Spatial Heterogeneity in Flood-Risk Capitalization: Evidence from the Thames Estuary Housing Market

Spatial Heterogeneity in Flood-Risk Capitalization: Evidence from the Thames Estuary Housing Market | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Spatial Heterogeneity in Flood-Risk Capitalization: Evidence from the Thames Estuary Housing Market Oluwaseun Damilola Ajayi, Tayo Odunsi, Paulinus Ugwu, Arti Rawat, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9213954/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Understanding whether housing markets internalize climate-related flood risk is central to assessing the economic consequences of coastal exposure. This study examines the capitalization of flood risk in residential property prices across the Thames Estuary housing market. Using transaction data for more than 73,000 residential properties, the analysis combines global spatial econometric models with geographically weighted regression to evaluate both the average and spatially varying effects of flood exposure on housing values. Global hedonic and spatial models indicate a positive average association between flood-zone exposure and property prices, suggesting that properties located within designated flood-risk areas command higher values. However, geographically weighted regression reveals substantial spatial heterogeneity in this relationship. In outer estuary communities, flood exposure is associated with significant price discounts, indicating that environmental risk is capitalized directly into housing values. In contrast, several urban waterfront submarkets display positive capitalization effects, where accessibility advantages and waterfront amenities appear to offset perceived environmental exposure. These findings demonstrate that the average flood-risk premium estimated by global models masks the coexistence of distinct spatial pricing regimes. Environmental risk is discounted in peripheral coastal markets but may be offset by amenity and redevelopment value in high-demand urban waterfront locations. The results highlight the importance of spatially heterogeneous modelling when evaluating climate-risk capitalization in housing markets and suggest that aggregate housing price signals may underestimate localized exposure within rapidly developing coastal urban systems. JEL Classification Codes: R31; R12; C21; Q54 Climate risk Flood risk capitalization Housing prices Spatial econometrics Geographically weighted regression Thames Estuary Figures Figure 1 Figure 2 Figure 3 1. Introduction Climate change is increasingly reshaping the spatial structure of housing markets through the capitalization of environmental risk into property values. Coastal and estuarine regions are particularly exposed as rising sea levels and intensifying flood events alter both the physical risk profile of housing assets and expectations about long-term protection infrastructure. A growing body of evidence shows that environmental hazards such as flood exposure, wildfire risk, and sea-level rise can generate measurable price discounts in residential markets (Bin and Polasky 2004 ; Beltrán et al. 2018 ; Bernstein et al. 2019 ). However, the magnitude and spatial distribution of these effects remain highly uneven across urban regions. Housing markets simultaneously capitalize multiple attributes including location, structural characteristics, and environmental amenities, while adaptation policies and protective infrastructure may partially offset perceived risk. As a result, the pricing of climate exposure is unlikely to be uniform across space. Instead, risk capitalization may vary substantially across local housing submarkets depending on geography, built form, and institutional flood governance regimes. A substantial literature examines how environmental risks are capitalized into housing prices using hedonic price models. Early studies document statistically significant discounts for properties located in flood-prone areas or subject to environmental hazards (Bin and Polasky 2004 ; Bin et al. 2008 ). More recent research shows that climate-related risks such as sea-level rise expectations, storm exposure, and insurance reforms can also influence housing market outcomes (Beltrán et al. 2018 ; Bernstein et al. 2019 ; Keys and Mulder 2020 ). While these studies provide important evidence that environmental risk affects housing prices, most empirical analyses rely on global regression frameworks that assume constant marginal effects across space. This assumption may be restrictive in coastal housing markets where risk exposure, topography, and flood protection infrastructure vary sharply across short geographic distances. If price responses to environmental risk differ across local housing submarkets, global hedonic estimates may mask important spatial variation in the capitalization process. A growing strand of spatial housing research therefore emphasizes the need for locally varying models capable of identifying heterogeneous price gradients across urban regions (Fotheringham et al. 2002 ; Páez et al. 2011 ). In this context, geographically weighted regression (GWR) provides a flexible framework for estimating location-specific relationships between property prices and environmental attributes, allowing the marginal effects of risk exposure to vary across space rather than imposing a single global parameter. This study examines the spatial capitalization of climate exposure in the housing market of the Thames Estuary, one of the most flood-exposed urban regions in Europe. Using transaction-level data for residential properties combined with elevation, river proximity, flood risk classifications, and structural housing characteristics, the analysis evaluates how environmental exposure influences property prices across local housing submarkets. Diagnostic tests reveal significant spatial dependence in hedonic model residuals, indicating that conventional global specifications fail to capture localized price dynamics. To address this issue, the study applies geographically weighted regression (GWR) to estimate spatially varying price relationships between housing values and climate exposure. The results show that flood risk and topographic elevation exhibit substantial spatial heterogeneity in their capitalization into housing prices. In low-lying estuarine corridors such as Canvey Island and Purfleet, flood exposure generates measurable price discounts, whereas proximity to the Thames produces localized price premiums in higher-elevation areas where amenity value dominates perceived risk. These findings demonstrate that environmental risk is not priced uniformly across the housing market but varies systematically across space depending on local exposure and geography. By identifying these spatially heterogeneous price responses, the study contributes to the literature on environmental risk capitalization and highlights the importance of spatial modeling approaches in climate-exposed housing markets. The remainder of the paper proceeds as follows. Section 2 reviews the literature on environmental risk capitalization and spatial heterogeneity in housing markets, with particular attention to flood exposure and coastal property valuation. Section 3 describes the study area, data sources, and construction of the housing, environmental, and structural variables used in the analysis. Section 4 presents the empirical methodology, beginning with a conventional hedonic price specification and extending to geographically weighted regression to estimate spatially varying price relationships. Section 5 reports the empirical results, including global model diagnostics and local parameter estimates that reveal spatial variation in climate risk capitalization across the Thames Estuary housing market. Section 6 concludes with the implications of these findings for housing market adjustment in climate-exposed urban regions. 2. Literature Review 2.1. Environmental Risk Capitalization in Housing Markets Environmental hazards can influence housing values when buyers incorporate expected exposure into property price formation. Early work on environmental capitalization demonstrates that housing markets often internalize localized risk through price adjustments when information about hazards becomes salient to buyers (Bin and Polasky, 2004 ). Evidence from flood-prone regions in the United States shows that properties exposed to flood hazards typically experience price discounts relative to comparable dwellings outside high-risk areas, although the magnitude of these discounts varies across locations and time periods (Atreya and Ferreira, 2015 ). Empirical studies increasingly document that capitalization depends on both the visibility of environmental risk and the availability of institutional risk signals such as flood maps, insurance pricing, or disaster events (Belanger and Bourdeau-Brien, 2018 ). Flood exposure presents a particularly relevant case because the valuation effects may operate through several channels. Physical vulnerability can reduce expected housing utility through anticipated damage or insurance costs, generating price discounts in exposed locations. At the same time, waterfront amenities can generate positive valuation effects, particularly where properties benefit from visual or recreational access to water bodies (Bin and Polasky, 2004 ). Housing markets therefore often reflect a trade-off between amenity value and environmental exposure. Empirical evidence indicates that these opposing forces may coexist within the same geographic market, producing complex valuation patterns rather than uniform price discounts (Dumm et al., 2016 ). Studies examining coastal and riverine housing markets show that proximity to water can increase property values in relatively safe areas while generating price penalties where flood exposure becomes more salient (Belanger and Bourdeau-Brien, 2018 ). Recent research has also emphasized that climate change may alter pricing dynamics by increasing the perceived probability of extreme events. Evidence suggests that climate-related risks such as sea-level rise or coastal flooding are beginning to influence housing markets in vulnerable regions, although the magnitude of price adjustments remains uneven across markets (Kahn et al., 2021 ). Limited capitalization of long-term climate risk has been documented in several contexts, suggesting that housing markets may respond slowly to emerging environmental threats when risks remain uncertain or institutionally mediated (Gourevitch et al., 2023 ). These findings highlight the importance of examining how climate exposure interacts with local geography and institutional contexts to influence housing valuation. Despite the growing literature on environmental risk capitalization, most empirical studies rely on global hedonic price models that estimate average price effects across entire housing markets. Such approaches implicitly assume that environmental risk influences property values uniformly across space. Urban housing markets, however, often exhibit substantial spatial heterogeneity due to differences in neighborhood characteristics, built environments, and localized exposure conditions. Global models may therefore obscure important spatial variation in how environmental risk is priced. Recent work in urban and regional science increasingly emphasizes the need for spatially explicit approaches capable of identifying geographically varying price relationships within housing markets (Pace and LeSage, 2004 ). The next strand of literature addresses this challenge by examining spatial heterogeneity in hedonic housing models. 2.2 Spatial Heterogeneity in Housing Price Formation Housing markets rarely operate as spatially homogeneous systems. Property values emerge from localized interactions between structural attributes, neighborhood characteristics, and accessibility conditions. Empirical research has long documented that the marginal value of housing attributes can vary substantially across space, reflecting differences in neighborhood composition, land-use patterns, and urban structure (Dubin, 1988 ; Sirmans et al., 2005 ). Conventional hedonic price models typically estimate global parameters that represent average relationships across an entire study area. Such specifications impose constant marginal effects and therefore assume that the influence of housing characteristics is spatially invariant. This assumption often conflicts with observed urban housing dynamics. Spatial clustering of property values and neighborhood-specific amenities can generate localized pricing regimes that differ across submarkets. Evidence from urban housing studies shows that structural and locational attributes may carry different price premiums depending on neighborhood context and market segmentation (Can, 1992 ; Osland, 2010 ). Ignoring these spatial variations can lead to biased estimates when the marginal willingness to pay for housing characteristics differs across locations. Spatial dependence in housing prices further complicates estimation because nearby properties tend to exhibit correlated price movements through neighborhood spillovers and shared environmental conditions (Pace and LeSage, 2004 ). Environmental exposure provides a particularly strong source of spatial heterogeneity in housing valuation. Flood risk, elevation, and river proximity vary across small geographic scales and therefore may influence property prices differently across neighborhoods within the same metropolitan region. Areas located within flood-prone corridors may exhibit price discounts due to perceived exposure, whereas nearby neighborhoods with higher elevation or protective infrastructure may experience little or no penalty. Empirical studies examining environmental risk have therefore increasingly recognized that spatially constant hedonic coefficients may obscure important local variation in price responses (Belanger and Bourdeau-Brien, 2018 ). Spatial econometric approaches have consequently become central tools for examining geographically varying relationships in housing markets. Recent advances in spatial housing research emphasize models that allow coefficients to vary across locations rather than imposing uniform global parameters. These approaches recognize that housing markets consist of overlapping submarkets where valuation mechanisms may differ depending on local conditions. Identifying such localized pricing structures is essential when analyzing environmental risks whose effects depend strongly on geographic context. The following section discusses geographically weighted regression as a modeling framework designed to capture these spatially varying price relationships. 2.3 Geographically Weighted Regression in Housing Market Analysis Spatially varying relationships in housing markets have motivated the development of local regression approaches that allow parameters to differ across geographic locations. Geographically weighted regression (GWR) provides a framework for estimating location-specific coefficients by calibrating local regressions using spatially weighted observations surrounding each property (Brunsdon et al., 1996 ). Unlike conventional global hedonic models that impose constant marginal effects, GWR permits the influence of housing characteristics to vary continuously across space. This flexibility makes the approach particularly suitable for analyzing housing markets characterized by localized price formation and heterogeneous neighborhood conditions. Applications of GWR in housing research demonstrate that the marginal effects of structural and locational attributes often vary significantly across urban regions. Empirical studies show that housing characteristics such as floor area, accessibility, and neighborhood amenities exhibit geographically differentiated price effects when local estimation techniques are employed (Fotheringham et al., 2002 ). These findings suggest that spatially constant hedonic coefficients may mask important variations in buyer preferences and market conditions across neighborhoods. Incorporating spatially varying parameters therefore improves the ability of housing models to capture localized valuation patterns. Environmental exposure represents a context in which geographically varying price effects are particularly likely to arise. Flood risk, elevation, and proximity to water bodies often vary at fine spatial scales, producing localized differences in perceived risk and amenity value. Housing markets located along rivers or coastlines may therefore display heterogeneous responses to environmental attributes depending on local topography and exposure conditions. Global models that estimate a single price discount for flood exposure cannot capture these geographically differentiated effects. Spatially adaptive models such as GWR allow the marginal influence of environmental variables to vary across locations, making it possible to identify where risk capitalization is strongest or weakest within a metropolitan region. Several studies have applied GWR to housing markets to examine spatial heterogeneity in price determinants. Evidence from urban housing markets shows that local estimation approaches can reveal geographically varying relationships that remain hidden in conventional hedonic regressions (Bitter et al., 2007 ; Cohen et al., 2019 ). Such approaches have been used to analyze spatial variation in the effects of neighborhood amenities, accessibility, and environmental quality on property values. However, relatively limited research has applied spatially varying regression techniques to examine the capitalization of climate-related risks in coastal housing markets. 3. Data 3.1 Study Area and Data Sources The empirical analysis focuses on residential housing markets located within the Thames Estuary corridor in southeast England. The estuary represents one of the most climate-exposed urban regions in Europe, with extensive areas situated at low elevations and subject to tidal flood risk. Major urban centres along the estuary include boroughs in east and southeast London as well as downstream municipalities such as Thurrock and Gravesham. These areas contain a diverse mix of residential housing types and represent an important segment of the wider London metropolitan housing market. Flood exposure within the region is managed through the Thames Estuary 2100 (TE2100) flood risk management strategy, which defines spatial policy management units used to guide long-term adaptation planning. Residential transaction data are obtained from the United Kingdom Land Registry Price Paid Data (PPD), which records property-level sales transactions across England and Wales. The dataset provides transaction prices, property types, and sale dates for individual residential properties. Only arm’s-length transactions of residential dwellings are retained for the analysis. Observations corresponding to non-market transfers or incomplete records are excluded to ensure consistency in price measurement. Transaction prices are transformed using the natural logarithm to reduce skewness and align with conventional hedonic modelling practices. Property-level structural characteristics are obtained from the national Energy Performance Certificate (EPC) register. The EPC database provides information on dwelling attributes including total floor area, number of habitable rooms, energy efficiency rating, and property type classifications. These structural characteristics are commonly used in hedonic housing models to capture differences in dwelling size and building attributes that influence market prices (Sirmans et al., 2005 ). Environmental exposure variables are constructed using geospatial data from multiple sources. Elevation data are derived from the Ordnance Survey OS Terrain 5 digital elevation model, which provides high-resolution ground elevation estimates across the study region. Distance to tidal frontage is calculated as the Euclidean distance from each property location to the nearest section of the River Thames using geographic information system (GIS) methods. Flood exposure classifications are obtained from the Environment Agency Flood Map for Planning, which identifies properties located within Flood Zone 2 and Flood Zone 3 according to their estimated annual probability of flooding. These spatial datasets are merged with the transaction data using georeferenced property locations to construct the final analytical dataset. The resulting dataset integrates housing transaction information with structural dwelling characteristics and environmental exposure variables, allowing analysis of how flood risk and topographic conditions influence residential property prices within the Thames Estuary housing market. 3.2 Variable Construction and Descriptive Statistics The dependent variable is the transaction price of residential properties. Following standard practice in hedonic housing models, the natural logarithm of sale price is used to reduce skewness and interpret estimated coefficients as approximate percentage changes in property value (Rosen, 1974 ; Sirmans et al., 2005 ). The use of log-transformed prices also improves the statistical properties of regression estimates when price distributions exhibit right skewness, which is typical in housing markets. Explanatory variables are grouped into three categories reflecting conventional hedonic modelling frameworks: structural characteristics, environmental exposure measures, and climate adaptation indicators. Structural housing attributes capture differences in dwelling size and type that influence buyer willingness to pay. These variables include total floor area, the number of habitable rooms, and categorical indicators for property type. Empirical housing research consistently identifies dwelling size as the dominant determinant of residential property values, reflecting the direct relationship between housing consumption and price formation (Dubin, 1988 ; Sirmans et al., 2005 ). Environmental exposure variables measure physical climate risk associated with the Thames Estuary floodplain. Elevation is measured as ground height above sea level using high-resolution digital elevation data. Lower elevations correspond to greater vulnerability to tidal flooding and storm surge events. Distance to river frontage is calculated as the Euclidean distance from each property to the nearest segment of the River Thames. River proximity may capture two opposing valuation channels. Close proximity can generate positive amenity value through waterfront views and recreational access, while simultaneously increasing perceived flood exposure. Prior research indicates that these competing effects often produce nonlinear price responses in coastal housing markets (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). Flood exposure is further captured using Environment Agency flood risk classifications. Binary indicators identify whether properties are located within Flood Zone 2 or Flood Zone 3, which correspond to moderate and high annual flood probability thresholds. These classifications represent regulatory signals of environmental exposure that may influence buyer perceptions of risk and insurance availability. Adaptive capacity variables capture institutional and structural factors that may mitigate perceived climate risk. Energy performance ratings are included as a proxy for building efficiency and structural resilience, while Policy Management Unit (PMU) indicators identify properties located within areas designated under the Thames Estuary 2100 flood management strategy. These spatial governance units reflect the long-term flood defence planning framework implemented across the estuary and therefore represent institutional signals that may influence expectations regarding future flood protection. Table 1 summarizes the definitions, transformations, and data sources of all variables included in the analysis. Table 1 Variable Definitions and Sources Category Variable Definition & Transformation Data Source Target Sale Price Transaction price of residential property (£); log-transformed for modelling UK Price Paid Data Climate Exposure Elevation (m) Ground elevation at property location (meters above sea level) Ordnance Survey OS Terrain 5 Distance to River Euclidean distance to nearest tidal frontage (meters); log-transformed Ordnance Survey Flood Zone 3 Binary indicator for high flood probability (> 1% annual probability) Environment Agency Flood Map Flood Zone 2 Binary indicator for moderate flood probability Environment Agency Flood Map Adaptive Capacity EPC Rating Energy performance score of dwelling EPC Register PMU Indicator for Thames Estuary 2100 Policy Management Unit TE2100 Plan Structural Total Floor Area Total internal floor area (sqm) EPC Register Habitable Rooms Number of habitable rooms in dwelling EPC Register Property Type Categorical: house, flat, bungalow, maisonette EPC Register 3.3 Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market Table 1 reports descriptive statistics for the housing transactions used in the empirical analysis, disaggregated by the Thames Estuary 2100 (TE2100) Policy Management Units. The dataset contains 73,039 residential property transactions distributed across multiple estuarine submarkets characterized by distinct housing structures, socioeconomic conditions, and exposure to tidal flood risk. Examining descriptive variation across these policy units provides an initial indication of the spatial heterogeneity that characterizes housing markets in large metropolitan regions (Dubin, 1988 ; Sirmans et al., 2005 ). Substantial variation in property prices is evident across policy units. Average transaction prices range from approximately £231,804 in Purfleet, Grays and Tilbury to more than £585,000 in London City. High-value areas such as Greenwich and Wandsworth to Deptford exhibit mean prices exceeding £500,000, reflecting their proximity to central London and strong demand for waterfront housing. Peripheral estuarine locations including North Kent and Canvey Island display substantially lower price levels, consistent with differences in accessibility, local economic conditions, and housing supply constraints. Similar price gradients between central urban areas and peripheral locations have been widely documented in metropolitan housing markets (Osland, 2010 ; Des Rosiers et al., 2011 ). Structural housing characteristics vary significantly across the TE2100 policy units. Average floor area ranges from approximately 72 m² in Purfleet, Grays and Tilbury to nearly 120 m² in Leigh Old Town and Southend-on-Sea. Property type composition also differs markedly across locations. Inner estuary districts such as Isle of Dogs and London City exhibit a high proportion of flats, reflecting dense urban development patterns and the vertical housing supply typical of central metropolitan areas. In contrast, outer estuarine zones such as Canvey Island and North Kent are dominated by detached and semi-detached houses. These structural differences represent fundamental determinants of housing prices in hedonic valuation models, as dwelling size and building form directly influence the consumption value of housing services (Rosen, 1974 ; Sirmans et al., 2005 ). Neighbourhood socioeconomic conditions exhibit comparable spatial variation. Mean household income ranges from approximately £30,400 in the Royal Docks to over £47,400 in Greenwich. Income disparities across neighbourhoods are a central driver of localized housing demand and frequently generate segmented housing markets within metropolitan regions (Can, 1992 ; Goodman and Thibodeau, 2003 ). Commuting accessibility displays similar spatial heterogeneity across the estuary corridor. Average travel times to employment centres range from approximately 3.5 minutes in Isle of Dogs to more than 10 minutes in Dartford and Swanscombe. Accessibility gradients of this type are commonly associated with spatial variation in housing prices because proximity to employment nodes influences household location decisions (Alonso, 1964 ; Glaeser et al., 2008 ). Environmental exposure variables highlight the climate vulnerability of the Thames Estuary housing market. More than half of the observations in the sample are located within high flood-risk zones. Certain policy units exhibit particularly concentrated exposure. Canvey Island and Isle of Grain consist entirely of properties classified as high flood-risk areas, reflecting their extremely low-lying coastal geography. Other districts such as Greenwich, Thamesmead, and Wandsworth to Deptford also contain a large share of properties located within high-risk flood zones. Previous studies have shown that such environmental hazards can influence housing prices when buyers incorporate expected exposure into their valuation decisions (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). Elevation