|
Valuing the green structure J.M.Halleux, Belgium; april 2002 Text of paper to the COST C 11 "Green structures and urban planning" 6th Management Committee Meeting and Working Group Meetings
|
|
Commented bibliography on the use of hedonic models to assess green structures influences on residential property values
Summary
This working document is based on a bibliographic review about the use of hedonic models to assess green structure influences on residential property values. Its general aim is therefore to assess a specific aspect of green structures performances. The document is structured in three sections. It begins with an introduction about the objectives of the researches taken into account. In section 2, some methodological issues are briefly commented, in particular the economic definition of the hedonic price function and the specific difficulties related to a hedonic model on green equipments. In section 3, the major results are presented and different performances are taken into account : aesthetic performances at the lot scale, aesthetic performances at the neighbourhood scale and recreational benefits.
Structure 1. Introduction 1.1. Research objectives : assessing the socio-economic value of ecological factors 1.2. Available economic valuation techniques 2. Methodological issues 2.1. The hedonic house price function 2.2. Methodological problems 3. Major results 3.1. Typology of green structures performances on the basis of HPM literature 3.2. Aesthetic performances at the lot scale 3.3. Aesthetic performances at the neighbourhood scale 3.4. Recreational benefits 4. Conclusion 5. Bibliography
1. INTRODUCTION 1.1 RESEARCH OBJECTIVES : ASSESSING THE SOCIO-ECONOMIC VALUE OF ECOLOGICAL FACTORS
1.1.1 Fundamental principle of the hedonic price method (HPM) A significant feature of the housing market is the fact that housing is not a homogeneous commodity. The prices at which houses and apartments are traded result in the consumer jointly acquiring a building, an accessibility potential as well as environmental characteristics tied together in a particular bundle. Therefore, by observing the determination of market prices, it is possible to infer the values attached to component characteristics in the bundle and, these values can then be used to evaluate the benefits of changes in the level of provision of these characteristics. The theoretical framework in which most of this work has taken place is that of the hedonic price theory (S. Rosen, 1974). The fundamental principle of the hedonic price method (HPM) was explicitly formulated by K. Lancaster (1966) : the utility provided by heterogeneous (composite) goods is based upon the utility yielded by their various attributes, rather than the good itself.
1.1.2 Interactions between economy and environmental characteristics Hedonic researches on interactions between green structures and residential property values are integrated into a broader theme related to interactions between economy and environmental characteristics. The general aim of this research topic is indeed to measure socio-economic values of ecological factors, as stated, for example, by L. Tyrväinen and J. Luttik : « Urban forests [Green structures] provide many environmental and social benefits, which are well documented in the literature (Robinette, 1972 ; Grey and Deneke, 1978 ; Laurie, 1979 ; Miller, 1988). Most of the values attached to urban forests [green structures] are non-priced environmental benefits. These values include those derived from pleasant landscape, clean air, peace and quiet and screening as well as potential recreational activities in wooded green spaces. Other benefits include reduced speed wind velocity, balanced microclimate, shading and erosion control. The low appreciation by planners of wooded green spaces in urban areas is often based on narrowly-defined economic reasons. One basic problem is that land-uses planning procedures do not include systematic assessment of urban forest [green structures] benefits. The costs of supplying urban forests [green structures] can be determined in a relative straightforward way, but the benefits are more difficult to estimate. In spite of the existing need, there has been little research seeking to determine the amenities values in monetary terms » (L. Tyrväinen, 1997, p. 212).
« If the socio-economic value of ecological factors can be demonstrated through a premium on house price, this strengthens the position of existing green areas in the policy decision process. It may thus act as a counterbalance in urban expansion plans, when urban development threatens green areas and open spaces. One of the most important reasons for the susceptibility of green areas and open spaces to urban pressures is that they are not articulated in monetary terms (More et al., 1988). Decision-makers compare economic factors like contribution to the tax base and employment or the value added to the local economy against the value of environmental factors. By expressing the latter in monetary terms they become comparable to the former. This will put more weight on environmental factors in the decision making process (although by no means all environmental values can be put into monetary terms) » (J. Luttik, 2000, pp. 161-162).
1.1.3 A central hypothesis : green equipments are a source of added value The general hypothesis of researches on interactions between green equipments and residential property values is, of course, that green has a positive effect on property levels. It is therefore hypothesised that green structures enter utility functions positively « and the model of household behaviour predicts that utility differences will be reflected in price differences among housing units. The hypothesis can be rejected if the usual statistical procedures do not find an association between [green structures] (air pollution) and property values. To be sure, observed associations do not prove causation. They may be due to chance or due to correlation between a third variable not included in the regressions and property values. But when such association are found in repeated statistical experiments with different data sets and different cities, they tend to support the hypothesis » (A.M. Freeman, 1979, p. 159).
1.1.4 How does green equipment enter utility function ? Beyond the central assumption of the positive effects of green structures on property levels, a major research aim is also to precise which environmental factors contribute the most to the quality of life, which is an asset in both, fundamental and applied researches : « The effect of an attractive environment on house prices may be intuitively evident, but it is unclear which environmental factors make a location pleasant to live in and of high quality, and how much people are willing to pay for providing these amenities » (L. Tyrväinen, 1999, p. 42). « Insight in the socio-economic value of green areas and open spaces will help to optimise socio-economic and ecological factors simultaneously. One of the issues where these insights may contribute is design of new areas, in particular where the distribution of green areas (including water bodies) and houses over new urban areas is concerned » (J. Luttik, 2000, p. 162).
