PreprintPDF Available
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Economic theory and extant research suggest that flood prone properties should attract a discount. This concept can be extended to properties affected by future sea level rise but there is limited information for purchasers to judge and make informed decisions about their investment. Using a comprehensive dataset comprising statutory rating valuation information and sales transactions for the period 2011-2016, a hedonic framework is applied in order to ascertain the implications of the existing flood discount and potential price effects of future sea level rise. The hedonic model identifies a price discount effect for properties affected by known flooded areas, whilst sea level rise has no notable effect on valuations or sales data. The results highlight that purchasers do not appear to price sea level rise risk and are under-prepared for the future challenges and implications sea level rise and the ancillary effects of future flooding, inundation and storm surge.
Content may be subject to copyright.
Working Paper
Sea Level Rise and House Price Capitalisation
Franz Fuerst
University of Cambridge, Department of Land Economy, 19 Silver Street, Cambridge, CB3
9EP, United Kingdom, ff274@cam.ac.uk
Georgia Warren-Myers
The University of Melbourne,Faculty of Architecture, Building and Planning
This version: 25th November 2019
Abstract
This research examines both the existence of a current flood stigma and potential capitalisation
of future sea level rise into current valuations and transaction prices in the housing market.
Using a comprehensive dataset comprising statutory rating valuation information and sales
transactions for the period 2011 2016, a hedonic framework is applied in order to ascertain
the implications of the existing flood discount and potential price effects of future sea level rise.
The hedonic model identifies a price discount effect for properties affected by known flooded
areas, whilst sea level rise has no notable effect on valuations or sales data. The lack of
information for purchasers on sea level rise in the purchase process creates an information
problem which might explain the current lack of discounting or consideration of sea level rise
in pricing. The results highlight that purchasers are not considering or pricing sea level rise
risk and are under-prepared for the future challenges and implications sea level rise and the
ancillary effects of future flooding, inundation and storm surge.
JEL Classification: D82, G12. G14, Q54, R31
Keywords: Climate change, residential real estate markets, asset prices, sea level rise, flood
risk
1. Introduction
Many of the worlds’ major cities are situated on coastlines and rivers systems which house
40% of the global population (in 1990), and it is predicted that 2.4 billion people will populate
these areas in 2050 with 80% living in cities (Kummu et al., 2016). These coastal cities will
likely be threatened directly or indirectly by sea level rise due to climate change (Neumann et
al. 2015). Approximately 10% of the world’s population are situated in low-elevation coastal
zones below 10 metres in elevation (McGranahan et al., 2007). Predictions for sea level rise
are uncertain, as many forecasts rely on a complex set of variables and assumptions, many of
which relate to the location and geographical feature, contribute to the variations and
uncertainty of predicting sea level rise. Current approaches to mitigation are not meeting global
targets and it is questionable if they will be met in the future given the general orientation
towards continued economic growth, rising middle class consumerism and population growth.
As a result, sea level rise is not necessarily an uncertain event, more a known event that is
occurring presently, albeit slowly. It is likely to increase more rapidly in the future and the
uncertainty centres on the implications of long-run pricing of risk. The exposure of cities means
that property will be both directly and indirectly affected; and consideration of the effects on
value is required.
Current information pertaining to climate change and sea level rise is often inconsistent and
subject to political influence (Desser et al., 2012). Further, different levels of governments and
their varying opinions have affected the strategic direction of responses and variation in action
and inaction (Jericho, 2019; Feldman, 2019). Although they may perceive climate change and
sustainability as important, they are inherently restrained on several levels from taking action
(Robert, 2011). When considering property stakeholders, many are ignorant of the risks to
consider the implications of climate change and sea level rise risk for their properties, as there
is limited detailed information being provided by statutory bodies (Geiger and Swim, 2016).
As a result, particularly in countries with limited current exposure, little has been undertaken
in terms of creating short and long-term mitigation and adaption plans for both private and
public owners in relation to climate change related threats or sea level rise, and are delaying
occupancy decisions and how long they endeavour to persist with living or occupying or
owning property in areas identified at risk (Kahn, 2016; Pankratz, 2019; Lawrence, 2019;
Hurlimann et al., 2018; Hurlimann, 2019; Barrage and Furst, 2019).
Incorporating climate change risk into property markets is likely to follow two major pathways;
firstly, by identifying properties at risk of complete inundation, and secondly, by increasing
awareness of market participants brought about by severe flooding incidents. Emerging
evidence is identifying implications of sea level rise threat and protection measures and the
effects on property prices (Walsh et al. 2019; Bernstein, 2019). Examining a more established
area of investigation, a meta-analysis of flood effects on property price literature by Beltran et
al. (2018) found floodplain identification entails a significant discounting to housing values.
As a consequence, the provision of better information to purchasers of sea level risk and future
flooding risk should be more accurately incorporated into pricing (Chivers and Flores, 2002).
The other discounting factor to consider is flooding event severity, reoccurrence and frequency,
which is now becoming a greater concern with reoccurring events, for example in the southern
United States (Ortega and Taspinar, 2018).
Both information provision and the increasing awareness of risk and actual events is likely to
propel the need for more investigation to assess risk and generate risk minimisation strategies
and adaption measures to minimise the future impact of flooding. In time, understanding of
risk, risk mitigation strategies and adoption approaches may influence investment and
occupation decisions and affect market values, insurable values, and the ability to obtain
security. However, as reported in Ortega and Taspinar (2018), recurrent events of inundation
and frequency will likely be the strongest drivers of discounting in the future. The presence of
sea level rise stands may be realised initially through increased flood events and severity of
those events.
This study sets out to demonstrate the relative implication of the lack of information provision
pertaining to risk of sea level rise and flooding for a case study area. We examine the discounts
associated with floodplain identified areas compared to unaffected areas and test to understand
whether expected sea level rise inundation is being factored into decision-making in property
transactions as evidenced by market prices. While floodplain information is readily available,
expected areas of future sea level rise are far more difficult to identify for stakeholders in the
real estate market.
This study uses a unique combination of GIS databases; planning and flood information; rating
authority valuation data; and residential sales data to investigate the consideration of sea level
rise and flood discounting in current value estimates and market pricing for housing. Further
we examine the assessments of statutory valuers as knowledgeable actors in the market and the
market perception of discounts associated with flood prone properties or stigma of known
areas. Firstly, we ascertain whether mandatory provision of risk information, or lack thereof,
has an effect on capital values (determined by statutory authorities) and sales prices
(determined by market participants, purchasers). Secondly, we examine whether properties
identified as being at risk of flooding are discounted compared to non-flood risk identified
properties. Finally and crucially, we identify if future sea level rise risk has an effect on
properties prices and values.
This research contributes to economic studies of sea level rise by considering the pricing
mechanisms behind current and future flood risk exposure. To this aim, it uses a unique and
very detailed database assembled from several sources and comprising one of the most
complete sets of property characteristics and flood risk identification employed in any study of
this subject. It also contributes to public policy formulation by demonstrating that purchasers
are only able to make well-informed decisions when information on potential hazards to the
property is readily available and sufficiently detailed.
2. Background
Housing globally is likely to be significantly affected by inundation from sea level rise and
increased periodic flooding from storm events and storm surges (Berman, 2019). Sea level rise
risk is substantially underestimated in the effects it will have on property. This is due to
modelling expectations focusing only on direct sea level rise, with a general assumption of flat
water and little consideration of wave setup, wave heights, tides, storms and storm surges let
alone the type of land affected and erosion considerations (Liu, 1997: Yohe et al, 1999; West
et al. 2001; Nicholls, 2002; Warren-Myers et al. 2018). As an example, Warren-Myers et al.
(2018a) found for a bayside municipality in Australia, that at a 0.8 metre sea level rise, only
0.24% of properties would be affected. However, given a conservative storm surge (of 0.5
metres), existing flood levels and high tide consideration this increased to approximately 40%
of the municipality being affected by potential flooding during a storm event. Further, a
modelled storm surge for the case study area investigated, using a storm surge of 1.5 metres,
would lead to close to 50% of the properties in the municipality affected (Climate Code Red,
2012; Warren-Myers et al. 2018a). Similar findings by Michael (2007) who determined in an
examination of Chesapeake Bay (United States) the damage losses from episodic flooding was
going to be nine times the loss from complete inundation at 3-foot sea level rise. Whilst
Bernstein et al. (2019) identified homes at risk of sea level rise were sold for 7% less than those
of similar distance to the beach but were not identified as being affected by sea level rise.
Further their study found that this discount has increased over time as public awareness grows.
Housing plays an important role in economic markets and the contraction or loss of values
within markets will have intense implications for individuals, businesses and broader financial
markets. As the risks and actual losses from sea level rise become more evident, and markets
price the flood risk, future changes will result in those that are least able to afford the risks will
be the most exposed, increasing issues of social justice (Pryce and Chen, 2011).
