Estimating and Forecasting the Probability of New Car Buying

Modeling household automobile ownership choices is a key component of travel behavior research and car demand analysis and forecasting. This article analyses the behavior of new car buyers based on household data. Therefore, the model of buying a new car was specified and estimated by logit method on raw household budget data for 1387. Then, the market demand for new cars has been simulated and forecasted for the period 1388-90 based on out of the sample prediction of the model.
Results show that household's real income, number of employed in household, the number of students per household, the age of household head, number of household members over 20 years, and dummy variable for urban/rural have a significant effect on buying a new car by the member of household. Among the explanatory variables, the number of people over 20 years has the largest effect and the household's real income has the lowest effect on new car ownership.