1.2 AVAILABLE ECONOMIC VALUATION TECHNIQUES
Apart from the HPM, different economic techniques can be used to value environmental amenities and green structures. Perhaps the most common are travel cost and contingent valuation (T.A. More et al., 1988, p. 140 ; F. Facchini, 1994, p. 376). « Travel cost, in its most basic form, asks park users how far they travelled to visit the park. It assumes that the economic value of the experience is the same to all users and that the user who travelled the greatest distance to reach the park is the marginal user. For this person, the economic value of the benefits is assumed to equal the cost of travel ; people who live beyond this point do not use the park because the costs presumably exceed the benefits. People who live closer to the park derive a « consumer surplus » from the park because they receive the same benefit, but have lower travel cost. Under the travel cost model, the consumer's surplus represents the total economic value of the park » (T.A. More et al., 1988, p. 140). However, in relation to urban planning, the travel cost method has actually limited applicability because, in the urban setting, there are often no travel or other expenses involved in accessing the green equipments (L. Tyrväinen (1997, p. 212). Contingent valuation « estimates consumer surplus directly by asking users what they would be willing to pay under various contingencies. The consumer surplus is the difference between what they would be willing to pay and what they are currently paying. As before, the economic value of the park is the aggregated consumer surplus » (T.A. More et al., 1988, p. 140).
2. METHODOLOGICAL ISSUES
2.1 THE HEDONIC HOUSE PRICE FUNCTION
2.1.1 Theory overview Let Z = (z1 ... zn) be a vector of housing attributes. In Rosen's model of implicit markets, the interaction of supply and demand for Z produces a market clearing function P(Z), which relates the vector of housing attributes to the composite price of the house itself, such that : P(Z) = P(z1 ... zn) P(Z) is the hedonic price function, and describes the house prices resulting from the interplay between housing supply and demand. Buyers and sellers take this price function as given in a competitive housing market. In other words, the hedonic function acts as the key parameter for a composite good like the price act for a homogeneous good in the traditional demand theory. The hedonic price function is, therefore, the equilibrium locus of both, bid prices and offer prices. As such, it tells us nothing about the underlying structure of the market. Since the price of a property is a consequence of the price of its housing attributes, P(Z) can be estimated from observations of the prices and attribute bundles of different houses. Moreover, the marginal implicit price of any attribute can be found by differentiating the hedonic price function with respect to that attribute :
P(Z) ) = aP(z) A(z)
2.1.2 Signification of the implicit price It is important to notice that the implicit price of a specific attribute does not depend on an intrinsic worth. Moreover, the marginal price derived from the hedonic function does not measure what a particular household is willing to pay for additional units of a housing characteristic. Rather, it is the value that results from demand and supply interactions on the entire market (R.B. Palmquist, 1991, p. 107 ; P. Le Goffe, 1996, pp. 189-191 ; S. Orford, 1999, p. 60). For instance (about view lot) : « just in the case of other economic goods, the price of a view lot is not dependent on its intrinsic worth, but rather on supply relative to demand at the margin. In some areas, view lots may be so abundant that at low population densities they may constitute a free good and therefore have little or no market value. At higher population densities, however, the supply of view lots may be greatly exceeded by the demand, and thus they would be able to command a premium » (Q. Gillard, 1981, p. 216).
2.1.3 Housing attributes integrated into the hedonic function When the HPM is applied to study housing markets, a distinction is made between two types of housing attributes : dwelling specific or structural attributes (size, number of rooms,...) and location specific attributes (R.K. Wilkinson, 1973). Within the location attributes, it has generally been accepted to distinguish accessibility and neighbourhood quality (J.R. Follain et E. Jemenez, 1985). Traditionally, accessibility represents the distance to the city-centre as proposed in the micro-economics literature by the measurement of the bid-rent curve (e.g. : W. Alonso, 1964 ; R.F. Muth, 1969). Neighbourhood quality is used to cover a large set of influences : measures of local amenities (school, shopping centre,...), measures of the socio-economic status of the neighbourhood, and measures of the quality of the residential environment, for example, view or access to open spaces such as parks and beaches.
2.2 METHODOLOGICAL PROBLEMS
2.2.1 General methodological problems related to the HPM issue There are several difficulties and problems in conducting a comprehensive HPM study. First, the method needs large and precise datasets from restricted time periods. In many European countries, these are difficult to collect. Beyond the data problem, there are also many technical problems in both, economic and geo-statistics terms. Economic problems are notably related to the question of market segmentation. As the hedonic price schedule represents a market's equilibrium, the way in which the market is defined is important. « For example, some environmental hedonic studies for housing have assumed that the scope of the appropriate market is national, whereas other studies have assumed that the scope of the appropriate market is as small as a census tract... If economists assume that there is a single market when it is actually segmented, their coefficients will be biased. On the other hand, if they assume that the markets are segmented when they are not, their estimates will be imprecise and they may have insufficient data in the segments » (R.B. Palmquist, 1991, p. 89).
Another fundamental issue in estimating the hedonic price function is choosing the functional form. Unfortunately, the economic theory does not suggest a particular functional form, as Rosen demonstrated. The hedonic price model was not developed within an economic framework appropriate to study spatial commodities. Indeed, this approach has been developed to analyse any integrated market with heterogeneous commodities, such as cars, washing machines or personal computers. Therefore, analysis on property goods (such as houses) has presented a fundamental problem : how to incorporate space into an aspatial economic model ? Neglecting the spatial element of property commodities will lead to violate the traditional assumptions (of independently, identically distributed error terms) in the econometric model, which can produce distortions (L. Anselin and D.A. Griffith, 1988). More precisely, the spatial effects of spatial autocorrelation and spatial heterogeneity explain many contradictory and counter-intuitive HPM results obtained with standard econometric methods (S. Orford, 1999).
