. Thus, for each of the covariates in x
i
, we compute expðx
0
i
^
bÞ at the median
and at the third quartile of the distribution of the covariate and report the relative
change in the expected number of Minergie buildings in the second column. The table
further displays the estimation results using the green parties’ share of votes instead of
the green index as indicator for environmentalism, as well as estimation results based
on cantonal ?xed effects (columns 3 and 4). As a robustness exercise, we also report the
results of a regression with cantonal ?xed effects (column 5).
Variable Description Mean SD P25 Median P75
AGE_0_19 Share of population less than 20 0.252 0.042 0.227 0.254 0.280
AGE_20_39 Share of population aged 20-40 0.271 0.038 0.250 0.273 0.294
AGE_40_59 Share of population aged 40-60 0.281 0.033 0.261 0.280 0.301
AGE_60_99 Share of population over 60 0.196 0.052 0.161 0.190 0.222
ALTITUDE Altitude above sea level (km) 0.804 0.481 0.484 0.613 0.930
DGERMAN German-speaking municipality
(yes ¼ 1) 0.607 0.488 – 1.000 1.000
GREEN_IND1 Share of green parties’ votes
a
0.088 0.094 0.057 0.127 0.161
GREEN_IND2 Green index 0.000 1.000 20.686 20.069 0.576
INCOME_LOW Share of taxpayers in low-income class 0.290 0.100 0.222 0.270 0.330
INCOME_MID Share of taxpayers in mid-income class 0.410 0.060 0.385 0.420 0.450
INCOME_HIGH Share of taxpayers in high-income class 0.300 0.100 0.232 0.290 0.360
MIN_BUILD
Number of Minergie buildings in
municipality 3.718 10.322 – 1.000 3.000
MIN_GRANT
Cantonal subsidy per Minergie
building (1,000 CHF) 8.687 10.497 – 4.416 10.936
NEW_BUILD
Number of new buildings in
municipality 68.963 97.551 13.000 38.000 89.000
Note:
a
Only available for 2,219 municipalities
Source: See Data Appendix
Table IV.
Descriptive statistics for
the 2,571 Swiss
municipalities
JFEP
3,1
94
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
t
h
e
g
r
e
e
n
i
n
d
e
x
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
g
r
e
e
n
p
a
r
t
i
e
s
’
s
h
a
r
e
o
f
v
o
t
e
s
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
c
a
n
t
o
n
a
l
?
x
e
d
e
f
f
e
c
t
s
P
a
r
a
m
e
t
e
r
E
s
t
i
m
a
t
e
E
f
f
e
c
t
s
a
E
s
t
i
m
a
t
e
E
f
f
e
c
t
s
a
E
s
t
i
m
a
t
e
I
n
t
e
r
c
e
p
t
2
4
.
0
1
1
(
0
.
7
8
9
)
–
2
5
.
0
4
3
(
0
.
8
1
5
)
–
2
4
.
0
2
1
(
0
.
8
6
1
)
N
u
m
b
e
r
o
f
n
e
w
b
u
i
l
d
i
n
g
s
i
n
m
u
n
i
c
i
p
a
l
i
t
y
0
.
6
6
4
(
0
.
0
3
4
)
*
*
*
0
.
4
0
3
0
.
6
3
0
(
0
.
0
3
5
)
*
*
*
0
.
4
1
4
0
.
6
1
1
(
0
.
0
3
4
)
*
*
*
G
r
e
e
n
i
n
d
e
x
0
.
1
7
9
(
0
.
0
3
2
)
*
*
*
0
.
1
2
2
n
/
a
n
/
a
0
.
2
0
5
(
0
.
0
3
6
)
*
*
*
G
r
e
e
n
p
a
r
t
i
e
s
’
s
h
a
r
e
o
f
v
o
t
e
s
(
%
)
n
/
a
n
/
a
3
.
6
0
8
(
0
.
6
8
8
)
*
*
*
0
.
1
5
5
n
/
a
S
h
a
r
e
o
f
t
a
x
p
a
y
e
r
s
i
n
m
i
d
-
i
n
c
o
m
e
c
l
a
s
s
3
.
9
6
0
(
0
.
7
8
3
)
*
*
*
0
.
1
2
0
4
.
4
5
8
(
0
.
8
9
1
)
*
*
*
0
.
1
3
4
4
.
1
9
5
(
0
.
7
9
3
)
*
*
*
S
h
a
r
e
o
f
t
a
x
p
a
y
e
r
s
i
n
h
i
g
h
-
i
n
c
o
m
e
c
l
a
s
s
4
.
2
5
1
(
0
.
4
9
5
)
*
*
*
0
.
3
2
4
4
.
1
0
3
(
0
.
5
5
2
)
*
*
*
0
.
3
2
0
3
.
1
3
6
(
0
.
5
2
4
)
*
*
*
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
a
g
e
d
2
0
-
4
0
2
.
6
1
9
(
1
.
1
6
8
)
*
0
.
0
5
8
3
.
4
1
8
(
1
.
2
6
5
)
*
*
0
.
0
7
5
3
.
1
1
2
(
1
.
2
0
3
)
*
*
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
a
g
e
d
4
0
-
6
0
0
.
2
7
7
(
1
.
2
5
9
)
0
.
0
0
6
0
.
4
5
1
(
1
.
3
3
5
)
0
.
0
0
9
1
.
3
4
3
(
1
.
3
4
2
)
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
o
v
e
r
6
0
1
.
8
3
5
(
0
.
8
6
0
)
*
0
.
0
6
1
2
.
9
5
6
(
0
.
8
6
0
)
*
*
*
0
.
0
9
1
3
.
0
7
4
(
0
.
8
9
3
)
*
*
*
A
l
t
i
t
u
d
e
a
b
o
v
e
s
e
a
l
e
v
e
l
(
k
m
)
0
.
1
4
2
(
0
.
0
7
9
)
0
.
0
4
6
0
.
4
1
4
(
0
.
0
9
1
)
*
*
*
0
.
1
9
1
0
.
0
4
5
(
0
.
0
9
7
)
G
e
r
m
a
n
-
s
p
e
a
k
i
n
g
m
u
n
i
c
i
p
a
l
i
t
y
(
y
e
s
¼
1
)
0
.
5
6
6
(
0
.
0
8
5
)
*
*
*
0
.
7
6
2
0
.
5
2
9
(
0
.
0
9
3
)
*
*
*
0
.
6
9
7
0
.
6
5
6
(
0
.
1
3
6
)
*
*
*
S
u
b
s
i
d
y
p
e
r
M
i
n
e
r
g
i
e
b
u
i
l
d
i
n
g
(
c
a
n
t
o
n
a
l
l
e
v
e
l
)
2
0
.
0
0
8
(
0
.
0
0
4
)
*
2
0
.
0
5
3
2
0
.
0
1
5
(
0
.
0
0
4
)
*
*
*
0
.
1
2
5
n
/
a
O
b
s
e
r
v
a
t
i
o
n
s
2
,
5
1
6
2
,
2
1
9
2
,
5
1
6
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
5
,
*
*
1
a
n
d
*
*
*
0
.
1
p
e
r
c
e
n
t
;
a
p
e
r
c
e
n
t
a
g
e
d
i
f
f
e
r
e
n
c
e
i
n
t
h
e
v
a
l
u
e
o
f
t
h
e
e
x
p
e
c
t
e
d
c
o
u
n
t
o
f
M
i
n
e
r
g
i
e
b
u
i
l
d
i
n
g
s
c
o
m
p
u
t
e
d
a
t
t
h
e
m
e
d
i
a
n
a
n
d
a
t
t
h
e
t
h
i
r
d
q
u
a
r
t
i
l
e
o
f
t
h
e
g
i
v
e
n
c
o
v
a
r
i
a
t
e
;
a
l
l
o
t
h
e
r
c
o
v
a
r
i
a
t
e
s
a
r
e
h
e
l
d
c
o
n
s
t
a
n
t
a
t
t
h
e
i
r
m
e
d
i
a
n
v
a
l
u
e
Table V.
Main estimation results,
ZINB model
“Green housing”
construction
95
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Per capita taxable income has a decisive impact on the number of Minergie buildings in a
municipality. Other things being equal, an increase in the proportion of taxpayers in the
highest income bracket from 29.0 percent (the median) to 36.0 percent (third quartile) is
associated with a 32.4 percent increase in the number of Minergie buildings. A share of
42.9 percent of taxpayers in the highest income bracket – corresponding to the
90th percentile, not shown in Table IV – is associated with an increase of 74 percent of
Minergie constructions.
High levels of environmentalism – as measured by the green index – are associated
with higher Minergie residential building densities. Amove fromthe median to the third
quartile of the index is followed by an increase of the green building density by
12.2 percent. A further move to the 90th percentile is associated with an increase of
27.5 percent in Minergie constructions. Substituting this measure of environmentalism
with the green parties’ voting share does not substantially alter the results (column 3 of
Table V). Raising the green voting share from 12.7 to 16.1 percent – which again
corresponds to a move fromthe median vote to the third quartile – leads to an increase in
the expected number of green buildings by 15.5 percent. A move to the 90th percentile
raises this effect to 24.4 percent.
Although the demographic structure of a municipality does affect the demand for
Minergie buildings, its impact is relatively small. It is further dif?cult to interpret. The
density of Minergie buildings increases with the share of 20-40 years old residents and
with the share of residents over 60 but is insensitive to the share of 40-60 years old[15].
The altitude, as a proxy for heating degree days, has a positive impact on the number
of green buildings, but its statistical signi?cance is quite sensitive to the model
speci?cation. It is highly signi?cant in the model of column 3, Table V, which is based on
a smaller sample. This is due to the fact that the cantons where the GPS did not run in the
2007 election are in majority located in the Swiss Alps and do not have a large share of
green buildings.