statistics further illustrate the geographic vulnerability of the region. Mean elevation values range between approximately 1.9 and 8.2 metres above sea level across policy units, with several areas including Canvey Island and Thamesmead situated at extremely low elevations. Low-lying topography increases exposure to tidal flooding and storm surge events, particularly in estuarine environments subject to sea-level rise. Distance to river frontage also varies substantially across the sample, capturing differences in both potential amenity value and environmental exposure. Empirical studies frequently find that proximity to water generates a complex valuation trade-off in housing markets, reflecting the coexistence of waterfront amenities and flood risk (Bin and Polasky, 2004 ; Dumm et al., 2016 ). The pronounced variation in housing prices, structural attributes, socioeconomic conditions, and environmental exposure across TE2100 policy units indicates that the capitalization of flood risk is unlikely to be spatially uniform across the Thames Estuary housing market. Conventional global hedonic models impose constant marginal effects across the study region and therefore cannot capture localized differences in environmental risk pricing. Spatial modelling approaches that allow housing price relationships to vary geographically are therefore necessary to identify how climate exposure influences property values across estuarine submarkets. The following section outlines the empirical framework used to estimate these spatially heterogeneous price effects. . Table 2 a — Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Structural Controls) Target Variable Structural Controls Price (£) Total Floor Area (m 2 ) Property Types – N (%) Mean (SD) (Min: Max) Mean (SD) (Min, Max) Bungalow Flat House Maisonette Barking & Dagenham 281,503.32 (148,046.9) (100: 9,345,000) 78.59 (24.95) (7.80: 297) 46 (0.6%) 1,467 (18%) 6,630 (80%) 165 (2.0%) Canvey Island 254,154.92 (93,459.44) (500: 1,300,000) 84.55 (38.68) (13: 875) 3,582 (47%) 86 (1.1%) 3,927 (52%) 27 (0.4%) Dartford & Erith 272,889.93 (332,545.9) (6,400: 21,812,419) 77.00 (25.70) (18: 395) 2 (< 0.1%) 1,952 (36%) 3,344 (62%) 125 (2.3%) East Tilbury & Mucking Mashes 275,865.39 (90,648.97) (55,000: 750,000) 87.03 (32.08) (36: 442) 0 (0%) 9 (2.3%) 384 (98%) 0 (0%) Greenwich 546,806.77 (280,278.1) (6,500: 3,500,000) 89.78 (33.99) (14.00: 304.00) 1 (< 0.1%) 673 (25%) 1,814 (67%) 226 (8.3%) Isle of Dogs & Lea Valley 455,572.21(452,208.17) (100: 17,220,000) 77.54 (34.95) (10.00: 811.43) 3 (< 0.1%) 1,919 (53%) 1,171 (32%) 519 (14%) Isle of Grain 420,000.00 () (420,000: 420,000) 168 () (168: 168) 0 (0%) 0 (0%) 1 (100%) 0 (0%) Leigh Old Town and Southend-on-Sea 424,062.29 (261,574.4) (4,500: 2,425,000) 119.80 (68.40) (5.24: 2,148.00) 580 (19%) 600 (19%) 1,856 (60%) 50 (1.6%) London City 585,690.23 (337,081.9) (29,465: 3,500,000) 74.44 (36.40) (23.00, 375.13) 0 (0%) 318 (50%) 237 (37%) 80 (13%) North Kent 247,879.84 (92,278.95) (10,000: 590,000) 82.20 (22.29) (25.10, 336.00) 59 (6.0%) 46 (4.6%) 870 (88%) 16 (1.6%) Purfleet, Grays & Tilbury 231,804.16 (723,185.7) (15,000:40,779,191) 72.09 (23.92) (18.47: 520.00) 60 (0.9%) 2,027 (32%) 4,213 (66%) 121 (1.9%) Rainham Marshes 314,959.77 (107,789.2) (26,730: 670,000) 86.57 (23.57) (33.48: 217.00) 434 (15%) 238 (8.1%) 2,155 (73%) 116 (3.9%) Royal Docks 322,365.66 (202,762.9) (100: 10,080,000) 79.24 (28.94) (8.04: 1,570.00) 41 (0.3%) 3,382 (25%) 9,554 (70%) 649 (4.8%) Shell Haven & Fobbing Marshes 237,751.64 (88,261.16) (5,000: 630,000) 76.35 (26.40) (36.00: 216.00) 4 (1.4%) 51 (18%) 220 (77%) 12 (4.2%) Swanscombe & Northfleet 359,346.91 (114,576.6) (64,995: 750,000) 99.86 (29.32) (37.10: 212.00) 0 (0%) 4 (5.8%) 63 (91% 2 (2.9%) Thamesmead 263,650.16 (192,867.7) (3,250: 11,915,135) 74.76 (24.23) (4.80: 503.00) 100 (1.6%) 1,450 (23%) 4,648 (73%) 160 (2.5%) Wandsworth to Deptford 523,421.35 (649,990.5) (500: 57,285,700) 80.83 (39.54) (4.70: 1,409.00) 10 (< 0.1%) 4,438 (42%) 4,929 (47%) 1,173 (11%) Overall (N = 73,039) 342,552.65 (397,875.3) (100: 57,285,700) 81.13 (34.55) (4.70: 2,148.00) 4,922 (6.7%) 18,660 (26%) 46,016 (63%) 3,441 (4.7%) Table 2 b — Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Neighbourhood Controls) Structural Controls Neighbourhood Controls Energy Ratings (A - G) – N(%) Net Annual Income A B C D E F G Mean (SD) (Min: Max) Barking & Dagenham 9 (0.1%) 544 (6.5%) 1,738 (21%) 4,264 (51%) 1,428 (17%) 213 (2.6%) 112 (1.3%) 31,946.13 (2,148.35) (25,985: 35,072) Canvey Island 5 (< 0.1%) 166 (2.2%) 1,218 (16%) 4,169 (55%) 1,685 (22%) 305 (4.0%) 74 (1.0%) 32,986.66 (1,093.06) (31,105: 34,650) Dartford & Erith 61 (1.1%) 3,100 (57%) 912 (17%) 1,020 (19%) 263 (4.8%) 54 (1.0%) 13 (0.2%) 34,554.92 (1,478.35) (31,592: 42,170) East Tilbury & Mucking Mashes 0 (0%) 32 (8.1%) 76 (19%) 209 (53%) 69 (18%) 4 (1.0%) 3 (0.8%) 36,118.90 (150.19) (36,088: 36,847) Greenwich 9 (0.3%) 389 (14%) 636 (23%) 1,128 (42%) 454 (17%) 61 (2.2%) 37 (1.4%) 47,401.68 (4,299.22) (32,610: 50,012) Isle of Dogs & Lea Valley 1 (< 0.1%) 271 (7.5%) 1,836 (51%) 1,225 (34%) 233 (6.5%) 40 (1.1%) 6 (0.2%) 36,363.32 (5,050.86) (26,517: 48,749) Isle of Grain 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (100%) 0 (0%) 0 (0%) 34,633.00 (NA) (34,633: 34,633) Leigh Old Town and Southend-on-Sea 0 (0%) 72 (2.3%) 377 (12%) 1,229 (40%) 1,028 (33%) 336 (11%) 44 (1.4%) 38,613.31 (6,840.70) (25,521: 43,932) London City 0 (0%) 14 (2.2%) 324 (51%) 235 (37%) 35 (5.5%) 19 (3.0%) 8 (1.3%) 44,937.11 (7,251.50) (33,051: 53,735) North Kent 0 (0%) 151 (15%) 158 (16%) 488 (49%) 164 (17%) 20 (2.0%) 10 (1.0%) 33,059.00 (0.00) (33,059: 33,059) Purfleet, Grays & Tilbury 33 (0.5%) 1,211 (19%) 1,771 (28%) 2,554 (40%) 698 (11%) 107 (1.7%) 47 (0.7%) 30,560.50 (2,176.55) (27,850: 35,752) Rainham Marshes 0 (0%) 384 (13%) 508 (17%) 1,340 (46%) 604 (21%) 83 (2.8%) 24 (0.8%) 38,792.09 (1,214.62) (38,297: 43,106) Royal Docks 3 (< 0.1%) 753 (5.5%) 3,520 (26%) 6,553 (48%) 2,205 (16%) 376 (2.8%) 216 (1.6%) 30,426.11 (4,919.76) (22,035: 46,930.00) Shell Haven & Fobbing Marshes 1 (0.3%) 1 (0.3%) 74 (26%) 111 (39%) 83 (29%) 14 (4.9%) 3 (1.0%) 35,735.44 (1,169.12) (31,872: 36,088) Swanscombe & Northfleet 0 (0%) 50 (72%) 2 (2.9%) 6 (8.7%) 8 (12%) 2 (2.9%) 1 (1.4%) 30,890.20 (2,480.23) (26,896: 32,408.00) Thamesmead 1 (< 0.1%) 203 (3.2%) 2,457 (39%) 2,635 (41%) 859 (14%) 165 (2.6%) 38 (0.6%) 34,340.62 (2,122.20) (30,572: 52,998.00) Wandsworth to Deptford 14 (0.1%) 686 (6.5%) 4,071 (39%) 4,130 (39%) 1,287 (12%) 257 (2.4%) 105 (1.0%) 40,797.65 (5,493.56) (33,608: 63,550.00) Overall (N = 73,039) 137 (0.2%) 8,027 (11%) 19,678 (27%) 31,296 (43%) 11,104 (15%) 2,056 (2.8%) 741 (1.0%) 34,844.72 (5,900.96) (22,035: 63,550.00) Table 2 c — Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Climate & Environment Variables) Neighbourhood Controls Climate and Environment Variables Total Time to Work (Min) Flood Risk Levels – N (%) Elevation (m) Distance to River (m) Mean (SD) (Min: Max) No Risk Mid Risk High Risk Mean (SD) (Min: Max) Mean (SD) (Min: Max) Barking & Dagenham 5.36 (2.82) (2.30, 12.22) 5,919 (71%) 444 (5.3%) 1,945 (23%) 5.14 (1.65) (-2.30: 11.90) 750.18 (400.96) (31.27: 2,076.7) Canvey Island 6.73 (3.22) (2.88, 16.27) 0 (0%) 0 (0%) 7,622 (100%) 1.87 (0.37) (-0.40: 3.80) 551.44(396.45) (10.31: 1,487.5) Dartford & Erith 10.76 (3.65) (2.98, 17.56) 2,085 (38%) 1,000 (18%) 2,338 (43%) 6.13 (1.71) (-2.20, 12.00) 425.61 (289.56) (27.49: 1,145.6) East Tilbury & Mucking Mashes 6.05 (3.92) (3.00, 12.77) 293 (75%) 100 (25%) 0 (0%) 6.77 (1.61) (-2.10, 9.90) 230.10 (146.48) (34.80: 914.40) Greenwich 4.84 (2.15) (1.66, 8.24) 777 (29%) 237 (8.7%) 1,700 (63%) 4.72 (2.03) (-2.30, 17.70) 602.12 (336.60) (31.97: 1,281.9) Isle of Dogs & Lea Valley 3.49 (1.74) (1.07, 9.35) 847 (23%) 387 (11%) 2,378 (66%) 4.10 (1.45) (-2.30, 9.60) 436.95 (234.03) (18.26, 1,042.34) Isle of Grain 9.71 (NA) (9.71, 9.71) 0 (0%) 0 (0%) 1 (100%) 5.50 (NA) (5.50, 5.50) 0.07 (NA) (0.07 0.07) Leigh Old Town and Southend-on-Sea 4.93 (1.78) (2.28, 9.49) 1,792 (58%) 892 (29%) 402 (13%) 5.79 (2.56) (0.90, 13.00) 660.91 (333.25) (13.98, 1,488.1) London City 4.80 (1.81) (3.10, 16.36) 69 (11%) 20 (3.1%) 546 (86%) 4.26 (1.05) (2.50, 9.70) 405.65 (116.00) (174.29, 614.37) North Kent 8.56 (3.70) (4.18, 12.74) 438 (44%) 28 (2.8%) 525 (53%) 5.83 (2.17) (2.80, 11.60) 798.98 (276.88) (166 1,264.65) Purfleet, Grays & Tilbury 5.95 (2.76) (2.05, 10.68) 2,347 (37%) 1,260 (20%) 2,814 (44%) 5.22 (4.22) (0.00, 16.50) 525.29 (275.20) (29.32, 1,211.8) Rainham Marshes 7.57 (3.39) (2.90, 12.44) 1,875 (64%) 297 (10%) 771 (26%) 5.08 (1.80) (1.50, 8.80) 481.38 (249.69) (19.70, 1,118.2) Royal Docks 3.77 (1.55) (1.28, 8.07) 5,164 (38%) 1,583 (12%) 6,879 (50%) 3.62 (1.97) (-2.30, 9.70) 1,251.34 (601.1) (27.49, 2,581.1) Shell Haven & Fobbing Marshes 9.25 (2.73) (3.92, 13.48) 228 (79%) 24 (8.4%) 35 (12%) 7.42 (2.28) (-2.10, 11.30) 113.11 (127.39) (20.23, 515.43) Swanscombe & Northfleet 10.06 (0.02) (10.05, 10.11) 65 (94%) 0 (0%) 4 (5.8%) 8.17 (1.38) (4.30, 9.30) 635.34 (90.02) (415.62, 718.92) Thamesmead 5.74 (3.11) (1.76, 12.33) 1,756 (28%) 236 (3.7%) 4,366 (69%) 3.17 (2.67) (-2.30, 12.30) 471.56 (272.36) (14.75, 1,278.6) Wandsworth to Deptford 4.59 (2.91) (1.07, 18.15) 1,620 (15%) 793 (7.5%) 8,137 (77%) 3.51 (1.41) (-2.30, 13.50) 1,334.57 (825.6) (27.67, 3,288.6) Overall (N = 73,039) 5.50 (3.18) (1.07, 18.15) 25,275 (35%) 7,301 (10.0%) 40,463 (55%) 4.17 (2.46) (-2.30, 17.70) 795.87 (604.67) (0.07, 3,288.60) 4 Empirical Strategy: Identifying Spatial Heterogeneity in Climate Risk Pricing 4.1 Baseline Hedonic Specification of Housing Price Formation The empirical analysis begins with a conventional hedonic price specification that relates residential property values to structural characteristics, neighbourhood attributes, and environmental exposure variables. Hedonic theory interprets housing prices as equilibrium outcomes reflecting the implicit prices of dwelling attributes in competitive housing markets (Rosen, 1974 ). Buyers select properties based on bundles of characteristics, and the observed transaction price reveals the marginal willingness to pay for these attributes. Structural features such as floor area and property type capture housing consumption value, while neighbourhood characteristics reflect local amenities, accessibility, and socioeconomic conditions (Sirmans et al., 2005 ). Environmental exposure variables are incorporated to examine whether housing markets internalize climate-related risks associated with the Thames Estuary floodplain. Flood risk indicators, elevation, and proximity to tidal waterways represent potential sources of environmental disamenity that may influence housing demand through expected damage risk, insurance costs, or perceived vulnerability. Previous research demonstrates that such environmental hazards may be capitalized into property values when buyers incorporate risk information into location decisions (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). The baseline hedonic specification is expressed as $$\:\text{l}\text{n}\left({P}_{i}\right)={\beta\:}_{0}+{\beta\:}_{1}{S}_{i}+{\beta\:}_{2}{N}_{i}+{\beta\:}_{3}{E}_{i}+{ϵ}_{i}$$ where \(\:{P}_{i}\) denotes the transaction price of property \(\:i\) , \(\:{S}_{i}\) represents structural housing attributes, \(\:{N}_{i}\) denotes neighbourhood characteristics, and \(\:{E}_{i}\) captures environmental exposure variables including flood zone classification, elevation, and distance to river frontage. The dependent variable is specified in logarithmic form to account for skewness in housing price distributions and to allow estimated coefficients to be interpreted as approximate percentage changes in property values (Malpezzi, 2002 ). Structural attributes include total floor area, number of habitable rooms, and property type indicators, which capture differences in dwelling size and building form. Neighbourhood characteristics incorporate measures of commuting accessibility and local income levels that reflect spatial variation in housing demand. Environmental variables represent both physical exposure to flood hazards and proximity to estuarine amenities. Distance to river frontage may therefore capture competing valuation effects, as waterfront proximity can simultaneously generate amenity benefits and increased exposure to flooding risk. Estimating the baseline hedonic model provides an initial assessment of how environmental exposure variables correlate with housing prices across the Thames Estuary housing market. However, the specification imposes constant marginal effects across the entire study region. This assumption may be restrictive in estuarine environments where flood exposure, topography, and neighbourhood characteristics vary significantly across short geographic distances. If the marginal impact of environmental variables differs across locations, global hedonic estimates may mask important spatial variation in climate risk capitalization. The next subsection evaluates whether spatial dependence is present in the housing price residuals and therefore whether spatial modelling approaches are required. 4.2 Testing for Spatial Dependence in Housing Price Residuals Housing markets are inherently spatial systems in which property values tend to cluster geographically. Properties located near one another frequently share similar neighbourhood characteristics, environmental conditions, and accessibility to amenities. These spatial interactions often generate dependence in housing prices that violates the independence assumptions underlying conventional regression models (Anselin, 1988 ; Pace and LeSage, 2004 ). When such spatial dependence is present, global hedonic models may produce inefficient estimates and mask localized price dynamics. To evaluate whether spatial dependence exists in the baseline hedonic model, the analysis examines the spatial autocorrelation structure of the regression residuals. Spatial autocorrelation measures the extent to which nearby observations exhibit similar values relative to those located further apart. Positive spatial autocorrelation occurs when geographically proximate properties exhibit similar pricing patterns, while negative spatial autocorrelation arises when neighbouring values differ systematically. The presence of spatial dependence is tested using Moran’s I statistic, which provides a global measure of spatial autocorrelation. The statistic is defined as $$\:I=\frac{n}{W}\frac{\sum\:_{i=1}^{n}\sum\:_{j=1}^{n}{w}_{ij}({e}_{i}-\stackrel{\prime }{e})({e}_{j}-\stackrel{\prime }{e})}{\sum\:_{i=1}^{n}({e}_{i}-\stackrel{\prime }{e}{)}^{2}}$$ where \(\:{e}_{i}\) denotes the residual from the baseline hedonic regression for property \(\:i\) , \(\:\stackrel{\prime }{e}\) represents the mean residual, \(\:{w}_{ij}\) is the spatial weight describing the proximity between observations \(\:i\) and \(\:j\) , \(\:n\) denotes the number of observations, and \(\:W\) is the sum of all spatial weights. Positive values of Moran’s I indicate spatial clustering of similar residual values, while values close to zero suggest spatial randomness. Significant spatial autocorrelation in the residuals would imply that the baseline hedonic specification fails to capture important spatial processes influencing housing prices. In the context of the Thames Estuary housing market, such spatial dependence may arise from localized flood exposure, neighbourhood-level amenities, or spatial spillovers in housing demand. Previous studies have shown that ignoring spatial dependence can bias estimates of housing price determinants and obscure localized price relationships (Pace and LeSage, 2004 ; Anselin and Rey, 2014 ). Evidence of spatial dependence motivates the use of modelling approaches capable of capturing spatial heterogeneity in housing price relationships. Global spatial models such as spatial lag or spatial error specifications account for spatial dependence but continue to impose constant parameter estimates across the study region. Estuarine housing markets characterized by substantial geographic variation in environmental exposure may require more flexible approaches that allow coefficients to vary locally across space. Geographically weighted regression provides such a framework by estimating location-specific parameter estimates using spatially weighted observations (Fotheringham et al., 2002 ). The following section introduces the GWR model used to identify spatial variation in the capitalization of climate risk within the Thames Estuary housing market. 4.3 Geographically Weighted Regression for Spatially Varying Housing Price Effects Evidence of spatial dependence in the residuals of the baseline hedonic model suggests that housing price relationships may vary geographically across the Thames Estuary region. Estuarine housing markets combine heterogeneous environmental exposure, varying neighbourhood conditions, and localized demand patterns. Under such conditions, the marginal effects of environmental attributes such as flood exposure or elevation are unlikely to remain constant across space. Global hedonic models impose uniform coefficients across the entire study region and therefore cannot capture localized differences in the capitalization of environmental risk. Geographically weighted regression (GWR) provides a modelling framework that allows regression coefficients to vary across geographic locations. The approach estimates a local regression equation at each observation point using spatially weighted neighbouring observations. This framework enables the identification of spatially varying relationships between housing prices and property characteristics while maintaining the interpretability of hedonic modelling structures (Fotheringham et al., 2002 ). Instead of producing a single global coefficient for each explanatory variable, GWR generates location-specific parameter estimates that reflect local housing market conditions. The GWR model can be expressed as: $$\:\text{l}\text{n}\left({P}_{i}\right)={\beta\:}_{0}({u}_{i},{v}_{i})+\sum\:_{k=1}^{K}{\beta\:}_{k}({u}_{i},{v}_{i}){X}_{ik}+{ϵ}_{i}$$ where \(\:{P}_{i}\) denotes the transaction price of property \(\:i\) , \(\:{X}_{ik}\) represents the set of explanatory variables including structural, neighbourhood, and environmental characteristics, and \(\:\left({u}_{i},{v}_{i}\right)\) denotes the geographic coordinates of property \(\:i\) . The parameters \(\:{\beta\:}_{k}({u}_{i},{v}_{i})\) vary across space, allowing the marginal impact of housing attributes to differ by location. This formulation enables the estimation of localized housing price functions that reflect spatial heterogeneity in market behaviour. Parameter estimation in GWR relies on spatial weighting schemes that assign greater influence to observations located closer to the regression point. Observations further away receive lower weights according to a kernel function. The weighting structure is defined by $$\:{w}_{ij}=K\left(\frac{{d}_{ij}}{b}\right)$$ where \(\:{w}_{ij}\) represents the spatial weight between observations \(\:i\) and \(\:j\) , \(\:{d}_{ij}\) denotes the geographic distance between the two observations, \(\:b\) is the bandwidth parameter controlling the spatial extent of the kernel, and \(\:K(\cdot\:)\) represents the kernel function. Kernel weighting ensures that parameter estimates at each location primarily reflect local housing market conditions rather than distant observations. Bandwidth selection is a critical component of the GWR estimation procedure. The bandwidth determines the spatial scale over which local relationships are estimated. Smaller bandwidths capture highly localized variation but may increase estimation variance, while larger bandwidths approximate global models by incorporating broader spatial information. The optimal bandwidth is typically selected using cross-validation or information criteria that balance model fit and parameter stability (Fotheringham et al., 2002 ). Applying GWR to the Thames Estuary housing market allows the empirical analysis to identify spatially varying capitalization patterns associated with flood exposure, elevation, and proximity to the river. Estuarine housing markets often exhibit localized trade-offs between waterfront amenities and environmental risk. Properties located close to the river may benefit from amenity value in some neighbourhoods while facing flood exposure penalties in others. Local parameter estimation enables the identification of these heterogeneous valuation effects that remain hidden in global regression models. Mapping the estimated coefficients from the GWR model provides insight into the geographic structure of climate risk capitalization across the study region. Spatial variation in the coefficients associated with flood risk variables can reveal where housing markets internalize environmental exposure more strongly and where such risks appear underpriced. These localized estimates therefore provide a spatially explicit perspective on how climate exposure influences housing values within the Thames Estuary. 5 Empirical Results: Spatial Capitalization of Flood Risk in the Thames Estuary 5.1 Spatial Dependence in Housing Prices Table 3 reports the global spatial diagnostics for the residuals of the baseline hedonic specification. The evidence rejects the view that the remaining pricing errors are spatially random. Moran’s \(\:I\) equals 0.1397 and is highly significant \(\:\left(z=75.653,\text{\hspace{0.25em}\hspace{0.05em}}p<0.001\right)\) , indicating strong positive residual autocorrelation. Properties located close to one another therefore share systematically similar unexplained price components, even after controlling for structural characteristics, neighbourhood conditions, and flood-related variables. This result is economically important. It implies that the baseline OLS model does not fully absorb the spatial organization of housing values in the Thames Estuary, where transaction prices are shaped by clustered amenities, local reputation effects, micro-location advantages, and environmental conditions that are not evenly distributed across space. Such residual clustering is a classic signal of misspecification in housing price models estimated on spatially embedded markets (Anselin, 1988 ; Pace and LeSage, 2004 ; LeSage and Pace, 2009 ). The Rao’s Score diagnostics show that this spatial structure is not confined to a single mechanism. Both the LM Error and LM Lag statistics are highly significant, and the robust versions of both tests remain highly significant as well. This pattern is more demanding than a simple finding of residual autocorrelation. A significant lag process suggests that prices in one location are partly anchored by prices observed in nearby locations, which is consistent with the role of comparable transactions, neighbourhood search, and localized market learning in residential price formation. A significant error process indicates that omitted influences are spatially clustered, such as local environmental quality, flood-defence credibility, waterfront redevelopment intensity, or neighbourhood-specific demand conditions. The coexistence of both processes is therefore consistent with a housing market in which spatial spillovers and omitted local fundamentals jointly shape valuation outcomes, a result that has been widely emphasized in spatial housing and urban econometric research (Can, 1992 ; Anselin and Rey, 2014 ; Holly, Pesaran and Yamagata, 2011 ) This matters directly for inference. Once residuals are spatially correlated, the classical OLS assumption of independently distributed disturbances is violated. Coefficient estimates may remain interpretable as conditional associations, but standard errors and significance tests become unreliable, and the model no longer provides a sufficient representation of the underlying price-generating process (Anselin, 1988 ; LeSage and Pace, 2009 ). In the present setting, the diagnostics indicate that global hedonic estimates are averaging across estuarine submarkets that differ sharply in flood exposure, topography, accessibility, and adaptation context. A single global coefficient on flood risk or river proximity is therefore unlikely to represent the true structure of capitalization across the study region. The diagnostics also have a substantive implication for the paper’s central argument. If the residual spatial process remains strong after controlling for flood zones, elevation, river distance, income, commuting accessibility, and housing structure, then climate-risk capitalization is unlikely to be spatially homogeneous. Global models may detect an average relationship, but they cannot reveal whether flood exposure is discounted in some parts of the estuary and offset by amenity or redevelopment pressures in others. Table 3 therefore does more than justify a technical correction. It shows that the Thames Estuary housing market is characterized by spatially differentiated price formation, which requires both global spatial econometric benchmarks and local parameter estimation. The next section turns to SEM and SDM estimates before the analysis proceeds to geographically weighted regression, where the local structure of flood-risk capitalization can be identified explicitly. Table 3 Spatial Autocorrelation Diagnostics Moran’s I Test for Residual Spatial autocorrelation Statistic Value Moran’s I 0.1397 Expected I -0.00038 Variance 0.00000343 z-value 75.653 p-value < 0.001 Rao’s Score (Lagrange Multiplier) Tests Test Statistic p-value LM Error 5674.51 < 0.001 LM Lag 5929.91 < 0.001 Robust LM Error 583.82 < 0.001 Robust LM Lag 839.23 < 0.001 Notes : Moran’s I tests for global spatial autocorrelation in the residuals of the baseline hedonic regression. Rao’s Score (Lagrange Multiplier) diagnostics examine the presence of spatial lag and spatial error dependence. Robust tests control for the presence of the alternative spatial process. All statistics are calculated using the spatial weights matrix \(\:\varvec{W}\) . 