Mode Choice Models

Aggregate and Disaggregate Approaches

Aggregate model choice models take into account many variables about each possible mode of transport the traveler has available for the trip, such as travel time and cost, and estimate the proportion of travelers that would use each mode of transport (for example, 70 percent by car and 30 percent by public transport). This is usually estimated for each origin destination pair, for a specific period of the day, and for a specific trip purpose.

To better understand travel behavior and the factors that will motivate individuals to shift to more environmentally friendly modes, most mode choice models are disaggregated discrete-choice models. These models have several important advantages over the aggregate approach because they explain why an individual makes a particular choice given her/his circumstances and are, therefore, better able to reflect changes in choice behavior because changes in individual characteristics and attributes of alternatives.

The aggregate approach, in contrast, rests primarily on statistical associations among relevant variables at a level other than that of the individual. The disaggregate approach is more suited for proactive policy analysis because it is causal, less tied to the estimation data, and more likely to include a range of relevant policy variables. Disaggregate discrete-choice models estimate the probabilities that a traveler making a specific trip from zone O to zone D in time period t for purpose p will choose a specific mode m of travel from a given set of available modes c.

In many practical applications, the model is estimated based on disaggregate data (for example, based on individual trips) and then applied in a semiaggregate approach. In this approach, given a trip table of a specific purpose and time of day for each origin–destination, the model will predict the shares of each mode for each socioeconomic group. Advance travel–demand models move more and more toward agent-based approaches with a full disaggregate application.

In the traditional four-step travel demand modeling, mode choice modeling is the third step following trip generation (How many trips are generated?) and trip distribution (How they are distributed in space?), and preceding trip assignment (Which routes they are using?). When time-of-day modeling precedes mode choice modeling, the latter can account for the level of service for the actual time of day.

...