In both the base and the alternative speci?cation, we ?nd a weakly negative
correlation between the amount of subsidy payments per buildingandthe number of new
Minergie buildings. If we exclude the cantons that do not grant any subsidies from the
base case regression, the correlation is even lower (coef?cient of 20.013 and standard
error 0.0054). We do not thinkthat subsidypayments have triggereda signi?cant number
of Minergie certi?cations. The payments were likely to be too small compared to the extra
cost associated with the green building construction and certi?cation cost. The extra cost
is estimated at 5-10 percent of conventional construction cost, i.e. CHF25,000-CHF 50,000
for a typical CHF500,000 construction. Incontrast, the mediansubsidypayment was only
CHF 4,416.
The language af?liation strongly correlates with the number of Minergie buildings.
Minergie buildingdensityfor German-speakingmunicipalities is 76.2 percent higher than
for comparable French-, Italian-, or Romansh-speaking municipalities. These results
suggest that cultural norms may exert in?uence on environmental choice that is different
from the choices expressed by political af?liation. Alternatively, this difference may
simply re?ect more extensive marketing activities of the Minergie association in the
German-speaking region. However, the case of the bilingual (German and French) Canton
of Valais further hints at a different sensitivity towards green building issues across the
language border. The share of green buildings inthe German-speaking part of the Canton
JFEP
3,1
96
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
is roughly ten percentage points higher than in the Southern, French-speaking part,
although they share a similar economic environment and the same cantonal laws.
We thus perform a robustness exercise and limit the estimation of the ?xed effects
model (column 5, Table V) to the cantons of Bern, Fribourg, Valais, and Graubu¨nden, the
only multilingual cantons. These cantons belong to the largest in terms of the number of
municipalities. They make up 1,063 of the 2,561 observations in the national sample. In
this setting, which includes cantonal ?xed effects, the estimated parameter for the
linguistic region is determined solely by intra-cantonal variation in green housing
construction. We obtain a parameter for the German-speaking indicator variable of
0.781 (standard error 0.142). This is even larger than the national estimate of 0.566. As a
further test, we then split the national sample in two, the ?rst sub-sample containing all
1,561 German speaking, the second containing only the French, Italian- and
Romansh-speaking municipalities (n ¼ 1,110). We then run the ZINB count
regression with the base case speci?cation (column 1, Table V). The results of the
estimation are listed in the ?rst two columns of Table VI.
While many of the variables lose statistical signi?cance, the coef?cients of the most
signi?cant ones do not change very much, when compared to the pooled results of
Table V. Indeed, they are similar across the two distinct regional samples. We also note
that neither income, demographic nor a lower share of votes for green parties can
possibly explain the large differences in the Minergie density between Western
Switzerland and the German-speaking part of the country[16].
Estimation results
in non-German-
speaking
municipalities
Estimation results
in German-
speaking
municipalities
Logistic regression on
Minergie shares
Parameter Estimate Estimate Estimate Effects
a
Intercept 23.086 (1.759) 23.954 (0.922)
* * *
25.196 (0.372)
* * *
n/a
Number of new buildings in
municipality 0.732 (0.084)
* * *
0.649 (0.038)
* * *
n/a n/a
Green index 0.137 (0.072)
*
0.187 (0.038)
* * *
0.233 (0.014)
* * *
0.188
Share of taxpayers in mid-
income class 2.155 (1.281) 5.150 (0.037)
* * *
1.666 (0.3901)
*
0.039
Share of taxpayers in high-
income class 3.679 (0.873)
* * *
4.989 (0.663)
* * *
1.713 (0.2416)
* * *
0.112
Share of population aged
20-40 3.699 (2.490) 2.317 (1.354) 1.2451 (0.489)
*
0.023
Share of population aged
40-60 20.539 (2.665) 20.017 (1.495) 0.3752 (0.652) 0.007
Share of population over 60 1.091
*
(1.976) 1.399 (0.972) 0.794 (0.372) 0.020
Altitude above sea
level (km) 0.147 (0.176) 0.138 (0.094) 0.128 (0.043) 0.028
Subsidy per Minergie
building (cantonal level) 20.014
*
(0.006) 20.005 (0.094) 0.003 (0.002) 0.023
German-speaking
municipality (yes ¼ 1)
n/a
1,110
n/a
1,561
0.555 (0.043)
* * *
1,449 0.696
Observations log likelihood 21,186 23,490 –
Note:
a
See the notes in Table V
Table VI.
Count regression
estimates for the
linguistic regions and
logistic regression results
for the communities with
at least one green
building
“Green housing”
construction
97
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As a ?nal robustness test, we perform a logistic regression on the share of the Minergie
buildings in the 1,449 communities with at least one green building. The results (column
3 and 4, Table VI) are consistent with the previous ?ndings. We notice the stronger
impact of the green index and the somewhat lesser effect of income differences on the
density of Minergie-certi?ed buildings. Still, the linguistic af?liation remains the by far
strongest predictor of green building activity.
5. Policy implications and conclusion
In recent years, there has been a global surge in interest in energy-ef?cient buildings.
In Switzerland, green building construction has been largely left to the initiative of
private property investors and owner occupiers. Their willingness to incur both the
certi?cation costs and the signi?cantly larger costs associated with higher
energy-ef?ciency standards has supported the Minergie label.
As Switzerland boasts one of the highest densities of green buildings, it offers a
congenial environment to examine the determinants of green building activity. This
paper presents one of the ?rst empirical analyses of “what drives the demand for green
housing”. The heterogeneous spatial distribution of green buildings in Switzerland
allows us to examine the impact of a comprehensive series of municipality level
attributes on green housing density. We develop and test six hypotheses to explain this
heterogeneity, including demographic, geographic, social, cultural, and political aspects.
We ?nd that differences in income levels and linguistic af?liation account for the largest
part of the systematic variation in green building activity across the municipalities. The
impact of environmentalism, as measured by voting data, is statistically signi?cant but
less important.
We pay particular attention to the effectiveness of government subsidies granted by
15 of the 26 Swiss cantons. Our empirical results show that higher subsidy payments
for newMinergie buildings are not associated with a larger number of certi?cations. Our
interpretation of this regression result is twofold.
First, the negative correlation may be due to reverse causation. It is possible that
authorities incantons with a larger number of homeowners already inclined to build green
see little or no need to offer further incentives. The canton of Zurich, for example, offers no
subsidies for new Minergie buildings but has the highest density of green buildings.
The reverse may be true in cantons with few green homes. More re?ned data would be
needed to investigate the causal effect of subsidy payments on green building activity.
Second, as the median subsidy payment accounts for just about a tenth of the extra
building cost associated with the Minergie certi?cation, we conjecture that the subsidies
are too small to trigger green construction. Accordingly, other factors must drive the
decision to build “green”, the most obvious being the private bene?ts of a Minergie
certi?cation. These bene?ts likely include the improved building quality and comfort, as
well as the hedge against rising energy prices. Ideology, while not decisive, does
contribute somewhat to green building activity.
In contrast to Kahn and Vaughn (2009), our results suggest that the willingness to
incur the extra cost is predominately related to income levels rather than
to environmental ideology. As such, they are amenable to an interpretation related to
the environmental Kuznets curve, the observation that environmental quality often
appears to improve as income grows beyond a certain level. Indeed, the claim that
pollutants involving very dispersed externalities – such as carbon emissions related
JFEP
3,1
98
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to energy-inef?cient buildings – could have no turning point is still actively discussed
(Galeotti et al., 2006). Our results would argue against this claim.
We conjecture that the strong correlation of green buildings with linguistic af?liation
is a result of the higher awareness of the Minergie label in the German-speaking part of
Switzerland and of the varying af?nity towards green technology among different
cultural groups.
The demand for green housing is likely to be the result of complex attitudes and
actions involving public good aspects (a better environment) and private bene?ts
(higher building quality). As shown by Delmas and Grant (2008) for the case of organic
wine, the decision to eco-certify and label a product additionally involves subtle
informational issues, both on the producers’ and the consumers’ side. For the case of
green buildings, it would be interesting to follow this lead to address the issue of price
discrimination against the renters and buyers of green property. This is left to further
research.
Notes
1. See Salvi et al. (2010) for an overview of the green building labels available in Switzerland.
2. All Minergie related ?gures are based on data and publications of the Verein Minergie,
see web site: www.minergie.ch/publications.478.html.
3. Directive on the EPBD 2002/91/EC of the European Parliament and Council). The limit of the
energy bound A is set at 32 kWh/a. The energy band B corresponds to an annual energy
consumption between 32 and 65 kWh/a.
4. The focus on residential properties is also dictated by the limits of the Swiss construction
statistics. Annual new constructions data are available only for housing.
5. As of the beginning of 2009, LEED had approximately 2,000 certi?ed units. Minergie, in the
roughly 40 times smaller Swiss market, counted seven times more (Beyeler et al., 2009).
6. We give the exact description of the data sources and discuss the issues related to the
construction of the variables in the Data Appendix.
7. This proposition, however, is disputed. See, e.g. Kristro¨m and Riera (1996) for evidence of an
income elasticity of environmental improvements less than one.
8. The ordinary least squares regression of the heating degree day index of 44 locations in
Switzerland on the respective altitude has a R
2
of 0.96.
9. Note that for the initiative “Right of Appeal of NGOs” (Verbandsbeschwerdeinitiative) the
“no” votes signals support for environmental issues.
10. The wording of the initiative may have confused many voters. In an exit poll, one-third of the
voters recognized to have cast a vote against their true voting intentions (GfS Bern, 2008).
Although the unintended yes and no votes have approximately leveled out each other, the
deviation from the true voting intention might partially explain the lower correlation of this
initiative to the other initiatives.
11. Elections for the National Council are held every four years. Each of the 26 cantons is a
constituency. The number of deputies of each constituency depends on the population of the
canton.
12. If we do not allow for individual heterogeneity, we obtain the standard Poisson regression
model.
13. As of 2008, the average Swiss municipality had 2,945 residents.
“Green housing”
construction
99
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
14. For the sake of a clear exposition, we do not tabulate the coef?cients of the logistic model in
Table V. We also estimated the general speci?cation using alternative processes including
the zero-in?ated Poisson model, the negative binomial model, and the standard Poisson
model. The zero-in?ated negative binomial was found to have the best ?t, especially for
municipalities with no or low number of green buildings. Comparative sensitivity tests may
be obtained from the authors.