Another complicating factor encountered in empirical HPM is the high correlation between variables (N.A. Powe et al., 1995). « An obvious interrelation, which is difficult to disentangle, is between social status and attractive location. People who can afford to do so have a tendency to choose attractive, green settings for their homes. As a consequence, certain towns or districts in attractive, green settings have become known as places for the rich. Are house buyers in these areas willing to pay a premium for the attractive environment setting or for the social status attached to living among the rich ? » (J. Luttik, 2000, p. 166). For example, in a recent French work on Lyon (C. Beckerich, 1999 and 2000), this interrelation between social and environmental attributes did not allow to test the hypothesis of the positive effects of green structures on apartment prices.
2.2.2 Methodological problems related to the green structure issue : green equipment heterogeneity Compared to other neighbourhood externalities, for instance noise pollution, green structures are extremely heterogeneous. In parallel, green structures performances are multidimen-sional, since many green equipments produce multiple benefits such as pleasant landscape, clean air or potential recreational activities... In practical terms, these two types of heterogeneity are a difficulty for the HPM researcher (L. Tyrväinen and A. Miettinen, 2000, p. 219). It is in fact difficult to select adequate variables when information on the values that residents attach to green structures remains insufficient. For instance, in comparison with the noise influence where the decibel scale can be used as a universal reference, green benefitsare difficult to classify and measure quantitatively. Therefore, there is not any simplequantitative scale to take the green influences into account. As the environmental variables are often correlated to themselves, the researcher has to choose those variables thought to measure different benefits. In many cases this is difficult and compromises have to be made between the multiple benefits, the number of variables, and their measurability. The heterogeneity issue is also a problem for the present exercise of writing a commented bibliography. As no robust green structure typology was available to researchers when thepapers were written, it is often difficult to apprehend the green equipments attributes (size of the park, tree species, landscape attributes...) or the performances that are actually taken into account. Nevertheless, as will be seen below, influence of a green equipment on residential property levels are clearly related to its attributes.
3. MAJOR RESULTS
3.1 TYPOLOGY OF GREEN STRUCTURES PERFORMANCES ON THE BASIS OF HPM LITTERATURE
3.1.1 Theoretical overview
When HPM is applied to study housing markets, as stated above, a traditional distinction is made between structural attributes (at the property scale) and location specific attributes (accessibility and neighbourhood quality). The literature survey has showed that green structures are actually related to both types of attributes. Indeed, although most of the work are related with the neighbourhood scale, several papers have also tackled the impact of the green at the scale of the residential property &endash; with the analysis on the impact of trees in the front yard.
To apprehend the important topic of the neighbourhood quality, the interpretation grid developed by M.M. Li and H.J. Brown (1980) is useful. Conversely to the traditional statement where land uses are characterised as either detrimental or benevolent to housing values, Li and Brown suggest that proximity to specific land uses is source of both, positive and negative impacts. Their formalisation is based on the classification of neighbourhood externalities within three categories : aesthetic attributes, pollution levels (such as congestion, noise,...) and proximity at the micro-scale. Following Li and Brown, it is useful to formalise this typology by taking the distance from the specific land use into account, which give a broad operational vision. It is then assumed that positive influence associated with accessibility and aesthetic performances declines with distance from the studied land use, whereas the negative influence associated with external diseconomies also diminishes with distance (Figure 1).
For a particular distance from the studied land use, the net effect is therefore determined by the three effects. This result can be illustrated by the vertical summation of the three curves (Figure 2).
Although pollution related to green equipments are generally negligible, some authors underline the fact that, in some circumstances, a green equipment can be a source of negative impacts. For instance T.A. More et al. (1988, p. 139) state that proper maintenance is essential to sustain the flow of green benefits as « a deteriorated park may become a social sore point within a neighbourhood, and may also prompt decision makers to consider it for non-park development options » (T.A. More et al., 1988, p. 139). J. Weicher and R. Zerbst (1973) also underline negative impacts when they conclude that « heavily used public parks may cause negative externalities to adjacent houses and may even decrease their prices » (L. Tyrväinen, 1999, p. 21). In this specific context, « pollution » effects are therefore more important than the positive effects associated with both, aesthetic attributes and proximity.
3.1.2 Green performances on the basis of HPM : aesthetic impacts and recreational benefits Although it is difficult to make clear distinction between performance types (in the study field as well as within the analysed publications), the bibliographic review has showed that two kinds of performances are taken into account when green structures influences are appraised by the HPM : aesthetic impacts and accessibility related to recreational benefits. For the whole of the researches taken into account, the view of green has been considered as a positive attribute. Of course, for some other land uses, negative impacts are likely (for instance industrial estates or high buildings). In parallel to aesthetic benefits, many researches have also tackled the accessibility issue. If accessibility is taken into account, it is the recreational benefits that are studied, as it will be
developed below. To precise the performance issue, it is useful to analyse how researchers choose the independent variables that have to be introduced in hedonic models. Two studies have been selected to illustrate this issue. Within European contexts, those two examples appear to be the most detailed. The first one, which was realised by L. Tyrväinen (L. Tyrväinen, 1997 ; L. Tyrväinen and H. Väänänen, 2000), focused on the impact of urban forests on the house prices in the Finnish towns of Joensuu and Salo. The second one, published by J. Luttik and M. Zijlstra (1997), explores the effect of different environmental factors on house prices within eight towns in the Netherlands.
