NYC-WTC_000187725

NYC 9/11 Public Portal Document IT

61

o As independent variables, the relative importance of transit access characteristics among physical site, neighborhood, and public access characteristics that influence land value. o As policy levers, variation in land value by distance from a rapid transit station stratified by differences in the following transit characteristics which are independent variables: - type of transit structure - type of transit service - year station opened - volume of train service - passenger volume - on-time performance - equipment on line - crowding on trains

How NYSTA is Designed The appropriate method of determining land value relationships in a built environment, when development has already taken place, is the use of multivariate regression analysis incorporating a wide array of explanatory variables in estimating a land price function. While the choice of the dependent variable in such an analysis is clear, that is, the unit price of land, the choice of independent variables necessitates a process of stepwise regression or factor analysis of all such explanatory factors. Before this choice is made, a critical decision occurs in the initial identification of independent variables, including dummy variables where appropriate. In this effort, we were guided by theory, prior related research studies, data availability, and tests for colinearity in variation between variables. Chief among omissions of potential independent variables were those quality measures of the built environment that are reflected in the type and condition of building materials used in existing structures. However, since these variables are not likely to covary with the transit access terms, it is doubtful that their effect on land value has been attributed to the measures of transit accessibility. Given the magnitude of the assembled data base of roughly one hundred thousand parcels recently sold in the study area in vacant, residential and commercial building classes, and the array of some 60 parcel-specific, neighborhood, and access-related factors of potential significance in explaining land value, a sampling method was adopted to extract a representative group of parcels from each building class. Because sampling preceded the measurement' of public access variables, it was not possible to stratify the sample based upon a range of distances to transit stations. Rather, considerations of borough and building class representation prevailed.

HR*

CITY_21MC100_006662292

NYC-WTC_000187792