5.2 Global Hedonic and Spatial Econometric Estimates Table 4 reports the baseline hedonic estimates together with the Spatial Error Model (SEM) and the Spatial Durbin Model (SDM). The results confirm that the fundamental structure of the housing price equation behaves consistently with the theoretical predictions of hedonic price models. Total floor area exhibits a strong and highly significant positive association with transaction prices across all specifications. Larger dwellings command systematically higher values, reflecting the fundamental role of interior living space in residential price formation. Property type effects follow expected patterns as well. Houses sell at a premium relative to the reference category, while flats exhibit negative coefficients in the global OLS specification. These patterns align with the long-established empirical literature on housing valuation, which consistently finds structural dwelling attributes to be among the most important determinants of property prices (Rosen, 1974 ; Sirmans et al., 2005 ). Neighbourhood attributes also display economically coherent relationships with housing prices. Local income levels show a positive and statistically significant association with transaction values across all models, indicating that higher-income areas command systematically higher property prices. This pattern reflects the role of neighbourhood purchasing power and local demand conditions in shaping residential land values, a central prediction of urban bid–rent theory (Alonso, 1964 ; Glaeser et al., 2008 ). Accessibility also plays a measurable role. Travel time to employment centres is negatively associated with property prices, suggesting that longer commuting distances impose a spatial discount on housing values. This finding is consistent with the long-standing observation that proximity to employment and transport networks remains a key driver of urban residential location choices (Muth, 1969 ; Gibbons and Machin, 2008 ). Environmental variables provide the first indication of climate-related capitalization in the global specification. Elevation exhibits a positive and statistically significant coefficient across all models, indicating that properties located at higher ground levels command measurable price premiums. This result is consistent with the hypothesis that buyers internalize physical flood exposure when forming housing valuations. Distance to river channels also enters positively, suggesting that properties located farther from tidal waterways tend to command higher values once structural and neighbourhood characteristics are controlled for. These effects imply that aspects of physical flood exposure are reflected in market prices even within the global specification. However, the coefficients for the flood-zone indicators reveal a more complex pattern. Properties located within both medium and high flood-risk zones exhibit positive and statistically significant coefficients in the global models. At face value, this suggests that properties exposed to flood risk are associated with higher transaction prices. Such a result appears counterintuitive relative to the expectation that environmental hazards should generate valuation discounts. Similar findings have occasionally been reported in coastal housing markets where the amenity value of waterfront proximity dominates perceived hazard exposure (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). In such settings, households may accept environmental risk in exchange for locational advantages such as waterfront views, proximity to urban cores, or redevelopment opportunities. The spatial econometric estimates confirm that housing price formation in the Thames Estuary exhibits strong spatial interaction. The SEM reports a highly significant spatial error coefficient \(\:\left(\lambda\:=0.1277\right)\) , indicating that unobserved spatially correlated factors influence housing prices beyond the explanatory variables included in the model. These factors may include localized environmental amenities, flood-protection infrastructure, neighbourhood reputation effects, or other spatially clustered determinants of residential desirability. The SDM similarly detects spatial dependence through the spatial autoregressive parameter, implying that housing prices are partially influenced by prices observed in nearby locations. Such spatial spillovers reflect the role of comparable transactions and local information diffusion in shaping residential market expectations (LeSage and Pace, 2009 ; Holly et al., 2011 ). Although these spatial econometric models address the residual spatial dependence identified in Table 3 , they retain an important restriction: marginal effects remain constant across space. In a coastal housing system characterized by heterogeneous flood exposure, elevation gradients, and localized adaptation policies, this assumption may obscure substantial variation in how environmental risk is capitalized into housing prices. The positive flood-risk coefficients observed in the global models may therefore reflect aggregation across submarkets in which flood exposure is discounted in some locations but offset by amenity and redevelopment effects in others. Identifying such spatial heterogeneity requires a modelling framework that allows price gradients to vary across locations. The following section therefore applies geographically weighted regression to estimate local housing price relationships and to examine how flood-risk capitalization differs across the Thames Estuary. Table 4 Global Hedonic and Spatial Econometric Model Estimates Variables OLS SEM SDM 10.740*** (0.0137) 10.840*** (0.0316) 10.894*** (0.0134) Total Floor Area 0.00547*** (0.000054) 0.00409*** (0.000059) 0.00548*** (0.000054) Property Type: Flat -0.0514*** (0.00703) -0.2309*** (0.00915) -0.0513*** (0.00464) House 0.1537*** (0.00654) 0.0858*** (0.00811) 0.1551*** (0.00406) Maisonette 0.0103 (0.00991) -0.1773*** (0.0115) 0.0116 Income After Housing 3.37e-05*** (3.03e-07) 3.58e-05*** (3.03e-07) 3.43e-05*** (3.03e-07) Travel Time to Employment -0.0143*** (0.000545) -0.0124*** (0.00136) -0.0145*** (0.000535) Flood Risk Medium 0.0529*** (0.00654) 0.0691*** (0.0146) 0.0523*** (0.00654) High 0.1001*** (0.00547) 0.1106*** (0.0129) 0.1004*** (0.00527) Elevation 0.0137*** (0.00102) 0.0163*** (0.00248) 0.0137*** (0.000971) Distance to River 0.000135*** (2.94e-06) 0.000143*** (7.27e-06) 0.000136*** (2.96e-06) Spatial Parameters & Model Diagnostics Spatial Coeff -0.00148*** (Rho, ρ) 0.1277*** (Lambda, λ) Log-Likelihood - -1999.95 -35.586.96 AIC 71,850 4,025.9 71,200 R² 0.382 - - Observations 64,947 64,947 64,947 Note : Dependent variable is the logarithm of transaction price. Standard errors are reported in parentheses. The Spatial Error Model (SEM) accounts for spatial correlation in the regression error term, while the Spatial Durbin/Lag Model incorporates spatial dependence in the dependent variable through the spatial autoregressive parameter \(\:\rho\:\) . Spatial weights are constructed using a symmetric binary matrix based on the k = 8 nearest neighbours to maintain model stability across 64,947 observations. Due to the Sparse Matrix (Matrix) estimation method, the SEM Log-Likelihood and AIC operate on a different numerical scale than the OLS; model selection is therefore guided by the significance of the spatial parameters and the Likelihood Ratio (LR) test. Significance levels: *** p < 0.01, ** p < 0.05, * p < 0.10. 5.3 Spatial Heterogeneity in Flood-Risk Capitalization Figure 1 maps the geographically weighted regression estimates of the local coefficient on high flood exposure across the Thames Estuary housing market. The results reveal substantial spatial heterogeneity in the capitalization of flood risk. While the global models in Table 4 produce a single positive coefficient for flood-zone exposure, the local estimates vary widely across space, ranging from strongly negative values in the outer estuary to positive values in parts of the western urban corridor. The global estimate therefore masks the coexistence of opposing valuation processes operating within different submarkets. The eastern estuary displays predominantly negative coefficients. Areas such as Leigh-on-Sea and the Isle of Grain exhibit some of the largest discounts associated with flood exposure. In these locations, the housing market appears to internalize environmental risk directly, with buyers requiring a measurable price concession to compensate for elevated flood probability. Such capitalization is consistent with the environmental risk literature, which shows that households discount property values when physical exposure to natural hazards becomes salient and difficult to insure or mitigate (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). In contrast, several western policy units exhibit positive coefficients, indicating that flood exposure is associated with higher property values. Locations along the Wandsworth–Deptford corridor and parts of Thamesmead display positive price effects despite regulatory classification within high-risk zones. These patterns suggest that waterfront amenities, transport accessibility, and urban redevelopment pressures may offset perceived environmental risk in high-demand urban submarkets. Housing markets frequently exhibit such amenity-risk trade-offs, particularly where coastal or riverfront locations combine environmental exposure with strong locational advantages (Glaeser et al., 2008 ; Belanger and Bourdeau-Brien, 2018 ). The spatial gradient observed in Fig. 1 therefore indicates that flood-risk capitalization is conditional on local market context rather than uniform across the estuary. In dense urban locations, amenity value and redevelopment expectations appear to dominate risk perception, producing positive capitalization effects. In peripheral coastal communities, physical exposure becomes the dominant factor shaping housing prices. Global hedonic and spatial econometric models cannot capture such variation because they impose constant marginal effects across space. Geographically weighted regression reveals that the average flood-risk coefficient reported in Table 4 reflects an aggregation of spatially heterogeneous pricing regimes rather than a uniform market response. This spatial heterogeneity has important implications for both empirical modelling and climate-risk policy. If environmental risk is capitalized differently across local housing markets, global price models may systematically misrepresent the distribution of climate exposure embedded in property values. The following section therefore examines spatial variation in model explanatory power to assess where the hedonic framework performs well and where additional unobserved factors may influence housing prices. 5.4 Spatial Variation in Model Explanatory Power Figure 2 maps the spatial distribution of the local coefficient of determination obtained from the geographically weighted regression. The results indicate substantial variation in the explanatory power of the hedonic specification across the Thames Estuary housing market. Local \(\:{R}^{2}\) values range from approximately 0.18 in several redevelopment zones to values exceeding 0.60 in established residential markets. Higher explanatory power is observed in mature residential areas such as Leigh-on-Sea and Wandsworth. In these locations, structural housing characteristics, neighbourhood income, commuting accessibility, and environmental variables jointly explain a large share of property price variation. Housing markets in these areas therefore appear to conform closely to the standard hedonic framework in which prices reflect the capitalization of structural and locational attributes (Rosen, 1974 ; Sirmans et al., 2005 ). The strong explanatory power in these zones suggests that market participants consistently price the observable attributes included in the model. Lower local \(\:{R}^{2}\) values emerge in areas undergoing rapid urban transformation, particularly within parts of the Royal Docks redevelopment corridor. In these locations, a substantial portion of housing price variation remains unexplained by the structural, neighbourhood, and environmental variables included in the model. One plausible explanation is the presence of additional valuation drivers not captured in the hedonic specification, such as speculative expectations linked to regeneration projects, future transport infrastructure, or major waterfront redevelopment schemes. Empirical studies of urban regeneration areas frequently document similar patterns, where large-scale planning interventions alter property prices in ways that are only partially explained by conventional housing attributes (Gibbons and Machin, 2008 ; Ahlfeldt et al., 2017 ). The spatial pattern of model fit therefore reinforces the argument that housing price formation in the Thames Estuary is heterogeneous across submarkets. In established residential areas, observable structural and environmental attributes explain most price variation. In rapidly transforming districts, unobserved development dynamics play a more prominent role. This spatial variation also helps explain the heterogeneous flood-risk capitalization identified in Fig. 1 . Where redevelopment expectations dominate price formation, environmental risk may exert a weaker influence on observed transaction prices. 5.5 Spatial Clustering of Flood-Risk Pricing Figure 3 presents the Local Indicators of Spatial Association (LISA) cluster map for the geographically weighted regression estimates of the flood-risk coefficient. The map identifies statistically significant spatial clusters in the relationship between flood exposure and property values across the Thames Estuary housing market. Unlike the coefficient map in Fig. 1 , which illustrates spatial variation in local price effects, the LISA analysis reveals whether similar pricing responses to flood risk are geographically concentrated. Two dominant spatial regimes emerge. First, the eastern estuary exhibits low–low clusters, where negative flood-risk coefficients are spatially concentrated. In these locations, flood exposure is systematically associated with lower property prices, indicating that the housing market capitalizes environmental risk in a consistent manner. Coastal communities such as Leigh-on-Sea and surrounding estuarine settlements fall within this regime. The clustering of negative coefficients suggests that buyers in these markets respond directly to physical exposure conditions rather than relying primarily on amenity valuation. This pattern is consistent with studies documenting hazard-induced price discounts in flood-exposed housing markets where risk salience is high and insurance or adaptation mechanisms are limited (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). Second, the western urban corridor contains high–high clusters, where positive flood-risk coefficients occur in spatially contiguous areas. These clusters indicate that properties exposed to flood risk command higher prices in nearby locations as well. Such outcomes suggest that waterfront amenities, accessibility advantages, and urban redevelopment expectations dominate risk perceptions in these high-demand submarkets. The presence of these clusters highlights the spatial concentration of amenity-driven valuation processes, particularly in dense metropolitan environments where environmental risk is perceived as secondary to locational advantages. The coexistence of these clusters reveals that flood-risk capitalization is not randomly distributed across the housing market. Instead, distinct spatial regimes characterize how environmental exposure enters housing price formation. In outer estuary markets, flood risk operates primarily as a negative externality that reduces property values. In central urban areas, the same environmental classification coexists with positive price effects due to the concentration of waterfront amenities and development pressures. Global hedonic models cannot capture such regime differentiation because they impose uniform marginal effects across space. Identifying these spatial clusters also has policy implications. If environmental risk is systematically discounted in some regions but not in others, property prices may fail to fully signal underlying climate exposure in high-demand urban submarkets. This creates the potential for localized misalignment between market valuation and environmental risk, particularly in redevelopment corridors where waterfront proximity remains a major determinant of housing demand. 5.6 Flood-Risk Capitalization across TE2100 Policy Management Units Table 5 summarizes the geographically weighted regression results at the level of the Thames Estuary 2100 (TE2100) Policy Management Units (PMUs). The table reports the median local flood-risk coefficient, the median local model fit, and the proportion of statistically significant coefficients within each policy unit. Aggregating the local estimates in this way allows the analysis to examine whether climate-risk capitalization differs systematically across the estuary’s flood-management governance structure. The results reveal substantial variation in flood-risk capitalization across policy units. In the eastern estuary, several units exhibit strongly negative median flood-risk coefficients. The Isle of Grain and Leigh Old Town display median coefficients of approximately − 0.30, indicating that properties located within high-risk flood zones are systematically discounted relative to comparable properties outside those zones. These findings suggest that housing markets in these coastal communities internalize environmental exposure directly into property prices. Such patterns are consistent with empirical evidence from coastal housing markets where the salience of flood risk generates measurable price discounts once hazard exposure becomes widely recognized by market participants (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). In contrast, several western and central estuary units display positive flood-risk coefficients. Policy units such as Dartford and Erith and Thamesmead exhibit positive median coefficients, implying that properties located within designated flood-risk zones command higher prices on average. These outcomes appear counterintuitive if flood risk is interpreted purely as an environmental disamenity. However, the spatial context of these locations provides an alternative explanation. These areas benefit from strong transport accessibility, ongoing urban redevelopment, and proximity to major employment centres. In such markets, the amenity value of waterfront proximity and redevelopment potential may dominate perceived environmental exposure, producing positive capitalization effects despite regulatory risk classification. The significance rates reported in Table 5 indicate that these patterns are not isolated observations but reflect consistent spatial market behaviour. Units such as Greenwich and Canvey Island exhibit high proportions of statistically significant local coefficients, suggesting that the estimated flood-risk effects represent systematic valuation responses rather than random variation across properties. The results therefore reinforce the spatial heterogeneity identified in the preceding sections: flood-risk capitalization is strongly conditioned by local economic context. These findings carry important implications for climate-risk assessment in coastal property markets. If environmental risk is discounted in some locations but offset by amenity or redevelopment value in others, global price models may obscure important spatial variation in how climate exposure is incorporated into property values. Policy frameworks that rely on aggregate price signals may therefore underestimate localized exposure in high-demand urban waterfront markets while overstating risk capitalization in peripheral coastal communities. Table 5 Summary of Local Flood-Risk Effects and Model Performance by Policy Management Unit Policy Units (pmus) N_Properties Median_FZ3_Beta Median_Local_R2 Significance Rate Isle of Grain (P4) 1 -0.2982 0.446692 100 Leigh Old Town and Southend-on-Sea (P4) 3086 -0.29231 0.640405 100 Canvey Island (P4) 7622 -0.24814 0.434998 94.01732 Greenwich (P5) 1441 -0.05275 0.541409 98.47328 East Tilbury & Mucking Marshes (P3) 393 -0.04961 0.525775 93.89313 Shell Haven & Fobbing Marshes (P3) 287 -0.04933 0.500102 40.41812 London City (P5) 634 -0.04878 0.708452 72.87066 Isle of Dogs & Lea Valley (P5) 2782 -0.03542 0.532862 69.12293 Purfleet, Grays & Tilbury (P4) 5780 -0.01973 0.617526 36.6263 Royal Docks (P4) 12362 -0.00951 0.227243 8.574664 Swanscombe & Northfleet (P4) 19 -0.00448 0.586845 0 North Kent Marshes (P3) 991 -0.00329 0.57967 0 Wandsworth to Deptford (P5) 9530 0.007242 0.637411 69.91605 Barking & Dagenham (P4) 7757 0.034057 0.336622 47.325 Rainham Marshes (P4) 2943 0.054027 0.412096 59.70099 Thamesmead (P4) 6343 0.109222 0.370902 87.04083 Dartford & Erith (P4) 2976 0.156295 0.542342 100 Note : Data represents median values for all properties within the specified Policy Management Unit. The "Significance Rate" indicates the percentage of properties within that unit where the local flood risk impact is statistically significant \(\:\left|t\right|>1.96\) 6 Discussion The empirical results indicate that the capitalization of flood risk in the Thames Estuary housing market is spatially heterogeneous rather than uniform. Global hedonic and spatial econometric models produce a positive average coefficient for flood-zone exposure, suggesting that properties located within designated flood-risk areas command higher prices. At face value, such a result appears inconsistent with the expectation that environmental hazards reduce property values. However, the geographically weighted regression estimates demonstrate that this global coefficient reflects the aggregation of opposing spatial processes operating across the estuary. In the outer estuary and several coastal communities, flood exposure is associated with clear price discounts. These markets appear to internalize environmental risk directly, with buyers requiring compensation for properties located in high-risk areas. This pattern aligns with the environmental hazard literature, which documents significant capitalization of flood risk when exposure becomes salient and when protective infrastructure or insurance coverage is limited (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). The spatial clustering of negative coefficients in the eastern estuary suggests that housing markets in these areas respond primarily to physical exposure conditions. In contrast, several urban submarkets closer to central London display positive flood-risk coefficients. These results indicate that the amenity value associated with waterfront proximity, urban redevelopment, and accessibility advantages can offset the perceived costs of environmental exposure. Similar trade-offs between environmental risk and amenity value have been documented in coastal housing markets where waterfront access generates strong demand despite the presence of natural hazards (Glaeser et al., 2008 ; Belanger and Bourdeau-Brien, 2018 ). In such cases, housing prices reflect the net valuation of multiple locational attributes rather than the isolated effect of flood risk. The coexistence of these spatial regimes highlights an important limitation of global housing price models. When environmental risk is capitalized differently across submarkets, a single global coefficient cannot represent the underlying price formation process. Instead, the estimated average effect may obscure local valuation patterns and produce misleading inferences regarding climate-risk pricing. The spatial econometric and geographically weighted regression results therefore suggest that climate-risk capitalization should be examined within a spatially heterogeneous modelling framework. The findings also carry implications for the interpretation of market signals in climate adaptation policy. Property prices are often used as indicators of how markets perceive and price environmental risk. However, the results suggest that such signals may vary systematically across space. In some housing markets, particularly in high-demand urban waterfront areas, prices may continue to rise despite environmental exposure. In other locations, particularly in peripheral coastal communities, flood risk appears to be directly capitalized into property values. Consequently, relying on aggregate housing market indicators may lead to an incomplete assessment of climate-risk exposure within coastal urban systems. 7 Conclusion This study examined how flood exposure is capitalized into residential property prices across the Thames Estuary housing market. Using transaction data covering more than seventy thousand properties and a combination of global and local spatial econometric models, the analysis evaluated whether climate-related environmental risk is reflected uniformly in housing values or whether capitalization varies across space. The empirical strategy proceeded from a baseline hedonic specification to spatial econometric models and ultimately to geographically weighted regression, allowing the relationship between flood exposure and housing prices to vary across locations. The results reveal that flood-risk capitalization in the Thames Estuary is spatially heterogeneous. Global models produce a positive coefficient for flood-zone exposure, suggesting that properties located within designated flood-risk areas command higher prices on average. Local estimates demonstrate that this result masks two opposing valuation regimes. In outer estuary markets and several coastal communities, flood exposure is associated with measurable price discounts, indicating that buyers internalize environmental risk when forming housing valuations. In contrast, several urban submarkets closer to central London exhibit positive capitalization effects, where the amenity value of waterfront proximity, accessibility advantages, and redevelopment expectations appear to offset perceived environmental exposure. The global premium observed in the spatial econometric models therefore reflects the aggregation of spatially differentiated pricing processes rather than a uniform market response. These findings contribute to the literature on environmental risk capitalization and spatial housing price modelling. Existing studies often estimate average effects of environmental hazards using global hedonic frameworks (Bin and Polasky, 2004 ; Belanger and Bourdeau-Brien, 2018 ). The present results demonstrate that such approaches may obscure important spatial variation in how environmental exposure is incorporated into property prices. Allowing price gradients to vary across space reveals that the capitalization of flood risk depends strongly on local market conditions, including urban development intensity, accessibility, and the amenity value of waterfront locations. The results also have implications for climate-risk assessment in coastal property markets. Property prices are frequently interpreted as signals of how markets perceive and incorporate environmental risk. However, the spatial heterogeneity identified in this study suggests that such signals may be uneven across housing markets. In high-demand urban waterfront areas, prices may continue to rise despite environmental exposure, potentially masking underlying climate vulnerability. In peripheral coastal communities, by contrast, environmental risk appears to be directly capitalized into housing values. Understanding this spatial differentiation is therefore essential for interpreting housing market responses to climate risk and for designing policies aimed at managing long-term coastal exposure. Future research could extend this analysis by examining how the capitalization of flood risk evolves over time as climate adaptation infrastructure, insurance regimes, and regulatory frameworks change. Incorporating temporal dynamics and forward-looking climate projections would provide additional insight into whether housing markets adjust gradually to environmental risk or whether price responses occur primarily after major hazard events. Declarations Author Contribution O.D.A. conceptualised the study, designed the research framework, provided leadership for spatial econometric analysis, interpreted the results, and wrote the original manuscript draft. T.O. provided structure and guide for the methodology of the original manuscript, and reviewed the draft. P.U. contributed to data preparation, statistical analysis, and assisted with the empirical implementation. A.R.R. supported the development of machine learning components, model evaluation, and contributed to interpretability analysis. F.M. provided expertise in quantitative modelling, supervised aspects of the empirical strategy, and contributed to critical revisions of the manuscript. All authors reviewed, edited, and approved the final version of the manuscript. References Ahlfeldt, G. M., Redding, S. J., Sturm, D. M., & Wolf, N. (2017). The economics of density: Evidence from the Berlin Wall. Econometrica , 85 (6), 2127–2189. Alonso, W. (1964). Location and land use: Toward a general theory of land rent . Harvard University Press. Anselin, L. (1988). Spatial econometrics: Methods and models . Kluwer Academic. Anselin, L., & Rey, S. J. (2014). Modern spatial econometrics in practice: A guide to GeoDa, GeoDaSpace and PySAL . GeoDa. Atreya, A., & Ferreira, S. (2015). Seeing is believing? Evidence from property prices in inundated areas. Risk Analysis , 35 (5), 828–848. Belanger, V., & Bourdeau-Brien, M. (2018). The impact of flood risk on the price of residential properties: The case of England. Journal of Real Estate Literature , 26 (1), 63–86. Beltrán, A., Maddison, D., & Elliott, R. J. R. (2018). Is flood risk capitalised into property values? Ecological Economics , 146 , 668–685. Bernstein, A., Gustafson, M., & Lewis, R. (2019). Disaster on the horizon: The price effect of sea level rise. Journal of Financial Economics , 134 (2), 253–272. Bin, O., & Polasky, S. (2004). Effects of flood hazards on property values: Evidence before and after Hurricane Floyd. Land Economics , 80 (4), 490–500. Bin, O., Kruse, J., & Landry, C. (2008). Flood hazards, insurance rates, and amenities: Evidence from the coastal housing market. Journal of Risk and Insurance , 75 (1), 63–82. Bitter, C., Mulligan, G. F., & Dall’erba, S. (2007). Incorporating spatial variation in housing attribute prices: A comparison of geographically weighted regression and the spatial expansion method. Journal of Geographical Systems , 9 (1), 7–27. Brunsdon, C., Fotheringham, A. S., & Charlton, M. (1996). Geographically weighted regression: A method for exploring spatial nonstationarity. Geographical Analysis , 28 (4), 281–298. Can, A. (1992). Specification and estimation of hedonic housing price models. Regional Science and Urban Economics , 22 (3), 453–473. Cohen, J. P., Coughlin, C. C., & Lopez, R. (2019). The boom and bust of US housing prices from various geographic perspectives. Journal of Housing Economics , 43 , 1–13. Des Rosiers, F., Thériault, M., & Villeneuve, P. (2011). Sorting out access and neighbourhood factors in hedonic price models. Journal of Property Research , 28 (4), 291–315. Dubin, R. A. (1988). Estimation of regression coefficients in the presence of spatially autocorrelated error terms. Review of Economics and Statistics , 70 (3), 466–474. Dumm, R. E., Sirmans, G. S., & Smersh, G. T. (2016). The capitalization of flood insurance premiums into house prices. Journal of Real Estate Finance and Economics , 52 (3), 316–339. Fotheringham, A. S., Brunsdon, C., & Charlton, M. (2002). Geographically weighted regression: The analysis of spatially varying relationships . Wiley. Gibbons, S., & Machin, S. (2008). Valuing school quality, better transport, and lower crime: Evidence from house prices. Oxford Review of Economic Policy , 24 (1), 99–119. Glaeser, E. L., Kolko, J., & Saiz, A. (2008). Consumer city. Journal of Economic Geography , 1 (1), 27–50. Goodman, A. C., & Thibodeau, T. G. (2003). Housing market segmentation and hedonic prediction accuracy. Journal of Housing Economics , 12 (3), 181–201. Gourevitch, J. D., Kahn, M. E., & Walsh, R. (2023). Climate adaptation and housing markets. Annual Review of Resource Economics , 15 , 327–347. Holly, S., Pesaran, M. H., & Yamagata, T. (2011). The spatial and temporal diffusion of house prices in the UK. Journal of Urban Economics , 69 (1), 2–23. Kahn, M. E., Mohaddes, K., Ng, R., Pesaran, M. H., Raissi, M., & Yang, J. C. (2021). Long-term macroeconomic effects of climate change: A cross-country analysis. Energy Economics , 104 , 105624. Keys, B. J., & Mulder, P. (2020). Neglected no more: Housing markets, mortgage lending, and sea level rise. Review of Financial Studies , 33 (7), 2902–2945. LeSage, J., & Pace, R. K. (2009). Introduction to spatial econometrics . CRC. Malpezzi, S. (2002). Hedonic pricing models: A selective and applied review. In T. O’Sullivan, & K. Gibb (Eds.), Housing economics and public policy . Blackwell. Muth, R. F. (1969). Cities and housing . University of Chicago Press. Osland, L. (2010). An application of spatial econometrics in relation to hedonic house price modeling. Journal of Real Estate Research , 32 (3), 289–320. Pace, R. K., & LeSage, J. P. (2004). Spatial statistics and models. In J. P. LeSage, & R. K. Pace (Eds.), Spatial econometrics . CRC. Páez, A., Long, F., & Farber, S. (2011). Moving window approaches for hedonic price estimation: An application to housing prices. Spatial Economic Analysis , 6 (3), 299–317. Rosen, S. (1974). Hedonic prices and implicit markets: Product differentiation in pure competition. Journal of Political Economy , 82 (1), 34–55. Sirmans, G. S., MacDonald, L., Macpherson, D., & Zietz, E. (2005). The composition of hedonic pricing models. Journal of Real Estate Literature , 13 (1), 3–43. