Question: Refer to Coshall, J., 2000, “Spectral Analysis of International Tourism Flows”,

Annals of Tourism Research, 27, 3, pp. 577-589.

Download the data file, FinalExam.xls. We are interested in analysing the

probability that a traveller from Sydney to Canberra decides to travel by air rather than alternative modes of transport, and include as exogenous variables the length of stay in Canberra (LENGTH), the age (AGE), gender (FEMALE) and income (INCOME) of the traveller in $’000, and the price of the airfare (PRICE). FLY is a binary variable, and is equal to 1 if the individual travelled by air and 0 otherwise.

(a) Estimate a linear probability model (LPM) of FLY on LENGTH, AGE, FEMALE, INCOMeand PRICE. Write down the regression equation.

(b) Estimate a probit model of FLY on LENGTH, AGE, FEMALE, INCOMeand PRICE. Write down the regression equation.

(c) Comment on the difference (if any) between the regression results in (a) and (b). What are the drivers of these differences?

(d) Suppose there is a male traveller who is planning on staying in Canberra for 4 days, he is 45 years of age, earns $45,000 per year, and flight tickets from Sydney to Canberra are currently priced at $150. Use the probit model in (b) to determine the probability that he will travel by air.

(e) Refer to (d). What is the marginal effect of a $5 increase in ticket prices on the above traveller’s probability of using air travel? Use the probit model estimates.