This leads to purchasers awareness and knowledge of the risks posed to the property, as a
purchaser will factor and consider the risks and probability of detrimental exposure of the
property to potential future losses or costs in the offering of a price for a property (McDonald,
1987; McClelland, 1990). This assumes rational decision-making and that humans, consider
the risks and probability of losses in determining the price they are willing to pay. Disaster
risks’ influence on property values is highly dependent on the information provided to the
purchaser, their awareness and subsequent due diligence. Consequently, decision-making for
housing choice is dependent on a cluster of factors that drive decisions, in regard to disclosed
disaster risk information; event frequency, impact and history; and decision-making factors
status quo bias, herd mentality and heuristics including anchoring, availability, and
representative heuristics.
Modelling sea level rise risk for property
Sea level rise risk research has focused on modelled loss of property and values or costs as a
result of inundation commonly based on LiDAR or height elevation studies (Titus, 1991;
Darwin et al., 2001 Bin et al. 2011; Scott et al., 2012; Fu et al. 2016; Warren-Myers et al.
2018). Further work also considered increased flooding in sea level rise scenarios (Michael,
2007; Pryce and Chen, 2011; Warren-Myers et al., 2018). In the economics and finance
literature, some research has focused on the house price implications as a result of sea level
rise protection measures or risk reduction (Hamilton, 2007; Walsh et al., 2019) as well as
investors’ and developers’ provisions for effectively seeking higher ground (Bunten and Kahn,
2017). More recently, Ortega and Taspinar (2018) demonstrated the increasing discounting
associated with properties in flood prone areas and the effect major events like hurricane Sandy
has on property values. Unlike the flood literature, the sea level rise scenario modelling
research estimates a ‘what if’ scenario, lacks actual events, and is highly implicated by
uncertainty and a dearth of information provided. However, evidence of flooding discounts for
flood prone or designated properties and research like Ortega and Taspinar (2018), suggest that
there is an underestimation of the price implications and subsequent risk to coastal properties
in relation to sea level rise and future flooding implications.
Action in regard to households taking mitigating or adaptive approaches to ownership and
investment in relation to sea level rise, is relatively unknown. Where extant decision-making
heuristics in relation to sea level rise are not triggered, perhaps because of: the ongoing
uncertainty debate of climate change, global warming and sea level rise (anchoring and
adjusting and representative heuristics); the lack of information provision of sea level rise risk
and limited events in which people are affected (availability heuristic), and a response to
maintain the status quo in buying habits (particularly in strong markets) and herd mentality
continuing of purchasing land at a premium in or near coastal area.
Property pricing models for inundation events and risk
Flooding studies have focused on these aspects steadily overtime, and a recent meta-analysis
by Beltran et al. (2018) specifically focused on flood plain discounts in house prices, where
information is provided about the ‘risk’ (a designated flood plain) and then further analysed in
the context of recent ‘events’. Consequently, the study demonstrated various heuristics, noted
above at play, and its implication in the conclusion of a 4.6% discount across the studies. For
example, knowledge of information (the flood plain designation) provides information on
which to anchor and adjust to; and the representativeness and availability heuristics indicated
by the event analysis and how this then affected values depended on the timing of events,
frequency of events and time since event. As demonstrated in Beltran’s analysis, various
aspects of the different heuristics have been tested to examine the implications on decision-
making in housing purchases and its relationship to flooding risk.
Botzen et al.’s (2008b) study of flood perception, willingness to pay and engagement in
strategies found evidence suggesting that consumer behaviour, perceptions and attitudes play
a major role in changes to properties’ market value. As does the importance of flood
information and disclosure as modelled by Votsis and Perrels (2016) indicating a rational
response to flood risk information. This is further supported by Beltran et al.’s (2018) meta-
analysis, examining floodplain exposed properties and event based research which provided a
thorough analysis of the different approaches to flood modelling and its implications for
property prices, being either value add demonstrated through the economic benefits of flood
protection either natural or man-made and willingness to pay for protection; and the value
discounting where reduction in the estimate of benefits for properties at risk of flood which is
then capitalised into prices. Many of these studies highlight the effect of time dependency and
frequency of event affects purchasers’ decision-making significantly. For example, in that a
recent disaster event might trigger a greater response initially but over time the market
demonstrates a level of recovery (Eves, 2002; Loomis, 2004), or where repeated events lead to
systemic long-term discounting (Eves, 2004a; Mueller, 2009). Suggesting, that reactions are
enhanced and pricing becomes more clearly factored in the long-term when events are
occurring regularly and of significant effect. When events occur more often particularly when
there is media attention, public awareness is heightened and this has longer term impacts on
value, conversely infrequency of events decreases awareness (Ortega and Taspinar 2018;
Beltran et al. 2018; Eves, 1999, 2001, 2002; Wilkinson and Eves, 2014).
The perception of risk for an asset is almost as significant as the actual risk to property market
values. In the UK only 50% of residents living in flood prone areas are actually aware of this
fact (Eves, 2004), as often purchasers are unaware their property is situated in a flood zone or
what the risks and probability of flood losses and costs are (Chivers and Flores, 2002). As an
informed purchaser will make decisions based on both actual and perceived risks, which do
change over time, as demonstrated Ortega and Taspinar (2018); who found a gradual and
persistent negative impact on houses identified in a flood zone. This was a result of existing
knowledge of flood zones with the added emphasis coming through a significant and
substantial flood event, demonstrating the influence of a significant event on existing
information, reinforcing and then creating a long-term effect on property values. This is further
supported by Cameron and Shah (2015) psychological experiment that found those exposed to
a recent flood event were more risk adverse than those not affected, and there was a greater
expectation that they may be affected again. However, limited evidence of discounting, may
be a result of expected future losses being underestimated, or not considered, as short-term
benefits and costs outweigh the long-term possibility of a flood (or sea level rise) and the
associated losses and costs that may never occur (Koning et al. 2017).
Two key factors of decision-making and price effects emerge from the literature. Firstly,
information and secondly event experience. In relation to sea level rise, the actual event
occurrence may still be some way off, so information plays a significant role in highlighting
risk identification. The question is what information is provided, how is it provided and how
informed are the decision-makers of the risks? The risk lies in the information about the
exposure of the property and how knowledgeable the purchasers are of the risks perceived or
actual. As found by Chivers and Flores (2002) the information asymmetry led to a market
failure, where rational purchasers did not have sufficient information to make well-informed
purchase decisions. Consequently, when considering the implications of sea level rise and
escalation of flooding events locally, nationally and globally, the short-term and long-term
effect on market values will be reliant upon information disclosure, processing of this
information and in time, and in time response to events.
This research investigates the relative implication of information provision on pricing for a
case study area, in Melbourne, Australia. At present, in Australia there are general mapping
applications and webpages with sea level rise information, but nothing is directly provided to
purchasers of properties relating to sea level rise. Yet, when purchasing a property in Victoria
purchasers are informed of flooding and inundation potential through planning maps contained
in a Section 32
1
. The research examines the discounting effect on sales prices and statutory
property values for floodplain identified properties and predicted sea level rise affected
properties. The floodplains are identified through planning maps, which are provided in the
Section 32. If a property is directly affected or in close proximity to a floodplain this
information is identified in the Section 32 by way of a map plan and commentary. At present,
there is no sea level rise inundation information or suggestion of, in the Section 32 or
information available directly from the local council. Consequently, this research aims to
highlight the implications of information disclosure and possible long-term implications this
1
Under Victorian legislation Sale of Land Act 1962 a Section 32 is a legal document provided by the vendor
and is intended to provide the purchaser with key information about the property that may have an effect on
their decision to sign the Contract of Sale. The Section 32 is separate from the Contract of Sale but is usually
handled in conjunction; and is usually created by a lawyer. If information in the Section 32 is incorrect or
missing; the purchaser has the right to withdraw from the sale and legal action can be taken.
may have for property values. This will have important consideration and implications for
current planning policies and what should be disclosed within a Section 32 Contract of Sale
and suggest local councils need to consider the inclusion of sea level rise risk information to
their constituents.
3. Research Strategy
This research uses a two-stage analysis: firstly, we use hedonic modelling to estimate if current
flood designated areas and areas within the 2100 sea level rise estimations are discounted in
local authority rating valuations. In the second stage, residential property transactions between
2011 and 2016 are analysed to investigate whether properties at risk of sea level rise trade at a
statistically significant discount. Three hypotheses will be tested in this context:
Hypothesis 1: Floodplain designated properties will be discounted compared to non-
floodplain identified properties
Hypothesis 2: Sea level rise risk identified properties will be discounted compared to
non-sea level rise risk identified properties
Hypothesis 3: Statutory valuers discount properties values in designated flood zones;
as educated market observers and trained professionals they are aware of the broader
disadvantages of flood risk and incorporate it adequately in their property appraisals.
The implicit price of housing characteristics is identified from the bundle of characteristics
associated with the value purchasers place on a dwelling and its integrated neighbourhood
features (Rosen, 1974). Any economic loss associated with flood identification and/or
anticipation of sea level rise inundation can be identified with a hedonic model as used by Fu
et al. (2016), Bin et al. (2008, 2011), Rambaldi et al. (2013), Hamilton , (2007) Michel (2007).