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9213954","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":611745310,"identity":"ac14360f-f63c-46d1-b2ed-b5241e512cf2","order_by":0,"name":"Oluwaseun Damilola Ajayi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEklEQVRIiWNgGAWjYHCCBAYGA5sEGI+xAcaSwK8ljTQtIHCYBC387AcefuYpOJ/HP/vswccVDHay/dOOP2D4UcOQOLMBuxbJnoRkaR6D28US5/KSDc8wJBvPuJ1jwNhzjCFxNg5bDA4kJIC0JDac4TGTbPzHnNhwO4eBgbeBIXEeLi3nHyT/5jE4lzj/DI/5zwaG+sT5t9MfMP7Fp+VGQhrQlgOJG4C2AL1+OHHD7QQDZpAtuBwmOeNBmuUcg+TEjWf4kiUbGI4bbwT65bDMMQljXN7n589JvvHmj13ivDO8Bz82MFTLzrud/vDhmxob2RkHcFjDwJPAxANhIMQO4I9I9gOMP9C1jIJRMApGwShABgDnJV9yKeUJaQAAAABJRU5ErkJggg==","orcid":"","institution":"Harper Adams University","correspondingAuthor":true,"prefix":"","firstName":"Oluwaseun","middleName":"Damilola","lastName":"Ajayi","suffix":""},{"id":611745312,"identity":"c7724774-87ee-4700-a33b-f24c427db59d","order_by":1,"name":"Tayo Odunsi","email":"","orcid":"","institution":"Build Africa","correspondingAuthor":false,"prefix":"","firstName":"Tayo","middleName":"","lastName":"Odunsi","suffix":""},{"id":611745313,"identity":"468f6b76-ef81-492b-b31f-b44d0dab25cc","order_by":2,"name":"Paulinus Ugwu","email":"","orcid":"","institution":"Pan-Atlantic University","correspondingAuthor":false,"prefix":"","firstName":"Paulinus","middleName":"","lastName":"Ugwu","suffix":""},{"id":611745314,"identity":"ae3c2c89-6316-4653-9f8c-700f233c5539","order_by":3,"name":"Arti Rawat","email":"","orcid":"","institution":"University of Bordeaux","correspondingAuthor":false,"prefix":"","firstName":"Arti","middleName":"","lastName":"Rawat","suffix":""},{"id":611745315,"identity":"0a0e6b93-8c5d-4be1-88a0-de0ed9a5639c","order_by":4,"name":"Farai Mlambo","email":"","orcid":"","institution":"University of the Witwatersrand","correspondingAuthor":false,"prefix":"","firstName":"Farai","middleName":"","lastName":"Mlambo","suffix":""}],"badges":[],"createdAt":"2026-03-24 15:38:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9213954/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9213954/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105488594,"identity":"849a145a-00ca-49c3-9148-26f194f1eac4","added_by":"auto","created_at":"2026-03-26 15:02:45","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1445375,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial Distribution of Local Flood-Risk Coefficients (GWR Estimates)\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9213954/v1/2cca4636b92b9b7e73238974.jpeg"},{"id":105488596,"identity":"18157273-08f8-4b38-9475-ac939765eb01","added_by":"auto","created_at":"2026-03-26 15:02:45","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1432318,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial Distribution of Local Model Fit (Local R²)\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9213954/v1/405886db35d67238dc322f4b.jpeg"},{"id":105488595,"identity":"2804119b-db74-4c92-b859-99ba66cb0666","added_by":"auto","created_at":"2026-03-26 15:02:45","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1447221,"visible":true,"origin":"","legend":"\u003cp\u003eLISA Cluster Map of Local Flood-Risk Pricing Effects\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9213954/v1/e2ce4eb355fb6ddb965d1a85.jpeg"},{"id":105565836,"identity":"bbba95e3-6d8b-4109-a97c-494115b58865","added_by":"auto","created_at":"2026-03-27 12:54:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6186143,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9213954/v1/bed6c595-48b4-4749-a110-ede25ec9d5f3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Spatial Heterogeneity in Flood-Risk Capitalization: Evidence from the Thames Estuary Housing Market","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eClimate change is increasingly reshaping the spatial structure of housing markets through the capitalization of environmental risk into property values. Coastal and estuarine regions are particularly exposed as rising sea levels and intensifying flood events alter both the physical risk profile of housing assets and expectations about long-term protection infrastructure. A growing body of evidence shows that environmental hazards such as flood exposure, wildfire risk, and sea-level rise can generate measurable price discounts in residential markets (Bin and Polasky \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Beltr\u0026aacute;n et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Bernstein et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, the magnitude and spatial distribution of these effects remain highly uneven across urban regions. Housing markets simultaneously capitalize multiple attributes including location, structural characteristics, and environmental amenities, while adaptation policies and protective infrastructure may partially offset perceived risk. As a result, the pricing of climate exposure is unlikely to be uniform across space. Instead, risk capitalization may vary substantially across local housing submarkets depending on geography, built form, and institutional flood governance regimes.\u003c/p\u003e \u003cp\u003eA substantial literature examines how environmental risks are capitalized into housing prices using hedonic price models. Early studies document statistically significant discounts for properties located in flood-prone areas or subject to environmental hazards (Bin and Polasky \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Bin et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). More recent research shows that climate-related risks such as sea-level rise expectations, storm exposure, and insurance reforms can also influence housing market outcomes (Beltr\u0026aacute;n et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Bernstein et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Keys and Mulder \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). While these studies provide important evidence that environmental risk affects housing prices, most empirical analyses rely on global regression frameworks that assume constant marginal effects across space. This assumption may be restrictive in coastal housing markets where risk exposure, topography, and flood protection infrastructure vary sharply across short geographic distances. If price responses to environmental risk differ across local housing submarkets, global hedonic estimates may mask important spatial variation in the capitalization process. A growing strand of spatial housing research therefore emphasizes the need for locally varying models capable of identifying heterogeneous price gradients across urban regions (Fotheringham et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; P\u0026aacute;ez et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In this context, geographically weighted regression (GWR) provides a flexible framework for estimating location-specific relationships between property prices and environmental attributes, allowing the marginal effects of risk exposure to vary across space rather than imposing a single global parameter.\u003c/p\u003e \u003cp\u003eThis study examines the spatial capitalization of climate exposure in the housing market of the Thames Estuary, one of the most flood-exposed urban regions in Europe. Using transaction-level data for residential properties combined with elevation, river proximity, flood risk classifications, and structural housing characteristics, the analysis evaluates how environmental exposure influences property prices across local housing submarkets. Diagnostic tests reveal significant spatial dependence in hedonic model residuals, indicating that conventional global specifications fail to capture localized price dynamics. To address this issue, the study applies geographically weighted regression (GWR) to estimate spatially varying price relationships between housing values and climate exposure. The results show that flood risk and topographic elevation exhibit substantial spatial heterogeneity in their capitalization into housing prices. In low-lying estuarine corridors such as Canvey Island and Purfleet, flood exposure generates measurable price discounts, whereas proximity to the Thames produces localized price premiums in higher-elevation areas where amenity value dominates perceived risk. These findings demonstrate that environmental risk is not priced uniformly across the housing market but varies systematically across space depending on local exposure and geography. By identifying these spatially heterogeneous price responses, the study contributes to the literature on environmental risk capitalization and highlights the importance of spatial modeling approaches in climate-exposed housing markets.\u003c/p\u003e \u003cp\u003eThe remainder of the paper proceeds as follows. Section \u003cspan refid=\"Sec2\" class=\"InternalRef\"\u003e2\u003c/span\u003e reviews the literature on environmental risk capitalization and spatial heterogeneity in housing markets, with particular attention to flood exposure and coastal property valuation. Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e3\u003c/span\u003e describes the study area, data sources, and construction of the housing, environmental, and structural variables used in the analysis. Section \u003cspan refid=\"Sec10\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents the empirical methodology, beginning with a conventional hedonic price specification and extending to geographically weighted regression to estimate spatially varying price relationships. Section \u003cspan refid=\"Sec14\" class=\"InternalRef\"\u003e5\u003c/span\u003e reports the empirical results, including global model diagnostics and local parameter estimates that reveal spatial variation in climate risk capitalization across the Thames Estuary housing market. Section \u003cspan refid=\"Sec21\" class=\"InternalRef\"\u003e6\u003c/span\u003e concludes with the implications of these findings for housing market adjustment in climate-exposed urban regions.\u003c/p\u003e"},{"header":"2. Literature Review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Environmental Risk Capitalization in Housing Markets\u003c/h2\u003e \u003cp\u003eEnvironmental hazards can influence housing values when buyers incorporate expected exposure into property price formation. Early work on environmental capitalization demonstrates that housing markets often internalize localized risk through price adjustments when information about hazards becomes salient to buyers (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Evidence from flood-prone regions in the United States shows that properties exposed to flood hazards typically experience price discounts relative to comparable dwellings outside high-risk areas, although the magnitude of these discounts varies across locations and time periods (Atreya and Ferreira, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Empirical studies increasingly document that capitalization depends on both the visibility of environmental risk and the availability of institutional risk signals such as flood maps, insurance pricing, or disaster events (Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Flood exposure presents a particularly relevant case because the valuation effects may operate through several channels. Physical vulnerability can reduce expected housing utility through anticipated damage or insurance costs, generating price discounts in exposed locations. At the same time, waterfront amenities can generate positive valuation effects, particularly where properties benefit from visual or recreational access to water bodies (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Housing markets therefore often reflect a trade-off between amenity value and environmental exposure. Empirical evidence indicates that these opposing forces may coexist within the same geographic market, producing complex valuation patterns rather than uniform price discounts (Dumm et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Studies examining coastal and riverine housing markets show that proximity to water can increase property values in relatively safe areas while generating price penalties where flood exposure becomes more salient (Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eRecent research has also emphasized that climate change may alter pricing dynamics by increasing the perceived probability of extreme events. Evidence suggests that climate-related risks such as sea-level rise or coastal flooding are beginning to influence housing markets in vulnerable regions, although the magnitude of price adjustments remains uneven across markets (Kahn et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Limited capitalization of long-term climate risk has been documented in several contexts, suggesting that housing markets may respond slowly to emerging environmental threats when risks remain uncertain or institutionally mediated (Gourevitch et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These findings highlight the importance of examining how climate exposure interacts with local geography and institutional contexts to influence housing valuation. Despite the growing literature on environmental risk capitalization, most empirical studies rely on global hedonic price models that estimate average price effects across entire housing markets. Such approaches implicitly assume that environmental risk influences property values uniformly across space. Urban housing markets, however, often exhibit substantial spatial heterogeneity due to differences in neighborhood characteristics, built environments, and localized exposure conditions. Global models may therefore obscure important spatial variation in how environmental risk is priced. Recent work in urban and regional science increasingly emphasizes the need for spatially explicit approaches capable of identifying geographically varying price relationships within housing markets (Pace and LeSage, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). The next strand of literature addresses this challenge by examining spatial heterogeneity in hedonic housing models.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Spatial Heterogeneity in Housing Price Formation\u003c/h2\u003e \u003cp\u003eHousing markets rarely operate as spatially homogeneous systems. Property values emerge from localized interactions between structural attributes, neighborhood characteristics, and accessibility conditions. Empirical research has long documented that the marginal value of housing attributes can vary substantially across space, reflecting differences in neighborhood composition, land-use patterns, and urban structure (Dubin, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Conventional hedonic price models typically estimate global parameters that represent average relationships across an entire study area. Such specifications impose constant marginal effects and therefore assume that the influence of housing characteristics is spatially invariant. This assumption often conflicts with observed urban housing dynamics. Spatial clustering of property values and neighborhood-specific amenities can generate localized pricing regimes that differ across submarkets. Evidence from urban housing studies shows that structural and locational attributes may carry different price premiums depending on neighborhood context and market segmentation (Can, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Osland, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Ignoring these spatial variations can lead to biased estimates when the marginal willingness to pay for housing characteristics differs across locations. Spatial dependence in housing prices further complicates estimation because nearby properties tend to exhibit correlated price movements through neighborhood spillovers and shared environmental conditions (Pace and LeSage, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eEnvironmental exposure provides a particularly strong source of spatial heterogeneity in housing valuation. Flood risk, elevation, and river proximity vary across small geographic scales and therefore may influence property prices differently across neighborhoods within the same metropolitan region. Areas located within flood-prone corridors may exhibit price discounts due to perceived exposure, whereas nearby neighborhoods with higher elevation or protective infrastructure may experience little or no penalty. Empirical studies examining environmental risk have therefore increasingly recognized that spatially constant hedonic coefficients may obscure important local variation in price responses (Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Spatial econometric approaches have consequently become central tools for examining geographically varying relationships in housing markets. Recent advances in spatial housing research emphasize models that allow coefficients to vary across locations rather than imposing uniform global parameters. These approaches recognize that housing markets consist of overlapping submarkets where valuation mechanisms may differ depending on local conditions. Identifying such localized pricing structures is essential when analyzing environmental risks whose effects depend strongly on geographic context. The following section discusses geographically weighted regression as a modeling framework designed to capture these spatially varying price relationships.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Geographically Weighted Regression in Housing Market Analysis\u003c/h2\u003e \u003cp\u003eSpatially varying relationships in housing markets have motivated the development of local regression approaches that allow parameters to differ across geographic locations. Geographically weighted regression (GWR) provides a framework for estimating location-specific coefficients by calibrating local regressions using spatially weighted observations surrounding each property (Brunsdon et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1996\u003c/span\u003e). Unlike conventional global hedonic models that impose constant marginal effects, GWR permits the influence of housing characteristics to vary continuously across space. This flexibility makes the approach particularly suitable for analyzing housing markets characterized by localized price formation and heterogeneous neighborhood conditions. Applications of GWR in housing research demonstrate that the marginal effects of structural and locational attributes often vary significantly across urban regions. Empirical studies show that housing characteristics such as floor area, accessibility, and neighborhood amenities exhibit geographically differentiated price effects when local estimation techniques are employed (Fotheringham et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). These findings suggest that spatially constant hedonic coefficients may mask important variations in buyer preferences and market conditions across neighborhoods. Incorporating spatially varying parameters therefore improves the ability of housing models to capture localized valuation patterns.\u003c/p\u003e \u003cp\u003eEnvironmental exposure represents a context in which geographically varying price effects are particularly likely to arise. Flood risk, elevation, and proximity to water bodies often vary at fine spatial scales, producing localized differences in perceived risk and amenity value. Housing markets located along rivers or coastlines may therefore display heterogeneous responses to environmental attributes depending on local topography and exposure conditions. Global models that estimate a single price discount for flood exposure cannot capture these geographically differentiated effects. Spatially adaptive models such as GWR allow the marginal influence of environmental variables to vary across locations, making it possible to identify where risk capitalization is strongest or weakest within a metropolitan region. Several studies have applied GWR to housing markets to examine spatial heterogeneity in price determinants. Evidence from urban housing markets shows that local estimation approaches can reveal geographically varying relationships that remain hidden in conventional hedonic regressions (Bitter et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Cohen et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Such approaches have been used to analyze spatial variation in the effects of neighborhood amenities, accessibility, and environmental quality on property values. However, relatively limited research has applied spatially varying regression techniques to examine the capitalization of climate-related risks in coastal housing markets.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Data","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Study Area and Data Sources\u003c/h2\u003e \u003cp\u003eThe empirical analysis focuses on residential housing markets located within the Thames Estuary corridor in southeast England. The estuary represents one of the most climate-exposed urban regions in Europe, with extensive areas situated at low elevations and subject to tidal flood risk. Major urban centres along the estuary include boroughs in east and southeast London as well as downstream municipalities such as Thurrock and Gravesham. These areas contain a diverse mix of residential housing types and represent an important segment of the wider London metropolitan housing market. Flood exposure within the region is managed through the Thames Estuary 2100 (TE2100) flood risk management strategy, which defines spatial policy management units used to guide long-term adaptation planning. Residential transaction data are obtained from the United Kingdom Land Registry Price Paid Data (PPD), which records property-level sales transactions across England and Wales. The dataset provides transaction prices, property types, and sale dates for individual residential properties. Only arm\u0026rsquo;s-length transactions of residential dwellings are retained for the analysis. Observations corresponding to non-market transfers or incomplete records are excluded to ensure consistency in price measurement. Transaction prices are transformed using the natural logarithm to reduce skewness and align with conventional hedonic modelling practices.\u003c/p\u003e \u003cp\u003eProperty-level structural characteristics are obtained from the national Energy Performance Certificate (EPC) register. The EPC database provides information on dwelling attributes including total floor area, number of habitable rooms, energy efficiency rating, and property type classifications. These structural characteristics are commonly used in hedonic housing models to capture differences in dwelling size and building attributes that influence market prices (Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Environmental exposure variables are constructed using geospatial data from multiple sources. Elevation data are derived from the Ordnance Survey OS Terrain 5 digital elevation model, which provides high-resolution ground elevation estimates across the study region. Distance to tidal frontage is calculated as the Euclidean distance from each property location to the nearest section of the River Thames using geographic information system (GIS) methods. Flood exposure classifications are obtained from the Environment Agency Flood Map for Planning, which identifies properties located within Flood Zone 2 and Flood Zone 3 according to their estimated annual probability of flooding. These spatial datasets are merged with the transaction data using georeferenced property locations to construct the final analytical dataset. The resulting dataset integrates housing transaction information with structural dwelling characteristics and environmental exposure variables, allowing analysis of how flood risk and topographic conditions influence residential property prices within the Thames Estuary housing market.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Variable Construction and Descriptive Statistics\u003c/h2\u003e \u003cp\u003eThe dependent variable is the transaction price of residential properties. Following standard practice in hedonic housing models, the natural logarithm of sale price is used to reduce skewness and interpret estimated coefficients as approximate percentage changes in property value (Rosen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1974\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). The use of log-transformed prices also improves the statistical properties of regression estimates when price distributions exhibit right skewness, which is typical in housing markets. Explanatory variables are grouped into three categories reflecting conventional hedonic modelling frameworks: structural characteristics, environmental exposure measures, and climate adaptation indicators. Structural housing attributes capture differences in dwelling size and type that influence buyer willingness to pay. These variables include total floor area, the number of habitable rooms, and categorical indicators for property type. Empirical housing research consistently identifies dwelling size as the dominant determinant of residential property values, reflecting the direct relationship between housing consumption and price formation (Dubin, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Environmental exposure variables measure physical climate risk associated with the Thames Estuary floodplain. Elevation is measured as ground height above sea level using high-resolution digital elevation data. Lower elevations correspond to greater vulnerability to tidal flooding and storm surge events. Distance to river frontage is calculated as the Euclidean distance from each property to the nearest segment of the River Thames. River proximity may capture two opposing valuation channels. Close proximity can generate positive amenity value through waterfront views and recreational access, while simultaneously increasing perceived flood exposure. Prior research indicates that these competing effects often produce nonlinear price responses in coastal housing markets (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFlood exposure is further captured using Environment Agency flood risk classifications. Binary indicators identify whether properties are located within Flood Zone 2 or Flood Zone 3, which correspond to moderate and high annual flood probability thresholds. These classifications represent regulatory signals of environmental exposure that may influence buyer perceptions of risk and insurance availability. Adaptive capacity variables capture institutional and structural factors that may mitigate perceived climate risk. Energy performance ratings are included as a proxy for building efficiency and structural resilience, while Policy Management Unit (PMU) indicators identify properties located within areas designated under the Thames Estuary 2100 flood management strategy. These spatial governance units reflect the long-term flood defence planning framework implemented across the estuary and therefore represent institutional signals that may influence expectations regarding future flood protection. