Abstract
This study develops nested logit models to explore the location choice of commercial developers in the Portland metropolitan area, a region that is widely known for its progressive urban growth management. The purposes are to study the dynamics of urban spatial structure from the supply side of the real estate market and to test the capacity of regional governments to manage urban growth by influencing the location choices of developers. The model results indicate that commercial development in the Portland region showed a clear trend of decentralization from its central city. Decentralized commercial developments were not generally dispersed in suburban areas either. Instead, commercial developers tended to locate them in locations within or in close proximity to employment subcenters. Among many control variables tested in the models, availability of vacant land zoned for commercial and industrial purposes proved to be the most important predictor for new commercial development.
Acknowledgements
The author wishes to thank John Gliebe, Tony Rufolo, James Strathman, Paul Waddell and anonymous referees for their extremely beneficial commentary on earlier drafts of this work.