15. The reference category is the share of residents aged 0-19.
16. Unfortunately, we cannot further differentiate the impact of language af?liation. Both the
Italian- and the Romansh-speaking communities are almost completely located within the
boarders of a single canton, Ticino for the former and Graubu¨nden for the latter. They are
thus nearly collinear with the cantonal ?xed effects or with the subsidy variable.
References
Banzhaf, S.H. and Walsh, R.P. (2008), “Do people vote with their feet? An empirical test of
Tiebout’s mechanism”, The American Economic Review, Vol. 98, pp. 843-63.
Beresteanu, A. and Li, S. (2008), “Gasoline prices, government support, and the demand for
hybrid vehicles in the US”, working paper, Duke University, Durham, January.
Beyeler, F., Beglinger, N. and Roder, U. (2009), “Minergie: the Swiss sustainable building
standard”, Innovations, Vol. 4 No. 4, pp. 241-4.
Delmas, M.A. and Grant, L.E. (2008), “Eco-labeling strategies: the eco-premium puzzle in the wine
industry”, American Association of Wine Economists Working Paper No. 13, March.
Galeotti, M., Lanza, A. and Pauli, F. (2006), “Reassessing the environmental Kuznets curve for
CO
2
emissions: a robustness exercise”, Ecological Economics, Vol. 57 No. 1, pp. 152-63.
GfS Bern (2008), “Analyse der eidgeno¨ssischen Abstimmungen vom 30”, November, available at:
www.polittrends.ch/abstimmungen/abstimmungsanalysen/vox-analysen/081130d.
html#3 (accessed 1 August 2010).
Hausman, J.A. (1979), “Individual discount rates and the purchase and utilization of energy-using
durables”, Bell Journal of Economics, Vol. 10, pp. 33-54.
Hersch, J. and Viscusi, K. (2005), “The generational divide in support for environmental policies:
European evidence”, NBER Working Paper No. 11859, Cambridge, MA, December.
Kahn, M.E. (2007), “Do greens drive Hummers? Environmental ideology as a determinant of
consumer choice”, Journal of Environmental Economics and Management, Vol. 54 No. 2,
pp. 129-45.
Kahn, M.E. and Vaughn, R.K. (2009), “Green market geography: the spatial clustering of hybrid
vehicles and LEED registered buildings”, The B.E. Journal of Economic Analysis & Policy,
Vol. 9 No. 2 (Contributions), Article 2.
Kotchen, M. and Moore, M. (2007), “Private provision of environmental public goods: household
participation in green-electricity programs”, Journal of Environmental Economics and
Management, Vol. 53, pp. 1-16.
Kotchen, M. and Moore, M. (2008), “Conservation behavior from voluntary restraint to a
voluntary price premium”, Environmental and Resource Economics, Vol. 48, pp. 195-210.
Kristro¨m, B. and Riera, P. (1996), “Is the income elasticity of environmental improvements less
than one?”, Environmental and Resource Economics, Vol. 7, pp. 45-55.
Linn, J. and Klier, T. (2009), “The price of gasoline and the demand for fuel ef?ciency: evidence
from monthly new vehicles sales data”, FRB of Chicago Working Paper No. 2009-15,
Chicago, IL, August.
JFEP
3,1
100
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
MacKay, D.J.C. (2008), Sustainable Energy – Without the Hot Air, available at: www.
withouthotair.com/download.html (accessed 1 August 2010).
Milton, K. (1996), Environmentalism and Cultural Theory: Exploring the Role of Anthropology in
Environmental Discourse, Routledge, London.
Salvi, M., Horehajova, A. and Neeser, J. (2010), “Der Minergie-Boom unter der Lupe”, Zu¨rcher
Kantonalbank, available at: www.ccrs.uzh.ch/index.php/publikationen (accessed 1 August
2010).
UK Department for Communities and Local Government (2010), “English Housing Survey
2008-2009”, Headline Report, available at: www.communities.gov.uk/publications/
housing/ehs200809headlinereport (accessed 1 August 2010).
Winkelmann, R. (2000), Econometric Analysis of Count Data, Springer, Berlin.
Further reading
Leire, C. and Thidell, A. (2005), “Product-related environmental information to guide consumer
purchases – a review and analysis of research on perceptions, understanding and use
among Nordic consumers”, Journal of Cleaner Production, Vol. 13, pp. 1061-70.
Appendix. Data Appendix
The Data Appendix provides additional information on the data sources and discusses some
issues related to the construction of the variables used in the paper. All data are available at the
municipality level with the exception of the government subsides to Minergie buildings,
available at the cantonal level only. The number of political municipalities in Switzerland
steadily decreased from 2,899 at the beginning of the year 2000 to 2,596 at the end of 2009. Our
cross-section consistently distinguishes between 2,571 municipalities. Municipalities that
merged during the investigated time period are added together for the full-time period.
Political voting results
The ?rst indicator (GREEN_IND1) is based on a factor analysis of the results of ?ve federal
initiatives on environmental issues, listed in Table III. Out of the 44 national initiatives
submitted to the vote between 2000 and 2009, we identi?ed ?ve issues that were suitable to
characterize the voters’ sentiments towards environmentalism. As detailed in the main text, we
also use results of the most recent (2007) election for the Swiss National Council (GREEN_IND2).
Election and voting data at the municipal level can be downloaded at the site of the Swiss Federal
Of?ce of Statistics at: www.bfs.admin.ch/bfs/portal/de/index/themen/17/03.html
Altitude
The average altitude of municipalities (ALTITUDE) is extracted from the RIMINI public use
map of the Swiss Federal Of?ce of Topography at: www.swisstopo.admin.ch/internet/swisstopo/
de/home/products/downloads/height/rimini.html
Income
The share of residents subject to the Federal income tax in seven income brackets is from the
Swiss Federal Tax Administration, 2006. We merged it to three classes, CHF 0-40,000
(INCOME_LOW), 40,000-75,000 (INCOME_MID), and above 75,000 (INCOME_HIGH), at: www.
estv.admin.ch/dokumentation/00075/00076/00701/index.html?lang ¼ de#sprungmarke0_8
Minergie data
Minergie provides address data for all certi?ed new and retro?tted buildings, including the year
of certi?cation. We obtain 11,555 new residential buildings (MIN_BUILD) that were certi?ed
from 2001 to 2008, at: www.minergie.ch/list-of-buildings.html
“Green housing”
construction
101
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Linguistic af?liation
The language spoken by the majority of the residents in a municipality is fromthe 2000 decennial
census. Each municipality is assigned to one of the four of?cial languages in Switzerland,
i.e. German, French, Italian, and Romansh. In the regressions, we distinguish German speaking
(DGERMAN) from non-German-speaking municipalities, at: www.bfs.admin.ch/bfs/portal/de/
index/infothek/lexikon/bienvenue___login/blank/zugang_lexikon.topic.1.html
Demographical data
The share of the population in four age classes, 0-19 (AGE_0_19), 20-39 (AGE_20_39), 40-59
(AGE_40_59), and over 60 (AGE_60_99) in 2000 is obtained from the Swiss Federal Of?ce of
Statistics, at: www.bfs.admin.ch/bfs/portal/de/index/infothek/lexikon/bienvenue___login/blank/
zugang_lexikon.topic.1.html
New residential buildings construction
The number of new residential buildings per municipality between 1998 and 2008
(NEW_BUILD) is from the Swiss Federal Of?ce of Statistics, at: www.bfs.admin.ch/bfs/portal/
de/index/infothek/onlinedb/superweb/presentation_generale.html
Government subsidies
The average grant for a new Minergie building in each canton (MIN_GRANT) is obtained from
the Swiss Federal Of?ce of Energy. The data cover the period 2003-2008. For 2001-2003, the
share of funding allocated to new Minergie buildings is available at the national level only. We
allocated this sum to the cantons in proportion to their share of payments between 2004 and
2008. www.bfe.admin.ch/dokumentation/publikationen
Corresponding author
Marco Salvi can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
JFEP
3,1
102
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_645798193.pdf
Switzerland boasts arguably the highest density of green properties in the world. In 2008,
more than 15 percent of total new construction received the Swiss energy building label Minergie. The
spatial distribution of these green buildings, however, is highly heterogeneous. In some regions, more
than half of the new dwellings are built according to the Swiss green building standard. In others, this
share is still negligible. The purpose of this paper is to identify the determinants of the distribution of
green housing.
Journal of Financial Economic Policy
What drives “green housing” construction? Evidence from Switzerland
Marco Salvi J uerg Syz
Article information:
To cite this document:
Marco Salvi J uerg Syz, (2011),"What drives “green housing” construction? Evidence from Switzerland",
J ournal of Financial Economic Policy, Vol. 3 Iss 1 pp. 86 - 102
Permanent link to this document:http://dx.doi.org/10.1108/17576381111116777
Downloaded on: 24 January 2016, At: 21:40 (PT)
References: this document contains references to 20 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 800 times since 2011*
Users who downloaded this article also downloaded:
Steve Swidler, (2011),"Homeownership: yesterday, today and tomorrow", J ournal of Financial Economic
Policy, Vol. 3 Iss 1 pp. 5-11http://dx.doi.org/10.1108/17576381111116786
Beverley Searle, (2011),"Recession and housing wealth", J ournal of Financial Economic Policy, Vol. 3 Iss 1
pp. 33-48http://dx.doi.org/10.1108/17576381111116867
Dag Einar Sommervoll, Gavin Wood, (2011),"Home equity insurance", J ournal of Financial Economic
Policy, Vol. 3 Iss 1 pp. 66-85http://dx.doi.org/10.1108/17576381111116768
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
What drives “green housing”
construction? Evidence from
Switzerland
Marco Salvi
Zu¨ rcher Kantonalbank and Department of Architecture,
Swiss Federal Institute of Technology Zu¨ rich, Zu¨ rich, Switzerland, and
Juerg Syz
Diener Syz Real Estate, Zollikon and Shanghai, and Universita¨ t Zu¨ rich,
Zu¨ rich, Switzerland
Abstract
Purpose – Switzerland boasts arguably the highest density of green properties in the world. In 2008,
more than 15 percent of total new construction received the Swiss energy building label Minergie. The
spatial distribution of these green buildings, however, is highly heterogeneous. In some regions, more
than half of the new dwellings are built according to the Swiss green building standard. In others, this
share is still negligible. The purpose of this paper is to identify the determinants of the distribution of
green housing.