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the definitions, transformations, and data sources of all variables included in the analysis.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eVariable Definitions and Sources\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCategory\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDefinition \u0026amp; Transformation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eData Source\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTarget\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSale Price\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTransaction price of residential property (\u0026pound;); log-transformed for modelling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eUK Price Paid Data\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClimate Exposure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eElevation (m)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGround elevation at property location (meters above sea level)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOrdnance Survey OS Terrain 5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDistance to River\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEuclidean distance to nearest tidal frontage (meters); log-transformed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOrdnance Survey\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFlood Zone 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBinary indicator for high flood probability (\u0026gt;\u0026thinsp;1% annual probability)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironment Agency Flood Map\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFlood Zone 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBinary indicator for moderate flood probability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEnvironment Agency Flood Map\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdaptive Capacity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEPC Rating\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEnergy performance score of dwelling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEPC Register\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePMU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIndicator for Thames Estuary 2100 Policy Management Unit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTE2100 Plan\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStructural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal Floor Area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTotal internal floor area (sqm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEPC Register\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHabitable Rooms\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of habitable rooms in dwelling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEPC Register\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProperty Type\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCategorical: house, flat, bungalow, maisonette\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEPC Register\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e reports descriptive statistics for the housing transactions used in the empirical analysis, disaggregated by the Thames Estuary 2100 (TE2100) Policy Management Units. The dataset contains 73,039 residential property transactions distributed across multiple estuarine submarkets characterized by distinct housing structures, socioeconomic conditions, and exposure to tidal flood risk. Examining descriptive variation across these policy units provides an initial indication of the spatial heterogeneity that characterizes housing markets in large metropolitan regions (Dubin, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Substantial variation in property prices is evident across policy units. Average transaction prices range from approximately \u0026pound;231,804 in Purfleet, Grays and Tilbury to more than \u0026pound;585,000 in London City. High-value areas such as Greenwich and Wandsworth to Deptford exhibit mean prices exceeding \u0026pound;500,000, reflecting their proximity to central London and strong demand for waterfront housing. Peripheral estuarine locations including North Kent and Canvey Island display substantially lower price levels, consistent with differences in accessibility, local economic conditions, and housing supply constraints. Similar price gradients between central urban areas and peripheral locations have been widely documented in metropolitan housing markets (Osland, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Des Rosiers et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Structural housing characteristics vary significantly across the TE2100 policy units. Average floor area ranges from approximately 72 m\u0026sup2; in Purfleet, Grays and Tilbury to nearly 120 m\u0026sup2; in Leigh Old Town and Southend-on-Sea. Property type composition also differs markedly across locations. Inner estuary districts such as Isle of Dogs and London City exhibit a high proportion of flats, reflecting dense urban development patterns and the vertical housing supply typical of central metropolitan areas. In contrast, outer estuarine zones such as Canvey Island and North Kent are dominated by detached and semi-detached houses. These structural differences represent fundamental determinants of housing prices in hedonic valuation models, as dwelling size and building form directly influence the consumption value of housing services (Rosen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1974\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNeighbourhood socioeconomic conditions exhibit comparable spatial variation. Mean household income ranges from approximately \u0026pound;30,400 in the Royal Docks to over \u0026pound;47,400 in Greenwich. Income disparities across neighbourhoods are a central driver of localized housing demand and frequently generate segmented housing markets within metropolitan regions (Can, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Goodman and Thibodeau, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Commuting accessibility displays similar spatial heterogeneity across the estuary corridor. Average travel times to employment centres range from approximately 3.5 minutes in Isle of Dogs to more than 10 minutes in Dartford and Swanscombe. Accessibility gradients of this type are commonly associated with spatial variation in housing prices because proximity to employment nodes influences household location decisions (Alonso, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1964\u003c/span\u003e; Glaeser et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Environmental exposure variables highlight the climate vulnerability of the Thames Estuary housing market. More than half of the observations in the sample are located within high flood-risk zones. Certain policy units exhibit particularly concentrated exposure. Canvey Island and Isle of Grain consist entirely of properties classified as high flood-risk areas, reflecting their extremely low-lying coastal geography. Other districts such as Greenwich, Thamesmead, and Wandsworth to Deptford also contain a large share of properties located within high-risk flood zones. Previous studies have shown that such environmental hazards can influence housing prices when buyers incorporate expected exposure into their valuation decisions (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Elevation statistics further illustrate the geographic vulnerability of the region.\u003c/p\u003e \u003cp\u003eMean elevation values range between approximately 1.9 and 8.2 metres above sea level across policy units, with several areas including Canvey Island and Thamesmead situated at extremely low elevations. Low-lying topography increases exposure to tidal flooding and storm surge events, particularly in estuarine environments subject to sea-level rise. Distance to river frontage also varies substantially across the sample, capturing differences in both potential amenity value and environmental exposure. Empirical studies frequently find that proximity to water generates a complex valuation trade-off in housing markets, reflecting the coexistence of waterfront amenities and flood risk (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Dumm et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The pronounced variation in housing prices, structural attributes, socioeconomic conditions, and environmental exposure across TE2100 policy units indicates that the capitalization of flood risk is unlikely to be spatially uniform across the Thames Estuary housing market. Conventional global hedonic models impose constant marginal effects across the study region and therefore cannot capture localized differences in environmental risk pricing. Spatial modelling approaches that allow housing price relationships to vary geographically are therefore necessary to identify how climate exposure influences property values across estuarine submarkets. The following section outlines the empirical framework used to estimate these spatially heterogeneous price effects.\u003c/p\u003e \u003cp\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ea \u0026mdash; Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Structural Controls)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eTarget Variable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c9\" namest=\"c4\"\u003e \u003cp\u003eStructural Controls\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003ePrice (\u0026pound;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eTotal Floor Area (m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c9\" namest=\"c6\"\u003e \u003cp\u003eProperty Types \u0026ndash; N (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(Min: Max)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(Min, Max)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBungalow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eFlat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eHouse\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMaisonette\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBarking \u0026amp; Dagenham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e281,503.32 (148,046.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(100: 9,345,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e78.59 (24.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(7.80: 297)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e46 (0.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1,467 (18%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e6,630 (80%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e165 (2.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanvey Island\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e254,154.92 (93,459.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(500: 1,300,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84.55 (38.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(13: 875)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e3,582 (47%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e86 (1.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3,927 (52%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e27 (0.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDartford \u0026amp; Erith\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e272,889.93 (332,545.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(6,400: 21,812,419)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e77.00 (25.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(18: 395)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1,952 (36%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3,344 (62%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e125 (2.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEast Tilbury \u0026amp; Mucking Mashes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e275,865.39 (90,648.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(55,000: 750,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e87.03 (32.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(36: 442)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e9 (2.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e384 (98%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreenwich\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e546,806.77 (280,278.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(6,500: 3,500,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e89.78 (33.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(14.00: 304.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e673 (25%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1,814 (67%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e226 (8.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Dogs \u0026amp; Lea Valley\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e455,572.21(452,208.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(100: 17,220,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e77.54 (34.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10.00: 811.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e3 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1,919 (53%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1,171 (32%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e519 (14%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Grain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e420,000.00 ()\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(420,000: 420,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e168 ()\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(168: 168)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeigh Old Town and Southend-on-Sea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e424,062.29 (261,574.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(4,500: 2,425,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e119.80 (68.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(5.24: 2,148.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e580 (19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e600 (19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1,856 (60%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e50 (1.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLondon City\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e585,690.23 (337,081.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(29,465: 3,500,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74.44 (36.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(23.00, 375.13)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e318 (50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e237 (37%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e80 (13%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth Kent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e247,879.84 (92,278.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(10,000: 590,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e82.20 (22.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(25.10, 336.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e59 (6.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e46 (4.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e870 (88%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e16 (1.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePurfleet, Grays \u0026amp; Tilbury\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e231,804.16 (723,185.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(15,000:40,779,191)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72.09 (23.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(18.47: 520.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e60 (0.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2,027 (32%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e4,213 (66%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e121 (1.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRainham Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e314,959.77 (107,789.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(26,730: 670,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e86.57 (23.57)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(33.48: 217.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e434 (15%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e238 (8.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2,155 (73%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e116 (3.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoyal Docks\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e322,365.66 (202,762.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(100: 10,080,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e79.24 (28.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(8.04: 1,570.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e41 (0.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3,382 (25%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e9,554 (70%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e649 (4.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShell Haven \u0026amp; Fobbing Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e237,751.64 (88,261.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(5,000: 630,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e76.35 (26.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(36.00: 216.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4 (1.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e51 (18%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e220 (77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e12 (4.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSwanscombe \u0026amp; Northfleet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e359,346.91 (114,576.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(64,995: 750,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e99.86 (29.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(37.10: 212.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4 (5.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e63 (91%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2 (2.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThamesmead\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e263,650.16 (192,867.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3,250: 11,915,135)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74.76 (24.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4.80: 503.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e100 (1.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1,450 (23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e4,648 (73%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e160 (2.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWandsworth to Deptford\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e523,421.35 (649,990.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(500: 57,285,700)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e80.83 (39.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4.70: 1,409.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e10 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4,438 (42%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e4,929 (47%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1,173 (11%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOverall (N\u0026thinsp;=\u0026thinsp;73,039)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e342,552.65 (397,875.3)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e(100: 57,285,700)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e81.13 (34.55)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e(4.70: 2,148.00)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e4,922 (6.7%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e18,660 (26%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e46,016 (63%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e3,441 (4.7%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eb \u0026mdash; Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Neighbourhood Controls)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"7\" nameend=\"c8\" namest=\"c2\"\u003e \u003cp\u003eStructural Controls\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e \u003cp\u003eNeighbourhood Controls\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"7\" nameend=\"c8\" namest=\"c2\"\u003e \u003cp\u003eEnergy Ratings (A - G) \u0026ndash; N(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e \u003cp\u003eNet Annual Income\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(Min: Max)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBarking \u0026amp; Dagenham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e544 (6.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,738 (21%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4,264 (51%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,428 (17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e213 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e112 (1.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e31,946.13 (2,148.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(25,985: 35,072)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanvey Island\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e166 (2.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,218 (16%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4,169 (55%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,685 (22%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e305 (4.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e74 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e32,986.66 (1,093.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(31,105: 34,650)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDartford \u0026amp; Erith\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e61 (1.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,100 (57%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e912 (17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,020 (19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e263 (4.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e54 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e13 (0.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e34,554.92 (1,478.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(31,592: 42,170)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEast Tilbury \u0026amp; Mucking Mashes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32 (8.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e76 (19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e209 (53%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e69 (18%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3 (0.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e36,118.90 (150.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(36,088: 36,847)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreenwich\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (0.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e389 (14%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e636 (23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,128 (42%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e454 (17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e61 (2.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e37 (1.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e47,401.68 (4,299.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(32,610: 50,012)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Dogs \u0026amp; Lea Valley\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e271 (7.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,836 (51%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,225 (34%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e233 (6.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e40 (1.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e6 (0.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e36,363.32 (5,050.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(26,517: 48,749)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Grain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e34,633.00 (NA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(34,633: 34,633)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeigh Old Town and Southend-on-Sea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72 (2.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e377 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,229 (40%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,028 (33%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e336 (11%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e44 (1.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e38,613.31 (6,840.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(25,521: 43,932)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLondon City\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14 (2.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e324 (51%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e235 (37%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e35 (5.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e19 (3.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e8 (1.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e44,937.11 (7,251.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(33,051: 53,735)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth Kent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e151 (15%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e158 (16%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e488 (49%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e164 (17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e20 (2.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e10 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e33,059.00 (0.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(33,059: 33,059)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePurfleet, Grays \u0026amp; Tilbury\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33 (0.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,211 (19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,771 (28%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2,554 (40%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e698 (11%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e107 (1.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e47 (0.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e30,560.50 (2,176.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(27,850: 35,752)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRainham Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e384 (13%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e508 (17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,340 (46%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e604 (21%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e83 (2.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e24 (0.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e38,792.09 (1,214.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(38,297: 43,106)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoyal Docks\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e753 (5.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,520 (26%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6,553 (48%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2,205 (16%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e376 (2.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e216 (1.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e30,426.11 (4,919.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(22,035: 46,930.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShell Haven \u0026amp; Fobbing Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (0.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (0.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74 (26%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e111 (39%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e83 (29%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e14 (4.