The general model of flood risk can be formulated as
 

-λσF
Pit is the transaction price or appraised value of a property, and Xn comprises a vector of hedonic
characteristics as specified in Table 2. The monetary value of partial or total damage to the
property in the event of flooding, denoted , is then subtracted from this hypothetical property
value after accounting for the likelihood of damage (taking a value between 0 and 1), the
perception of risk by the buyer based on information known that the time and representing
the discount rate by which monetised future damage is discounted to present value. Since the
values of these parameters cannot be retrieved directly, the following hedonic model
specification is used for the empirical estimation:
(1)
 
 S
In this specification, Li and Si are binary variables indicating if the property is located in a
designated current flood area or if it is located in an area likely to be affected by future sea
level rise. For the control sample of properties affected by neither of these designations, these
variables take the value of zero. To test our Hypotheses 1 and 2, we expect negative and
significant values of βL and βS as supportive of the respective hypothesis. Hypothesis 3 is then
tested by comparing the coefficients of interest βL and βS across the estimations for valuation .
There are several challenges to examining the implications of floodplain identification where
the hedonic price effect might be the error term or unobservable factors may correlate due to
the endogeneity and omitted variable bias (OMV). In order to overcome these issues, we have
taken a number of steps to ensure better matching and reducing the probability of OMV by
limiting the area to the case study area, focusing on particular residential property types, and
including a large number of relevant control variables, particularly those that are suspected to
be correlated with a flood zone location, coastal or canal/river frontage. Bin et al., (2008),
Hallstrom & Smith (2005), Bin and Polasky (2004) and Shultz and Fridgen, (2001), all report
a negative impact of flood zone location on the price of a dwelling. Further, the collection
statutory valuation data has provided additional characteristics about dwellings not generally
available, like date built, renovated, view quality and any other improvements. The statutory
valuation data information for each property has been matched to individual sales; in which
the sales information and commentary have been further data mined to provide further housing
characteristics. This provides a comprehensive dataset with an extensive list of features for
each property.
As with many flood-based analyses, the disentangling of the positive benefits of being located
near to the ocean, sea or stretch of water and the negative association with potential flooding
is required. Bin et al. (2008), emphasise the need to control for positive and negative pricing
factors of locations adjacent to water. To disentangle these factors for this study, a number of
controls are put in place. Firstly, all properties that are identified as having potential for amenity
associated with the bay waterfront are identified with a binary variable; this is extended to
properties with canal frontage as it is perceived as an amenity in the municipality. Secondly,
further controls for the identification of waterfront and distance from the waterfront are
included. Thirdly, as suggested by Daniel et al. (2009b), it is important to provide controls for
views/vistas. This information is contained in the valuation dataset used in this analysis and is
dummy-coded to control bay views, city or lake views separately and is also graded on quality.
Finally, a continuous elevation variable was used as, for example, suggested by Ragjapaksa et
al. (2016).
(2)
4. Data and Study Area
The data collected for this study comes from a number of sources providing vital and detailed
information at different levels that endeavour to minimise the limitations observed in other
studies. In particular, thoroughness in identifying as many descriptive characteristics of the
properties, their spatial connectedness, socio-economic, flood disclosure information, flood
events and sea level rise information. Table 1 specifies the data type and sources in this study.
Table 1. Data sources
Element
Values
Statutory valuation
data
Capital improved value, site value,
property type, bedrooms, dwelling size,
land size, improvements and quality
measures
Sales price data
Property prices, property type, age,
bedrooms, bathrooms, garage, and land
size, date of transaction
Floodplain
identification
Planning maps and Melbourne Water
mapping (general height considered at
flood risk 1.6m AHD)
Sea Level Rise
information
0.2 metres, 0.5 metres, 0.8 metres,
1 metres, 1.1 metres
Flood levels
1.6 metres
Highest
astronomical tide for
Port Phillip Bay
0.5 metres
Storm surge
0.5 metres and 1.0 metres
0.9 metres
Case study area information and property data collection
The data collected from the case study municipality comprised information held as part of the
statutory rating valuation procedure. The case study area is situated close to the central business
district in Melbourne and abuts the bay. The municipality had a population in 2015 of 107,142
people (Australian Bureau of Statistics, 2017). The profile of this community according to
SEIFA information (which is an index used to reflect disadvantage and consider low income,
low employment, level of unskilled occupants and limited educational attainment); the
municipality is listed at 1069 on the SEIFA index scale in 2016, which by comparison rates as
one of the least disadvantaged areas in Melbourne (ID Community, 2018a). Which is further
demonstrated in the comparison of the median sales prices in the municipality and that of
greater Melbourne, shown in Figure 1.
Source: ID Community (2018b) City of Port Phillip Housing Prices
Figure 1. Median sales for municipality and Greater Melbourne
The information about the dwellings within the case study area were obtained through the
Valuer General Victoria (with approval of the local municipal council). The data provided
comprised the statutory valuation information used to assess the statutory values for the
municipality. The level of detail available goes well beyond many of the other studies that have
examined flood prone property. In particular, the data pertaining to the dwellings comprises
the usual bedrooms and key features, but also dwelling age and renovation year, a quality of
style code, quality of condition code, size of the dwelling, land size, views and additional
improvements to the dwelling or land. Thus, providing a comprehensive dataset of housing
characteristics for analysis. In addition, this dataset also comprised the assessed values for the
property on a site value, capital improved and net annual value basis for the 2016 rating year.
The second property database we utilised is transaction prices for the municipality for January
2011 to December 2016, purchased from Australian Property Monitor. The sales prices were
matched to the property information from the statutory valuation set, so to provide a second
dataset with the same dwelling characteristics. Further data mining of the marketing
information allowed for further features of the property to be gathered and used in the analysis.
Geolocational attributes for the data
The dwelling data set was then geolocated in order to undertake further analysis and layering
of additional data for each dwelling. The information and datasets used comprised the planning
maps from local government, geospatial data, and GeoScience LiDAR elevation data. These
datasets were connected to identify whether the property was situated within a floodplain area;
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2011 2012 2013 2014 2015 2016 2017
Median House Price Municipality Median House PriceGreater Melbourne
Median Unit Price Municaplity Median Unit Price Greater Melbourne
the elevation of each property and spatial calculations identifying distance to the
coast/beachfront; the lake and parklands, schools, train stations and public transport stops and
key retail destination areas within the municipality. This process was performed in RStudio
and ArcGIS and calculated using Euclidean distance and shown in metres.
Flood risk identification
The risk of flooding has been captured through the statutory planning overlays, which specifies
location of probable flooding in a 1:100 year storm event. These properties situated within a
flood plain were identified using a binary tag, and used as a dummy variable. In addition, for
those properties not in a flood plain area, but situated close to, distances were mapped using
GIS and Euclidean distance measures from the closest flood plain.
The planning overlays that indicate a flood plain or flood risk are determined by the local water
authority (Melbourne Water) in conjunction with the local government authority based on
1:100 year flood plain information and recent flood events. Although these have not been
updated since the early 2000’s, and are currently being reviewed and updated they represent
for this study the most recent publicly available information relating to flood plain and flood
risk identification. The planning scheme maps and descriptions are included in the Section 32
(Sale of Land Act 1962, Victoria), which communicate with potential purchasers the salient
statutory legal parameters, including the zoning of the property and any overlays which also
affect the property. The overlay which identifies potential flood risk and provides potential
purchasers with the information. An overlay is also highlighted in the Section 32 if the property
is not affected directly by the overlay but lies in proximity to the flood overlay.
Sea level rise risk identification
This study utilises the data and information generated in the paper by Warren-Myers et al.
(2018) which provides a full and detailed discussion of the modelling approach to ascertain sea
level rise; future flood risk; storm surges and high tides. A brief overview of the key approach
is outlined to provide a narrative of the approach used to estimate sea level rise risk for
properties.
Firstly, LiDAR data was acquired from GeoScience Australia (2015a) to ascertain the
topography for the area; utilising Geographic Information Systems and RStudio this digital
elevation model was combined with property parcel maps. Next, the well-known and used
‘bathtub’ approach for modelling sea level rise was applied (Geoscience Australia 2015a,
2015b; McInnes et al. 2015; and Hauer et al., 2016). The digital elevation model created then
creates a cut of the topography at a particular level (for the purposes of this study, the cut was
made for sea level rise predictions of 0.2 metres; 0.5 metres; 0.8 metres; 1 metre and 1.1
metres), this digital elevation model then assumes the land below is subject to sea level rise;
whilst that above is not. This model does not take into account soil type, erosion, barriers, sea
walls or other factors; it is a blunt approach assuming all land is the same, just affected at
different heights. The objective of Warren-Myers et al. (2018) was to identify the relevant
hazard zones for sea level rise prone land. Utilising the same approach, additional measures
were added to incorporate and consider the effects of future flooding due to changes in the
extant flood levels (considered by Melbourne Water, 2017 to be 1.6 metres above); storm
surges (McInnes et al, 2009; 2011; Tonkin and Taylor; 2014) and Ministry for the
Environment, 2008) and astronomical high tides in Port Phillip bay (GeoScience Australia,
2015b). This meant that properties identified as under the 0.8 metre sea level rise were
completely submerged; however, the inclusion of additional factors like storm surge and
existing flood levels suggested substantially more properties faced future flooding likelihood.
