scispace - formally typeset
Open AccessJournal ArticleDOI

Examining the spatial relationship between environmental health factors and house prices: NO2 problem?

TLDR
In this article, the authors present a spatial analysis of air quality and noise pollution and their association with house prices, using 2,501 sale transactions for the period 2013, using three different spatial modelling approaches, namely, ordinary least squares using spatial dummies, a geographically weighted regression (GWR) and a spatial lag model (SLM).
Abstract
Purpose Air quality, noise and proximity to urban infrastructure can arguably have an important impact on the quality of life. Environmental quality (the price of good health) has become a central tenet for consumer choice in urban locales when deciding on a residential neighbourhood. Unlike the market for most tangible goods, the market for environmental quality does not yield an observable per unit price effect. As no explicit price exists for a unit of environmental quality, this paper aims to use the housing market to derive its implicit price and test whether these constituent elements of health and well-being are indeed capitalised into property prices and thus implicitly priced in the market place. Design/methodology/approach A considerable number of studies have used hedonic pricing models by incorporating spatial effects to assess the impact of air quality, noise and proximity to noise pollutants on property market pricing. This study presents a spatial analysis of air quality and noise pollution and their association with house prices, using 2,501 sale transactions for the period 2013. To assess the impact of the pollutants, three different spatial modelling approaches are used, namely, ordinary least squares using spatial dummies, a geographically weighted regression (GWR) and a spatial lag model (SLM). Findings The findings suggest that air quality pollutants have an adverse impact on house prices, which fluctuate across the urban area. The analysis suggests that the noise level does matter, although this varies significantly over the urban setting and varies by source. Originality/value Air quality and environmental noise pollution are important concerns for health and well-being. Noise impact seems to depend not only on the noise intensity to which dwellings are exposed but also on the nature of the noise source. This may suggest the presence of other externalities that arouse social aversion. This research presents an original study utilising advanced spatial modelling approaches. The research has value in further understanding the market impact of environmental factors and in providing findings to support local air zone management strategies, noise abatement and management strategies and is of value to the wider urban planning and public health disciplines.

read more

Content maybe subject to copyright    Report

Journal of European Real Estate Research
Examining the Spatial relationship between environmental
health factors and house prices: NO2 problem?
Journal:
Journal of European Real Estate Research
Manuscript ID
JERER-01-2018-0008.R1
Manuscript Type:
Research Paper
Keywords:
housing markets, environmental health factors, air quality, noise pollution,
Geographically weighted regression, Spatial Lag Model
Journal of European Real Estate Reserach
'This author accepted manuscript is deposited under a Creative Commons Attribution Non-commercial 4.0
International (CC BY-NC) licence. This means that anyone may distribute, adapt, and build upon the work for non-
commercial purposes, subject to full attribution. If you wish to use this manuscript for commercial purposes, please
contact permissions@emerald.com.'

Journal of European Real Estate Research
1
Examining the Spatial relationship between environmental health factors and house prices: NO
2
problem?
Abstract
Purpose: The impact of both air quality, noise and proximity to urban infrastructure can arguably
have an important impact on the quality of life. Environmental quality (the price of good health) has
become a central tenet for consumer choice in urban locales when deciding on a residential
neighbourhood. Unlike the market for most tangible goods, the market for environmental quality does
not yield an observable per unit price effect. As no explicit price exists for a unit of environmental
quality, this paper utilizes the housing market to derive its implicit price and test whether these
constituent elements of health and wellbeing are indeed capitalised into property prices and thus
implicitly priced in the market place.
Design: A considerable number of studies have used hedonic pricing models incorporating spatial
effects to assess the impact of air quality, noise and proximity to noise pollutants on property
market pricing. This study presents a spatial analysis of air quality and noise pollution and their
association with house prices, using 2,501 sale transactions for the period 2013. To assess the impact of
the pollutants, three different spatial modelling approaches are employed, namely, an OLS using spatial
dummies, a Geographically Weighted Regression and a Spatial Lag Model.
Findings: The findings suggest that air quality pollutants have an adverse impact on house prices which
fluctuates across the urban area. The analysis suggests that the noise level does matter, although this
varies significantly over the urban setting and varies by source.
Originality/value: Air quality and environmental noise pollution are important concerns for health
and wellbeing. Noise impact seems to depend not only on the noise intensity to which dwellings are
exposed but also on the nature of the noise source. This may suggest the presence of other
externalities that arouse social aversion. This research presents an original study utilising advanced
spatial modelling approaches. The research has value in further understanding the market impact of
environmental factors and in providing findings to support local air zone management strategies,
noise abatement and management strategies and is of value to the wider urban planning and public
health disciplines.
KEY WORDS: housing markets, environmental health factors, air quality, noise pollution, house prices,
GWR, SLM
Introduction
All locations experience an array of impacts from environmental factors, comprising both positive and
negative dis(amenity) effects. However, some environmental factors, in addition to their pot
ential to cause
nuisance or loss of amenity, can have profound implications for the health and wellbeing of individuals
living within affected areas. Existing research has tended to illustrate that prolonged exposure to poor air
and noise quality comprises an adverse impact upon health in society in a number of ways, ranging from
simple annoyance (Ouis, 2001; Ohrstrom et al., 2007; de Kluizenaar et al., 2013), sleep disturbance (Net,
Page 1 of 47 Journal of European Real Estate Reserach
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

