Academic journal article Journal of Real Estate Literature


Academic journal article Journal of Real Estate Literature


Article excerpt


This paper develops four novel measures of a residential property's view using purely topographical data, evaluates each measure's contribution to property prices, and compares the efficacy of these new variables with manually-generated measures of view used in the literature. Using a set of semi-logarithmic and spatial autoregressive hedonic regression models for Sydney, Australia, this paper finds that water views contribute 2.30%-216.2% to property prices and land views contribute -5.94%-69.7%. These new measures confer an improvement in out-of-sample prediction accuracy of up to 6.2%.

(ProQuest: ... denotes formulae omitted.)

As the real estate agent motto proclaims, location (and more location) is a fundamental determinant of how the housing market values a residential property; but what are the individual features of location that impart value upon a particular property? Proximity to desirable localities, neighborhood characteristics, demographics, and visual aspect are all spatially distributed attributes that contribute to a residential property's market value. Of all these features, a property's visual aspect, or 'view,' is one facet that is not often associated with a property's geographical location. However, views are one of the most distinctly spatial attributes a property can have.

View is a characteristic that is purely determined by a property's position, its surrounding topography, and the geometric relationships between them. The existing methodology for appraising a residential property's view and determining how that view contributes to its final sale price can be separated into four main categories: (1) binary, (2) multiple binaries, (3) qualitative, and (4) geographic information systems (GIS). Within the economic literature, the first three categories are the overwhelmingly predominant methods used to investigate the impact of views on house prices (Bourassa, Hoesli, and Sun, 2004) and each of these three measures suffer from a high degree of subjectivity and are labor intensive to generate.

In this paper, four new spatial GIS techniques are developed for quantifying view, all measured in continuous metric units using three-dimensional topographical data. Each of these new measures can be generated by an individual using commonly available GIS software and topographic data and are shown to be more accurate, consistent, and have more desirable properties than either the binary, qualitative or previously used GIS methods. Although automated GIS measures of view could never substitute for the ability of a manual inspection to identify all of the idiosyncratic features that may impact an individual property's actual view, such as trees, telegraph poles or balcony positioning, the intention of this paper is to compare the contribution of these automated measures with that of the manually-generated measures of view used in the literature. Using a series of semi-logarithmic and mixed spatial autoregressive hedonic regression models, the results demonstrate that these new spatial measures produce a distinct improvement in model accuracy and out-of-sample prediction, as well as a better understanding of the complex role views play in the market value of a residential property.


From the four categories of view studies discussed previously, the literature has yielded a variety of estimated returns to views, from being the largest contributor toward a property's value (Benson, Hansen, Schwartz, and Smersh, 1998) to statistical and economic insignificance (Brown and Pollakowski, 1977; Paterson and Boyle, 2002). A recommended place to begin any survey of the housing view literature is Bourassa, Hoesli, and Sun (2004), who provide a thorough review up until 2003 of the major studies that have investigated the impact of views on residential property prices.

The binary category of the view literature includes those studies that use a single binary variable to indicate the existence versus the non-existence of a view from any particular property. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.