In her first study on the urban forests of Joensuu, L. Tyrväinen chose to construct three variables (L. Tyrväinen, 1997, pp. 215-216). The first variable, the distance to the nearest wooded recreation area, was chosen to describe recreation opportunities. Indeed, wooded recreation areas are large areas (tens of hectares) with facilities as trails and lights to allow practices such as ski, jogging or walking. The two other variables (the relative amount of green areas in the housing districts and the distance to the nearest small forested area) were built to assess landscape amenities. Therefore, they were produced to measure aesthetic performances. The differentiation between aesthetic performances and recreational possibilities was also used by J. Luttik and M. Zijlstra to select independent variables (1997, p. 13). By visiting each of the 3 000 houses of their sample, differentiation was made between, on the one hand, view of various types of attractive environments and, on the other hand, proximity to recreational equipments.
3.1.3 Preference changes depending on the context An important feature of the HPM is that they are local, i.e. based on local housing markets, town structures and cultural preferences. Great caution is therefore required to generalise HPM results as, for both performances, aesthetic and recreational, demand greatly varies across Europe. For instance, it is stated by L. Tyrväinen (1999, p. 35) that it is risky to transfer the Finnish results on forest aesthetic performances to central Europe, where demands for quality in urban greens differ as a result of cultural differences and history on land use. For all that, her results probably apply to other Nordic countries, where wooded environment is also generally appreciated. Furthermore as tastes and practices vary, there are also differences in the characteristics that make a recreational site attractive in each country or region. For instance, the recreational use of Italian forests &endash; measured as the average number of visits &endash; is considerably lower than in Finland (L. Tyrväinen, 1999, p. 35).
3.2 AESTHETIC PERFORMANCES AT THE LOT SCALE In the USA, researches on trees influences on residential sites have been conducted since the 1970s. As previously stated, this type of research is related to structural attributes of the properties rather than to location specific attributes (neighbourhood quality). In terms of typology performance, it is mostly the aesthetic of green which is assessed as we expect trees to raise the value of residential property mainly for aesthetic reasons (trees make the property look more attractive). However, trees in the front yard also provide shade, noise abatements, privacy, wildlife habitat, and wind reduction (L.M. Anderson and H.K. Cordell, 1985, p. 162).
3.2.1 Major results In a paper published in 1973, B.R. Payne reported that trees contribute substantially to the value of residential property. His results showed that hypothetical sales prices were increased by an average of 7 % (from 5 % to 15 %) for houses landscaped with trees. The estimates were received from realtors and owners who compared photographs of similar houses on lots with a differing number of trees. In 1976, Morales looked at 14 variables influencing the price of suburban houses in Manchester, Connecticut. After standardising for other variables affecting house prices, such as the number of rooms, bedrooms, baths, floor area, garages, age, plot size, etc…, trees were estimated to add $ 2 686, or 6 % of the total, to the value of the houses observed, a figure very close to Payne's estimate. In another study, Morales et al. (1983) applied multiple regression to predict the selling price of property in Greece, New York. In this study, houses on lots with tree cover sold at an average of 15 % more than those on untreed lots. The untreed lots sold for an average of $ 51 108 and the tree covered lots sold for $ 60 164. Therefore, it was assumed that $ 9 500 would be the value of the trees.
A study by L.M. Anderson and H.K. Cordell (1984 and 1988) of 844 single family residential property sales in Athens, Georgia, indicated that landscaping with trees was associated with a 3 to 5 % increase in sales prices, not greatly different from the increase of 7 % in Payne's study (1973) and 6 % in Morales' study (1976). As stated by the authors, the smaller estimate of the price increase « may be attributed in part to the relative abundance of trees and other urban vegetation in most residential areas in Athens and surrounding counties » (L.M. Anderson and H.K. Cordell, 1984, p. 165). This conclusion is in line with the theoretical signification of the implicit price. As stated above, the implicit price is not actually dependent upon an intrinsic worth, but on the result of demand and supply interactions. If urban vegetation is abundant, it is therefore coherent to asses lower market value, although intrinsic performances &endash; in terms of quality of life &endash; will continue to be present.
3.2.2 House builders behaviours It is interesting to notice that, in the USA, as wooded lots sell on higher prices, but also more quickly than houses on cleared lots, house builders are aware of the added value generated by trees (L.M. Anderson and H.K. Cordell, 1988, p. 162). Therefore, they have abandoned the practice of clearing all trees from the lot before construction begins. Trees are often left in relatively undisturbed buffer zones between properties, ensuring that a greater proportion survive the construction process. However, only time will tell whether the trees remaining on these lots will survive the numerous stresses and abuses they suffer during the construction process (L.M. Anderson and H.K. Cordell, 1988, p. 162). 3.2.3 Urban sprawl and aesthetic performances at the lot scale As stated, for example, by the Progress Report 2001 (B. Duhem, 2001), the topic « green structure and urban planning » appeared as an important one in the general debate about urban sprawl. It is therefore useful to analyse green aesthetic performances at the light of
3.2.4 Tree species' influence In the study realised by L.M. Anderson and H.K. Cordell (1988, p. 161), tree species were taken into account to analyse the valuation difference between hardwoods and pines. The regression coefficients indicated that hardwoods are slightly more valuable than pines, but that each contributes substantially to property values. In the UK, similar conclusions were found by G. Garrod and K. Willis (1992b, p. 725). By studying the effect of countryside characteristics on property prices, they found a significant positive relationship between broad-leaved woodlands and house prices, and a significant negative relationship between mature coniferous forests and house prices. They concluded that environmental benefits could be increased substantially by decreasing the relative proportion of mature conifers more than 40 years old near residential areas. In the Finnish study on the urban forests of Joensuu, L. Tyrväinen also found similar conclusions as the direct distance to nearest forested park had a negative influence on apartment prices. According to the author, this situation can be understood by the notion that dense, mature coniferous forests may not be appreciated close to a house in these latitudes (L. Tyrväinen, 1997, p. 220). On the basis of the reference grid developed by M.M. Li and H.J. Brown (1980) (see above), the three considered studies leads to the idea that the shading effect can be understood as a « source of pollution ». The direct conclusion is that this effect of trees has to be mitigated by proper management, for example by planting deciduous and low trees near houses.