As indicated in Warren-Myers et al (2018) the real risk of sea level rise is underestimated due
to the factors of storm surges, high tides and storm events not being considered in sea level rise
models; consequently, this research utilises these additional risk factors as a future flooding
event.
A limitation of this approach is again, the concept that the land is equal, and doesn’t take into
account drainage patterns; land and soil types; erosion and any barriers like sea walls; thus the
extent of damage is assumed to be the same and to affect the whole property, which may not
necessarily always be the case, but without further in-depth modelling these are the limitations
of the modelling. This ignores additional factors like wave set-up and how different spatial or
infrastructure type barriers or accessways could accelerate or reduce the speed of water egress.
In order to identify and compensate for the model’s blunt approach, floodplain mapping
developed by Melbourne Water and the local authority were then cross-referenced and overlaid
to identify further flood risk potential within the area. For more detail on the model
development and iterations of the model please see Warren-Myers et al. (2018).
In summary, land prone to possible sea level rise and future flooding has been identified as a
binary variable at various heights and are dummy variables. The key modelled scenario focuses
on a 1 metre sea level rise measure (IPCC, 2019; Warren-Myers, 2018). In addition, a coniuous
variable for height (according to the Australian Height Datum) has been utilised in the
modelling. Whilst existing flood risk areas denoted in the floodplain overlays have been also
treated as a binary dummy variable for the purposes of the analysis. Primarily, the identification
of properties likely at risk in 2100 from sea level rise and future flooding have been adopted
from Warren-Myers et al. (2018).
Summary Statistics
Summary information for the dataset is shown in Table 2, comprising the information from the
valuation dataset and sold properties in the municipality. The tables reports the means, with
extended tables with mean and standard deviations of the salient hedonic characteristics used
in the appendix 1 for the valuation data set and appendix 2 for the sales data set. Dummy
variables are applied for risk associated with flooding or sea level rise. In addition, other control
variables include whether the properties are located on the beachfront, and distance to the
beach, parks, full-service grocery store and retail areas, schools, healthcare and transport.
Table 2. Summary statistics for the valuation data
Total sample
Houses
Apartments and Units
means
Values
Sales
Values
Sales
Values
Sales
Sales Price
$1,010,709
$905,899
$1,543,536
$1,318,038
$575,528
$625,588
Construction Year
1950
1945
1916
1923
1979
1970
Building area (m2)
112
148
149
149
82
82
Building Condition (scale 1 - 10)
5.39
5.64
5.66
5.96
5.18
5.28
Quality (scale 1 - 5)
3.07
3.10
3.05
3.09
3.10
3.11
Bedroom (number)
2.26
1.41
2.84
1.57
1.78
1.30
Bathrooms2
2.22
2.82
1.81
Car space (1=yes)
0.78
1.07
0.62
1.09
0.93
1.06
Sea level rise risk (1=yes)
0.01
0.01
0.00
0.01
0.01
0.01
Sea level rise & flood risk (1=yes)
0.28
0.25
0.42
0.34
0.17
0.20
Flood overlay (1=yes)
0.14
0.18
0.16
0.23
0.13
0.15
Beach front (1=yes)
0.05
0.07
0.02
0.04
0.09
0.10
Canal front (1=yes)
0.01
0.01
0.01
0.01
0.01
0.01
Site is 400 metres from park (1=yes)
0.83
0.87
0.85
0.86
0.81
0.87
Distance to beach (metres)
1064.71
935.66
956.15
904.47
1148.30
956.87
Distance to grocery and retail (metres)
571.55
540.44
575.19
582.66
571.69
511.78
Distance to nearest childcare facility (metres)
475.53
505.34
475.52
487.80
475.72
517.26
Distance to nearest pool or public recreation
facility (metres)
1648.75
1689.47
1451.58
1630.94
1818.21
1729.21
Distance to nearest healthcare facility (metres)
668.58
668.76
652.89
633.61
683.79
692.62
Distance to nearest public Secondary School
(metres)
1809.27
1703.32
1630.09
1587.80
1959.82
1781.76
Distance to nearest public Primary School
(metres)
949.24
979.44
847.38
886.61
1033.74
1042.47
Distance to nearest rail station (metres)
6088.90
2169.99
4411.45
2347.55
6752.34
2050.01
Distance to nearest tram (metres)
425.04
445.14
442.68
473.27
411.61
426.05
Distance to nearest pharmacy (metres)
572.69
543.36
642.93
619.83
516.22
491.44
N
37008
3950
16639
1599
19535
2351
5. Results
To test for price effects of current flood zone and future SLR exposure, we regress the
logarithm of the 2016 Capital Values as appraised by the municipal valuers against a
comprehensive set of property and neighbourhood characteristics as detailed in the previous
section. The analysis presented in Table 3 is based on a representative sample of 37,008
residential properties in the municipality. The naïve or baseline model only uses flood zone
and SLR exposure and explains only about 8.4% of the variation in the natural logarithm of
home prices. The large positive coefficients obtained in Model 1 thus demonstrate that a
superficial analysis might come to the conclusion that properties under threat of current or
future flooding command a premium rather than a discount. However, in an urban setting such
as the one we study, the true effects might be masked by differences in location and/or property
features which necessitates the use of a more complete hedonic model. Indeed, Models 2-4
show vastly improved predictive power, with R2 values of up to 94% when views, beachfront
2
Valuation dataset does not have information on the number of bathrooms
location and neighbourhood amenities are incorporated into the model along with fixed effects
for location, construction material and quality of building.
As expected, larger properties and a higher the number of bedrooms command higher capital
values while year of construction is negative, i.e. older properties command a premium over
more recently built properties. The mean age of houses in the municipality is 1915, which is
reflective of the historic and heritage nature of the area; and apartments and units have a mean
age of 1978. The observed premia might be due to the local value of heritage properties in the
area, and the premium placed upon the older housing stock more generally, if in good condition
(Field, 2019). It is possible that these older properties carry larger prestige among home buyers
as they are associated with a number of high-quality design features that the more recent homes
lack. Also, this ‘age premium’ is found to be stronger in houses than it is in apartment buildings.
Model 2 excludes neighbourhood characteristics as well as canal and beachfront identification
identifies a significant and negative effect on capital values if the property is identified as being
in a flood risk area. However, there is no apparent discount or negative effect in relation to
future sea level rise. The extended models take into account attractive views and whether the
property is beachfront (3). Our full model (4) incorporates all variables and controls for
beachfront, canal front and proximity to amenities including the beach. Again, current flood
zone overlay is associated with a statistically significant discount. It is interesting to note that,
once the attractive features of proximity to water are appropriately controlled for, the discount
on current flood zone properties nearly doubles. Since the full model effectively disentangles
positive and negative effects of proximity to water, a major limitation of previous studies on
this topic, it is likely that the coefficients obtained in Model 4 are close to the actual effects
after controlling for confounders. Exposure to future SLR is insignificant in the full model
which suggests that valuers might not take into account this factor, either explicitly or
implicitly, when conducting property appraisals.
Table 3. Regression results for Capital Values
Value (ln)
(1)
(2)
(3)
(4)
Naive
Full model (no
beach or canal front
or neighbourhood
characteristics)
Full model (no
beach or canal
front
Full model
Under water in SLR scenario
0.434***
0.00794**
0.000854
0.000203
Current flood zone overlay
0.143***
-0.00668*
-0.0101***
-0.0110***
Year of construction (ln)
No
-2.273***
-2.370***
-2.381***
Building area sqm (ln)
No
0.611***
0.607***
0.606***
Number of bedrooms (ln)
No
0.133***
0.135***
0.136***
Land area sqm (ln)
No
0.0583***
0.0605***
0.0598***
Beachfront
No
No
No
0.0285***
Canal front
No
No
No
0.0253**
Dwelling type
No
Yes
Yes
Yes
Construction materials
No
Yes
Yes
Yes
Quality of Style
No
Yes
Yes
Yes
Condition rating
No
Yes
Yes
Yes
Building Features
No
Yes
Yes
Yes
Neighbourhood
No
No
Yes
Yes
Location (suburb)
No
Yes
Yes
Yes
Views
No
No
Yes
Yes
adj. R2
0.084
0.938
0.94
0.94
N
37,008
36,330
36,330
36,330
* p < 0.05, ** p < 0.01, *** p < 0.001
In the next step, the hedonic model is applied to empirical evidence from sales transactions as
shown in Table 4. This estimation, although similar to the valuation sample in terms of
geographic area and time frame, is based on a considerably smaller sample (n=5,206), which
is to be expected as only a fraction of properties that are regularly assessed by a valuer were
traded during the study period. However, the model still explains about 66% of the variation
in the natural log of sales prices in the full model (4). As the sales data is collected over a period
of time, the timing of sale is controlled for in the regression analysis. The results reflect
comparable direction of effect and significance to the valuation sample; however, construction
year has a stronger negative effect compared to the capital value analysis. .
In a similar vein to the capital value analysis, the sea level rise risk shows a positive and
significant effect on sales price in the naïve model (1), yet drops in significance in the full
models. This suggests a lack of awareness of risk and potentially other factors driving the
positive result. This is confirmed by Models 2-4 which successively incorporate
neighbourhood characteristics and then also distance from the beach and other factors.
