NYC-WTC_000187725

NYC 9/11 Public Portal Document

44 li that B = 1.5. This means that workers have an extra time endowment per day of four hours for extra potential income. The minimization is: H MinimiEja*(JiT^*) ^-p (1.5y,j-AG\y„) ]" (17)

By trial and error, b = 0.3 gives a good result. A = 500, by the assumption that there are 250 working days in a year. Minimizing, we find a^^ = 177.81, aj = 144.26, a^ = 207.17 for single family, condominiums and cooperatives, and rental apartments respectively. In the third step, the occupancy model given by (10) is calibrated. This is done by using evidence from CATLAS [see Anas and Duann (1985)] where it was estimated that a 1% increase in a zone’s average rent resulted in a 0.24% increase in the occupied housing supply of that zone.^ Thus, the weighted average rent elasticity of housing supply is set at 0.24 and X is computed as:

(18) X = 0.2 4 --------

where denotes the observed housing occupancy rate in each housing submarket (j,k) and is calculated by dividing the number of households residing in zone j with the total housing stock in zone j. This gives X = 0.00118785. Then, the dummy variables djj^, in (5) are calculated as follows:

(19) , for each (j,k)

In the fourth step, the shopping expenditure per trip, z, is calibrated, together with <o, the exponent of floor space in the definition of the a^^'s. First, set z = 0 and all = o, and calibrate w by means of a least squares procedure. Note that given z and u, the predicted shopping trips arriving at zone 1 are:

Although this number was obtained for Chicago, since then the occupancy model has also been estimated for Pittsburgh, Houston and San Diego with highly similar results. Still, it is important to calibrate this model more thoroughly in the future for each of the New York boroughs.

L CITY_21MC100_006662275

NYC-WTC_000187775