Testing the Temporal Stability of Accessibility Value in Residential Hedonic Prices

Purpose – This paper bridges the gap between, on the one hand, supply-driven (urban form and transportation networks) and demand-driven (action-based) accessibility to urban amenities and, on the other hand, house price dynamics as captured through hedonic modelling. It aims at assessing temporal changes in the valuation of accessibility, while ordering households’ priorities among access to labour market, schools and shopping outlets. Design/methodology/approach – Several indexes are built using a methodology developed by Thériault et al. (2005). They integrate car-based travel time on the road network (using GIS), distribution of opportunities (activity places) within the city, and willingness of persons to travel in order to reach specific types of activity places (mobility behaviour). While some measure centrality (potential attractiveness considering travel time from population to opportunities) others consist of action-based indexes using fuzzy logic and capture the willingness to travel in order to reach actual specific activity places (workplaces, schools, shopping centres, groceries). They summarise suitable opportunities available from each neighbourhood. Rescaled indices (worst 0 to 100 best) are inserted simultaneously into a multiplicative hedonic model of single-family houses and condominiums sold in Quebec City during years 1986, 1991 and 1996 (10,269 transactions). Manipulations of accessibility indexes are developed for ordering their relative impact on sale prices and isolate effects of each index on the variation of house values, thus providing proxies of households’ priorities. Moreover, the parameter expansion modelling approach developed by Casetti (1972 and 1986) is used to control for changes (temporal drift and market-oriented adjustment) in the valuation of each property-specific, taxation or accessibility attribute during the study period. Findings – This original approach proves efficient in isolating the cross-effects of urban centrality from accessibility to several types of amenities, while controlling for multicollinearity and heteroscedasticity. Results are in line with expectations. While only a few property-specific attributes experience a change in their marginal contribution to house value during the study period, all relative accessibility indexes do. Every single accessibility index has a much stronger effect on house values than centrality (which is still significant). When buying their home, households put more emphasis on access to schools than they put on access to the labour market, which in turn, prevail over accessibility to either shopping centres or, finally, groceries. The ordering is rather stable but the actual valuation of a specific amenity may change over time and among market segments (single houses versus condos). Practical implications – Better understanding the effect of accessibility to amenities on house values provides guidelines for choosing among a set of new neighbourhoods to develop in order to generate optimal fiscal effects for municipalities. It could also provide guidelines for decision making when improving transportation networks or locating new activity centres.