do have views of downtown St. Paul, but this analysis identified only properties in close proximity to the downtown area as having these views. As properties near the downtown area tend to have low property values because of a number of factors, among them high crime rates, this would also explain this negative result.
The coefficient for view richness, which measures the number of land types visible, was also negative and significant. The marginal implicit price for increasing the richness of a view by 10% evaluated at the mean home sale price indicated a price decrease of $2834. This suggests a preference on the part of homeowners for views with low diversity, that is, a low number of lands cover types in view. We had expected this variable to positively impact home sale prices based on the results of a previous study that found increased diversity in views to be highly valued (Bastian et al., 2002). However, that study was conducted in Wyoming, a rural land market, where increased diversity likely corresponds to an increase in natural and agricultural land cover types visible in a view. In the Ramsey County area, where most land use and land cover classes are urban, higher view richness likely increases the number of different urban land use classes in view, making it more likely that undesirable urban land uses will be visible from a home.
The signs of all open space access variables were as expected, although not all variables were significant. This indicates that, in general, decreasing the distance to the nearest open space feature increases home sale prices. Homes located near both lakes and streams have significantly increased sale values. Evaluated at the mean home sale price and an initial distance of 1000m, the marginal implicit price for reducing the distance to the nearest lake by 100 m produces a $216 increase in home sale value, the highest marginal implicit price of all open space feature types. Proximity to streams influenced home values to a lesser degree. The marginal implicit price for reducing the distance between a home and the closest stream by 100 m evaluated at the mean home sale price and a starting distance of 1000 m suggests a home sale price increase of $127. Thus, although it is desirable to live near a stream, it is more desirable to live near a lake.
Our results indicate that parks and trails are highly desirable features. The marginal implicit price for decreasing the road distance to the closest park by 100 m, when evaluated at an initial distance of 1000m and the mean home sale value, indicates an increase in home sale value of $136 while the marginal implicit price for decreasing the Euclidean distance from a home to the closest trail evaluated in the same manner indicates an increase of $119. Thus, close proximity to parks on roads and to trails as-the-crow-flies increases home values. This suggests that individuals may be more inclined to access parks by roads, either by driving or walking, and perceive proximity to them on a road network. Conversely, individuals appear to consider proximity to trails in terms of straightline distance and may not access them via roads, possibly because these features have more continuously located access points than parks. Thus, although both parks and trails do increase home sale prices, people appear to perceive and access them differently and this is reflected in the amount they pay to live near them.