Journal of European Real Estate Research
2
2004), increasing risk of stroke
1
(Sørensen et al., 2011), hypertension
1
(Jarup et al., 2008; Bodin et al.,
2009), myocardial infarction (Babisch et al., 2005), and neuro-degenerative conditions (Rückerl et al.,
2011; Chen et al., 2017). Indeed, in their recent study, Chen et al. (2017) found that residing adjacent to a
c
ongested road (within 50 metres) adversely affects cognition and the likelihood of higher incidence of
dementia of up to 7%, with 11% of cases of dementia linked to air pollution.
It is trite that humans are adversely affected by ex
pose to pollutants in ambient air. In response, the UK,
in common with all EU member states, has an extensive environmental protection regime which has
p
roduced substantial improvements in environmental quality over the last seventy years. Urban areas for
the most part, no longer suffer from sulphur dioxide and smog from coal burning which produced ver
y
visible pollution. Whilst various sources of visible air pollution have been largely
remediated, by the
implementation of clean air legislation and the uptake of the use of natural
gas for home heating and
energy generation, air pollution especially from road transport (nitrogen dioxide and particulate matter
(
PM
2.5
)) and energy production continues to have major health and quality of life impacts, particularly
exposure in urban settings. The increasingly problem of air quality
has been highly publicised in popular
media reports which showcase ongoing problems and challenges pertaining
to both air and traffic
pollution. Pertinently, across the UK, air pollution is estimated to contribute 40,000 deaths per annum.
Indeed, recent reports have highlighted that combined the effects of long-term exposure to nitrogen
dioxide (NO
2
) and a particulate matter (PM2.5) in the UK’s largest conurbation, London, is linked to
5,900 and 3,500 deaths respectively
2
. Indeed, a White paper issued by the London Assembly Health and
Environment Committee (2012)
3
indicates that up to 9% of deaths in London are caused by air-borne
man-made particles.
Under EU law, health based standards and objectives stipulate that
the average hourly level of NO
2
in the
air must not exceed 200 micrograms (per cubic metre) more than 18 times in a year
4
. Nonetheless,
research has shown that EU-set limits on key pollutants have been frequently breached over the last
decade and in large conurbations the stipulated EU annual air pollution limits have been breached in a
matter of days
5
(London AQN, 2017)
6
. This has been witnessed on the legal front with the High Court
ruling that existing approaches to tackle pollution are not sufficient and ordering urgent changes to
regulate and ‘clean’ London’s air. This has once again led to
calls for sharper policy responses and
solutions, with recommendations including the ‘phasing out’ of diesel vehicles, the creation of Ultra Low
Emission Zones. The Government response has been to commit to channeling more funding resources
into tackling air pollution in the UK of circa £875 million over the next five years
7
.
In the UK, air quality is monitored by both central and local government with a regulatory system of air
quality management and assessment setting air (environmental) quality standards and objectives for
specific pollutants (DEFRA, 2007:13). The Air Quality Standards Regulations set out the responsibilities
1 Traffic noise greater than 60dBA increases higher risk for stroke (Sørensen et al., 2011). Traffic noise [24-hour average] of 55
dBA @ a higher risk for hypertension (Bodin et al., 2009)
2
http://www.bbc.co.uk/news/uk-england-london-33536989.
3
London Assembly Health and Environment Committee (2012) Air Pollution in London, Issues Paper, December 2012.
4
http://ec.europa.eu/environment/air/quality/standards.htm
5
Annual air pollution limit breached on 19 occasions within a 5 day period for a south London road. At one point NO
2
levels
were nearly double the legal limit. Putney High Street, which was the first London road to exceed its legal limit last year, went on
to exceed the hourly limit more than 1,100 times in 2016.
6
http://www.londonair.org.uk/LondonAir/General/research.aspx.
7
https://tfl.gov.uk/info-for/media/press-releases/2017/june/mayor-launches-plan-to-improve-air-quality-on-the-tube
Page 2 of 47Journal of European Real Estate Reserach
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

