Solving for A in (2) and substituting the results back into (1), we can eliminate the scale parameter and ensure that the demand function passes through the point (d_{0},P_{0}, T_{0}). The resulting equation is:

j i T i P i j j i i i i i T T T T P P d d , 0 0 0 0 ε ε ε ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ T ⎝ ⎛ =

(3)

Then, if a policy or program changes the transportation costs and travel times, the new number of vehicle trips is obtained by substituting the new costs and travel times into equation (3), giving:

d_{i }

j i T i P i j j i i i i i T T T T P P d , 0 1 0 1 0 1 0 ε ε ε ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ T ⎝ ⎛ =

(4)

Finally, what we are interested in is the change in the number of vehicle trips, which is given by:

Δd_{i }= d_{i1 }− d_{i0 }= d_{i0 }

⎢ ⎢ ⎣ ⎡ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ , 0 1 0 1 0 1 j i T i P i j j i i i T T T T P P T ε ε ε

−1

⎥ ⎥ ⎦ ⎤

(5)

This last formula constitutes the approach to model the change in demand brought about by program or policies affecting the perceived cost of travel, both monetary and non monetary. Equation (5) can be simplified or expanded to include additional cost factors and to comprise cross relationships with one or more modes as shown in the next section of the report.

# Advantages and Constraints

The previous section of this report observed that there exist different ways of providing a simple, yet powerful and robust approach to estimating the impacts of alternative

strategies employs a

at

a

sketch

planning

level.

For

example,

the

EPA

COMMUTER

model

pivot point approach which relies on the theoretical underpinnings of discrete

choice governing travel demand decision making process among opposing transportation modes.

estimation. The EPA choice relies on mutually exclusive alternatives, namely This is what is usually assumed by

an individual different and all predictive

evaluation methods herein is capable of not only from SOV

currently available for TDM assessment. The approach proposed estimating mode share changes. It can predict mode share changes to other alternatives, but among these alternatives themselves. It

35