destination accessibility had the strongest effect on motorized and nonmotorized travel, while density had the weakest association. Due to the nature of meta-analysis, the study is general in scope and is limited to the targets and data laid out in the studies it aggregates. It does not analyze its own standardized set of data within an exhaustive range of target cities. Lane (2008) examined a suite of variables which may affect a city's propensity for adding a light rail (LRT) system, including residential and employment density, as well as sprawl and total roadway length measures. Residential and employment density were both significant variables. The resulting model accurately predicted the introduction of LRT in 33 of 35 cities, but was limited in scope to systems in the United States. Mees (2009) discussed the relationship between urban densities and transit use, and after he calculated urban densities for residential and non-residential land, compared the relationship between urban density and method of travel to work in US, Canadian, Australian, and English cities. Kuby et al. (2004) analyzed the factors which influence light rail boardings in nine cities in the United States, developing a model for predicting demand for light rail systems. The study found several factors which positively affect light rail ridership, including employment and population density, as well as station location. However, in these studies, service levels for urban public transport have not been considered. Kim et al. (2007) investigated the factors impacting rider mode choice in access to light rail stations, finding a significant influence of land use type on passenger mode choice. As it concentrated on external urban structural factors which influence mode choice, this study also did not consider factors within the transit system itself, such as service level. Concerning Japanese studies in the same field, Miyata et al. (1993) evaluated the changes of populations in cities, towns and villages along railway lines caused by the closure of local railways, and examined the effect on local societies. Nakagawa et al. (1993) verified the effect of railway development on the populations in local cities, towns and villages from transition of population and the timings of railway development. In the above research, analyses were limited to the municipal level, and neither study included consideration for urban structure. Tsuji et al. (1999), targeting cities having tramways in Japan, verified the relationship between the compactness of cities and tramways, but did not look at service levels of those trams, and concentrated only on Japanese cities with tramway systems. Oba et al. (2008) clarified that populations around railway stations are decreasing and populations in the areas separated from the stations are increasing in local cities in Japan. Matsunaka et al. (2008) compared special characteristics of tram corridors in Japan and France with overall urban characteristics, and Nakamichi et al. (2007) clarified the relationship between ...