Journal of European Real Estate Research
3
of local authorities in relation to air quality management and require that air quality reviews be conducted
t
o assess the quality of air within local authority districts and that Air Quality Management Areas
(AQMA) be declared where it appears that standards or objectives are not being achieved. In locations
where the air quality objectives are not met, local authorities are required to produce an air quality action
plan setting out how the standards will be met within a specified period. Yet despite these policy
objectives and promises, in Northern Ireland, ambient air quality continues to pose detrimental effects on
health and quality of life. The Department of Environment Farming and Rural Affairs (DEFRA, 2016)
stated that recent evidence on the health impact of exposure to nitrogen dioxide has strengthened
s
ignificantly, with reports persistently showing health warnings as a consequence of high levels of air
pollution
8
. Indeed, evidence has been released directly linking Nitrogen Dioxide exposure to mortality
rates. Additionally, in light of the fact that many of the sources of Nitrogen Dioxide are also sources of
particulate matter (PM), the impact of exposure to particulate matter pollution (PM2.5) is estimated to
have an effect on mortality equivalent to nearly 29,000 deaths (DEFRA, 2016:7). As a consequence, there
are currently (as of 2017) 26 AQMA’s declared in Northern Ireland; 19 of which are for Nitrogen Dioxide
emissions from road traffic sources and the remainder principally for PM10 from domestic (solid fuel
burning) sources. As a result, four Air Quality Managements Areas have been declared in the Belfast area
in relation to Nitrogen Dioxide emissions from vehicles.
In a similar vein, whilst exposure to noise is inescapable and usually has limited impact on quality of life,
in certain instances it can be so intrusive as to cause significant adverse effects on health (McKay and
Murray, 2017). In a general sense, noise may be characterised as “environmental noise - noise from
transportation and industrial sources; and neighbourhood noise - noise arising from within the community
such as from entertainment premises, trade and business premises, construction noise and noise in the
street (DoE NI, 2014:3). Nonetheless, noise is also a material planning consideration and “the planning
system has a role to play in minimising the potential for adverse impact upon health and well-being
through noise, by means of its influence on the location, layout and design of new development and
consideration of the amenity impacts” (DoE NI, 2015:117). While nuisance legislation deals with noise
from specific sources, the Environmental Noise Directive 2002/49/EC provides the framework to identify
noise pollution levels and trigger the necessary action across the EU – generally implemented through the
Environmental Noise Regulations within the UK. In Northern Ireland, the Environmental Noise
Regulations 2006 (NI) which require a number of actions: namely the assessment of exposure to
environmental noise; the provision of information on environmental noise and its effects on the public;
preventing and reducing environmental noise and preserving environmental noise quality where it is
good. These assessments have necessitated the production of strategic noise maps identifying areas which
have roads with more than 3 million vehicle movements annually, railways with more than 30,000 train
passages annually, airports with more than 50,000 movements annually and urban areas with more than
100,000 inhabitants.
Arguably, the exposure to air and noise pollution remains a major cause of ill health and mortality within
the UK. Despite this acknowledgement, an important aspect of assessing the effectiveness of
environmental policies that address the improvement of environmental air and noise quality is obtaining a
quantitative measure of the economic value of any accrued benefits or negatives across geographic
8
https://www.irishnews.com/news/northernirelandnews/2016/11/25/news/air-pollution-health-warnings-in-belfast-and-derry-
801996/
Page 3 of 47 Journal of European Real Estate Reserach
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