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e35,735.44 (1,169.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(31,872: 36,088)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSwanscombe \u0026amp; Northfleet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e50 (72%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2 (2.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6 (8.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e8 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2 (2.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1 (1.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e30,890.20 (2,480.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(26,896: 32,408.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThamesmead\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (\u0026lt;\u0026thinsp;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e203 (3.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,457 (39%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2,635 (41%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e859 (14%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e165 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e38 (0.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e34,340.62 (2,122.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(30,572: 52,998.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWandsworth to Deptford\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 (0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e686 (6.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4,071 (39%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4,130 (39%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,287 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e257 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e105 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e40,797.65 (5,493.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(33,608: 63,550.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOverall (N\u0026thinsp;=\u0026thinsp;73,039)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e137 (0.2%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e8,027 (11%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e19,678 (27%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e31,296 (43%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e11,104 (15%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e2,056 (2.8%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e741 (1.0%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e34,844.72 (5,900.96)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e(22,035: 63,550.00)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ec \u0026mdash; Descriptive Statistics: Spatial and Structural Variation in the Thames Estuary Housing Market (Climate \u0026amp; Environment Variables)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eNeighbourhood Controls\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"7\" nameend=\"c10\" namest=\"c4\"\u003e \u003cp\u003eClimate and Environment Variables\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eTotal Time to Work (Min)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c6\" namest=\"c4\"\u003e \u003cp\u003eFlood Risk Levels \u0026ndash; N (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c8\" namest=\"c7\"\u003e \u003cp\u003eElevation (m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e \u003cp\u003eDistance to River (m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(Min: Max)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo Risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMid Risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eHigh Risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(Min: Max)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(Min: Max)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBarking \u0026amp; Dagenham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.36 (2.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.30, 12.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5,919 (71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e444 (5.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,945 (23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.14 (1.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30: 11.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e750.18 (400.96)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(31.27: 2,076.7)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanvey Island\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.73 (3.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.88, 16.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e7,622 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.87 (0.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-0.40: 3.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e551.44(396.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(10.31: 1,487.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDartford \u0026amp; Erith\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.76 (3.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.98, 17.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,085 (38%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,000 (18%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2,338 (43%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.13 (1.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.20, 12.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e425.61 (289.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(27.49: 1,145.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEast Tilbury \u0026amp; Mucking Mashes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.05 (3.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.00, 12.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e293 (75%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e100 (25%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.77 (1.61)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.10, 9.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e230.10 (146.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(34.80: 914.40)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreenwich\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.84 (2.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.66, 8.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e777 (29%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e237 (8.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,700 (63%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4.72 (2.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30, 17.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e602.12 (336.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(31.97: 1,281.9)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Dogs \u0026amp; Lea Valley\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.49 (1.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.07, 9.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e847 (23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e387 (11%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2,378 (66%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4.10 (1.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30, 9.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e436.95 (234.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(18.26, 1,042.34)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Grain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.71 (NA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(9.71, 9.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.50 (NA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(5.50, 5.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.07 (NA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(0.07 0.07)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeigh Old Town and Southend-on-Sea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.93 (1.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.28, 9.49)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,792 (58%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e892 (29%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e402 (13%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.79 (2.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(0.90, 13.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e660.91 (333.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(13.98, 1,488.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLondon City\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.80 (1.81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.10, 16.36)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e69 (11%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e20 (3.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e546 (86%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4.26 (1.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(2.50, 9.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e405.65 (116.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(174.29, 614.37)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth Kent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.56 (3.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(4.18, 12.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e438 (44%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e28 (2.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e525 (53%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.83 (2.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(2.80, 11.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e798.98 (276.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(166 1,264.65)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePurfleet, Grays \u0026amp; Tilbury\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.95 (2.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.05, 10.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,347 (37%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,260 (20%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2,814 (44%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.22 (4.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(0.00, 16.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e525.29 (275.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(29.32, 1,211.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRainham Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.57 (3.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.90, 12.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,875 (64%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e297 (10%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e771 (26%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.08 (1.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(1.50, 8.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e481.38 (249.69)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(19.70, 1,118.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoyal Docks\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.77 (1.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.28, 8.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5,164 (38%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,583 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e6,879 (50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.62 (1.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30, 9.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1,251.34 (601.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(27.49, 2,581.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShell Haven \u0026amp; Fobbing Marshes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.25 (2.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.92, 13.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e228 (79%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e24 (8.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e35 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e7.42 (2.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.10, 11.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e113.11 (127.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(20.23, 515.43)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSwanscombe \u0026amp; Northfleet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.06 (0.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(10.05, 10.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e65 (94%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4 (5.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8.17 (1.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(4.30, 9.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e635.34 (90.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(415.62, 718.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThamesmead\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.74 (3.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.76, 12.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,756 (28%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e236 (3.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4,366 (69%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.17 (2.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30, 12.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e471.56 (272.36)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(14.75, 1,278.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWandsworth to Deptford\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.59 (2.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.07, 18.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,620 (15%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e793 (7.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e8,137 (77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.51 (1.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e(-2.30, 13.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1,334.57 (825.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e(27.67, 3,288.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOverall (N\u0026thinsp;=\u0026thinsp;73,039)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e5.50 (3.18)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e(1.07, 18.15)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e25,275 (35%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e7,301 (10.0%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e40,463 (55%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e4.17 (2.46)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e(-2.30, 17.70)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e795.87 (604.67)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e(0.07, 3,288.60)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4 Empirical Strategy: Identifying Spatial Heterogeneity in Climate Risk Pricing","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Baseline Hedonic Specification of Housing Price Formation\u003c/h2\u003e \u003cp\u003eThe empirical analysis begins with a conventional hedonic price specification that relates residential property values to structural characteristics, neighbourhood attributes, and environmental exposure variables. Hedonic theory interprets housing prices as equilibrium outcomes reflecting the implicit prices of dwelling attributes in competitive housing markets (Rosen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1974\u003c/span\u003e). Buyers select properties based on bundles of characteristics, and the observed transaction price reveals the marginal willingness to pay for these attributes. Structural features such as floor area and property type capture housing consumption value, while neighbourhood characteristics reflect local amenities, accessibility, and socioeconomic conditions (Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Environmental exposure variables are incorporated to examine whether housing markets internalize climate-related risks associated with the Thames Estuary floodplain. Flood risk indicators, elevation, and proximity to tidal waterways represent potential sources of environmental disamenity that may influence housing demand through expected damage risk, insurance costs, or perceived vulnerability. Previous research demonstrates that such environmental hazards may be capitalized into property values when buyers incorporate risk information into location decisions (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe baseline hedonic specification is expressed as\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{l}\\text{n}\\left({P}_{i}\\right)={\\beta\\:}_{0}+{\\beta\\:}_{1}{S}_{i}+{\\beta\\:}_{2}{N}_{i}+{\\beta\\:}_{3}{E}_{i}+{ϵ}_{i}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{i}\\)\u003c/span\u003e\u003c/span\u003edenotes the transaction price of property \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{i}\\)\u003c/span\u003e\u003c/span\u003erepresents structural housing attributes, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{i}\\)\u003c/span\u003e\u003c/span\u003edenotes neighbourhood characteristics, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{E}_{i}\\)\u003c/span\u003e\u003c/span\u003ecaptures environmental exposure variables including flood zone classification, elevation, and distance to river frontage. The dependent variable is specified in logarithmic form to account for skewness in housing price distributions and to allow estimated coefficients to be interpreted as approximate percentage changes in property values (Malpezzi, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Structural attributes include total floor area, number of habitable rooms, and property type indicators, which capture differences in dwelling size and building form. Neighbourhood characteristics incorporate measures of commuting accessibility and local income levels that reflect spatial variation in housing demand. Environmental variables represent both physical exposure to flood hazards and proximity to estuarine amenities. Distance to river frontage may therefore capture competing valuation effects, as waterfront proximity can simultaneously generate amenity benefits and increased exposure to flooding risk.\u003c/p\u003e \u003cp\u003eEstimating the baseline hedonic model provides an initial assessment of how environmental exposure variables correlate with housing prices across the Thames Estuary housing market. However, the specification imposes constant marginal effects across the entire study region. This assumption may be restrictive in estuarine environments where flood exposure, topography, and neighbourhood characteristics vary significantly across short geographic distances. If the marginal impact of environmental variables differs across locations, global hedonic estimates may mask important spatial variation in climate risk capitalization. The next subsection evaluates whether spatial dependence is present in the housing price residuals and therefore whether spatial modelling approaches are required.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Testing for Spatial Dependence in Housing Price Residuals\u003c/h2\u003e \u003cp\u003eHousing markets are inherently spatial systems in which property values tend to cluster geographically. Properties located near one another frequently share similar neighbourhood characteristics, environmental conditions, and accessibility to amenities. These spatial interactions often generate dependence in housing prices that violates the independence assumptions underlying conventional regression models (Anselin, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Pace and LeSage, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). When such spatial dependence is present, global hedonic models may produce inefficient estimates and mask localized price dynamics. To evaluate whether spatial dependence exists in the baseline hedonic model, the analysis examines the spatial autocorrelation structure of the regression residuals. Spatial autocorrelation measures the extent to which nearby observations exhibit similar values relative to those located further apart. Positive spatial autocorrelation occurs when geographically proximate properties exhibit similar pricing patterns, while negative spatial autocorrelation arises when neighbouring values differ systematically. The presence of spatial dependence is tested using Moran\u0026rsquo;s I statistic, which provides a global measure of spatial autocorrelation. The statistic is defined as\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:I=\\frac{n}{W}\\frac{\\sum\\:_{i=1}^{n}\\sum\\:_{j=1}^{n}{w}_{ij}({e}_{i}-\\stackrel{\\prime }{e})({e}_{j}-\\stackrel{\\prime }{e})}{\\sum\\:_{i=1}^{n}({e}_{i}-\\stackrel{\\prime }{e}{)}^{2}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{i}\\)\u003c/span\u003e\u003c/span\u003edenotes the residual from the baseline hedonic regression for property \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{\\prime }{e}\\)\u003c/span\u003e\u003c/span\u003erepresents the mean residual, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{w}_{ij}\\)\u003c/span\u003e\u003c/span\u003eis the spatial weight describing the proximity between observations \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003eand \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:n\\)\u003c/span\u003e\u003c/span\u003edenotes the number of observations, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:W\\)\u003c/span\u003e\u003c/span\u003eis the sum of all spatial weights. Positive values of Moran\u0026rsquo;s I indicate spatial clustering of similar residual values, while values close to zero suggest spatial randomness. Significant spatial autocorrelation in the residuals would imply that the baseline hedonic specification fails to capture important spatial processes influencing housing prices. In the context of the Thames Estuary housing market, such spatial dependence may arise from localized flood exposure, neighbourhood-level amenities, or spatial spillovers in housing demand. Previous studies have shown that ignoring spatial dependence can bias estimates of housing price determinants and obscure localized price relationships (Pace and LeSage, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Anselin and Rey, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eEvidence of spatial dependence motivates the use of modelling approaches capable of capturing spatial heterogeneity in housing price relationships. Global spatial models such as spatial lag or spatial error specifications account for spatial dependence but continue to impose constant parameter estimates across the study region. Estuarine housing markets characterized by substantial geographic variation in environmental exposure may require more flexible approaches that allow coefficients to vary locally across space. Geographically weighted regression provides such a framework by estimating location-specific parameter estimates using spatially weighted observations (Fotheringham et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The following section introduces the GWR model used to identify spatial variation in the capitalization of climate risk within the Thames Estuary housing market.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Geographically Weighted Regression for Spatially Varying Housing Price Effects\u003c/h2\u003e \u003cp\u003eEvidence of spatial dependence in the residuals of the baseline hedonic model suggests that housing price relationships may vary geographically across the Thames Estuary region. Estuarine housing markets combine heterogeneous environmental exposure, varying neighbourhood conditions, and localized demand patterns. Under such conditions, the marginal effects of environmental attributes such as flood exposure or elevation are unlikely to remain constant across space. Global hedonic models impose uniform coefficients across the entire study region and therefore cannot capture localized differences in the capitalization of environmental risk. Geographically weighted regression (GWR) provides a modelling framework that allows regression coefficients to vary across geographic locations. The approach estimates a local regression equation at each observation point using spatially weighted neighbouring observations. This framework enables the identification of spatially varying relationships between housing prices and property characteristics while maintaining the interpretability of hedonic modelling structures (Fotheringham et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Instead of producing a single global coefficient for each explanatory variable, GWR generates location-specific parameter estimates that reflect local housing market conditions.\u003c/p\u003e \u003cp\u003eThe GWR model can be expressed as:\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:\\text{l}\\text{n}\\left({P}_{i}\\right)={\\beta\\:}_{0}({u}_{i},{v}_{i})+\\sum\\:_{k=1}^{K}{\\beta\\:}_{k}({u}_{i},{v}_{i}){X}_{ik}+{ϵ}_{i}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{i}\\)\u003c/span\u003e\u003c/span\u003edenotes the transaction price of property \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{ik}\\)\u003c/span\u003e\u003c/span\u003erepresents the set of explanatory variables including structural, neighbourhood, and environmental characteristics, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left({u}_{i},{v}_{i}\\right)\\)\u003c/span\u003e\u003c/span\u003edenotes the geographic coordinates of property \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e. The parameters \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\beta\\:}_{k}({u}_{i},{v}_{i})\\)\u003c/span\u003e\u003c/span\u003evary across space, allowing the marginal impact of housing attributes to differ by location. This formulation enables the estimation of localized housing price functions that reflect spatial heterogeneity in market behaviour.\u003c/p\u003e \u003cp\u003eParameter estimation in GWR relies on spatial weighting schemes that assign greater influence to observations located closer to the regression point. Observations further away receive lower weights according to a kernel function. The weighting structure is defined by\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:{w}_{ij}=K\\left(\\frac{{d}_{ij}}{b}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{w}_{ij}\\)\u003c/span\u003e\u003c/span\u003erepresents the spatial weight between observations \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003eand \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{d}_{ij}\\)\u003c/span\u003e\u003c/span\u003edenotes the geographic distance between the two observations, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:b\\)\u003c/span\u003e\u003c/span\u003eis the bandwidth parameter controlling the spatial extent of the kernel, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:K(\\cdot\\:)\\)\u003c/span\u003e\u003c/span\u003erepresents the kernel function. Kernel weighting ensures that parameter estimates at each location primarily reflect local housing market conditions rather than distant observations.\u003c/p\u003e \u003cp\u003eBandwidth selection is a critical component of the GWR estimation procedure. The bandwidth determines the spatial scale over which local relationships are estimated. Smaller bandwidths capture highly localized variation but may increase estimation variance, while larger bandwidths approximate global models by incorporating broader spatial information. The optimal bandwidth is typically selected using cross-validation or information criteria that balance model fit and parameter stability (Fotheringham et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Applying GWR to the Thames Estuary housing market allows the empirical analysis to identify spatially varying capitalization patterns associated with flood exposure, elevation, and proximity to the river. Estuarine housing markets often exhibit localized trade-offs between waterfront amenities and environmental risk. Properties located close to the river may benefit from amenity value in some neighbourhoods while facing flood exposure penalties in others. Local parameter estimation enables the identification of these heterogeneous valuation effects that remain hidden in global regression models. Mapping the estimated coefficients from the GWR model provides insight into the geographic structure of climate risk capitalization across the study region. Spatial variation in the coefficients associated with flood risk variables can reveal where housing markets internalize environmental exposure more strongly and where such risks appear underpriced. These localized estimates therefore provide a spatially explicit perspective on how climate exposure influences housing values within the Thames Estuary.