Property value effects of linear river ferries that service multiple stops in cities are under-explored. The Brisbane CityCat, CityHopper, and CityFerries combine to form a ferry system with 24 terminals. A geographically weighted regression (GWR) approach is used to determine property value effects of the system. Cross-sectional property data is used in combination with a set of neighborhood variables derived from 2011 census data, spatial feature location, and transport datasets (roads, busway and train station locations) for the city. The preferred global model had a good fit and showed expected signs for all parameters, showing that property prices tended to decline with distance from ferry terminals, when controlling for other variables. For every kilometer close a location is to a ferry terminal, there is an expected price increase of 4 percent on average, across the study area. The GWR local model also had good fit and suggested property value gains around specific terminals. Visual inspection suggests that locations where more ferry-oriented development opportunities have been taken in recent decades are the sites with the greatest positive property value effects. The implications are that land developers are justified in seeking ferry terminals to service their developments. Keywords: land value uplift, ferry, geographically weighted regression, accessibility of public transport
IntroductionCities in both the developed and developing worlds have recently introduced linear passenger ferry systems that provide relatively frequent services along rivers or parallel to shorelines, servicing multiple stops (Thompson, Burroughs, and Smythe
Pseudopanel data have been increasingly applied in travel demand analysis to investigate the long-run travel demand when genuine panel data are unavailable. However, conventional estimation techniques have typically been used without a careful consideration of some unique properties of pseudopanel data. This paper shows that ignoring these properties potentially leads to estimation bias or inefficiency not observed in genuine panel data. The method used is a Monte Carlo experiment with scenarios designed to generate various data possessing pseudopanel data characteristics under conditions of limited observations; the performance of various estimator is evaluated with the use of the simulation results. This research found that the large between-group variation of the exogenous variable and the variance of unobserved group effects in pseudopanel data are the primary causes of estimation bias and inefficiency. Other factors such as cohort sizes and nonspherical errors have a smaller effect on the estimators’ performance. An empirical application using Sydney Household Travel Survey data is also presented to illustrate the simulation findings.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.