Journal of European Real Estate Research
4
neighbourhoods (Freeman 2003). In the absence of an explicit market for clean air (quality), and noise
pollution, this paper empirically assesses whether these environmental amenities such as (perceived) good
air quality and reduced noise levels are capitalised into property prices.
There are a variety of spatial based modelling frame
works in existence for examining house prices and
housing markets. As discussed by Khalid (2015), a vast number of contemporary studies are
incorporating explicit consideration of spatial effects in the estimation of hedonic price functions. This
paper therefore examines three differing spatial methodologies to account for potential missing spatial
variables, spatial trends and spatial heterogeneity when considering the effect of environmental concerns
on house prices. We focus on three methods and consider some methodological is
sues associated with the
estimation of an implicit price for clean air and noise by including a number of parameters and distance
bands within the hedonic modelling frameworks. The rationale behind this approach is that, ceteris
paribus, houses located in areas with cleaner air or reduced noise will hav
e this benefit capitalized into
their value - reflected in a higher or lower sales price. Interestingly, if environmental factors such as air
and noise pollution, in addition to their recognised impacts on the health and wellbeing, also have an
association with the price of residential property across a range of exposure levels (natural consequence)
this perhaps illustrates a quadratic trade-off.
This is key for informing regional and local policy
as to the (dis)amenity effect of noise and air quality,
undoubtedly helping urban renewal and revitalisation strategies (intensification) and providing evidence
base for the cost of ‘good’ or ‘poor’ planning. The paper proceeds as follows, Section 2 reviews the
relevant literature related to house prices and the role of environmental quality – specifically air and noise
attributes within housing markets. This is followed by the data and methodology section with results and
a discussion presented in Section 4. Finally, conclusions are offered.
Literature
In the context of housing literature, amenities and environment effects are key considerations, and
hedonic methods with spatial analyses have gained popularity to provide estimates of the proximity
“effect” of a variety of positive and negative environment-specific externalities on property prices
(McConnell and Walls, 2005). Numerous studies have examined proximate locational externalities
(Kauko, 2003) demonstrating added or destroyed value based on the urban environment (Des Rosiers et
al., 2002), neighbourhood style and distance and accessibility to amenities (Brunauer et al., 2013; Liao
and Chen, 2013; Reed, 2013; Dube¨ et al., 2014) displaying mixed pricing effects. There is a long and
evolutionary history investigating the relationship between air (quality) and noise (pollution) and property
prices within hedonic price schedules, with the approach becoming an established methodology in
environmental economics (Anselin and Lozano-Gracia, 2008).
Seminal studies examined the relationship between air pollution (dosages) and real estate values (Crocker,
1968; Reid, 1962; Ridker, 1967; Anderson and Crocker, 1971), illustrating the fundamental thesis that a
portion of air pollution damage to artifacts and organisms is capitalised negatively into the value of land
and immobile durable improvements thereon. The studies of Ridker and Henning (1967) and Harrison
and Rubinfeld (1978) served to generate a voluminous literature base scrutinising the theoretical,
methodological and empirical aspects of air quality (Anselin and Lozano-Gracia, 2008) suggesting that
property value differences are a result of variations in air pollution. This is rife in numerous international
b
ased studies. In the Latin American context, a number of studies have examined the relationship between
air quality and property value (Filippini and Martínez-Cruz, 2016). In Columbia, and primarily the
Page 4 of 47Journal of European Real Estate Reserach
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

Citations
More filters
Posted Content

A Spatial Difference-in-Differences Estimator to Evaluate the Effect of Change in Public Mass Transit Systems on House Prices

TL;DR: In this article, a spatial Difference-in-Differences (SDID) estimator is proposed to account for possible spatial spillover effects, and an empirical application of the SDID estimator based on the development of a new commuter rail transit system for the suburban agglomeration of Montreal (Canada).
Journal ArticleDOI

Take a Look Around: Using Street View and Satellite Images to Estimate House Prices

TL;DR: In this article, the authors proposed a pipeline that uses a deep neural network model to automatically extract visual features from images to estimate house prices in London, UK and found that learning to characterize the urban quality of a neighborhood improves house price prediction, even when generalizing to previously unseen London boroughs.
Journal ArticleDOI

Take a Look Around: Using Street View and Satellite Images to Estimate House Prices

TL;DR: In this paper, the authors proposed a pipeline that uses a deep neural network model to automatically extract visual features from images to estimate house prices in London, UK and found that learning to characterize the urban quality of a neighborhood improves house price prediction, even when generalizing to previously unseen London boroughs.
Journal ArticleDOI

Does environmental noise affect housing rental prices in developing countries? Evidence from Ecuador

TL;DR: In this paper, the effects of environmental noise on the housing rental price have been studied in developed countries; however, there are no similar studies in developing countries; the main reason for this is the lack of public information about environmental noise levels since the governments of turn have shown little interest in this type of data.
Journal ArticleDOI

Modelling Housing Rents Using Spatial Autoregressive Geographically Weighted Regression: A Case Study in Cracow, Poland

TL;DR: In-depth analysis of rent determinants using the GWR-SAR model exposed the complexity of the rental market in Cracow and revealed some determinants that are ubiquitous for almost the entire city.
References
More filters
Book

Geographically Weighted Regression: The Analysis of Spatially Varying Relationships

TL;DR: In this paper, the basic GWR model is extended to include local statistics and local models for spatial data, and a software for Geographically Weighting Regression is described. But this software is not suitable for the analysis of large scale data.
Book

Model-based Geostatistics

TL;DR: An overview of model-based geostatistics can be found in this paper, where a generalized linear model is proposed for estimating geometrical properties of geometrically constrained data.
Journal ArticleDOI

Hedonic housing prices and the demand for clean air

TL;DR: In this article, the authors investigated the methodological problems associated with the use of housing market data to measure the willingness to pay for clean air, using a hedonic housing price model and data for the Boston metropolitan area.
Journal ArticleDOI

Geographically Weighted Regression

TL;DR: In this article, a technique for exploring this phenomenon, geographically weighted regression, is introduced, and a related Monte Carlo significance test for spatial non-stationarity is also considered, using limiting long-term illness data from the 1991 UK census.
Related Papers (5)
Frequently Asked Questions (13)
Q1. What are the contributions in this paper?

