The central objective of this study is to operationalize a wage-based model of residential site value transfer, with an associated land development component. Site value transfer represents the process, whereby commuting cost savings in relation to marginal expenditures are transferred into "differential rents" on more advantageously located sites. These periodic rent payments are then capitalized into a site value increment of total housing price at a given point in time.

For the calibration and testing phases, of this study, extensive cross-sectional data have been extracted from Metropolitan Toronto;​ for the period 1961-71. The data file consists of several variable groupings, that is land use, housing, transportation, socio-economic and demographic.

The model employs an iterative algorithm that interrelates commuting cost savings, site value transfer, land use develop- ment, population-employment allocation and the transportation planning process. A preliminary step to the calibration of a site value transfer. expression concerns the isolation of the site value component from the total housing package market value, using a hedonic price index with factor analytic procedures. This index is established based on transaction specific Multiple Listing Service/data for 1961 and 1971. Step one of the calibration phase involves the structuring of a mathematical expression that relates the site value component from the hedonic price index with commuting cost savings over marginal conditions for each district at a given point in time. In the second calibration step, an abstract land development score is estimated for 1961 and 1971 census districts based on a weighted mathematical expression com- prising a number of statistically significant normalized land use factor scores. A non-discrete land use state transition expression of a deterministic nature is then calibrated on a census district basis for the transition period 1961-71.. In the final calibration step, site value transfers, with associated land development transitions, are connected to the transpor- tation planning process, via a standard trip generation and attraction, trip distribution, modal split and network assignment sequence.

Within a limited test framework a number of transportation policies are analyzed through a simulated application of the site value transfer model on an abstracted representation of the Toronto transportation system. The major objectives of this policy test are to identify, the spatial distribution of site value changes, to identify associated land development effects, to measure modal sensitivity and to establish ་ convergence.

The results of this study indicate that transportation investments may cause both positive and negative site value impacts over urban space simultaneously, so that both value recovery and compensation must be considered in any equitable valuation formula. Regardless of transportation policy, site value effects caused by land development changes appear to be significant. These changes are strongly affected by base year conditions and may enhance or retard initial changes in the absence of shifts in population and employment. In this simulation exercise convergence is readily established for all policy inputs.