3.3 AESTHETIC PERFORMANCES AT THE NEIGHBOURHOOD SCALE About aesthetic performances at the neighbourhood scale, the literature review leads to the differentiation between, on the one hand, performances related to « concentrated elements » and, on the other hand, performances related to « diffuse elements ». By concentrated elements, we mean punctual or linear equipments, such as parks, rivers or shorelines. On the opposite, by diffuse elements, we mean general landscape characteristics that can be described by the use of landscape metrics.
3.3.1 Concentrated aesthetic attributes
View of green Unfortunately, in many publications on the impact of environmental amenities on house prices, aesthetic performances are not clearly differentiated from other performances. The first exception to this rule is the study conducted by Q. Gillard (1981) in the Los Angeles area. This analysis, which is based on 1970 data, suggests, ceteris paribus, that a view of parks or open spaces adds some $ 3 887 to the price of residential property, or a little more than the presence of a fireplace and somewhat less than a swimming pool. Given the fact that the mean selling price of the housing units in the sample was $ 42 128, the view-lot premium represents an added value of ± 9 %. In the study on the Finnish town of Salo, realised by L. Tyrväinen and A. Miettinen (2000, p. 212), the view from the dwelling window was also specifically taken into account &endash; in addition to access. This leads to the result that « dwellings with a view onto forest are on average 4,9 % more expensive that dwellings with otherwise similar characteristics » (L. Tyrväinen and A. Miettinen, 2000, pp. 215-216). this phenomenon.The fact that green aesthetics at the lot scale has a positive effects on house prices perfectly correlates with residential mobility analysis where the search for natural amenities is considered as a dynamic leading to sprawl processes. Indeed, when the search for nature is solved by individualistic solutions related to a specific house plot, this can very easily lead to low density and sprawl.
The specific influence of a view on green strips and parks was also tested by J. Luttik and M. Zijlstra (J. Luttik and M. Zijlstra, 1997 ; J. Luttik, 2000). This analysis on several Dutch towns sheds some doubts on the financial impacts of trees aesthetic. Indeed, this hypothesis has only been statistically verified on 5 tests out of 14 case studies. This ambiguity must probably be related with theoretical developments of HPM and it must be reminded that the implicit price is not dependent upon an intrinsic worth, but on the result of demand and supply interactions. In the Netherlands, where green structures have been widely developed &endash; as it was stated during the Cost C11 meeting in Breda &endash;, it is therefore likely that the abundance of urban green leads to minimise their market value (J. Luttik and M. Zijlstra, 1997, pp. 27- 28). From this point of view, a null impact result can be the consequence of effective planning and absence of green shortage. In parallel, it is important to notice that, despite the decreasing market value, green intrinsic performances related to the quality of urban life are still effective.
View of blue In the Dutch study realised by J. Luttik and M. Zijlstra, a major conclusion was that the most influential attribute is the presence of water features. For instance, the largest increase in house prices &endash; up to 28 % &endash; was found for houses with a garden facing water, especially if this water was connected to a sizeable lake. The study of P.B. McLeod (1984) on local amenities in Perth (Western Australia) also underlined the strong influence of water. In this study, it was found that, in addition to recreational sites access, a view of the Swan River commands also a substantial premium of 28 %. According to J. Luttik (2000, p. 166), this « corresponds with findings from landscape psychologist. As is stressed by Kaplan and Kaplan (1989) : "Water is a highly prized element in the landscape" ». On the basis of this conclusion, J. Luttik recommends to include water bodies in town development programmes.
The role of landscape design The major aim of the analysis conducted by T.A. More et al. (1988) on the valuation of urban parks in Worcester, Massachusetts, was to precise how parks' attributes influence property values. Their major conclusion is that parks emphasising open-space may be more effective at maximising property value benefits than parks that offer developed sports facilities. This result, which supports those of J. Weicher and R. Zerbst (1973), leads to underline the role of thoughtful landscaping and designing. Indeed, design is necessary in order to optimise on-site recreational benefits and external aesthetic benefits. « Of particular concern is the zone of interaction where the park and its surroundings meet. Heren buffers of natural vegetation can screen high-use facilities from surrounding properties to both reduce the negative impacts of use and enhance the experiential quality for users by screening out traffic sights and sounds » (T.A. More et al., 1988, p. 150).
3.3.2 Diffuse aesthetic attributes To our knowledge, two studies have tackled the hypothesis that the value of a house is affected by the pattern of surrounding land uses, not just by features concentrated in specific locations. In those studies, landscape indices developed by landscape ecologists were thus used as independent variables.
Major results In her doctoral dissertation focused on remote sensing methods, I. Reginster (1998) built a variable on the global greening within the two Belgian agglomerations of Namur and Charleroi. This variable, which was available at the pixel scale (10 m x 10 m), was, thereafter, correlated with residential land prices. Although significant, this variable seemed to have a weak impact on prices. A deeper analysis focussing on land market mechanisms was conducted by J. Geoghegan et al. (1997). This was realised through a transaction analysis within Washington DC outskirts. In this study, several variables were tested in two buffers surrounding each housing transaction. A first buffer, within a 0,1 km radius, was built to capture the immediate neighbourhood influence (i.e. what can be seen from one's house). The second buffer, within 1 km radius, encompass what is within an easy walk from the house.