As no explicit price exists for a unit of environmental quality, this paper utilizes the housing market to derive its implicit price and test whether these constituent elements of health and wellbeing are indeed capitalised into property prices and thus implicitly priced in the market place. This study presents a spatial analysis of air quality and noise pollution and their association with house prices, using 2,501 sale transactions for the period 2013. This research presents an original study utilising advanced spatial modelling approaches. The findings suggest that air quality pollutants have an adverse impact on house prices which fluctuates across the urban area. The analysis suggests that the noise level does matter, although this varies significantly over the urban setting and varies by source. This may suggest the presence of other externalities that arouse social aversion. The research has value in further understanding the market impact of environmental factors and in providing findings to support local air zone management strategies, noise abatement and management strategies and is of value to the wider urban planning and public health disciplines. 

In consolidating these findings future work will seek to utilise data which is envisaged to become available in terms of nuances in air quality data which may permit a difference in difference methodology to be adopted, to more robustly capture the change in pricing and air quality using a two-step hedonic framework. Future work will seek to address limitations implicit herein: Results are restricted to one urban area, therefore it would be beneficial to compare peri-urban areas to examine or reflect the changing density of housing market and urban form – extending the analysis. Also, a number of aspects of estimation were not taken into consideration and remain the subject of future work. The strong evidence of remaining heterogeneity and spatial correlation would suggest that perhaps a different scale of analysis might be more appropriate. 

low road noise has a detrimental impact on property value in the more built-up urban environment towards the CBD, with high road noise associated with increasing value. 

The Low air quality estimates show significant variation remaining negative until the 3rd quartile of the coefficient value, thereby inferring that low air quality (high levels of NO2) impact negatively upon prices, with the exception of well-established upmarket housing areas towards the South- South East of the city reflective of the utility trade-off between level of air pollution and desirable living locales. 

According to their estimates, a variation of 1 g/m3 of particulate matter causes an increase of 0.20 percentage points in the average value of houses. 

In terms of adjacency, properties located within 250 metres of arailway seemingly are negatively impacted in terms of their prices when scrutinising the linear OLSmodel, however only the distance bands up to 125m are significant at the 10% level. 

Turning to the air pollution variables of interest, examination of the NO2 coefficient (Model 3) reveals it to have a negative impact on property prices (β = -0.190, p<.001),inferring that the higher the NO2 level, the stronger the spatial autocorrelation of house prices, as indicated by the positive coefficient of the interaction term ( /01P* ∗ NO = 0.017). 

Early studies by Ball (1973) and Berry and Bedarz (1975) presented arguments for the importance of including spatial variables in valuation and house price models – concluding that a traditional ordinary least squares (OLS) model that treats all locations equally is flawed; error terms will likely fluctuate across submarkets, and will also be correlated with similar, nearby properties, therefore violating the assumption of a constant error variance (in residuals) which may occur due to structural instability of parameters across space, modelled functional forms that are not spatially representative, or missing variables (Anselin, 1988). 

This has once again led to calls for sharper policy responses andsolutions, with recommendations including the ‘phasing out’ of diesel vehicles, the creation of Ultra LowEmission Zones. 

The rail noise coefficient is negative illustrating it to comprise a statistically significanteffect on property prices (β = -0.908, p<.001), and indicating that the higher the noise level, the strongerthe spatial autocorrelation of house prices, as displayed by the positive coefficient of the interaction term/01a* ∗ j kl. 

As a result, heteroscedasticity, orspatial heterogeneity inherent in the property price data may also represent differences in the urbanenvironment which need to be modelled more reliably by the spatially varying GWR coefficients. 

The need for spatial consideration within hedonic pricing models has long been a concern within the valuation arena as both supply and demand of real estate will vary across a given location as tastes, preferences, willingness, and abilities to buy flucutate. 

This is also highlighted by Muller and Loomis (2008) who also caution that the gap between coefficients corrected and uncorrected for spatial dependence may not always be economically significant – inferring that the inefficiency attributable to spatial influences may not be large enough to cause critical errors in policy decisions.