Design/methodology/approach – For 2,571 Swiss municipalities, the author computes the green
building share of new residential buildings. Data are collected for several variables measuring
demographic, geographic, social, cultural, and political aspects that – according to the authors’
hypothesis – may in?uence green building activity. Count regression is used to estimate the impact of
these variables on the demand for green buildings.
Findings – It is found that differences in income levels and cultural af?liation between Swiss
municipalities account for the largest part of the variation in green building activity. The impact of
homeowners’ stance on environmentalism is highly signi?cant but less important. Government
subsidies do not seem to trigger additional green housing activity.
Originality/value – The paper presents one of the ?rst empirical analyses regarding the
determinants of green building activity. Thanks to a comprehensive dataset, the authors are able to
investigate the impact of potential drivers of “green housing” construction activity. The regional
variation in governmental incentives is analysed and delivers valuable insight for policymakers
interested in spurring the development of green buildings.
Keywords Environmental regulations, Residential homes, Switzerland
Paper type Research paper
1. Introduction
The Swiss property market is an ideal playground to examine the determinants of the
demand for green properties. Indeed, Switzerland has one of the highest densities of
energy-ef?cient buildings in the world (Salvi et al., 2010). By mid-2010, more than 16,000
new and retro?tted buildings had received the Swiss green building label “Minergie”.
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The authors are indebted to Andrea Horeha´jova´, Julie Neeser, and Andreas Bro¨hl for helpful
comments and research assistance. The authors would also like to thank Steven Swidler,
Erika Meins, and Philippe Thalmann for their precious help and encouragement. The comments
of two anonymous reviewers are gratefully acknowledged.
JFEP
3,1
86
Journal of Financial Economic Policy
Vol. 3 No. 1, 2011
pp. 86-102
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381111116777
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In 2008, roughly 15 percent of newly constructed buildings successfully completed the
Minergie certi?cation process. This paper draws on data collected from the Swiss
market to investigate the question of “who builds green houses and why”. Previous
research has shown that homeowners do value the expected future cost savings
generated by investments in energy-ef?cient buildings. Several authors have
documented the incentive effects of higher energy prices on the demand for
energy-ef?cient technologies, see, e.g. Hausman, 1979; Beresteanu and Li, 2008; Linn and
Klier, 2009. However, this paper argues that more moderate utility bills alone do not
explain the demand for green properties.
We observe that the spatial distribution of green buildings in Switzerland is highly
heterogeneous. In some cities, more than half of the new constructions are built
according to the Swiss green building standard. In other regions, this share is negligible,
suggesting that there are more subtle drivers of the demand for green buildings than
energy cost savings. To detect these drivers, we investigate the determinants of green
housing activity that lead to regional clusters in Switzerland. Our approach is closely
related to Kahn and Vaughn (2009) who study clusters of Leadership in Energy and
Environmental Design (LEED) registered buildings and hybrid cars in the USA.
However, in their paper, these authors analyze just 10,000 registered LEED buildings
(765 of themcerti?ed) scattered across the USA. Inthis paper, we explore a market where
the density of certi?ed green buildings is by two orders of magnitude higher. Moreover,
thanks to its decentralized political structures and to the intensive use of direct
democratic instruments at the federal, cantonal, and municipal level, Switzerland offers
an ideal situation for studying the impact of environmentalism and government
subsidies on green building activity. We use a unique dataset, including all newly built,
Minergie labeled residential buildings in Switzerland. We relate the green housing
density in Swiss municipalities to corresponding demographic, geographic, social,
cultural, and political attributes. We include a measure of environmentalism based on
voting data as well as government subsidies offered at the level of the 26 Swiss cantons.
We ?nd that – among all investigated variables – differences in income across
municipalities account for the largest part of the explained variation in green building
activity. Linguistic af?liation, as a proxy for cultural norms, turns out to have a strong
impact on the regional distribution of Minergie residential buildings as well. The
in?uence of political af?liation, as measured by voting data, is statistically signi?cant
but less important. Government subsidies for green buildings do not appear to have any
positive impact on the clustering of green properties.
The paper is organized as follows. In Section 2, we describe the characteristics of the
Swiss green building standard Minergie and its rapid propagation. We also document
the large regional differences of green building activity. In Section 3, we develop six
hypotheses for potential drivers leading to the observed regional clusters of green
housing activity. We present the regression results with regard to the correlates of
green housing adoption in Section 4. We draw our conclusions in Section 5.
2. Green buildings in Switzerland
2.1 The Swiss green building standard “Minergie”
Minergie is the leading eco-label for energy-ef?cient buildings in Switzerland[1].
Anon-pro?t private association, Minergie is supported by its members, which include the
federal government, the cantons, schools, companies, individuals, andvarious associations.
“Green housing”
construction
87
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Minergie offers both eco-labeling and eco-certi?cation. Third parties, usually a cantonal
authority, certify Minergie buildings. There are three levels of Minergie building
certi?cations. The basic Minergie certi?cation is used broadly for new and retro?tted
buildings. To attain the standard, the building must achieve a reduction of at least
25 percent in general energy consumption in comparison to the average conventional
building. In addition to this requirement, fossil-fuel consumption has to be less than half of
that of the average conventional building. Minergie-Pis a stricter certi?cation that requires
very low energy consumption and is especially demanding with regard to heating energy
demand. This standard broadly corresponds to the German Passivhaus Standard. Finally,
Minergie-ECO involves an additional certi?cation that veri?es the use of
environmental-friendly building materials. In this paper, we do not differentiate between
the sub-labels as the basic Minergie label covers 95 percent of the certi?ed buildings[2].
The list of the certi?ed buildings is publicly available on the Minergie web site. Real estate
agents routinely advertise the presence of the label as a part of the sale process.
As with other green building labels, the implicit assumption of Minergie is that the
energy consumption of a dwelling is a function of its building standard. Energy
consumption estimates are basedonthe characteristics of the materials applied and used
to assess whether a new or retro?tted building quali?es for the Minergie label. As of
today (2010), the requirements of the basic Minegie standard described above set a limit
of 38 kWh per square meter of ?oor area and year. This corresponds roughly to the lower
bound of the energy rating “B” of the European Energy Performance of Buildings
Directive (EPBD)[3]. The use of active ventilation is mandatory to obtain certi?cation.
Since its launch in 1998, the Minergie label has been quite successful. In principle,
properties of all types – be it of?ce buildings or residential housing – can be certi?ed if
they meet the label’s criteria. In practice, however, private homeowners are at the
forefront of green building activity in Switzerland, as most green properties belong to
residential owner occupiers and private owners of residential multi-family buildings.
As of August 2009, 11,555 or 91 percent of all certi?ed buildings are residential
buildings, whereof 68 percent are single family and 32 percent multi-family homes.
Of the 9 percent of non-residential units, about 70 percent are owned by the public sector.
Schools, sports facilities or of?ce buildings make up the larger part of this category.
Because of the predominant share of residential buildings, we focus our analysis on this
segment[4]. Table I summarizes the distribution of certi?ed buildings by property type.
The number of new Minergie buildings tripled between 2004 and 2009. While at
the beginning of this period, only 5 percent of the new buildings received the label, this
proportionincreasedto15percent in2008. Figure 1showsthe number of Minergie-certi?ed
Number of Minergie-certi?ed buildings Percentage of total
Single-family homes 7,810 62
Multi-family homes 3,745 29
Others, whereof 1,101 9
Administration of?ces 432
Schools 308
Sport facilities 73
Total 12,656 100
Source: See Data Appendix
Table I.
Minergie-certi?ed
buildings in Switzerland
as of mid-2009
JFEP
3,1
88
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
new residential buildings and their share of total new residential buildings since 1998.
However, newconstructionrepresents onlya small part of the total built stock. Hence, only
about 1 percent of the existing Swiss buildings have been certi?ed so far. Nevertheless, to
the best of our knowledge, the rate of green buildings is higher in Switzerland than in
comparable countries. Indeed, Minergie’s penetration rate in Switzerland is roughly
280 times higher than LEED’s rate in the USA, where it represents the most widely used
green building label[5]. In England, the number of residential buildings in the
energy-ef?ciency rating bands “A” and “B” represented only 0.3 percent of the housing
stock in 2008 (UK Department for Communities and Local Government, 2010).
2.2 The spatial distribution of green housing
The geographical distribution of Minergie buildings in Switzerland is highly clustered.
Most Minergie houses are located in the northern and northeastern part of the country,
as well as in the cities of Bern and Geneva. To compare green building activity between
municipalities, we divide the number of Minergie-certi?ed new buildings by the
number of total residential buildings constructed between 1998 and 2008. Figure 2
shows a map of the spatial distribution of this share. The city of Zurich stands out with
a share of approximately 20 percent, followed by regions in the agglomeration of
Zurich. Alpine touristic resorts such as Davos and Zermatt also stand out for their high
share of energy-ef?cient buildings. In these resorts, roughly one new building in ten
has received the Minergie label. At the other end of the scale, the cantons of Ticino and
Jura and the area around the lake of Geneva exhibit very low green housing activity.
There, the share of new green buildings accounted for less than 2 percent.
On ?rst inspection, the share of Minergie-certi?ed houses seems to mirror the
linguistic regions in Switzerland. In the German-speaking part, every ?fth new
residential building completed in 2008 obtained the Minergie certi?cation, while in
French-speaking Romandy and in the Italian-speaking region only one in 12,
respectively, one in 14 did. On the other hand, French-speaking Geneva tops the list of
green building activity among Swiss cities, as shown in Table II. The heterogeneous
distribution of green building activity raises the question of the drivers of the demand
for green buildings. We address this question in the remainder of the paper.
Figure 1.