Variables such as « percent of open space » (percent forestry and agriculture) and « percent of low density residential' land uses » were first tested. An interesting result to note is that for the percent open space variable. « For the smaller buffer, the marginal contribution of more open space is positive and significant, so more open space in one's immediate neighbour-hood is valued. However, for the larger buffer, the percent open space variable is negative and significant. All else held constant, more forestry and agriculture in this larger measure of area around a housing transaction leads to a decrease in selling price » (J. Geoghegan et al., 1997, pp. 258-259). This result tends to prove that, within a suburban context, it is the immediate neighbourhood's characteristics that influence utility function. In J. Geoghegan et al.'s work, landscape indices such as diversity and fragmentation were also tested 1 . In this study, which links landscape indices to human values, the hypothesis is that diversity and fragmentation have a negative impact on house prices, which was verified at the 0,1 km buffer. Therefore, it is verified that increasing fragmentation and diversity is undesirable. It seems to be synonymous with a checkered landscape, more potentially conflicting edges and therefore a higher potential for negative externalities (J. Geoghegan et al., 1997, p. 257).
Urban sprawl and diffuse aesthetic attributes Following the example of aesthetic performances at the lot scale, diffuse aesthetic results can also be analysed at the light of urban sprawl. From this point of view, it is interesting to notice that, while suburban settlements are activated by the search for natural amenities and homogeneous green landscapes, new out-of-town developments tend to drastically affect those landscape features. Indeed, as it was showed by M. Antrop (1998), initial rural landscapes of suburban districts have been marked, in the last decades, by a typical evolution leading to both, urbanisation and fragmentation, i.e. characteristics that tend to enter utility functions negatively.
1 The diversity variable measures heterogeneity of land uses based on Shannon index (e.g. M. Antrop, 1998). It describes whether land uses taken into account are concentrated in a few categories or distributed among many categories. The fragmentation index is the perimeter to area ratio. This ratio increases as land becomes more subdivided or as patches of a contiguous land use become more dispersed.
3.4 RECREATIONAL BENEFITS
3.4.1 Relationships between green equipment's attributes, catchment area and price impact The catchment area of a green equipment is dependent on its attributes : the more the equipment is attractive, the further people will travel to visit the site. In order to explain the maximum distance people agree to travel to a green equipment, the size is often taken into account. For instance, on the basis of both, Belgian and French surveys (M. Deconinck, 1982 ; P. Lecroat, 1992), some maximal distances have been estimated : ± 1 000 meters for a park larger than 30 hectares ; ± 500 meters for a park between 10 and 30 hectares ; ± 250 meters for a park smaller than 10 hectares. Theoretically, other attributes also determine the attractiveness of each recreation area, for instance landscape features, facilities, accessibility and available substitute areas (L. Tyrväinen, 1999, p. 33). Following the hypothesis that the spatial configuration of the price influence due to a green equipment is related to the extent of its catchment area, it is therefore logical to consider that the price influence will expand with the attractiveness of the park (measured for instance by the size). To our knowledge, studies do not exist on the double relationships between green equipment's attributes and the spatial configuration of both, catchment area and house price influence. Nevertheless, some HPM results confirm the likely hypothesis that bigger parks have wider influences. For instance, the results presented by S. Orford (1999) about the Cardiff case study are in line with this hypothesis. As presented by Figure 3, the impacts of green equipments on house prices is indeed dependent upon the size of the parks, with the biggest park of the city (Bute Park) having a much more important impact than the smaller ones. Figure 3 : Valuing green structures in Cardiff Bute Park price impact Small parks price impact
The precise analysis realised on Cardiff by S. Orford is also interesting concerning the influence of the smallest parks. It is indeed showed that very small parks have only positive externalities with those properties with a view of them (S. Orford, 1999, p. 170). As the small parks do not actually offer recreation possibilities, this conclusion validates the idea that it is necessary to distinguish aesthetic performances from recreational performances.
3.4.2 Maximal extent of the price impact Some HPM results on the maximal extent of the financial impact of a green equipment can be underlined. As stated above, the maximal extent is of course dependent on several factors. For all that, Dutch (A.T. Fennema, 1996, p. 35 ; J. Luttik, 2000, p. 166) as well as Finnish (L. Tyrväinen, 1999) results suggest that, for a local recreational site, the distance to a green equipment has a price effect as long as the areas are within walking distances from home, which means between 400 metres and 600 meters. Previously, T.A. More et al. (1988) also estimated urban parks to influence property prices up to 600 metres from houses. These results are « in line with many recreation studies which report that the most intensive use of the areas occurs near the home environment (Gåsdal 1993, Sievänen 1993, Jensen 1998). The distance must not exceed a walking distance from home, if the [green equipment] forest is to be used frequently (Kardell 1985, Bussey 1996) » (L. Tyrväinen, 1999, p. 33).