The lack of a significant premium in the sales model implies that residential buyers during the
time period considered for this analysis did not significantly discount their bids in response to
the threat of current or future sea level rise in the area while the statutory valuers appear to
price the risk of a current flood zone overlay, even though both buyers and valuers appear to
disregard future SLR exposure in the area. purchasers are pricing in the risk of flooding more
so than the statutory valuers. However, when neighbourhood characteristics, in particular
beachfront and canal frontage area controlled for the significance is lost in the sales although
still demonstrating stronger effects and the sign of the coefficient is in line with other studies
considering flood risk perspectives.
Table 4. Sales price regressions
Sales Price (ln)
(1)
(2)
(3)
(4)
Naive
Full model (no beach
or canal front or
neighbourhood
characteristics)
Full model (no beach
or canal front)
Full model
Under water in SLR scenario
0.360***
0.0408
0.0408
0.0391
Current flood zone overlay
0.168***
-0.0213
-0.0213
-0.0222
Year of construction (ln)
No
-4.410***
-4.410***
-4.458***
Building area sqm (ln)
No
0.132***
0.132***
0.133***
Number of bedrooms (ln)
No
0.514***
0.514***
0.513***
Land area sqm (ln)
No
0.00981
0.00981
0.00872
Beachfront
No
No
No
0.0247
Canal front
No
No
No
0.109
Dwelling type
No
Yes
Yes
Yes
Construction materials
No
Yes
Yes
Yes
Quality of Style
No
Yes
Yes
Yes
Condition rating
No
Yes
Yes
Yes
Building Features
No
Yes
Yes
Yes
Location (suburb)
No
Yes
Yes
Yes
Neighbourhood
No
No
Yes
Yes
Views
No
Yes
Yes
Yes
Year of sale
No
Yes
Yes
Yes
adj. R2
0.063
0.662
0.662
0.662
N
5206
3001
3001
3001
* p < 0.05, ** p < 0.01, *** p < 0.001
Robustness checks
Although our estimations of both appraised values and actual transaction prices utilise a very
extensive set of building and neighbourhood features, we cannot rule out with absolute
certainty that there may still be unobserved features that bias the coefficients of interest.
Consequently, we perform additional robustness checks to minimise the risk of omitted
variable bias in our main results presented above. Our first robustness check is to collect and
analyse additional data on land values as land should not be affected by any unobserved
features that may be inherent in properties. The land values for the sales dataset are derived
from estimating the value of the capital improvements as per the statutory information and by
extrapolating an implicit land price, referred to hereafter as ‘land price’. The valuation dataset
already had listed site values, which are referred to as ‘land value’ hereafter. Table 6 shows a
good model fit with 81% explained in the variation of the natural logarithm of the land values;
whilst 83% is explained through the variation of the natural logarithm of the implicit land
prices. Both sets control for land size and large residential development sites are not considered
as part of the analysis. Similarly, land for apartment blocks is excluded to ensure a like-for-like
comparison.
Both Models 1 and 2 in Table 5 find a significant and negative relationship for land prices and
values that are situated within a flood risk overlay. This supports the earlier models that overall
also found a similar negative relationship, yet the relationship is significant for both land values
and land prices and by a larger amount. By concentrating on the land alone, we can establish
that the discount factor for flood risk properties is established in the underpinning site values
and land prices even when location and neighbourhood characteristics are considered and
controlled for. Again, there was no indication of a negative relationship with sea level rise risk.
Beachfront still commanded a significant premium in both models, of almost comparable
levels, and the distance to the beach was again negative and significant, indicating a desire to
live in proximity to the beach.
Table 5. Regression of land value and land price
(1)
(2)
Land value (ln)
Land Price (ln)
Under water in SLR scenario
-0.0014
0.00917
Current flood zone overlay
-0.0355***
-0.0508***
Land Area
0.657***
0.628***
Beachfront
0.304***
0.329***
Distance to beach
-0.000106***
-0.000112***
Neighbourhood characteristics
Yes
Yes
Year of sale
Yes
Yes
Land use
Yes
Yes
Location (suburb)
Yes
Yes
Land area<1,000m2
Yes
Yes
adj. R2
0.811
0.823
N
16,433
1,475
Our next robustness check explores if properties that are subject to a current or future threat of
flooding are generally less well maintained than those that are unencumbered by this risk. As
discussed, lack of information on the state of repair and maintenance of a property has plagued
many previous studies on the price impact of flooding and similar disamenities. Although this
study controls for this factor, closer scrutiny in a separate estimation seems warranted. Table 6
presents the results of an ordered logistic regression with the building condition rating recorded
in the valuation of the property serving as the dependent variable regressed on the vector of
hedonic characteristics as previously. The results of the full model (3) underline the importance
of controlling for building condition as it appears that properties in zones liable to flooding are
generally in significantly worse state of repair and maintenance. If confirmed for other studies,
this could mean that some of the previously found large discounts for potential flood exposure
are inflated by the additional negative factor of systematically inferior condition of the
properties in at-risk areas.
Table 6. Estimation of building condition (to find out if flood exposure is associated with inferior building condition)
Ordered logistic regression
(1)
(2)
(3)
Naïve model
Full model (no beach or canal front)
Full model
Under water in SLR scenario
-0.0939***
0.186***
-0.170***
Current flood zone overlay
0.360***
0.0986**
-0.195***
Year of construction (ln)
0
-29.22***
-16.92***
Building area sqm (ln)
0
3.430***
0.834***
Number of bedrooms (ln)
0
-0.594***
-0.500***
Land area sqm (ln)
0
-1.402***
-0.190***
Beachfront
0
-0.000983
0.456***
Canalfront
0
0.14
-0.362*
Dwelling type
No
Yes
Yes
Construction materials
No
Yes
Yes
Quality of Style
No
Yes
Yes
Building Features
No
Yes
Yes
Location (suburb)
No
Yes
Yes
Views
No
Yes
Yes
pseudo R2
0.03
0.08
0.08
N
36918
16400
19251
* p < 0.05, ** p < 0.01, *** p < 0.001
A further robustness check varies that level of sea level rise to see if a significant effect can be
detected if a moderate sea level rise of only 1 metre is assumed. We re-estimate the full model
of the values and sales regression in Tables 3 and 4 with an SLR scenario of 1 metre. Table 7
shows that while height above sea level is generally associated with a price and valuation
discount, the flood zone discount is similar to the main results, the SLR scenario shows a
premium in the values whilst the market displays a discount although not significant. In any
case, it is confirmed that no discount exists even for areas that will be affected with a relatively
high degree of probability.
Table 7. Robustness checks with continuous SLR and 1 metre SLR
Variations of SLR measure
(1)
(2)
(3)
(4)
Sales Price
Capital Value
Continuous
SLR
Continuous
SLR 1 metre
Continuous
SLR
Continuous
SLR 1 metre
Height above sea level (continuous)
-0.000832
-0.00141***
Underwater if SLR=1
-0.118
0.0694***
Current flood zone overlay
-0.0356
-0.0339
-0.0141***
-0.0109***
FULL CONTROLS
YES
YES
YES
YES
R2
0.673
0.673
0.94
0.94
adj. R2
0.662
0.662
0.94
0.94
AIC
2899.6
2899.3
-30358.6
-30355.8
BIC
3458.2
3457.9
-29712.5
-29709.8
F
.
.
7964
7953
N
3001
3001
36330
36330
* p < 0.05, ** p < 0.01, *** p < 0.001
6. Discussion
One of the main results emerging from the hedonic pricing models presented above is that flood
risk designated properties are generally discounted for the risk of partial or complete damage
of the property due to flooding. The model specification controls for a very large number of
dwelling characteristics not normally included in residential regression analysis due to data
limitations; locational and neighbourhood amenity elements, a large number of building
specification variables, including waterfront location, attractive vistas and a condition rating of
the buildings, all of which are known confounders in property pricing studies of flood risk. The
results provide clear empirical evidence in support of Hypothesis 1, i.e. that floodplain
designated properties will be discounted compared to non-floodplain identified properties.
More specifically, an inverse statistical relationship was found between flood risk designated
properties and capital values, sales prices, site values and land prices, which concurs with the
findings of, for example, Beltran et al. (2018) of discounts applicable to flood risk or floodplain
designated land. Beltran et al. (2018) in their meta analysis found an average 4.6% discount
for floodplain designated properties; the results of our study demonstrate a discount between 1
3% for properties and between 2 5% discount in land value for properties identified in a
flood risk area through the statutory authority planning overlays. Further, it would appear in
our analysis that the market are discounting the flood risk higher than the valuers within the
municipality.
The second hypothesis of this study, that sea level rise risk identified properties will be
discounted compared to non-sea level rise risk identified properties, was found to have no
significant results. In some of the estimates presented above, properties within the identified
sea level rise risk areas even had a positive and significant association with price. This suggests
that sea level rise is either deliberately not factored into purchasing and appraisal decisions in
the study area because the problem is deemed to be too far ahead in the future, or the market
and purchasers are unaware of the risks.
As far as discounts for flood risk from professional real estate valuers is concerned, we find
support for our third hypothesis, particularly vis-à-vis the evidence from market transactions.
