NYC-WTC_000187725

NYC 9/11 Public Portal Document

33 relate the q^'s to the sizes of the shopping zones.

The next step is to plug the optimal trip frequencies given by (5) into the utility function (1) , and obtain the indirect (optimized) utility function for each (j,k,m), given each i:

1- (1~P) "*■ P i"*" i' 1 22 (7)

V j, k, l,n

The final step describes the way in which households are distributed among the residential bundles (j,k,m) given their workplaces i by means of a logit model. The logit probabilities are;

P =(ff } _____________exp (6 jE[ j] )_____________ (8) E (F„) «exp(6jE[v;^,J )+exp(6^C7j ' rxs

and,

E (lf„) ■’exp (6 texp (8 jU,) rx8

where is an exogenous utility level for residential choices outside the study area (to be represented by a peripheral superzone o) . The probabilities sum to one over (j,k,m) plus o for each i. is the stock of type k housing units in zone j. The 6^ measure the degree of idiosyncratic taste homogeneity among households employed at i. As approaches infinity, all households employed at i are identical in idiosyncratic tastes and choose the same residential choice (j,k,m). As <?£ approaches zero, all households are extremely heterogeneous and all alternatives have an equal probability of being chosen. 1-c measures the degree of correlation in the idiosyncratic utilities attributed to the Hjjj dwellings in the same '(j,k,m). If these dwellings are totally

CITY_21MC100_006662264

NYC-WTC_000187764