4. CONCLUSION According to the literature review on hedonic models used to assess green structures influences on residential property values, the general hypothesis about positive impacts of green is confirmed. More precisely, the fact that green structures enter utility functions positively is related to two major kinds of performances : aesthetic performances and recreational performances. Of course, this result strengthens the position of green structure in the policy decision process. However, in the Netherlands for instance, this general hypothesis has only been partially confirmed, which can probably be related with the fact that green planning has been significant in this country. Indeed, as the implicit price is the result of demand and supply interactions, abundance of urban green leads to minimise their market values, although intrinsic performances will, of course, continue to be present. Beyond the central hypothesis of positive effects of green structures on property levels, this literature review also enables us to underline some specific environmental factors that contribute to the quality of urban life. For instance, as stated in different researches, the presence of water features seems to be particularly appreciated, which leads to recommend integration of water bodies in town development programmes. The role of park design can also be pinpointed. As also showed by different HPM researches, it is actually useful to create buffers of natural vegetation to reduce negative impacts due to sport and recreational uses. A third green management recommendation is related to the shading effect as it is showed by HPM studies that coniferous should not be planted near houses.
HPM results can also be analysed at the light of urban sprawl. Firstly, it is interesting to notice that the positive effects of green structures on residential prices confirm the relationship between suburbanisation and searches for natural amenities. Clearly, as households are willing to live in a green environment, this can very easily lead to sprawl if green qualities cannot be found within traditional urban fabrics. In parallel, it is also essential to notice that, while suburban settlements are activated by the search of natural amenities and homogeneous green landscapes, new out-of-town developments tend to drastically affect those features as they lead to both, fragmentation and decrease in the percent of open space.
Bibliography : Literature on green structure and residential prices Anderson L.M. and Cordell H.K., 1985, « Residential Property Values Improved by Lanscaping with Trees », Southern Journal of Applied Forestry, n°9, 162-166. Anderson L.M. and Cordell H.K., 1988, « Influence of Trees on Residential Property Values in Athens, Georgia (U.S.A.) : A survey based on Actual Sales Prices », Landscape and Urban Planning, n°15, 153-164. Beckerich C., 2000, Biens publics et valeurs immobilières, Thèse de Doctorat de Sciences Economiques, Université LUMIERE Lyon, inédit. Beckerich C., 2001, Biens publics et valeurs immobilières, ADEF, Paris. Correll M.R., Lillydahl J.H. and Singell L.D., 1978, « The effects of Greenbelts on Residential Property Values : Some Findings on the Political Economy of Open Space », Land Economics, n°54, 207-217. Darling A.H., 1973, « Measuring Benefits Generated by Urban Water Parks », Land Economics, n°49, 22-34. Fennema A.T., Veeneklaas F.R., Vreke J., 1996, « Meerwaarde woningen door nabijheid van groen », Stedebouw en Ruimtelijke Ordening, n°3, 33-35. Gardner M.B., Pollakowski H.O., 1977, « Economic valuation of shoreline », Review of Economics and Statistics, Vol. 59, 272-278. Garrod G.D., 1994, « Using the hedonic pricing model to value landscape features », Landscape Research, Vol. 19(1), 26-28. Garrod G.D. and Willis K.G., 1992a, « Valuing Goods' Characteristics : an Application of the Hedonic Price Method to Environmental Attributes », Journal of Environmental Managment, n°34, 59-76. Garrod G.D. and Willis K.G., 1992b, « The environment economic impact of woodland : a two-stage hedonic price model of the amenity value of forestry in Britain », Applied Economics, n°24, 715-728. Geoghegan J., Wainger L.A. and Bockstael N.E., 1997, « Spatial landscapes indices in a hedonic framework : an ecological economics analysis using GIS », Ecological Economics, n°23, 251-264. Gillard Q., 1981, « The effect of environmental amenities on house values : the example of a view lot », Professional Geographer, Vol. 33(2), 216-220. Lake I.R., Andrew A.L., Bateman I.J. and Langford I.H., 1998, « Modelling environmental influences on property prices in an urban environment », Computer, Environment and Urban Systems, Vol. 22, n°2, 121-136. Lansford N.H. and Jones L.L., 1995, « Recreational and Aesthetic Value of water using Hedonic Price Analysis », Journal of Agricultural and Resource Economics, Vol. 20(2), 341- 355. Li M.M. and Brown H.J., 1980, « Micro-Neighborhood Externalities and Hedonic Housing Prices », Land Economics, Vol. 56, n°2, 125-141. Luttik J., 1998, « The value of trees, water and open space as reflected by house prices in the Netherlands », Paper presented at the World Congress of Environmental and Resource Economists, Venice, Italy, June 25-27, 1998. http://www.feem.it/worldcongress/index.html Luttik J., 2000, « The value of trees, water and open spaces as reflected by house prices in the Netherlands », Landscape and Urban Planning, n°48, 161-167. Luttik J. and Zijlstra M., 1997, Woongenot heeft een prijs. Het waardeverhogend effect van een groene en waterrijke omgeving op de huizenprijs, DLO, Staring Centrum, Rapport 562, Wageningen. McLeod P.B., 1984, « The demand for local amenity : an hedonic price analysis », Environment and Planning A, Vol. 16, 389-400. Morales D.J., Frederick R.M. and Weber R.L., 1983, « Two