This finding diverges from some extant studies, notably Harrison et al (2001) who found that
tax assessors slightly over-assessed properties located in flood zones, relative to those in other
areas. This may be indicative of a stronger stigma related to flood risk reflected in the price
paid for a property in our study area, when the purchasers are presented information that
demonstrates the flood risk.
The wider implications of our findings are twofold. Firstly, mandatory provision of risk
information on SLR may precipitate the hypothesised discounting effect on price. From our
analysis, we cannot discern if the lack of this effect is due to a lack of information, awareness
or hyperbolic time discounting of future losses. In the Australian state of Victoria, all properties
at point of sale, are issued prior to the signing of the contract with pertinent information about
the property; at present this includes a map and description issued by the statutory authority,
of whether the property is within or near a flood risk area. Consequently, purchasers are made
aware of the implied flood risk of the property; and it is expected that a rationale purchaser
would factor in and price the risk into the price paid (Beltran et al., 2018). The analysis in our
study demonstrates a discount associated with flood risk designated property; suggesting that
purchasers are making pricing decisions in relation to the perceived risk. Concurring with
Votsis and Perrels (2016), in that there is a strong importance of disclosure of information;
there is at least the possibility that such mandatory information provision may have a
subsequent impact on property values and prices.
When examining whether SLR has an effect on prices and capital values; there is at present no
empirical evidence to suggest there is a discounting occurring in relation to sea level rise. This
is understandable in the context of sales prices, because at present buyers are not provided with
any information that might affect their decision-making in relation to sea level rise and the
perceived risk to their property. Furthermore, the lack of current information available to both
potential purchasers, owners and valuers within the municipality could create future liability
and responsibility issues.
To protect owners, occupiers, investors and public authorities of coastal properties from future
financial losses and potentially greater losses both direct and indirect; action may be needed to
provide these stakeholders with a clearer understanding of the possible risks associated with
property in coastal areas. Further, early awareness of the issue may provoke greater response
from residents to consider not only mitigation action, but planning for adaptive actions that
may take years and considerable sums of money to fund.
7. Conclusions
What has been apparent in studies of the Australian environment has been, that exposure to
adverse environmental events, be it flood, fire or drought, the increasing frequency of
substantial precipitation, extreme storms and winds or heat waves, all contribute to short-term
discounts and long-term impingement on capital growth. Consequently, it can be considered
the effect of sea level rise is likely to have a significant effect on property value. Areas that are
directly affected might face heavy discounting and even subsequent total loss; and properties
not directly inundated will face the costs and losses associated with increased flooding. Future
changes to regulation, legislation or even environmental considerations will affect prevailing
land uses, development opportunity and costs; insurance premiums, financing costs;
depreciation, changes in tenant demand and occupancy, and increases in maintenance, statutory
reporting and refurbishment costs; all of which will ultimately affect property values. Further,
in consideration of the anticipation of sea level rise and increased flooding will likely result in
increases in the number of uninsurable properties and increases in insurance premiums and
regulatory measures for those partially affected or within the region. To gain a greater
understanding of the likelihood of the impact on property, measures could be put in place to
identify, ascertain and quantify risks in order to demonstrate a stronger reasoning for
implementing mitigation and adaption strategies for property assets. By connecting the value
to the profiling of sea level rise risk identification process, this can be considered by property
stakeholders and governments and result in subsequent action. A necessary condition for this
to occur seems to be that stakeholders need to be able to understand and quantify the risks
posed.
The present study highlights that property markets identify and price the risk profiles of
properties. Future work, for example working with quasi-natural experiments or randomised
control studies, may seek to disentangle the effect of incomplete information from hyperbolic
discounting to ascertain whether information provision in itself may be sufficient for bringing
about the expected price impacts or whether future financial and physical damage from sea
level rise may still be deemed too uncertain by the majority of market participants to enter the
pricing decisions of individual property purchases.
Acknowledgements
We are grateful to Thrive Research Hub, the Faculty of Architecture, Building and Planning,
and the University of Melbourne for providing funding for data; to the Valuer General
Victoria, and the local bayside Municipality for providing assistance and access to data for
the analysis in this project; and to Gideon Aschwanden, Andy Krause and Matthew Palm for
valuable advice and assistance in the collating, mapping, geocoding and analysing of datasets.
References
Atreya, A., Ferreira, S., Kriesel, (2013) Forgetting the Flood? An Analysis of the Flood Risk
Discount over Time, Land Economics, 89(4) 577-596
Atreya, A., Ferreria, S. (2015) Seeing is Believing? Evidence from Proeprty Prices in Inundated
Areas, Risk Analysis, 35(5), 828-848
Barrage, L., & Furst, J. (2019). Housing investment, sea level rise, and climate change
beliefs. Economics Letters, 177, 105-108.
Beltran, A., Maddison, D., Elliott, R.J.R., (2018) Is Flood Risk Capitalised Into Property
Values, Ecological Economics, 146 (2018), 668 - 685
Bernstein, A., Gustafson, M. T., & Lewis, R. (2019). Disaster on the horizon: The price effect
of sea level rise. Journal of Financial Economics, 134 (2019), 253-272.
Bin, O., Kruse, J.B., Landry, C.E., (2008) Flood Hazards, Insurance Rats and Amenities:
Evidence from the Coastal Housing Market, Journal of Risk and Insurance, 75(1) 63 82
Bin, O., Crawford, T.W., Kruse, J.B., Landry, C.E., (2008) Viewscapes and Flood Hazard:
Coastal Housing Market Response to Amenities and Risk, Land Economics 84(3) 434-448
Bin, O., Kruse, J.B., (2006) Real Estate Markets Response to Coastal Flood Hazards, Natural
Hazards Review, 7(4) 137 144
Bin, O., Landry, C.E., (2013) Changes in implicit flood risk premiums: empirical evidence
from the housing market, Journal of Environmental Economics and Management, 65 (2013)
361-378
Bin, O., Poulter, B., Dumas, C. F., & Whitehead, J. C. (2011). Measuring the impact of sea
level rise on coastal real estate: A hedonic property model approach. Journal of Regional
Science, 51(4), 751-767.
Brouwer, R., van Ek, R., (2004) Integrated ecological, economic and social impact assessment
of alternative flood control policies in the Netherlands, Ecological economics, 50 (2004) 1
21
Bunten, D., Kahn, M.E. (2017) Optimal real estate capital durability and localized climate
change disaster risk, Journal of Housing Economics, 36(2017) 1 7
Cameron, L., Shah, M., (2015) Risk-taking behaviour in the wake of natural disasters, Journal
of Human Resources, 50(2), 484-515
Chen, Y., Fingelton, B., Pryce, G., Chen, A.S., Djordjevic, S., (2013) Implications of rising
flood-risk for employment locations: a GMM spatial model with agglomeration and
endogenous house price effects, Journal of Property Research, 30(4) 298-323
Chivers, J., & Flores, N. E. (2002). Market failure in information: the national flood insurance
program. Land Economics, 78(4), 515-521.
Daniel, V.E., Florax, J.G.M., Rietveld, P., (2009a) Flooding risk and housing values: an
economic assessment of environmental hazard, Ecological Economics, 69(2009) 355-365
Daniel, V.E., Florax, J.G.M., Rietveld, P (2009b). Floods and Residential Property Values: A
Hedonic Price Analysis for the Netherlands. Built Environment (1978-), 35(4), 563-576.
Retrieved from http://www.jstor.org/stable/23290004
Department of Environment, Land, Water and Planning. (2015). Climate-Ready Victoria:
Victoria November 2015.Melbourne: Department of Environment, Land,Water and Planning.
Department of Environment, Land,Water and Planning. (2018). Victorian Planning Provisions,
Clause 44.05, Special Building Overlay. Available online: http://planning-
schemes.delwp.vic.gov.au/schemes/vpps (accessed on 5 April 2018)
Deser, C., Phillips, A., Bourdette, V., & Teng, H. (2012). Uncertainty in climate change
projections: the role of internal variability. Climate dynamics, 38(3-4), 527-546.
Feldman, H. (2019). Young people won’t accept inaction on climate change, and they’ll be
voting in droves. The Conversation. May 2, 2019. Accessed 30/1/2019 from
http://theconversation.com/young-people-wont-accept-inaction-on-climate-change-and-
theyll-be-voting-in-droves-116361
Field, N. (2019). History has a price when it comes to heritage homes. Money. August 21,
2019. Accessed 10 September 2019 from https://www.moneymag.com.au/heritage-homes
Fu, X., Song, J., Sun, B., Pen, Z. (2016). “Living on the edge”: Estimating the economic cost
of sea level rise on coastal real estate in the Tampa Bay region, Florida, Ocean and Coastal
Management, 133(2016) 11 17.
Geiger, N., & Swim, J. K. (2016). Climate of silence: Pluralistic ignorance as a barrier to
climate change discussion. Journal of Environmental Psychology, 47, 79-90.