Number of Minergie
certi?ed new residential
buildings and their share
of total new residential
buildings, 1998 to 2008
3,000
2,500
2,000
1,500
1,000
500
0% 0
1999 2000 2001
Number of Minergie new homes
Share of Minergie new homes of total new homes
2002 2003 2004 2005 2006 2007 2008
2%
4%
6%
8%
10%
12%
14%
16%
18% “Green housing”
construction
89
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. The drivers of green building activity
3.1 Hypothesis development
Even though at least one Minergie building is present in more than half of the Swiss
municipalities, the share of green buildings varies widely across the municipalities. We
develop six hypotheses to explain this heterogeneity. Our list of likely correlates of green
housingactivityincludes demographic, geographic, social, cultural, andpolitical aspects[6].
3.1.1 Income. If green buildings are superior goods, their demand will be strongly
positively related to income[7]. We test the hypothesis that Minergie buildings are
more likely in richer municipalities:
H1. Green building demand increases with income.
3.1.2 Age. Some researchers have argued that the willingness to pay for environmentally
friendly goods declines with age (Hersch and Viscusi, 2005). We test this hypothesis by
including the municipalities’ age distribution in our regressions:
Rank City
Percentage of
all new buildings
Number of Minergie-certi?ed
new buildings
1 Geneva 34.5 39
2 Zurich 33.2 249
3 Bern 19.8 26
4 Winterthur 15.1 97
5 Lucerne 14.7 22
6 Basel 8.5 9
7 St Gall 8.4 18
8 Lugano 3.1 10
9 Lausanne 1.2 4
Source: See Data Appendix
Table II.
Recent green construction
activity in Switzerland’s
largest cities (2004-2008)
Figure 2.
Minergie’s share of new
residential buildings in
Swiss regions, 1998-2008
1998-2008
0% - 2%
2% - 4%
4% - 6%
6% - 9%
9% - 12%
12% - 15%
15% - 20%
JFEP
3,1
90
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H2. Green building demand is negatively related to age.
3.1.3 Cultural norms. At a more general level, cultural differences may be important in
addressing environmental problems (Milton, 1996). This may also in?uence green
building activity. As a multilingual country, Switzerland has natural cultural
boundaries within its national borders. We investigate whether the different linguistic
regions vary in their af?nity towards green housing:
H3. Green building demand varies with linguistic af?liation.
3.1.4 Geography. The energy consumption of a building depends crucially on heating
demand, which is a function of the difference between internal and external temperature
(MacKay, 2008). Outside temperature may thus affect the demand for energy-ef?cient
buildings. Unfortunately, average local temperature and heating degrees data is only
available for the limited number of communities in which a meteorological station is
located. In Switzerland, however, differences in temperature are strongly correlated with
altitude, which can be easily obtained for every community[8]. We thus include the
average height above sea level as a proxy for this demand driver:
H4. Green building demand is positively related to altitude.
3.1.5 Government subsides. We are interested in measuring the impact played by
governmental subsides on regional green building activity. The Swiss cantons have a
wide discretion when ?xing the amount of the subsidies to be granted to energy-ef?cient
buildings. As a result, payments for Minergie-certi?ed newbuildings vary considerably
fromcanton to canton. In 2008, the canton of Bern paid subsidies totaling CHF2.2 million
while 11 out of the 26 Swiss cantons, including the canton of Zurich, did not make any
subsidy payments. Among the cantons supporting Minergie new buildings, payments
ranged from CHF 3,700 per building in canton Ticino to CHF 31,340 in the canton of
Valais. We use the average subsidy payment per new Minergie building in each canton
to investigate the effectiveness of governmental programs:
H5. Green building demand is positively related to government subsidies.
3.1.6 Environmental activity. Environmentalists tend to be more likely to purchase
green products and may be willing to pay more for environmentally friendly products
(Kotchen and Moore, 2007, 2008; Kahn and Vaughn, 2009). We suppose that
municipalities with a large share of the population supportive of green ideas are more
likely to have a higher share of green buildings. We measure environmentalismbased on
revealed preference political data:
H6. Green building demand is positively related to the degree of
environmentalism in the municipality.
3.2 Measuring environmentalism
Following Kahn (2007) and Kahn and Vaughn (2009), we construct two indicators of
environmentalism at the municipal level. The basic rationale is that people “vote with
their feet” to ?nd the community that provides their optimal bundle of environmental
public goods and taxes (Banzhaf and Walsh, 2008). In Switzerland, voters have the
additional opportunity to decide directly on the bundle of environmental public goods.
Through ballot initiatives and referenda at the federal, cantonal, and municipal level,
“Green housing”
construction
91
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
they can propose or refuse particular changes in the legislation. At the municipal level,
voters are routinely asked to vote on the ?nancing of local infrastructure projects such as
schools or new of?ce buildings for the administration. The outcome of the voting on
these initiatives can be informative about the voters’ stance on environmental issues.
We base our ?rst indicator, the “greenindex”, onthe results of ?ve federal initiatives on
environmental issues, listed in Table III. All ?ve federal initiatives were rejected, most of
themby a wide margin[9]. To construct our ?rst aggregate measure of environmentalism,
we run a factor analysis on the voting results of the federal initiatives. The correlation of
the pro-environmental vote across the ?ve initiatives is high. The bivariate correlation
coef?cients range between 0.25 and 0.75. The 2008 federal initiative “Right of appeal of
NGOs” (Verbandsbeschwerdeinitiative) has the lowest correlation with other initiatives.
In the factor analysis, it receives the lowest factor loading (0.177)[10]. The factor loadings
of the other initiatives range between 0.23 and 0.32.
By this account, the list of the “greenest” communities in Switzerland closely
matches the list of the main cities with Zurich, Geneva, and Basel among the 5 percent
of the Swiss municipalities with the highest green index values. Of the ten largest
Swiss cities, only Lugano, situated in the Italian-speaking canton Ticino, does not
appear in the 10 percent of municipalities with the highest green index score.
The second measure of local environmentalism is based on the results of the 2007
election for the Swiss National Council, Switzerland’s lower house of parliament[11]. We
count the percentage of votes cast at the municipal level for the global positioning system
(GPS) and the 2004 founded Green Liberal Party (GLP). Environmental issues and the
promotion of renewable energies are at the core of both parties’ platforms. Accordingly,
the GPS endorsed all pro-environment initiatives listed in Table III. Both the GPS and the
GLP recommended the rejection of the “Right of appeal of NGOs” – initiative. On social
issues, the GPS is in general allied with left wing parties, whereas the GLP is positioned at
the center of the political spectrum. In the 2007 elections, the GLP won 1.4 percent of the
popular vote nationwide and three out of 200 seats. The GPS won 9.6 percent of the votes
and 20 seats. Unsurprisingly, green parties are strongest in the main urban areas. Six out
Ballot title Year Main aim Yes votes (%)
Participation
rate (%)
“Cut traf?c by half” 2000 Reduction of road traf?c by half over
a ten-year period
21.3 42.3
“Solar cent” 2000 Introduction of a tax of CHF 0.005 per
kWh on non-renewable energy
sources. Half of the tax ear-marked for
solar power uses
31.3 44.7
“Steering tax on non-
renewable energy”
2000 Introduction of a Pigou tax of CHF
0.02/kWh on non-renewable energy
sources
44.5 44.9
“Electricity without
nuclear power”
2003 Gradual abandonment of nuclear
energy
32.7 49.7
“Right of appeal for
NGOs”
2008 Curtailing of the rights of
environmental organizations to appeal
construction projects
33.0 47.4
Source: See Data Appendix
Table III.
Federal initiatives on
general environmental
issues in the period
2000-2008
JFEP
3,1
92
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
of the tenlargest Swiss cities belongto the 5 percent communities withthe highest share of
green parties votes. Again, Lugano stands out among this group with a share of only 4.6
percent of green votes, lower than the median at 8.5 percent. The Spearman rank-order
correlation between the two measures of environmentalism is 0.36.
The green parties did not run for election in the smaller, mostly rural constituencies.
Hence, this direct measure of environmentalism is not available for nine out of
26 cantons. Data on federal initiatives, however, are available for all municipalities.
We test both measures of environmentalism in the following regression analysis.
4. Empirical results
4.1 Model selection
We test the six hypotheses stated in Section 3 with data available at the municipal level.
We run count regressions on the number of Minergie-labeled properties built between
1998 and 2008 in each municipality. We take into account the non-negative
integer-valued aspect of the dependent variable. Speci?cally, we assume that the
conditional expectation of y
i
, the number of Minergie buildings in municipality i, is:
Eð y
i
jx
i
; t
i
Þ ¼ m
i
t
i
¼ expðx
0
i
b þ 1
i
Þ; ð1Þ
where x
i
is the vector of covariates and t
i
¼ exp 1
i
ð Þ is an heterogeneity factor
independent of x
i
. Taking the exponential ensures that the mean parameter is
non-negative; adding t
i
allows for unobserved heterogeneity between municipalities that
is not fully accounted for by the covariates[12]. It can be shown (Winkelmann, 2000) that
the distribution of y
i
conditional on x
i
and t
i
is Poisson distributed with:
gð y
i
Þ ¼ PðY
i
¼ y
i
jx
i
; t
i
Þ ¼
e
m
i
m
i
y
i
!
; i ¼ 0; 1; 2; . . . : ð2Þ
The heterogeneity factor, t
i
, can be integrated out of this conditional distribution under
the assumption that it is gamma distributed. This solution is called the negative binomial
model. It is more general than the Poisson regression. Its use is widespread because,
unlike the standard Poisson model, the conditional variance can exceed the conditional
mean. As such, it can accommodate over-dispersion resulting fromneglected unobserved
heterogeneity.
