The paper describes the development of an air passenger demand model for the Baltimore–Washington metropolitan region that was undertaken as part of a recently concluded ACRP project that explored the use of disaggregated socioeconomic data in air passenger demand studies. The model incorporated a variable reflecting the change in household income distribution, together with more traditional aggregate causal variables: population, employment, average household income, and airfares as measured by the average U.S. airline yield, as well as several year-specific dummy variables. The model was estimated on annual data for the period 1990 to 2010 and obtained statistically significant estimated coefficients for all variables, including both the average household income and the household income distribution variable. Including household income distribution in the model resulted in a significant change to the estimated coefficient for average household income, giving a much higher estimated elasticity of demand with respect to average household income compared with a model that does not consider changes in household income distribution. This has important implications for the use of such demand models for forecasting, as household income distribution and average household income may change in the future in quite different ways, which would affect the future levels of air passenger travel projected by the models.
A Method for Private Car Transportation Dispatching Based on a Passenger Demand Model