Geoscience Australia, (2015a). ‘OzCoasts: Australian Online Coastal Information’. Accessed
7 June 2017. http://www.ozcoasts.gov.au/index.jsp
Geoscience Australia, (2015b). ‘OzCoasts FAQ’. Accessed 20 September 2017.
http://www.ozcoasts.gov.au/climate/sd_fqa.jsp
Hauer, M. E., J. M. Evans and D. R. Mishra. (2016). ‘Millions Projected to be at Risk from
Sea-Level Rise in the Continental United States’. Nature Climate Change 6, 69195.
doi:10.1038/nclimate2961
Harrison, D.M.m, Smersh, G.T., Schwartz, A.L., (2001) Environmental Determinants of
Housing Prices: The impact of flood zone status, Journal of Real Estate Research, 21(1/2), 3
20.
Hamilton, J. M. (2007). Coastal landscape and the hedonic price of accommodation. Ecological
Economics, 62(3-4), 594-602.
Ho, M. C., Shaw, D., Lin, S., & Chiu, Y. C. (2008). How do disaster characteristics influence
risk perception?. Risk Analysis: An International Journal, 28(3), 635-643.
Hurlimann, A. C., Browne, G. R., Warren-Myers, G., & Francis, V. (2018). Barriers to
climate change adaptation in the Australian construction industryImpetus for regulatory
reform. Building and Environment, 137, 235-245.
Hurlimann, A. C., Warren-Myers, G., & Browne, G. R. (2019). Is the Australian construction
industry prepared for climate change?. Building and Environment, 153, 128-137.
Kahn, M. E. (2016). The climate change adaptation literature. Review of Environmental
Economics and Policy, 10(1), 166-178.
de Koning, K., Filatova, T., & Bin, O. (2017). Bridging the Gap Between Revealed and Stated
Preferences in Flood-prone Housing Markets. Ecological economics, 136, 1-13.
Kummu, M., De Moel, H., Salvucci, G., Viviroli, D., Ward, P. J., & Varis, O. (2016). Over the
hills and further away from coast: global geospatial patterns of human and environment over
the 20th21st centuries. Environmental Research Letters, 11(3), 034010.
Jericho, G. (2019). Parlimentarians deserve our wrath for 30 years of inaction, not climate
protesters. The Guardian. 13th October 2019. Accessed 30/11/2019, from
https://www.theguardian.com/business/grogonomics/2019/oct/12/parliamentarians-deserve-
our-wrath-for-30-years-of-inaction-not-climate-protesters
Lawrence, J., Bell, R., Blackett, P., Stephens, S., & Allan, S. (2018). National guidance for
adapting to coastal hazards and sea-level rise: Anticipating change, when and how to change
pathway. Environmental science & policy, 82, 100-107.
Lichter, M., & Felsenstein, D. (2012). Assessing the costs of sea-level rise and extreme
flooding at the local level: a GIS-based approach. Ocean & coastal management, 59, 47-62.
MelbourneWater. 2017. Planning for Sea Level Rise Guidelines: Port Phillip andWestern Port.
Available online: https://www.melbournewater.com.au/sites/default/files/Planning-for-sea-
levels.pdf (accessed on 15 June 2017).
Melbourne Water (2017) Floods: a recurring natural hazard, accessed 4th April 2019,
https://shaping.banyule.vic.gov.au
McCoy, S. J., & Zhao, X. (2018). A City under Water: A Geospatial Analysis of Storm
Damage, Changing Risk Perceptions, and Investment in Residential Housing. Journal of the
Association of Environmental and Resource Economists, 5(2), 301-330.
McGranahan, G., Balk, D., & Anderson, B. (2007). The rising tide: assessing the risks of
climate change and human settlements in low elevation coastal zones. Environment and
urbanization, 19(1), 17-37.
McInnes, K., J. Church, D. Monselesan, J. R. Hunter, J. G. O’Grady, I. D. Haigh and X. Zhang.
(2015). ‘Information for Australian Impact and Adaption Planning in Response to Sea-level
Rise’. Australian Meteorological and Oceanographic Journal 65, no. 1 (October): 12749.
Michael, J. A. (2007). Episodic flooding and the cost of sea-level rise. Ecological economics,
63(1), 149-159.
Neumann, B., Vafeidis, A. T., Zimmermann, J., & Nicholls, R. J. (2015). Future coastal
population growth and exposure to sea-level rise and coastal flooding-a global assessment.
PloS one, 10(3), e0118571.
Nicholls, R. J., Hanson, S., Herweijer, C., Patmore, N., Hallegatte, S., Corfee-Morlot, J.,
Chateau, J. & Muir-Wood, R. (2007). Ranking of the world’s cities most exposed to coastal
flooding today and in the future. Report Organiseation for Economic Co-=operation and
Development, Paris, France.
Ortega, F., & Taspinar, S. (2018). Rising Sea Levels and Sinking Property Values: The Effects
of Hurricane Sandy on New York's Housing Market, Journal of Urban Economics, 106 (2018)
81 100.
Pankratz, N.M. C. and Bauer, R. Derwall, J. (2019) Climate Change, Firm Performance, and
Investor Surprises (May 21, 2019). Available at SSRN: https://ssrn.com/abstract=3443146
Pryce, G., Chen, Y., (2011) Flood risk and the consequences for housing of a changing climate:
An international perspective, Risk Management, 13(4). Risk Sustainability and the
Environment (October 2011) pp. 228-246
Rambaldi, A. N., Fletcher, C. S., Collins, K., & McAllister, R. R. (2013). Housing shadow
prices in an inundation-prone suburb. Urban Studies, 50(9), 1889-1905.
Gifford, R. (2011). The dragons of inaction: Psychological barriers that limit climate change
mitigation and adaptation. American psychologist, 66(4), 290.
Scott, D., Simpson, M. C., & Sim, R. (2012). The vulnerability of Caribbean coastal tourism
to scenarios of climate change related sea level rise. Journal of Sustainable Tourism, 20(6),
883-898.
Titus, J.G., Park, R.A., Leatherman, S.P., Weggel, J.R., Greene, M.S., Mausel, P.W., Brown,
S., Gaunt, C., Trehan, M. and Yohe, G., (1991). Greenhouse effect and sea level rise: the cost
of holding back the sea. Coastal Management, 19(2), 171-204.
Votsis, A., & Perrels, A. (2016). Housing prices and the public disclosure of flood risk: a
difference-in-differences analysis in Finland. The Journal of Real Estate Finance and
Economics, 53(4), 450-471.
Walsh, P., Griffiths, C., Guignet, D., & Klemick, H. (2019). Adaptation, sea level rise, and
property prices in the Chesapeake Bay watershed. Land economics, 95(1), 19-34.
Warren-Myers, G., Fuerst, F., Aschwanden, G., (2018) Estimating the Potential Risks of Sea
Level Rise for Public and Private Property Ownership, Occupation and Management, Risks, 6,
37-58.