In contrast to standard logit regression, the use of count regression allows us the take
directly into account the fact that in 40.5 percent of all municipalities no Minergie house
has been built between 1998 and 2008. Zero-in?ated count models provide a simple way
of modeling so-called excess zeros (Winkelmann, 2000, p. 109). We thus explicitly model
the production of zero counts by specifying a Bernoulli trial that has g(y
i
) as outcome
with probability w
i
, or zero otherwise. This gives rise to a zero-in?ated negative binomial
(ZINB) model, where the probability of a non-zero event depends on the characteristics
of a municipality z
i
, speci?ed as:
w
i
¼ F
i
¼ Fðz
0
i
gÞ; ð3Þ
where the link function F is a logistic function. The standard estimator for the negative
binomial model is the maximum likelihood estimator. Estimates of the coef?cient
vectors b and g are found by minimization of the corresponding log-likelihood function
(Winkelmann, 2000). We next present estimates for various speci?cation of the ZINB
model. We discuss the estimation results for simpler models, such as the Poisson model,
the negative binomial model, and the zero-in?ated Poisson model.
“Green housing”
construction
93
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.2 Regression results
In the base model, the number of Minergie houses is regressed on several covariates
related to the hypotheses developed in Section 3. The covariates include the share of
residents in four age classes, the average altitude in the municipality, the majority
language spoken in the municipality, two indicators of environmentalism, the share of
residents in each of three income brackets, the amount of subsidies for Minergie
buildings and the total number of new buildings in the municipality. Again, we refer to
the Data Appendix for the data sources and the exact de?nition of the variables.
TableIVpresentssummarystatistics. Between1998and2008anaverageof 3.7Minergie
buildings and about 70 new buildings were completed in each municipality[13]. We notice
that there is not much variation in the population distribution by age. The median Swiss
municipality is situatedat an altitude of 613 meters above sea level. In the 25 percent richest
communities, at least 42 percent of the tax income payers are in the highest income bracket.
The ?rst column of Table V presents our base case estimation results for the vector
of parameters b in equation (1). In this speci?cation, we use the ZINB model and the
green index as indicator for environmentalism[14]. The coef?cients for the green index,
the income variables, and the language af?liation are highly signi?cant. To illustrate
the economic effect of the estimated coef?cients, we report the change in the expected
number of Minergie buildings per municipality that is associated with a given change
of a covariate. Thus, for each of the covariates in x
i
, we compute expðx
0
i
^
bÞ at the median
and at the third quartile of the distribution of the covariate and report the relative
change in the expected number of Minergie buildings in the second column. The table
further displays the estimation results using the green parties’ share of votes instead of
the green index as indicator for environmentalism, as well as estimation results based
on cantonal ?xed effects (columns 3 and 4). As a robustness exercise, we also report the
results of a regression with cantonal ?xed effects (column 5).
Variable Description Mean SD P25 Median P75
AGE_0_19 Share of population less than 20 0.252 0.042 0.227 0.254 0.280
AGE_20_39 Share of population aged 20-40 0.271 0.038 0.250 0.273 0.294
AGE_40_59 Share of population aged 40-60 0.281 0.033 0.261 0.280 0.301
AGE_60_99 Share of population over 60 0.196 0.052 0.161 0.190 0.222
ALTITUDE Altitude above sea level (km) 0.804 0.481 0.484 0.613 0.930
DGERMAN German-speaking municipality
(yes ¼ 1) 0.607 0.488 – 1.000 1.000
GREEN_IND1 Share of green parties’ votes
a
0.088 0.094 0.057 0.127 0.161
GREEN_IND2 Green index 0.000 1.000 20.686 20.069 0.576
INCOME_LOW Share of taxpayers in low-income class 0.290 0.100 0.222 0.270 0.330
INCOME_MID Share of taxpayers in mid-income class 0.410 0.060 0.385 0.420 0.450
INCOME_HIGH Share of taxpayers in high-income class 0.300 0.100 0.232 0.290 0.360
MIN_BUILD
Number of Minergie buildings in
municipality 3.718 10.322 – 1.000 3.000
MIN_GRANT
Cantonal subsidy per Minergie
building (1,000 CHF) 8.687 10.497 – 4.416 10.936
NEW_BUILD
Number of new buildings in
municipality 68.963 97.551 13.000 38.000 89.000
Note:
a
Only available for 2,219 municipalities
Source: See Data Appendix
Table IV.
Descriptive statistics for
the 2,571 Swiss
municipalities
JFEP
3,1
94
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
t
h
e
g
r
e
e
n
i
n
d
e
x
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
g
r
e
e
n
p
a
r
t
i
e
s
’
s
h
a
r
e
o
f
v
o
t
e
s
E
s
t
i
m
a
t
i
o
n
r
e
s
u
l
t
s
u
s
i
n
g
c
a
n
t
o
n
a
l
?
x
e
d
e
f
f
e
c
t
s
P
a
r
a
m
e
t
e
r
E
s
t
i
m
a
t
e
E
f
f
e
c
t
s
a
E
s
t
i
m
a
t
e
E
f
f
e
c
t
s
a
E
s
t
i
m
a
t
e
I
n
t
e
r
c
e
p
t
2
4
.
0
1
1
(
0
.
7
8
9
)
–
2
5
.
0
4
3
(
0
.
8
1
5
)
–
2
4
.
0
2
1
(
0
.
8
6
1
)
N
u
m
b
e
r
o
f
n
e
w
b
u
i
l
d
i
n
g
s
i
n
m
u
n
i
c
i
p
a
l
i
t
y
0
.
6
6
4
(
0
.
0
3
4
)
*
*
*
0
.
4
0
3
0
.
6
3
0
(
0
.
0
3
5
)
*
*
*
0
.
4
1
4
0
.
6
1
1
(
0
.
0
3
4
)
*
*
*
G
r
e
e
n
i
n
d
e
x
0
.
1
7
9
(
0
.
0
3
2
)
*
*
*
0
.
1
2
2
n
/
a
n
/
a
0
.
2
0
5
(
0
.
0
3
6
)
*
*
*
G
r
e
e
n
p
a
r
t
i
e
s
’
s
h
a
r
e
o
f
v
o
t
e
s
(
%
)
n
/
a
n
/
a
3
.
6
0
8
(
0
.
6
8
8
)
*
*
*
0
.
1
5
5
n
/
a
S
h
a
r
e
o
f
t
a
x
p
a
y
e
r
s
i
n
m
i
d
-
i
n
c
o
m
e
c
l
a
s
s
3
.
9
6
0
(
0
.
7
8
3
)
*
*
*
0
.
1
2
0
4
.
4
5
8
(
0
.
8
9
1
)
*
*
*
0
.
1
3
4
4
.
1
9
5
(
0
.
7
9
3
)
*
*
*
S
h
a
r
e
o
f
t
a
x
p
a
y
e
r
s
i
n
h
i
g
h
-
i
n
c
o
m
e
c
l
a
s
s
4
.
2
5
1
(
0
.
4
9
5
)
*
*
*
0
.
3
2
4
4
.
1
0
3
(
0
.
5
5
2
)
*
*
*
0
.
3
2
0
3
.
1
3
6
(
0
.
5
2
4
)
*
*
*
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
a
g
e
d
2
0
-
4
0
2
.
6
1
9
(
1
.
1
6
8
)
*
0
.
0
5
8
3
.
4
1
8
(
1
.
2
6
5
)
*
*
0
.
0
7
5
3
.
1
1
2
(
1
.
2
0
3
)
*
*
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
a
g
e
d
4
0
-
6
0
0
.
2
7
7
(
1
.
2
5
9
)
0
.
0
0
6
0
.
4
5
1
(
1
.
3
3
5
)
0
.
0
0
9
1
.
3
4
3
(
1
.
3
4
2
)
S
h
a
r
e
o
f
p
o
p
u
l
a
t
i
o
n
o
v
e
r
6
0
1
.
8
3
5
(
0
.
8
6
0
)
*
0
.
0
6
1
2
.
9
5
6
(
0
.
8
6
0
)
*
*
*
0
.
0
9
1
3
.
0
7
4
(
0
.
8
9
3
)
*
*
*
A
l
t
i
t
u
d
e
a
b
o
v
e
s
e
a
l
e
v
e
l
(
k
m
)
0
.
1
4
2
(
0
.
0
7
9
)
0
.
0
4
6
0
.
4
1
4
(
0
.
0
9
1
)
*
*
*
0
.
1
9
1
0
.
0
4
5
(
0
.
0
9
7
)
G
e
r
m
a
n
-
s
p
e
a
k
i
n
g
m
u
n
i
c
i
p
a
l
i
t
y
(
y
e
s
¼
1
)
0
.
5
6
6
(
0
.
0
8
5
)
*
*
*
0
.
7
6
2
0
.
5
2
9
(
0
.
0
9
3
)
*
*
*
0
.
6
9
7
0
.
6
5
6
(
0
.
1
3
6
)
*
*
*
S
u
b
s
i
d
y
p
e
r
M
i
n
e
r
g
i
e
b
u
i
l
d
i
n
g
(
c
a
n
t
o
n
a
l
l
e
v
e
l
)
2
0
.
0
0
8
(
0
.
0
0
4
)
*
2
0
.
0
5
3
2
0
.
0
1
5
(
0
.
0
0
4
)
*
*
*
0
.
1
2
5
n
/
a
O
b
s
e
r
v
a
t
i
o
n
s
2
,
5
1
6
2
,
2
1
9
2
,
5
1
6
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
5
,
*
*
1
a
n
d
*
*
*
0
.
1
p
e
r
c
e
n
t
;
a
p
e
r
c
e
n
t
a
g
e
d
i
f
f
e
r
e
n
c
e
i
n
t
h
e
v
a
l
u
e
o
f
t
h
e
e
x
p
e
c
t
e
d
c
o
u
n
t
o
f
M
i
n
e
r
g
i
e
b
u
i
l
d
i
n
g
s
c
o
m
p
u
t
e
d
a
t
t
h
e
m
e
d
i
a
n
a
n
d
a
t
t
h
e
t
h
i
r
d
q
u
a
r
t
i
l
e
o
f
t
h
e
g
i
v
e
n
c
o
v
a
r
i
a
t
e
;
a
l
l
o
t
h
e
r
c
o
v
a
r
i
a
t
e
s
a
r
e
h
e
l
d
c
o
n
s
t
a
n
t
a
t
t
h
e
i
r
m
e
d
i
a
n
v
a
l
u
e
Table V.
