Examining Accessibility and Proximity-Related Effects of Bogota's Bus Rapid System Using Spatial Hedonic Price Models

Abstract

With several successful cases worldwide, bus rapid transit (BRT) has re-emerged as a cost-effective transportation alternative for providing urban mobility. Despite the resurgence of BRT, there is a worldwide paucity of research examining its accessibility and proximity-related effects, such as noise and air pollution. A recent World Bank (2000) study observed, "the impacts of busways on land use and city structure have been little researched." The limited research available is dated, and suggests that the land use impacts of BRT systems have been negligible. The paucity exists despite the fact that a careful examination of BRT impacts is key for the transportation planning process and for transportation public policy makers. Potential misperceptions of these impacts can lead to significant facility redesign, construction delays, or compensation to affected parties, all of which may amount to millions of dollars. Furthermore, the success of innovative infrastructure financing tools, such as value-capture, hinges on understanding whether or not positive infrastructure impacts are capitalized into land values. A cursory review of current BRT systems in operation worldwide suggests that some BRTs have resulted in environmental degradation of bus corridors, arising from heavy bus flows, high vehicle emissions, and noise. In other cases, such as Curitiba, BRTs have played an integral role in the successful articulation of an integrated land use and transport strategy. However, the extremely limited number of ex-post BRT evaluations has been limited to the examination of busway impacts in a handful of North American cases. As a result, by using spatial hedonic price models, this project determines the extent to which the accessibility and proximity-related disamenities of a BRT in Bogota, Colombia are capitalized into land rents. In these models, a measure of land value is regressed against discrete characteristics of a property, such as its structure, neighborhood amenities and disamenities, and overall location. Because of particularities in the application of hedonic models, and of the data collected, the coefficient estimates are determined using maximum likelihood methods that account for the spatial autocorrelation present in the model residuals. Data collected between January and April 2002 are used to estimate several model specifications. Local accessibility effects of the busway are measured by walking access distance to the nearest station, while regional accessibility effects are measured as line-haul travel distance from the nearest busway station to three regional attractors of activities. Proximity-related effects are measured as the straight distance from the property to the busway right-of-way. Results from the preferred functional form specification suggest that for every 0.1 km from the BRT station, the advertised rental price of a property decreases by 3.71%, after controlling for structural, neighborhood, and environmental variables. Similarly, as the distance from the busway right-of-way increases by 0.1 km, the value of the property increases by 7.00%. These results translate into an elasticity of -0.31 for local accessibility and 0.44 for proximity-related effects. Based on a comprehensive literature review, this is the first empirical study in recent decades that consistently estimates evidence regarding the nature and magnitude of accessibility and proximity-related impacts of BRTs. Such evidence has a wide range of practical applications, from determining the usefulness of innovative land-based tax instruments that hinge on the capitalization of positive busway effects, to informing policy-makers about the land development consequences of transportation infrastructure alternatives. Keywords: Bus Rapid Transit (BRT), busways, accessibility, proximity-related effects, hedonic price models, spatial autocorrelation, Bogota.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 5 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.