Appendix 1: Summary information for valuation dataset
Total sample
Houses
Apartments and
Units
Capital Improved Value
$1,010,709
$1,543,536
$575,528
($817,648)
($881,379)
($365,976)
Construction Year
1950
1916
1979
(46)
(41)
(26)
Building area (m2)
112
149
82
(63)
(66)
(40)
Building Condition (scale 1 - 10)
5.39
5.66
5.18
(1.67)
(1.68)
(1.64)
Quality (scale 1 - 5)
3.07
3.05
3.10
(0.45)
(0.47)
(0.44)
Bedroom (number)
2.26
2.84
1.78
(0.90)
(0.79)
(0.67)
Car space (1=yes)
0.78
0.62
0.93
(0.80)
(0.70)
(0.86)
Sea level rise risk (1=yes)
0.01
0.00
0.01
(0.09)
(0.01)
(0.12)
Sea level rise & flood risk (1=yes)
0.28
0.42
0.17
(0.45)
(0.49)
(0.38)
Flood overlay (1=yes)
0.14
0.16
0.13
(0.35)
(0.36)
(0.33)
Beachfront (1=yes)
0.05
0.02
0.09
(0.23)
(0.13)
(0.28)
Canal front (1=yes)
0.01
0.01
0.01
(0.09)
(0.09)
(0.09)
Site is 400m from park (1=yes)
0.83
0.85
0.81
(0.38)
(0.36)
(0.39)
Distance to beach (metres)
1064.71
956.15
1148.30
(736.50)
(627.10)
(803.90)
Distance to grocery and retail (metres)
571.55
575.19
571.69
(298.70)
(275.10)
(318.10)
Distance to nearest childcare facility (metres)
475.53
475.52
475.72
(252.40)
(257.70)
(248.30)
Distance to nearest pool or public recreation facility
(metres)
1648.75
1451.58
1818.20
(1032.80)
(1066.90)
(969.30)
Distance to nearest healthcare facility (metres)
668.58
652.89
683.79
(302.70)
(302.80)
(303.10)
Distance to nearest public Secondary School (metres)
1809.27
1630.09
1959.82
(856.20)
(817.10)
(858.30)
Distance to nearest public Primary School (metres)
949.24
847.38
1033.74
(380.90)
(380.80)
(357.80)
Distance to nearest rail station (metres)
6088.90
4411.45
6752.34
(63109.70)
(43948.80)
(70388.50)
Distance to nearest tram (metres)
425.04
442.68
411.61
(334.90)
(338.30)
(332.50)
Distance to nearest pharmacy (metres)
572.69
642.93
516.22
(331.80)
(379.80)
(273.10)
N
37008
16639
19535
mean coefficients; sd in parentheses
Appendix 2: Summary information for sales dataset
Total sample
Houses
Apartments and
Units
Sales Price
$905,899
$1,318,038
$625,588
$819,198
$1,044,575
($437,666)
Construction Year
1945
1923
1970
(50)
(46)
(41)
Building area (m2)
148
149
82
(239)
(98)
(42)
Building Condition (scale 1 - 10)
6
6
5
(2)
(2)
(2)
Quality (scale 1 - 5)
3.10
3.09
3.11
(0.44)
(0.45)
(0.42)
Bedroom (number)
1.41
1.57
1.30
(0.63)
(0.74)
(0.51)
Bathrooms
2.22
2.82
1.81
(0.96)
(1.00)
(0.67)
Car space (1=yes)
1.07
1.09
1.06
(0.78)
(1.00)
(0.60)
Sea level rise risk (1=yes)
0.01
0.01
0.01
(0.09)
(0.10)
(0.08)
Sea level rise & flood risk (1=yes)
0.25
0.34
0.20
(0.44)
(0.47)
(0.40)
Flood overlay (1=yes)
0.18
0.23
0.15
(0.38)
(0.42)
(0.35)
Beach front (1=yes)
0.07
0.04
0.10
(0.26)
(0.19)
(0.30)
Canal front (1=yes)
0.01
0.01
0.01
(0.09)
(0.09)
(0.09)
Site is 400m from park (1=yes)
0.87
0.86
0.87
(0.34)
(0.35)
(0.34)
Distance to beach (metres)
936
904
957
(705)
(622)
(756)
Distance to grocery and retail (metres)
540
583
512
(294)
(282)
(300)
Distance to nearest childcare facility (metres)
505
488
517
(243)
(251)
(236)
Distance to nearest pool or public recreation facility (metres)
1689
1631
1729
(993)
(1102)
(910)
Distance to nearest healthcare facility (metres)
669
634
693
(348)
(324)
(362)
Distance to nearest public Secondary School (metres)
1703
1588
1782
(853)
(829)
(861)
Distance to nearest public Primary School (metres)
979
887
1042
(403)
(376)
(410)
Distance to nearest rail station (metres)
2170
2348
2050
(1058)
(1092)
(1018)
Distance to nearest tram (metres)
445
473
426
(357)
(376)
(341)
Distance to nearest pharmacy (metres)
543
620
491
(339)
(375)
(301)
N
3950
1599
2351
mean coefficients; sd in parentheses
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The estimation of future sea level rise (SLR) is a major concern for cities near coastlines and river systems. Despite this, current modelling underestimates the future risks of SLR to property. Direct risks posed to property include inundation, loss of physical property and associated economic and social costs. It is also crucial to consider the risks that emerge from scenarios after SLR. These may produce one-off or periodic events that will inflict physical, economic and social implications, and direct, indirect and consequential losses. Using a case study approach, this paper combines various forms of data to examine the implications of future SLR to further understand the potential risks. The research indicates that the financial implications for local government will be loss of rates associated with total property loss and declines in value. The challenges identified are not specific to this research. Other municipalities worldwide experience similar barriers (i.e., financial implications, coastal planning predicaments, data paucity, knowledge and capacity, and legal and political challenges). This research highlights the need for private and public stakeholders to co-develop and implement strategies to mitigate and adapt property to withstand the future challenges of climate change and SLR.
Article
Coastal communities are facing the dual threat of increasing sea level rise (SLR) and swelling populations, causing challenging policy problems. To help inform policy makers, this paper explores the property price impact of structures that help protect against SLR, using a novel and spatially explicit dataset of coastal features. Results indicate that adaptation structures can have a significant positive impact on waterfront home prices, with the most vulnerable homes seeing the largest impacts. The Chesapeake Bay is facing increasing pressure from SLR, and this is one of the first papers to report that local property markets are incorporating that threat. © 2019 by the Board of Regents of the University of Wisconsin System.
Article
Research examining the construction industry demonstrates that it is the source of a large percentage of global greenhouse gas emissions, subject to future climate change risks, and that there is limited evidence of adaptation. This paper explores the preparedness of the Australian construction industry to adapt to climate change risks, through twenty-one key stakeholder interviews. The interviews addressed participants' perceptions of the industry's: climate change risk awareness, analytical capacity to address climate change risk, and current actions to address these risks. The main risk participants identified was extreme weather during construction (e.g. heat and wind) with impacts for occupational health and safety, supply chain reliability, project delivery delays, and profit. The impact of these risks was somewhat down played. While a high level of perceived analytical capacity to respond to climate change was observed, real action was limited by systemic, outdated regulations, lack of client motivation, and perceived costs. The climate change actions which were undertaken were predominantly climate change mitigation activities, largely focussing on organisational processes, rather than adaptations to built form. The results indicate that if Australia's built form is to be well adapted to climate change, greater preparedness to facilitate tangible and systemic actions to address climate change risk in the Australian construction industry is needed. Specifically, interviewees responses imply that increased awareness of the impacts of climate change, better translation of existing solutions into practice, and regulatory reform are needed.
Article
We investigate the association of new housing construction with sea level rise exposure and climate change beliefs. We combine U.S. Census construction permits data, sea level rise projections, county-level climate change belief estimates (Howe et al., 2015), and standard housing controls. The results suggest that (i) sea level rise vulnerability is associated with significantly reduced construction in areas with high climate change beliefs, but that (ii) this relationship is significantly attenuated in more skeptical areas, suggesting that climate skepticism may be delaying adaptation.
Article
It is increasingly recognised that the risks associated with climate change must be addressed through both mitigation and adaptation. Buildings are vulnerable to climate change risk and are also the source of a significant proportion of greenhouse gas emissions which contribute to climate change. The construction industry has significant potential to facilitate adaptation through actions that both reduce its contribution to greenhouse gas emissions across the construction and building lifecycle, and through physical adaptation of buildings and settlements to withstand present and future changes. However, there is limited evidence of significant adaptive action to date, and little is known about existing barriers to adaptation actions in the construction industry. This research explores barriers to climate change adaptation in the Australian construction industry through qualitative interviews with twenty-one key stakeholders. The barriers identified included: the use of inconsistent and unclear language, limited regulation, perceived unaffordability of initiatives, lack of awareness of climate change, and lack of client demand to implement initiatives. Recommendations to facilitate strategies for adaptation to climate change in the construction industry are provided. These focus on the need to a) address climate change through regulatory reform, and b) address the structure of the construction industry and its interrelationship with other built environment professions and processes.
Article
Economic theory suggests that, other things being equal, properties located within a floodplain should suffer a price discount. A survey of the existing evidence nonetheless reveals that this price discount lies anywhere between − 75.5% to a + 61.0% price premium. In this paper we summarise and explore the wide variation in the results to obtain ‘best’ estimates with which to guide policy. Results from our meta-analysis comprising 37 published works and 364 point estimates indicate marked differences between studies according to when and where they were conducted. For coastal regions the results show that properties located in the floodplain command higher prices; this finding is however likely to be caused by a high correlation between omitted coastal amenities and flood risk. There is moreover, evidence that publication bias affects the coastal flooding literature. Results from meta-regression analyses intended to uncover sources of heterogeneity confirm that controlling for time elapsed since the most recent flood is especially important. For inland flooding the price discount associated with location in the 100-year floodplain is − 4.6%. Although other estimates are defensible, we suggest this figure be used as a rule of thumb to determine the benefits of flood relief projects to households.
Article
Sea-level rise challenges public policy-making because existing planning frameworks and methods are designed to promote certainty using static and time-bound planning and legal instruments. Sea-level rise is a dynamic and uncertain process, which is deeply uncertain towards the latter part of this century and beyond. Communities require decision making approaches that can enable adjustments to policies ahead of damage, without entrenching current exposure to hazards or incurring larger than necessary adjustment costs in the future. We first discuss the nature of the sea-level problem, the policy context that creates decision-making challenges and how they have been typically addressed through policy and practice. Secondly, we show how an assessment and planning approach, designed to address uncertainty and change (the Dynamic Adaptive Policy Pathways (DAPP) planning approach), has been integrated into national guidance for coastal hazard and climate change decision-making in New Zealand. The Guidance integrates hazard and sea-level rise assessments with uncertainty type and with the scale and scope of activity. It is underpinned with values-based community engagement, and uses signals and decision triggers for monitoring and adjusting pathways to meet objectives over time. The applicability of the approach in the Guidance for other policy problems involving uncertainty, is also discussed.
Article
Homes exposed to sea level rise (SLR) sell for approximately 7% less than observably equivalent unexposed properties equidistant from the beach. This discount has grown over time and is driven by sophisticated buyers and communities worried about global warming. Consistent with causal identification of long-horizon SLR costs, we find no relation between SLR exposure and rental rates and a 4% discount among properties not projected to be flooded for almost a century. Our findings contribute to the literature on the pricing of long-run risky cash flows and provide insights for optimal climate change policy.