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 Empirical Results: Spatial Capitalization of Flood Risk in the Thames Estuary","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Spatial Dependence in Housing Prices\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e3\u003c/span\u003e reports the global spatial diagnostics for the residuals of the baseline hedonic specification. The evidence rejects the view that the remaining pricing errors are spatially random. Moran\u0026rsquo;s \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:I\\)\u003c/span\u003e\u003c/span\u003eequals 0.1397 and is highly significant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left(z=75.653,\\text{\\hspace{0.25em}\\hspace{0.05em}}p\u0026lt;0.001\\right)\\)\u003c/span\u003e\u003c/span\u003e, indicating strong positive residual autocorrelation. Properties located close to one another therefore share systematically similar unexplained price components, even after controlling for structural characteristics, neighbourhood conditions, and flood-related variables. This result is economically important. It implies that the baseline OLS model does not fully absorb the spatial organization of housing values in the Thames Estuary, where transaction prices are shaped by clustered amenities, local reputation effects, micro-location advantages, and environmental conditions that are not evenly distributed across space. Such residual clustering is a classic signal of misspecification in housing price models estimated on spatially embedded markets (Anselin, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Pace and LeSage, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; LeSage and Pace, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). The Rao\u0026rsquo;s Score diagnostics show that this spatial structure is not confined to a single mechanism. Both the LM Error and LM Lag statistics are highly significant, and the robust versions of both tests remain highly significant as well. This pattern is more demanding than a simple finding of residual autocorrelation. A significant lag process suggests that prices in one location are partly anchored by prices observed in nearby locations, which is consistent with the role of comparable transactions, neighbourhood search, and localized market learning in residential price formation. A significant error process indicates that omitted influences are spatially clustered, such as local environmental quality, flood-defence credibility, waterfront redevelopment intensity, or neighbourhood-specific demand conditions. The coexistence of both processes is therefore consistent with a housing market in which spatial spillovers and omitted local fundamentals jointly shape valuation outcomes, a result that has been widely emphasized in spatial housing and urban econometric research (Can, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Anselin and Rey, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Holly, Pesaran and Yamagata, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2011\u003c/span\u003e)\u003c/p\u003e \u003cp\u003eThis matters directly for inference. Once residuals are spatially correlated, the classical OLS assumption of independently distributed disturbances is violated. Coefficient estimates may remain interpretable as conditional associations, but standard errors and significance tests become unreliable, and the model no longer provides a sufficient representation of the underlying price-generating process (Anselin, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; LeSage and Pace, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). In the present setting, the diagnostics indicate that global hedonic estimates are averaging across estuarine submarkets that differ sharply in flood exposure, topography, accessibility, and adaptation context. A single global coefficient on flood risk or river proximity is therefore unlikely to represent the true structure of capitalization across the study region. The diagnostics also have a substantive implication for the paper\u0026rsquo;s central argument. If the residual spatial process remains strong after controlling for flood zones, elevation, river distance, income, commuting accessibility, and housing structure, then climate-risk capitalization is unlikely to be spatially homogeneous. Global models may detect an average relationship, but they cannot reveal whether flood exposure is discounted in some parts of the estuary and offset by amenity or redevelopment pressures in others. Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e3\u003c/span\u003e therefore does more than justify a technical correction. It shows that the Thames Estuary housing market is characterized by spatially differentiated price formation, which requires both global spatial econometric benchmarks and local parameter estimation. The next section turns to SEM and SDM estimates before the analysis proceeds to geographically weighted regression, where the local structure of flood-risk capitalization can be identified explicitly.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSpatial Autocorrelation Diagnostics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eMoran\u0026rsquo;s I Test for Residual Spatial autocorrelation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMoran\u0026rsquo;s I\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e0.1397\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExpected I\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e-0.00038\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e0.00000343\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ez-value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e75.653\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eRao\u0026rsquo;s Score (Lagrange Multiplier) Tests\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLM Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5674.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLM Lag\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5929.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRobust LM Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e583.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRobust LM Lag\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e839.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003e\u003cb\u003eNotes\u003c/b\u003e: Moran\u0026rsquo;s I tests for global spatial autocorrelation in the residuals of the baseline hedonic regression. Rao\u0026rsquo;s Score (Lagrange Multiplier) diagnostics examine the presence of spatial lag and spatial error dependence. Robust tests control for the presence of the alternative spatial process. All statistics are calculated using the spatial weights matrix \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{W}\\)\u003c/span\u003e\u003c/span\u003e.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Global Hedonic and Spatial Econometric Estimates\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e4\u003c/span\u003e reports the baseline hedonic estimates together with the Spatial Error Model (SEM) and the Spatial Durbin Model (SDM). The results confirm that the fundamental structure of the housing price equation behaves consistently with the theoretical predictions of hedonic price models. Total floor area exhibits a strong and highly significant positive association with transaction prices across all specifications. Larger dwellings command systematically higher values, reflecting the fundamental role of interior living space in residential price formation. Property type effects follow expected patterns as well. Houses sell at a premium relative to the reference category, while flats exhibit negative coefficients in the global OLS specification. These patterns align with the long-established empirical literature on housing valuation, which consistently finds structural dwelling attributes to be among the most important determinants of property prices (Rosen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1974\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Neighbourhood attributes also display economically coherent relationships with housing prices. Local income levels show a positive and statistically significant association with transaction values across all models, indicating that higher-income areas command systematically higher property prices. This pattern reflects the role of neighbourhood purchasing power and local demand conditions in shaping residential land values, a central prediction of urban bid\u0026ndash;rent theory (Alonso, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1964\u003c/span\u003e; Glaeser et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Accessibility also plays a measurable role. Travel time to employment centres is negatively associated with property prices, suggesting that longer commuting distances impose a spatial discount on housing values. This finding is consistent with the long-standing observation that proximity to employment and transport networks remains a key driver of urban residential location choices (Muth, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1969\u003c/span\u003e; Gibbons and Machin, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Environmental variables provide the first indication of climate-related capitalization in the global specification. Elevation exhibits a positive and statistically significant coefficient across all models, indicating that properties located at higher ground levels command measurable price premiums. This result is consistent with the hypothesis that buyers internalize physical flood exposure when forming housing valuations. Distance to river channels also enters positively, suggesting that properties located farther from tidal waterways tend to command higher values once structural and neighbourhood characteristics are controlled for. These effects imply that aspects of physical flood exposure are reflected in market prices even within the global specification.\u003c/p\u003e \u003cp\u003eHowever, the coefficients for the flood-zone indicators reveal a more complex pattern. Properties located within both medium and high flood-risk zones exhibit positive and statistically significant coefficients in the global models. At face value, this suggests that properties exposed to flood risk are associated with higher transaction prices. Such a result appears counterintuitive relative to the expectation that environmental hazards should generate valuation discounts. Similar findings have occasionally been reported in coastal housing markets where the amenity value of waterfront proximity dominates perceived hazard exposure (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). In such settings, households may accept environmental risk in exchange for locational advantages such as waterfront views, proximity to urban cores, or redevelopment opportunities. The spatial econometric estimates confirm that housing price formation in the Thames Estuary exhibits strong spatial interaction. The SEM reports a highly significant spatial error coefficient \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left(\\lambda\\:=0.1277\\right)\\)\u003c/span\u003e\u003c/span\u003e, indicating that unobserved spatially correlated factors influence housing prices beyond the explanatory variables included in the model. These factors may include localized environmental amenities, flood-protection infrastructure, neighbourhood reputation effects, or other spatially clustered determinants of residential desirability. The SDM similarly detects spatial dependence through the spatial autoregressive parameter, implying that housing prices are partially influenced by prices observed in nearby locations. Such spatial spillovers reflect the role of comparable transactions and local information diffusion in shaping residential market expectations (LeSage and Pace, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Holly et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Although these spatial econometric models address the residual spatial dependence identified in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e3\u003c/span\u003e, they retain an important restriction: marginal effects remain constant across space. In a coastal housing system characterized by heterogeneous flood exposure, elevation gradients, and localized adaptation policies, this assumption may obscure substantial variation in how environmental risk is capitalized into housing prices. The positive flood-risk coefficients observed in the global models may therefore reflect aggregation across submarkets in which flood exposure is discounted in some locations but offset by amenity and redevelopment effects in others. Identifying such spatial heterogeneity requires a modelling framework that allows price gradients to vary across locations. The following section therefore applies geographically weighted regression to estimate local housing price relationships and to examine how flood-risk capitalization differs across the Thames Estuary.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eGlobal Hedonic and Spatial Econometric Model Estimates\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOLS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSEM\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSDM\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.740*** (0.0137)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10.840*** (0.0316)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10.894*** (0.0134)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal Floor Area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00547*** (0.000054)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00409*** (0.000059)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.00548*** (0.000054)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eProperty Type:\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFlat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0514*** (0.00703)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.2309*** (0.00915)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.0513*** (0.00464)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHouse\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.1537*** (0.00654)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0858*** (0.00811)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.1551*** (0.00406)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaisonette\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0103 (0.00991)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.1773*** (0.0115)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.0116\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIncome After Housing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.37e-05*** (3.03e-07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.58e-05*** (3.03e-07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.43e-05*** (3.03e-07)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTravel Time to Employment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0143*** (0.000545)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.0124*** (0.00136)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.0145*** (0.000535)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eFlood Risk\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedium\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0529*** (0.00654)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0691*** (0.0146)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.0523*** (0.00654)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.1001*** (0.00547)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.1106*** (0.0129)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.1004*** (0.00527)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eElevation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0137*** (0.00102)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0163*** (0.00248)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.0137*** (0.000971)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDistance to River\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.000135*** (2.94e-06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.000143*** (7.27e-06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.000136*** (2.96e-06)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eSpatial Parameters \u0026amp; Model Diagnostics\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpatial Coeff\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.00148*** (Rho, ρ)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.1277*** (Lambda, λ)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLog-Likelihood\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-1999.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-35.586.96\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAIC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e71,850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4,025.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e71,200\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.382\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e64,947\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64,947\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64,947\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cb\u003eNote\u003c/b\u003e: Dependent variable is the logarithm of transaction price. Standard errors are reported in parentheses. The Spatial Error Model (SEM) accounts for spatial correlation in the regression error term, while the Spatial Durbin/Lag Model incorporates spatial dependence in the dependent variable through the spatial autoregressive parameter \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\rho\\:\\)\u003c/span\u003e\u003c/span\u003e. Spatial weights are constructed using a symmetric binary matrix based on the k\u0026thinsp;=\u0026thinsp;8 nearest neighbours to maintain model stability across 64,947 observations. Due to the Sparse Matrix (Matrix) estimation method, the SEM Log-Likelihood and AIC operate on a different numerical scale than the OLS; model selection is therefore guided by the significance of the spatial parameters and the Likelihood Ratio (LR) test. Significance levels: *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, * p\u0026thinsp;\u0026lt;\u0026thinsp;0.10.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e5.3 Spatial Heterogeneity in Flood-Risk Capitalization\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e maps the geographically weighted regression estimates of the local coefficient on high flood exposure across the Thames Estuary housing market. The results reveal substantial spatial heterogeneity in the capitalization of flood risk. While the global models in Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e4\u003c/span\u003e produce a single positive coefficient for flood-zone exposure, the local estimates vary widely across space, ranging from strongly negative values in the outer estuary to positive values in parts of the western urban corridor. The global estimate therefore masks the coexistence of opposing valuation processes operating within different submarkets. The eastern estuary displays predominantly negative coefficients. Areas such as Leigh-on-Sea and the Isle of Grain exhibit some of the largest discounts associated with flood exposure. In these locations, the housing market appears to internalize environmental risk directly, with buyers requiring a measurable price concession to compensate for elevated flood probability. Such capitalization is consistent with the environmental risk literature, which shows that households discount property values when physical exposure to natural hazards becomes salient and difficult to insure or mitigate (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). In contrast, several western policy units exhibit positive coefficients, indicating that flood exposure is associated with higher property values. Locations along the Wandsworth\u0026ndash;Deptford corridor and parts of Thamesmead display positive price effects despite regulatory classification within high-risk zones. These patterns suggest that waterfront amenities, transport accessibility, and urban redevelopment pressures may offset perceived environmental risk in high-demand urban submarkets. Housing markets frequently exhibit such amenity-risk trade-offs, particularly where coastal or riverfront locations combine environmental exposure with strong locational advantages (Glaeser et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The spatial gradient observed in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e therefore indicates that flood-risk capitalization is conditional on local market context rather than uniform across the estuary. In dense urban locations, amenity value and redevelopment expectations appear to dominate risk perception, producing positive capitalization effects. In peripheral coastal communities, physical exposure becomes the dominant factor shaping housing prices. Global hedonic and spatial econometric models cannot capture such variation because they impose constant marginal effects across space. Geographically weighted regression reveals that the average flood-risk coefficient reported in Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e4\u003c/span\u003e reflects an aggregation of spatially heterogeneous pricing regimes rather than a uniform market response.\u003c/p\u003e \u003cp\u003eThis spatial heterogeneity has important implications for both empirical modelling and climate-risk policy. If environmental risk is capitalized differently across local housing markets, global price models may systematically misrepresent the distribution of climate exposure embedded in property values. The following section therefore examines spatial variation in model explanatory power to assess where the hedonic framework performs well and where additional unobserved factors may influence housing prices.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e5.4 Spatial Variation in Model Explanatory Power\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e maps the spatial distribution of the local coefficient of determination obtained from the geographically weighted regression. The results indicate substantial variation in the explanatory power of the hedonic specification across the Thames Estuary housing market. Local \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003evalues range from approximately 0.18 in several redevelopment zones to values exceeding 0.60 in established residential markets.\u003c/p\u003e \u003cp\u003eHigher explanatory power is observed in mature residential areas such as Leigh-on-Sea and Wandsworth. In these locations, structural housing characteristics, neighbourhood income, commuting accessibility, and environmental variables jointly explain a large share of property price variation. Housing markets in these areas therefore appear to conform closely to the standard hedonic framework in which prices reflect the capitalization of structural and locational attributes (Rosen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1974\u003c/span\u003e; Sirmans et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). The strong explanatory power in these zones suggests that market participants consistently price the observable attributes included in the model.\u003c/p\u003e \u003cp\u003eLower local \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003evalues emerge in areas undergoing rapid urban transformation, particularly within parts of the Royal Docks redevelopment corridor. In these locations, a substantial portion of housing price variation remains unexplained by the structural, neighbourhood, and environmental variables included in the model. One plausible explanation is the presence of additional valuation drivers not captured in the hedonic specification, such as speculative expectations linked to regeneration projects, future transport infrastructure, or major waterfront redevelopment schemes. Empirical studies of urban regeneration areas frequently document similar patterns, where large-scale planning interventions alter property prices in ways that are only partially explained by conventional housing attributes (Gibbons and Machin, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Ahlfeldt et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe spatial pattern of model fit therefore reinforces the argument that housing price formation in the Thames Estuary is heterogeneous across submarkets. In established residential areas, observable structural and environmental attributes explain most price variation. In rapidly transforming districts, unobserved development dynamics play a more prominent role. This spatial variation also helps explain the heterogeneous flood-risk capitalization identified in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Where redevelopment expectations dominate price formation, environmental risk may exert a weaker influence on observed transaction prices.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e5.5 Spatial Clustering of Flood-Risk Pricing\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents the Local Indicators of Spatial Association (LISA) cluster map for the geographically weighted regression estimates of the flood-risk coefficient. The map identifies statistically significant spatial clusters in the relationship between flood exposure and property values across the Thames Estuary housing market. Unlike the coefficient map in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, which illustrates spatial variation in local price effects, the LISA analysis reveals whether similar pricing responses to flood risk are geographically concentrated. Two dominant spatial regimes emerge. First, the eastern estuary exhibits low\u0026ndash;low clusters, where negative flood-risk coefficients are spatially concentrated. In these locations, flood exposure is systematically associated with lower property prices, indicating that the housing market capitalizes environmental risk in a consistent manner. Coastal communities such as Leigh-on-Sea and surrounding estuarine settlements fall within this regime. The clustering of negative coefficients suggests that buyers in these markets respond directly to physical exposure conditions rather than relying primarily on amenity valuation. This pattern is consistent with studies documenting hazard-induced price discounts in flood-exposed housing markets where risk salience is high and insurance or adaptation mechanisms are limited (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSecond, the western urban corridor contains high\u0026ndash;high clusters, where positive flood-risk coefficients occur in spatially contiguous areas. These clusters indicate that properties exposed to flood risk command higher prices in nearby locations as well. Such outcomes suggest that waterfront amenities, accessibility advantages, and urban redevelopment expectations dominate risk perceptions in these high-demand submarkets. The presence of these clusters highlights the spatial concentration of amenity-driven valuation processes, particularly in dense metropolitan environments where environmental risk is perceived as secondary to locational advantages.\u003c/p\u003e \u003cp\u003eThe coexistence of these clusters reveals that flood-risk capitalization is not randomly distributed across the housing market. Instead, distinct spatial regimes characterize how environmental exposure enters housing price formation. In outer estuary markets, flood risk operates primarily as a negative externality that reduces property values. In central urban areas, the same environmental classification coexists with positive price effects due to the concentration of waterfront amenities and development pressures. Global hedonic models cannot capture such regime differentiation because they impose uniform marginal effects across space. Identifying these spatial clusters also has policy implications. If environmental risk is systematically discounted in some regions but not in others, property prices may fail to fully signal underlying climate exposure in high-demand urban submarkets. This creates the potential for localized misalignment between market valuation and environmental risk, particularly in redevelopment corridors where waterfront proximity remains a major determinant of housing demand.