methods of valuating trees on residential sites », Journal of Arboriculture, Vol. 9, n°1, 21-24. More A.T., Stevens T. and Allen P.G., 1988, « Valuation of Urban Parks », Landscape and Urban Planning, n°15, 139-152. Orford S., 1999, Valuing the Built Environment. GIS and house price analysis, Ashgate, Aldershot - Brookfield USA - Singapore - Sydney. Payne B.R., 1973, « The twenty-nine tree home improvement plan », Natural History, Vol. 82, n°9, 74-75. Powe N.A., Garrod G.D., Brundson C.F. and Willis K.G., 1997, « Using a geographic information system to estimate an hedonic price model of the benefits of woodland access », Forestry, Vol. 70, n°2, 139-149. Powe N.A., Garrod G.D., and Willis K.G., 1995, « Valuation of urban amenities using an hedonic price model », Journal of Property Research, n°12, 137-147. Reginster I., 1998, Qualité de l'environnement urbain et choix de localisation résidentielle. Construction d'une méthodologie et analyse, Dissertation présentée en vue de l'obtention du grade de Docteur en Sciences, Université Catholique de Louvain, unpublished. Tyrväinen L., 1997, « The amenity value of the urban forest : an application of the hedonic pricing method », Landscape and Urban Planning, n°37, 211-222. Tyrväinen L., 1999, « Monetary valuation of urban forest amenities in Finland », Research Paper, Finnish Forest Research Institute, n°739. Tyrväinen L. and Miettinen A., 2000, « Property prices and Urban Forests Amenities », Journal of Environmental Economics and Managment, n°39, 205-223. Weicher J.C. and Zerbst R.H., 1973, « The externalities of neighbourhood parks : an empirical investigation », Land Economics, n°49, pp. 99-105. Willis K.G. and Garrod G.D., 1993, « The contribution of trees and woodland to the value of property », Arboricultural Journal, Vol. 17, 211-219. Other Literature on green structure Antrop M., 1998, « Landscape change : Plan or chaos ? », Landscape and Urban Planning, n°41, 155-161. Antrop M. and Van Eetvelde V., 2000, « Holistic aspects of suburban landscapes : visual image interpretation and landscapes metrics », Landscape and Urban Planning, n°50, 43-58. Deconinck M., 1982, Etude de la répartition des espaces publics de loisir de plein air dans les agglomérations urbaines. Analyse de leur rayonnement, de leurs fonctions et de leurs utilisateurs. Le cas de l'agglomération bruxelloise, Thèse présentée en vue de l'obtention du grade de Docteur en Sciences Université Libre de Bruxelles, unpublished. Duhem B., 2001, COST C11 - Green structure and urban planning, Progress report 2001. Lecroat P., 1992, « La nature au coin de la rue », Cahiers de l'IAURIF, n°102, pp. 49-59. Theoretical literature on HPM and urban economics Alonso W., 1964, Location and land use, toward a general theory of land rent, Harvard University Press, Cambridge. Amigues J.-P., 1997, « Enjeux et limites de l'évaluation des actifs naturels », Economie et Sociétés, n°35, 35-55. Anselin L. and Griffith D.A., 1988, « Do spatial effects really matter in regression analy-sis ? », Papers of the Regional Science Association, Vol. 65, 11-34. Facchini F., 1994, « L'évaluation du paysage : revue critique de la littérature », Revue d'Economie Régionale et Urbaine, n°3, 375-402. Facchini F., 1995, « Economie et paysage : la place de la gestion politique », L'Espace Géographique, n°4, 319-327. Follain J.R. and Jimenez E., 1985, « Estimating the demand fotr housing characteristics : A critique and survey », Regional Science and Urban Economics, Vol. 15, 77-107. Lancaster K.J., 1966, « A new approach of consumer theory », Journal of Political Economy, Vol. 74, 132-157. Le Goffe P., 1996, « La méthode des prix hédonistes : principes et application à l'évaluation des biens environnementaux », Cahiers d'économie et sociologie rurales, n°39-40, 180-198. Muth R.F., 1969, Cities and Housing, The University of Chicago Press, Chicago. Palmquist R.B., 1991, « Hedonic Methods », Measuring the Demand for Environmental Quality (Braden J.B. and Kolstad C.D. Editors), Elsevier Science Publishers B.V. (North-Holland), 77-119. Rosen S., 1974, « Hedonic prices and implicit markets : product differenciation in pure competition », Journal of Political Economy, n°1, 347-363. Wilkinson R.K., 1973, « Measuring the determinants of relative house prices », Environment and Planning, Vol. 5, 357-367. Literature quoted by cited authors Bussey S.C., 1996, Public uses, preferences and perceptions of urban woodlands in Reddittch, PhD Thesis, University of central England, Birmingham. Gåsdal O., 1993. « Uteliv i byen » (outdoor recreation in towns), Vårt friluftsliv : aktiviteter, miljokrav og forvaltningsebehov (B.P. Kaltenborg and M. Vorkinn (Eds)), Norsk insitutt for naturforkning, Temahefte 3, pp. 33-41. Grey G. and Deneke F., 1978, Urban Forestry, John Wiley and Sons, New York. Jensen F.S., 1998, Forest recreation in Denmark from the 1970s to the 1990s, PhD dissertation, Danish Forest and Landscape Research Institute, Department of Park and Landscape. Kaplan R. and Kaplan S., 1989, The experience of Nature : A psychological Perspective, Cambridge University Press, New York. Kardell L., 1985, « Recreation forest - a new silviculture concept », Ambio (14) 3, pp. 139- 147. Laurie I., 1979, Nature in Cities. The Natural Environment in the Design and Development of Urban Green Space, John Wiley,Chichester. Miller R.W., 1988, Urban Forestry. Planning and Managing Urban Green Spaces, Prentice Hall, Englewood Cliffs, NJ. Robinette C., 1972, Plants, people and environment quality, Department of the Interior. National Park Service, Washington DC. Sievänen T., 1993, « Kaupunkiväestön ulkoilukäyttäytyminen ja ulkoilualueiden käyttö. Hämeenlinnan ulkoilututkimus » (Outdoor recreation household survey in the City of Hämeenlinna), Folia Forestalia 824.
|
updated 25 oct 2002