STRATEGY SESSION: Estimating the Demand for Amtrak Rail Passenger Business Amtrak [the national passenger railroad in the United 1. Explain the rationale for each explanatory vari- States) has used an aggregate demand model to able appearing in the model and what sign (positive forecast system-wide passenger revenues. The first or negative] you expect for its regression coefficient. part of its model was a multivariate linear regression 2. How did Amtrak use this forecast of system- that forecast its system-wide passenger miles-the wide passenger miles to obtain its estimate of sys- dependent variable la passenger mile is one pas- tem-wide passenger revenues? senger moved one mile). Explanatory variables were 3. What is Amtrak's estimate of income elasticity of disposable personal income; Amtrak's average demand for passenger service [based on the income fare; the ratio of Amtrak's fare to the airlines' aver- level of the economy as a whole]? If U.S. disposable age fare; retail gasoline prices; dummy variables to personal income per capita increases from $27,000 reflect such events as weather, holidays, strikes, and to $28,000, what's your prediction for the increase in derailments; and dummy variables to reflect sea- train passenger miles in the new situation fall other sonal variation. independent variables remaining constant)? Will this The most important determinant of rail passenger prediction be 100% accurate? Why or why not? miles was disposable personal income la proxy vari- 4. Even though disposable personal income rose able for the strength of the U.S. economy). From the re- from 2000 through 2008 lon a real income basis by gression, a 1% increase in disposable personal income 13.26%). Amtrak's market share of U.S. passenger was expected to yield a 1.8% increase in system-wide miles traveled has remained virtually constant (0. 1 105% passenger miles. in 2000 and 0.1120% in 2008). Can you explain this