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e5.6 Flood-Risk Capitalization across TE2100 Policy Management Units\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e5\u003c/span\u003e summarizes the geographically weighted regression results at the level of the Thames Estuary 2100 (TE2100) Policy Management Units (PMUs). The table reports the median local flood-risk coefficient, the median local model fit, and the proportion of statistically significant coefficients within each policy unit. Aggregating the local estimates in this way allows the analysis to examine whether climate-risk capitalization differs systematically across the estuary\u0026rsquo;s flood-management governance structure. The results reveal substantial variation in flood-risk capitalization across policy units. In the eastern estuary, several units exhibit strongly negative median flood-risk coefficients. The Isle of Grain and Leigh Old Town display median coefficients of approximately\u0026thinsp;\u0026minus;\u0026thinsp;0.30, indicating that properties located within high-risk flood zones are systematically discounted relative to comparable properties outside those zones. These findings suggest that housing markets in these coastal communities internalize environmental exposure directly into property prices. Such patterns are consistent with empirical evidence from coastal housing markets where the salience of flood risk generates measurable price discounts once hazard exposure becomes widely recognized by market participants (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn contrast, several western and central estuary units display positive flood-risk coefficients. Policy units such as Dartford and Erith and Thamesmead exhibit positive median coefficients, implying that properties located within designated flood-risk zones command higher prices on average. These outcomes appear counterintuitive if flood risk is interpreted purely as an environmental disamenity. However, the spatial context of these locations provides an alternative explanation. These areas benefit from strong transport accessibility, ongoing urban redevelopment, and proximity to major employment centres. In such markets, the amenity value of waterfront proximity and redevelopment potential may dominate perceived environmental exposure, producing positive capitalization effects despite regulatory risk classification. The significance rates reported in Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e5\u003c/span\u003e indicate that these patterns are not isolated observations but reflect consistent spatial market behaviour. Units such as Greenwich and Canvey Island exhibit high proportions of statistically significant local coefficients, suggesting that the estimated flood-risk effects represent systematic valuation responses rather than random variation across properties. The results therefore reinforce the spatial heterogeneity identified in the preceding sections: flood-risk capitalization is strongly conditioned by local economic context.\u003c/p\u003e \u003cp\u003eThese findings carry important implications for climate-risk assessment in coastal property markets. If environmental risk is discounted in some locations but offset by amenity or redevelopment value in others, global price models may obscure important spatial variation in how climate exposure is incorporated into property values. Policy frameworks that rely on aggregate price signals may therefore underestimate localized exposure in high-demand urban waterfront markets while overstating risk capitalization in peripheral coastal communities.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of Local Flood-Risk Effects and Model Performance by Policy Management Unit\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePolicy Units (pmus)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eN_Properties\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eMedian_FZ3_Beta\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMedian_Local_R2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSignificance Rate\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Grain (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.2982\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.446692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeigh Old Town and Southend-on-Sea (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3086\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.29231\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.640405\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanvey Island (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.24814\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.434998\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e94.01732\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreenwich (P5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1441\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.05275\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.541409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e98.47328\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEast Tilbury \u0026amp; Mucking Marshes (P3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e393\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.04961\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.525775\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e93.89313\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShell Haven \u0026amp; Fobbing Marshes (P3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e287\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.04933\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.500102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e40.41812\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLondon City (P5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e634\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.04878\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.708452\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e72.87066\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIsle of Dogs \u0026amp; Lea Valley (P5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2782\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.03542\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.532862\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e69.12293\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePurfleet, Grays \u0026amp; Tilbury (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.01973\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.617526\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e36.6263\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoyal Docks (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.00951\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.227243\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8.574664\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSwanscombe \u0026amp; Northfleet (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.00448\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.586845\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth Kent Marshes (P3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e991\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e-0.00329\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.57967\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWandsworth to Deptford (P5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9530\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e0.007242\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.637411\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e69.91605\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBarking \u0026amp; Dagenham (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e0.034057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.336622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e47.325\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRainham Marshes (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2943\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e0.054027\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.412096\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e59.70099\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThamesmead (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6343\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e0.109222\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.370902\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e87.04083\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDartford \u0026amp; Erith (P4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2976\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003e0.156295\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003e0.542342\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003e\u003cb\u003eNote\u003c/b\u003e: Data represents median values for all properties within the specified Policy Management Unit. The \"Significance Rate\" indicates the percentage of properties within that unit where the local flood risk impact is statistically significant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left|t\\right|\u0026gt;1.96\\)\u003c/span\u003e\u003c/span\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"6 Discussion","content":"\u003cp\u003eThe empirical results indicate that the capitalization of flood risk in the Thames Estuary housing market is spatially heterogeneous rather than uniform. Global hedonic and spatial econometric models produce a positive average coefficient for flood-zone exposure, suggesting that properties located within designated flood-risk areas command higher prices. At face value, such a result appears inconsistent with the expectation that environmental hazards reduce property values. However, the geographically weighted regression estimates demonstrate that this global coefficient reflects the aggregation of opposing spatial processes operating across the estuary.\u003c/p\u003e \u003cp\u003eIn the outer estuary and several coastal communities, flood exposure is associated with clear price discounts. These markets appear to internalize environmental risk directly, with buyers requiring compensation for properties located in high-risk areas. This pattern aligns with the environmental hazard literature, which documents significant capitalization of flood risk when exposure becomes salient and when protective infrastructure or insurance coverage is limited (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The spatial clustering of negative coefficients in the eastern estuary suggests that housing markets in these areas respond primarily to physical exposure conditions.\u003c/p\u003e \u003cp\u003eIn contrast, several urban submarkets closer to central London display positive flood-risk coefficients. These results indicate that the amenity value associated with waterfront proximity, urban redevelopment, and accessibility advantages can offset the perceived costs of environmental exposure. Similar trade-offs between environmental risk and amenity value have been documented in coastal housing markets where waterfront access generates strong demand despite the presence of natural hazards (Glaeser et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). In such cases, housing prices reflect the net valuation of multiple locational attributes rather than the isolated effect of flood risk.\u003c/p\u003e \u003cp\u003eThe coexistence of these spatial regimes highlights an important limitation of global housing price models. When environmental risk is capitalized differently across submarkets, a single global coefficient cannot represent the underlying price formation process. Instead, the estimated average effect may obscure local valuation patterns and produce misleading inferences regarding climate-risk pricing. The spatial econometric and geographically weighted regression results therefore suggest that climate-risk capitalization should be examined within a spatially heterogeneous modelling framework.\u003c/p\u003e \u003cp\u003eThe findings also carry implications for the interpretation of market signals in climate adaptation policy. Property prices are often used as indicators of how markets perceive and price environmental risk. However, the results suggest that such signals may vary systematically across space. In some housing markets, particularly in high-demand urban waterfront areas, prices may continue to rise despite environmental exposure. In other locations, particularly in peripheral coastal communities, flood risk appears to be directly capitalized into property values. Consequently, relying on aggregate housing market indicators may lead to an incomplete assessment of climate-risk exposure within coastal urban systems.\u003c/p\u003e"},{"header":"7 Conclusion","content":"\u003cp\u003eThis study examined how flood exposure is capitalized into residential property prices across the Thames Estuary housing market. Using transaction data covering more than seventy thousand properties and a combination of global and local spatial econometric models, the analysis evaluated whether climate-related environmental risk is reflected uniformly in housing values or whether capitalization varies across space. The empirical strategy proceeded from a baseline hedonic specification to spatial econometric models and ultimately to geographically weighted regression, allowing the relationship between flood exposure and housing prices to vary across locations. The results reveal that flood-risk capitalization in the Thames Estuary is spatially heterogeneous. Global models produce a positive coefficient for flood-zone exposure, suggesting that properties located within designated flood-risk areas command higher prices on average. Local estimates demonstrate that this result masks two opposing valuation regimes. In outer estuary markets and several coastal communities, flood exposure is associated with measurable price discounts, indicating that buyers internalize environmental risk when forming housing valuations. In contrast, several urban submarkets closer to central London exhibit positive capitalization effects, where the amenity value of waterfront proximity, accessibility advantages, and redevelopment expectations appear to offset perceived environmental exposure. The global premium observed in the spatial econometric models therefore reflects the aggregation of spatially differentiated pricing processes rather than a uniform market response.\u003c/p\u003e \u003cp\u003eThese findings contribute to the literature on environmental risk capitalization and spatial housing price modelling. Existing studies often estimate average effects of environmental hazards using global hedonic frameworks (Bin and Polasky, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belanger and Bourdeau-Brien, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The present results demonstrate that such approaches may obscure important spatial variation in how environmental exposure is incorporated into property prices. Allowing price gradients to vary across space reveals that the capitalization of flood risk depends strongly on local market conditions, including urban development intensity, accessibility, and the amenity value of waterfront locations. The results also have implications for climate-risk assessment in coastal property markets. Property prices are frequently interpreted as signals of how markets perceive and incorporate environmental risk. However, the spatial heterogeneity identified in this study suggests that such signals may be uneven across housing markets. In high-demand urban waterfront areas, prices may continue to rise despite environmental exposure, potentially masking underlying climate vulnerability. In peripheral coastal communities, by contrast, environmental risk appears to be directly capitalized into housing values. Understanding this spatial differentiation is therefore essential for interpreting housing market responses to climate risk and for designing policies aimed at managing long-term coastal exposure.\u003c/p\u003e \u003cp\u003eFuture research could extend this analysis by examining how the capitalization of flood risk evolves over time as climate adaptation infrastructure, insurance regimes, and regulatory frameworks change. Incorporating temporal dynamics and forward-looking climate projections would provide additional insight into whether housing markets adjust gradually to environmental risk or whether price responses occur primarily after major hazard events.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eO.D.A. conceptualised the study, designed the research framework, provided leadership for spatial econometric analysis, interpreted the results, and wrote the original manuscript draft. T.O. provided structure and guide for the methodology of the original manuscript, and reviewed the draft. P.U. contributed to data preparation, statistical analysis, and assisted with the empirical implementation. A.R.R. supported the development of machine learning components, model evaluation, and contributed to interpretability analysis. F.M. provided expertise in quantitative modelling, supervised aspects of the empirical strategy, and contributed to critical revisions of the manuscript. All authors reviewed, edited, and approved the final version of the manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAhlfeldt, G. M., Redding, S. J., Sturm, D. M., \u0026amp; Wolf, N. (2017). The economics of density: Evidence from the Berlin Wall. \u003cem\u003eEconometrica\u003c/em\u003e, \u003cem\u003e85\u003c/em\u003e(6), 2127\u0026ndash;2189.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlonso, W. (1964). \u003cem\u003eLocation and land use: Toward a general theory of land rent\u003c/em\u003e. Harvard University Press.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnselin, L. (1988). \u003cem\u003eSpatial econometrics: Methods and models\u003c/em\u003e. Kluwer Academic.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnselin, L., \u0026amp; Rey, S. J. (2014). \u003cem\u003eModern spatial econometrics in practice: A guide to GeoDa, GeoDaSpace and PySAL\u003c/em\u003e. GeoDa.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAtreya, A., \u0026amp; Ferreira, S. (2015). Seeing is believing? Evidence from property prices in inundated areas. \u003cem\u003eRisk Analysis\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(5), 828\u0026ndash;848.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelanger, V., \u0026amp; Bourdeau-Brien, M. (2018). The impact of flood risk on the price of residential properties: The case of England. \u003cem\u003eJournal of Real Estate Literature\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(1), 63\u0026ndash;86.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeltr\u0026aacute;n, A., Maddison, D., \u0026amp; Elliott, R. J. R. (2018). Is flood risk capitalised into property values? \u003cem\u003eEcological Economics\u003c/em\u003e, \u003cem\u003e146\u003c/em\u003e, 668\u0026ndash;685.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBernstein, A., Gustafson, M., \u0026amp; Lewis, R. (2019). Disaster on the horizon: The price effect of sea level rise. \u003cem\u003eJournal of Financial Economics\u003c/em\u003e, \u003cem\u003e134\u003c/em\u003e(2), 253\u0026ndash;272.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBin, O., \u0026amp; Polasky, S. (2004). Effects of flood hazards on property values: Evidence before and after Hurricane Floyd. \u003cem\u003eLand Economics\u003c/em\u003e, \u003cem\u003e80\u003c/em\u003e(4), 490\u0026ndash;500.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBin, O., Kruse, J., \u0026amp; Landry, C. (2008). Flood hazards, insurance rates, and amenities: Evidence from the coastal housing market. \u003cem\u003eJournal of Risk and Insurance\u003c/em\u003e, \u003cem\u003e75\u003c/em\u003e(1), 63\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBitter, C., Mulligan, G. F., \u0026amp; Dall\u0026rsquo;erba, S. (2007). Incorporating spatial variation in housing attribute prices: A comparison of geographically weighted regression and the spatial expansion method. \u003cem\u003eJournal of Geographical Systems\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(1), 7\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrunsdon, C., Fotheringham, A. S., \u0026amp; Charlton, M. (1996). Geographically weighted regression: A method for exploring spatial nonstationarity. \u003cem\u003eGeographical Analysis\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 281\u0026ndash;298.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCan, A. (1992). Specification and estimation of hedonic housing price models. \u003cem\u003eRegional Science and Urban Economics\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(3), 453\u0026ndash;473.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCohen, J. P., Coughlin, C. C., \u0026amp; Lopez, R. (2019). The boom and bust of US housing prices from various geographic perspectives. \u003cem\u003eJournal of Housing Economics\u003c/em\u003e, \u003cem\u003e43\u003c/em\u003e, 1\u0026ndash;13.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDes Rosiers, F., Th\u0026eacute;riault, M., \u0026amp; Villeneuve, P. (2011). Sorting out access and neighbourhood factors in hedonic price models. \u003cem\u003eJournal of Property Research\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 291\u0026ndash;315.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDubin, R. A. (1988). Estimation of regression coefficients in the presence of spatially autocorrelated error terms. \u003cem\u003eReview of Economics and Statistics\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(3), 466\u0026ndash;474.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDumm, R. E., Sirmans, G. S., \u0026amp; Smersh, G. T. (2016). The capitalization of flood insurance premiums into house prices. \u003cem\u003eJournal of Real Estate Finance and Economics\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(3), 316\u0026ndash;339.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFotheringham, A. S., Brunsdon, C., \u0026amp; Charlton, M. (2002). \u003cem\u003eGeographically weighted regression: The analysis of spatially varying relationships\u003c/em\u003e. Wiley.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibbons, S., \u0026amp; Machin, S. (2008). Valuing school quality, better transport, and lower crime: Evidence from house prices. \u003cem\u003eOxford Review of Economic Policy\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(1), 99\u0026ndash;119.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlaeser, E. L., Kolko, J., \u0026amp; Saiz, A. (2008). Consumer city. \u003cem\u003eJournal of Economic Geography\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(1), 27\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoodman, A. C., \u0026amp; Thibodeau, T. G. (2003). Housing market segmentation and hedonic prediction accuracy. \u003cem\u003eJournal of Housing Economics\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(3), 181\u0026ndash;201.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGourevitch, J. D., Kahn, M. E., \u0026amp; Walsh, R. (2023). Climate adaptation and housing markets. \u003cem\u003eAnnual Review of Resource Economics\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e, 327\u0026ndash;347.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolly, S., Pesaran, M. H., \u0026amp; Yamagata, T. (2011). The spatial and temporal diffusion of house prices in the UK. \u003cem\u003eJournal of Urban Economics\u003c/em\u003e, \u003cem\u003e69\u003c/em\u003e(1), 2\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKahn, M. E., Mohaddes, K., Ng, R., Pesaran, M. H., Raissi, M., \u0026amp; Yang, J. C. (2021). Long-term macroeconomic effects of climate change: A cross-country analysis. \u003cem\u003eEnergy Economics\u003c/em\u003e, \u003cem\u003e104\u003c/em\u003e, 105624.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeys, B. J., \u0026amp; Mulder, P. (2020). Neglected no more: Housing markets, mortgage lending, and sea level rise. \u003cem\u003eReview of Financial Studies\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(7), 2902\u0026ndash;2945.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeSage, J., \u0026amp; Pace, R. K. (2009). \u003cem\u003eIntroduction to spatial econometrics\u003c/em\u003e. CRC.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalpezzi, S. (2002). Hedonic pricing models: A selective and applied review. In T. O\u0026rsquo;Sullivan, \u0026amp; K. Gibb (Eds.), \u003cem\u003eHousing economics and public policy\u003c/em\u003e. Blackwell.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuth, R. F. (1969). \u003cem\u003eCities and housing\u003c/em\u003e. University of Chicago Press.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOsland, L. (2010). An application of spatial econometrics in relation to hedonic house price modeling. \u003cem\u003eJournal of Real Estate Research\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(3), 289\u0026ndash;320.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePace, R. K., \u0026amp; LeSage, J. P. (2004). Spatial statistics and models. In J. P. LeSage, \u0026amp; R. K. Pace (Eds.), \u003cem\u003eSpatial econometrics\u003c/em\u003e. CRC.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026aacute;ez, A., Long, F., \u0026amp; Farber, S. (2011). Moving window approaches for hedonic price estimation: An application to housing prices. \u003cem\u003eSpatial Economic Analysis\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(3), 299\u0026ndash;317.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosen, S. (1974). Hedonic prices and implicit markets: Product differentiation in pure competition. \u003cem\u003eJournal of Political Economy\u003c/em\u003e, \u003cem\u003e82\u003c/em\u003e(1), 34\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSirmans, G. S., MacDonald, L., Macpherson, D., \u0026amp; Zietz, E. (2005). The composition of hedonic pricing models. \u003cem\u003eJournal of Real Estate Literature\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 3\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Climate risk, Flood risk capitalization, Housing prices, Spatial econometrics, Geographically weighted regression, Thames Estuary","lastPublishedDoi":"10.21203/rs.3.rs-9213954/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9213954/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnderstanding whether housing markets internalize climate-related flood risk is central to assessing the economic consequences of coastal exposure. This study examines the capitalization of flood risk in residential property prices across the Thames Estuary housing market. Using transaction data for more than 73,000 residential properties, the analysis combines global spatial econometric models with geographically weighted regression to evaluate both the average and spatially varying effects of flood exposure on housing values. Global hedonic and spatial models indicate a positive average association between flood-zone exposure and property prices, suggesting that properties located within designated flood-risk areas command higher values. However, geographically weighted regression reveals substantial spatial heterogeneity in this relationship. In outer estuary communities, flood exposure is associated with significant price discounts, indicating that environmental risk is capitalized directly into housing values. In contrast, several urban waterfront submarkets display positive capitalization effects, where accessibility advantages and waterfront amenities appear to offset perceived environmental exposure. These findings demonstrate that the average flood-risk premium estimated by global models masks the coexistence of distinct spatial pricing regimes. Environmental risk is discounted in peripheral coastal markets but may be offset by amenity and redevelopment value in high-demand urban waterfront locations. The results highlight the importance of spatially heterogeneous modelling when evaluating climate-risk capitalization in housing markets and suggest that aggregate housing price signals may underestimate localized exposure within rapidly developing coastal urban systems.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eJEL Classification Codes:\u003c/strong\u003e R31; R12; C21; Q54\u003c/p\u003e","manuscriptTitle":"Spatial Heterogeneity in Flood-Risk Capitalization: Evidence from the Thames Estuary Housing Market","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-26 15:02:40","doi":"10.21203/rs.3.rs-9213954/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"167fe548-bc15-4556-a1ed-f551d145de99","owner":[],"postedDate":"March 26th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-26T15:02:40+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-26 15:02:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9213954","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9213954","identity":"rs-9213954","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}