Main estimation results,
ZINB model
“Green housing”
construction
95
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Per capita taxable income has a decisive impact on the number of Minergie buildings in a
municipality. Other things being equal, an increase in the proportion of taxpayers in the
highest income bracket from 29.0 percent (the median) to 36.0 percent (third quartile) is
associated with a 32.4 percent increase in the number of Minergie buildings. A share of
42.9 percent of taxpayers in the highest income bracket – corresponding to the
90th percentile, not shown in Table IV – is associated with an increase of 74 percent of
Minergie constructions.
High levels of environmentalism – as measured by the green index – are associated
with higher Minergie residential building densities. Amove fromthe median to the third
quartile of the index is followed by an increase of the green building density by
12.2 percent. A further move to the 90th percentile is associated with an increase of
27.5 percent in Minergie constructions. Substituting this measure of environmentalism
with the green parties’ voting share does not substantially alter the results (column 3 of
Table V). Raising the green voting share from 12.7 to 16.1 percent – which again
corresponds to a move fromthe median vote to the third quartile – leads to an increase in
the expected number of green buildings by 15.5 percent. A move to the 90th percentile
raises this effect to 24.4 percent.
Although the demographic structure of a municipality does affect the demand for
Minergie buildings, its impact is relatively small. It is further dif?cult to interpret. The
density of Minergie buildings increases with the share of 20-40 years old residents and
with the share of residents over 60 but is insensitive to the share of 40-60 years old[15].
The altitude, as a proxy for heating degree days, has a positive impact on the number
of green buildings, but its statistical signi?cance is quite sensitive to the model
speci?cation. It is highly signi?cant in the model of column 3, Table V, which is based on
a smaller sample. This is due to the fact that the cantons where the GPS did not run in the
2007 election are in majority located in the Swiss Alps and do not have a large share of
green buildings.
In both the base and the alternative speci?cation, we ?nd a weakly negative
correlation between the amount of subsidy payments per buildingandthe number of new
Minergie buildings. If we exclude the cantons that do not grant any subsidies from the
base case regression, the correlation is even lower (coef?cient of 20.013 and standard
error 0.0054). We do not thinkthat subsidypayments have triggereda signi?cant number
of Minergie certi?cations. The payments were likely to be too small compared to the extra
cost associated with the green building construction and certi?cation cost. The extra cost
is estimated at 5-10 percent of conventional construction cost, i.e. CHF25,000-CHF 50,000
for a typical CHF500,000 construction. Incontrast, the mediansubsidypayment was only
CHF 4,416.
The language af?liation strongly correlates with the number of Minergie buildings.
Minergie buildingdensityfor German-speakingmunicipalities is 76.2 percent higher than
for comparable French-, Italian-, or Romansh-speaking municipalities. These results
suggest that cultural norms may exert in?uence on environmental choice that is different
from the choices expressed by political af?liation. Alternatively, this difference may
simply re?ect more extensive marketing activities of the Minergie association in the
German-speaking region. However, the case of the bilingual (German and French) Canton
of Valais further hints at a different sensitivity towards green building issues across the
language border. The share of green buildings inthe German-speaking part of the Canton
JFEP
3,1
96
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
is roughly ten percentage points higher than in the Southern, French-speaking part,
although they share a similar economic environment and the same cantonal laws.
We thus perform a robustness exercise and limit the estimation of the ?xed effects
model (column 5, Table V) to the cantons of Bern, Fribourg, Valais, and Graubu¨nden, the
only multilingual cantons. These cantons belong to the largest in terms of the number of
municipalities. They make up 1,063 of the 2,561 observations in the national sample. In
this setting, which includes cantonal ?xed effects, the estimated parameter for the
linguistic region is determined solely by intra-cantonal variation in green housing
construction. We obtain a parameter for the German-speaking indicator variable of
0.781 (standard error 0.142). This is even larger than the national estimate of 0.566. As a
further test, we then split the national sample in two, the ?rst sub-sample containing all
1,561 German speaking, the second containing only the French, Italian- and
Romansh-speaking municipalities (n ¼ 1,110). We then run the ZINB count
regression with the base case speci?cation (column 1, Table V). The results of the
estimation are listed in the ?rst two columns of Table VI.
While many of the variables lose statistical signi?cance, the coef?cients of the most
signi?cant ones do not change very much, when compared to the pooled results of
Table V. Indeed, they are similar across the two distinct regional samples. We also note
that neither income, demographic nor a lower share of votes for green parties can
possibly explain the large differences in the Minergie density between Western
Switzerland and the German-speaking part of the country[16].
Estimation results
in non-German-
speaking
municipalities
Estimation results
in German-
speaking
municipalities
Logistic regression on
Minergie shares
Parameter Estimate Estimate Estimate Effects
a
Intercept 23.086 (1.759) 23.954 (0.922)
* * *
25.196 (0.372)
* * *
n/a
Number of new buildings in
municipality 0.732 (0.084)
* * *
0.649 (0.038)
* * *
n/a n/a
Green index 0.137 (0.072)
*
0.187 (0.038)
* * *
0.233 (0.014)
* * *
0.188
Share of taxpayers in mid-
income class 2.155 (1.281) 5.150 (0.037)
* * *
1.666 (0.3901)
*
0.039
Share of taxpayers in high-
income class 3.679 (0.873)
* * *
4.989 (0.663)
* * *
1.713 (0.2416)
* * *
0.112
Share of population aged
20-40 3.699 (2.490) 2.317 (1.354) 1.2451 (0.489)
*
0.023
Share of population aged
40-60 20.539 (2.665) 20.017 (1.495) 0.3752 (0.652) 0.007
Share of population over 60 1.091
*
(1.976) 1.399 (0.972) 0.794 (0.372) 0.020
Altitude above sea
level (km) 0.147 (0.176) 0.138 (0.094) 0.128 (0.043) 0.028
Subsidy per Minergie
building (cantonal level) 20.014
*
(0.006) 20.005 (0.094) 0.003 (0.002) 0.023
German-speaking
municipality (yes ¼ 1)
n/a
1,110
n/a
1,561
0.555 (0.043)
* * *
1,449 0.696
Observations log likelihood 21,186 23,490 –
Note:
a
See the notes in Table V
Table VI.
Count regression
estimates for the
linguistic regions and
logistic regression results
for the communities with
at least one green
building
“Green housing”
construction
97
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As a ?nal robustness test, we perform a logistic regression on the share of the Minergie
buildings in the 1,449 communities with at least one green building. The results (column
3 and 4, Table VI) are consistent with the previous ?ndings. We notice the stronger
impact of the green index and the somewhat lesser effect of income differences on the
density of Minergie-certi?ed buildings. Still, the linguistic af?liation remains the by far
strongest predictor of green building activity.
5. Policy implications and conclusion
In recent years, there has been a global surge in interest in energy-ef?cient buildings.
In Switzerland, green building construction has been largely left to the initiative of
private property investors and owner occupiers. Their willingness to incur both the
certi?cation costs and the signi?cantly larger costs associated with higher
energy-ef?ciency standards has supported the Minergie label.
As Switzerland boasts one of the highest densities of green buildings, it offers a
congenial environment to examine the determinants of green building activity. This
paper presents one of the ?rst empirical analyses of “what drives the demand for green
housing”. The heterogeneous spatial distribution of green buildings in Switzerland
allows us to examine the impact of a comprehensive series of municipality level
attributes on green housing density. We develop and test six hypotheses to explain this
heterogeneity, including demographic, geographic, social, cultural, and political aspects.
We ?nd that differences in income levels and linguistic af?liation account for the largest
part of the systematic variation in green building activity across the municipalities. The
impact of environmentalism, as measured by voting data, is statistically signi?cant but
less important.
We pay particular attention to the effectiveness of government subsidies granted by
15 of the 26 Swiss cantons. Our empirical results show that higher subsidy payments
for newMinergie buildings are not associated with a larger number of certi?cations. Our
interpretation of this regression result is twofold.
First, the negative correlation may be due to reverse causation. It is possible that
authorities incantons with a larger number of homeowners already inclined to build green
see little or no need to offer further incentives. The canton of Zurich, for example, offers no
subsidies for new Minergie buildings but has the highest density of green buildings.
The reverse may be true in cantons with few green homes. More re?ned data would be
needed to investigate the causal effect of subsidy payments on green building activity.
Second, as the median subsidy payment accounts for just about a tenth of the extra
building cost associated with the Minergie certi?cation, we conjecture that the subsidies
are too small to trigger green construction. Accordingly, other factors must drive the
decision to build “green”, the most obvious being the private bene?ts of a Minergie
certi?cation. These bene?ts likely include the improved building quality and comfort, as
well as the hedge against rising energy prices. Ideology, while not decisive, does
contribute somewhat to green building activity.
In contrast to Kahn and Vaughn (2009), our results suggest that the willingness to
incur the extra cost is predominately related to income levels rather than
to environmental ideology. As such, they are amenable to an interpretation related to
the environmental Kuznets curve, the observation that environmental quality often
appears to improve as income grows beyond a certain level. Indeed, the claim that
pollutants involving very dispersed externalities – such as carbon emissions related
JFEP
3,1
98
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to energy-inef?cient buildings – could have no turning point is still actively discussed
(Galeotti et al., 2006). Our results would argue against this claim.
We conjecture that the strong correlation of green buildings with linguistic af?liation
is a result of the higher awareness of the Minergie label in the German-speaking part of
Switzerland and of the varying af?nity towards green technology among different
cultural groups.
The demand for green housing is likely to be the result of complex attitudes and
actions involving public good aspects (a better environment) and private bene?ts
(higher building quality). As shown by Delmas and Grant (2008) for the case of organic
wine, the decision to eco-certify and label a product additionally involves subtle
informational issues, both on the producers’ and the consumers’ side. For the case of
green buildings, it would be interesting to follow this lead to address the issue of price
discrimination against the renters and buyers of green property. This is left to further
research.
Notes
1. See Salvi et al. (2010) for an overview of the green building labels available in Switzerland.
2. All Minergie related ?gures are based on data and publications of the Verein Minergie,
see web site: www.minergie.ch/publications.478.html.
3. Directive on the EPBD 2002/91/EC of the European Parliament and Council). The limit of the
energy bound A is set at 32 kWh/a. The energy band B corresponds to an annual energy
consumption between 32 and 65 kWh/a.
4. The focus on residential properties is also dictated by the limits of the Swiss construction
statistics. Annual new constructions data are available only for housing.
5. As of the beginning of 2009, LEED had approximately 2,000 certi?ed units. Minergie, in the
roughly 40 times smaller Swiss market, counted seven times more (Beyeler et al., 2009).
6. We give the exact description of the data sources and discuss the issues related to the
construction of the variables in the Data Appendix.
7. This proposition, however, is disputed. See, e.g. Kristro¨m and Riera (1996) for evidence of an
income elasticity of environmental improvements less than one.
8. The ordinary least squares regression of the heating degree day index of 44 locations in
Switzerland on the respective altitude has a R
2
of 0.96.
9. Note that for the initiative “Right of Appeal of NGOs” (Verbandsbeschwerdeinitiative) the
“no” votes signals support for environmental issues.
10. The wording of the initiative may have confused many voters. In an exit poll, one-third of the
voters recognized to have cast a vote against their true voting intentions (GfS Bern, 2008).
Although the unintended yes and no votes have approximately leveled out each other, the
deviation from the true voting intention might partially explain the lower correlation of this
initiative to the other initiatives.
11. Elections for the National Council are held every four years. Each of the 26 cantons is a
constituency. The number of deputies of each constituency depends on the population of the
canton.
12. If we do not allow for individual heterogeneity, we obtain the standard Poisson regression
model.
13. As of 2008, the average Swiss municipality had 2,945 residents.
“Green housing”
construction
99
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
14. For the sake of a clear exposition, we do not tabulate the coef?cients of the logistic model in
Table V. We also estimated the general speci?cation using alternative processes including
the zero-in?ated Poisson model, the negative binomial model, and the standard Poisson
model. The zero-in?ated negative binomial was found to have the best ?t, especially for
municipalities with no or low number of green buildings. Comparative sensitivity tests may
be obtained from the authors.
15. The reference category is the share of residents aged 0-19.
16. Unfortunately, we cannot further differentiate the impact of language af?liation. Both the
Italian- and the Romansh-speaking communities are almost completely located within the
boarders of a single canton, Ticino for the former and Graubu¨nden for the latter. They are
thus nearly collinear with the cantonal ?xed effects or with the subsidy variable.
References
Banzhaf, S.H. and Walsh, R.P. (2008), “Do people vote with their feet? An empirical test of
Tiebout’s mechanism”, The American Economic Review, Vol. 98, pp. 843-63.
Beresteanu, A. and Li, S. (2008), “Gasoline prices, government support, and the demand for
hybrid vehicles in the US”, working paper, Duke University, Durham, January.
Beyeler, F., Beglinger, N. and Roder, U. (2009), “Minergie: the Swiss sustainable building
standard”, Innovations, Vol. 4 No. 4, pp. 241-4.
Delmas, M.A. and Grant, L.E. (2008), “Eco-labeling strategies: the eco-premium puzzle in the wine
industry”, American Association of Wine Economists Working Paper No. 13, March.
Galeotti, M., Lanza, A. and Pauli, F. (2006), “Reassessing the environmental Kuznets curve for
CO
2
emissions: a robustness exercise”, Ecological Economics, Vol. 57 No. 1, pp. 152-63.
GfS Bern (2008), “Analyse der eidgeno¨ssischen Abstimmungen vom 30”, November, available at:
www.polittrends.ch/abstimmungen/abstimmungsanalysen/vox-analysen/081130d.
html#3 (accessed 1 August 2010).
Hausman, J.A. (1979), “Individual discount rates and the purchase and utilization of energy-using
durables”, Bell Journal of Economics, Vol. 10, pp. 33-54.
Hersch, J. and Viscusi, K. (2005), “The generational divide in support for environmental policies:
European evidence”, NBER Working Paper No. 11859, Cambridge, MA, December.
Kahn, M.E. (2007), “Do greens drive Hummers? Environmental ideology as a determinant of
consumer choice”, Journal of Environmental Economics and Management, Vol. 54 No. 2,
pp. 129-45.
Kahn, M.E. and Vaughn, R.K. (2009), “Green market geography: the spatial clustering of hybrid
vehicles and LEED registered buildings”, The B.E. Journal of Economic Analysis & Policy,
Vol. 9 No. 2 (Contributions), Article 2.
Kotchen, M. and Moore, M. (2007), “Private provision of environmental public goods: household
participation in green-electricity programs”, Journal of Environmental Economics and
Management, Vol. 53, pp. 1-16.
Kotchen, M. and Moore, M. (2008), “Conservation behavior from voluntary restraint to a
voluntary price premium”, Environmental and Resource Economics, Vol. 48, pp. 195-210.
Kristro¨m, B. and Riera, P. (1996), “Is the income elasticity of environmental improvements less
than one?”, Environmental and Resource Economics, Vol. 7, pp. 45-55.
Linn, J. and Klier, T. (2009), “The price of gasoline and the demand for fuel ef?ciency: evidence
from monthly new vehicles sales data”, FRB of Chicago Working Paper No. 2009-15,
Chicago, IL, August.
JFEP
3,1
100
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
MacKay, D.J.C. (2008), Sustainable Energy – Without the Hot Air, available at: www.
withouthotair.com/download.html (accessed 1 August 2010).
Milton, K. (1996), Environmentalism and Cultural Theory: Exploring the Role of Anthropology in
Environmental Discourse, Routledge, London.
Salvi, M., Horehajova, A. and Neeser, J. (2010), “Der Minergie-Boom unter der Lupe”, Zu¨rcher
Kantonalbank, available at: www.ccrs.uzh.ch/index.php/publikationen (accessed 1 August
2010).
UK Department for Communities and Local Government (2010), “English Housing Survey
2008-2009”, Headline Report, available at: www.communities.gov.uk/publications/
housing/ehs200809headlinereport (accessed 1 August 2010).
Winkelmann, R. (2000), Econometric Analysis of Count Data, Springer, Berlin.
Further reading
Leire, C. and Thidell, A. (2005), “Product-related environmental information to guide consumer
purchases – a review and analysis of research on perceptions, understanding and use
among Nordic consumers”, Journal of Cleaner Production, Vol. 13, pp. 1061-70.
Appendix. Data Appendix
The Data Appendix provides additional information on the data sources and discusses some
issues related to the construction of the variables used in the paper. All data are available at the
municipality level with the exception of the government subsides to Minergie buildings,
available at the cantonal level only. The number of political municipalities in Switzerland
steadily decreased from 2,899 at the beginning of the year 2000 to 2,596 at the end of 2009. Our
cross-section consistently distinguishes between 2,571 municipalities. Municipalities that
merged during the investigated time period are added together for the full-time period.
Political voting results
The ?rst indicator (GREEN_IND1) is based on a factor analysis of the results of ?ve federal
initiatives on environmental issues, listed in Table III. Out of the 44 national initiatives
submitted to the vote between 2000 and 2009, we identi?ed ?ve issues that were suitable to
characterize the voters’ sentiments towards environmentalism. As detailed in the main text, we
also use results of the most recent (2007) election for the Swiss National Council (GREEN_IND2).
Election and voting data at the municipal level can be downloaded at the site of the Swiss Federal
Of?ce of Statistics at: www.bfs.admin.ch/bfs/portal/de/index/themen/17/03.html
Altitude
The average altitude of municipalities (ALTITUDE) is extracted from the RIMINI public use
map of the Swiss Federal Of?ce of Topography at: www.swisstopo.admin.ch/internet/swisstopo/
de/home/products/downloads/height/rimini.html
Income
The share of residents subject to the Federal income tax in seven income brackets is from the
Swiss Federal Tax Administration, 2006. We merged it to three classes, CHF 0-40,000
(INCOME_LOW), 40,000-75,000 (INCOME_MID), and above 75,000 (INCOME_HIGH), at: www.
estv.admin.ch/dokumentation/00075/00076/00701/index.html?lang ¼ de#sprungmarke0_8
Minergie data
Minergie provides address data for all certi?ed new and retro?tted buildings, including the year
of certi?cation. We obtain 11,555 new residential buildings (MIN_BUILD) that were certi?ed
from 2001 to 2008, at: www.minergie.ch/list-of-buildings.html
“Green housing”
construction
101
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Linguistic af?liation
The language spoken by the majority of the residents in a municipality is fromthe 2000 decennial
census. Each municipality is assigned to one of the four of?cial languages in Switzerland,
i.e. German, French, Italian, and Romansh. In the regressions, we distinguish German speaking
(DGERMAN) from non-German-speaking municipalities, at: www.bfs.admin.ch/bfs/portal/de/
index/infothek/lexikon/bienvenue___login/blank/zugang_lexikon.topic.1.html
Demographical data
The share of the population in four age classes, 0-19 (AGE_0_19), 20-39 (AGE_20_39), 40-59
(AGE_40_59), and over 60 (AGE_60_99) in 2000 is obtained from the Swiss Federal Of?ce of
Statistics, at: www.bfs.admin.ch/bfs/portal/de/index/infothek/lexikon/bienvenue___login/blank/
zugang_lexikon.topic.1.html
New residential buildings construction
The number of new residential buildings per municipality between 1998 and 2008
(NEW_BUILD) is from the Swiss Federal Of?ce of Statistics, at: www.bfs.admin.ch/bfs/portal/
de/index/infothek/onlinedb/superweb/presentation_generale.html
Government subsidies
The average grant for a new Minergie building in each canton (MIN_GRANT) is obtained from
the Swiss Federal Of?ce of Energy. The data cover the period 2003-2008. For 2001-2003, the
share of funding allocated to new Minergie buildings is available at the national level only. We
allocated this sum to the cantons in proportion to their share of payments between 2004 and
2008. www.bfe.admin.ch/dokumentation/publikationen
Corresponding author
Marco Salvi can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
JFEP
3,1
102
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
0
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_645798193.pdf