The Relationship Between Car Ownership and Income

Car ownership is the number of cars per 1 000 people and passenger cars refer to motor vehicles which are more than two-wheelers. It is used to measure the growth in passenger transport. Car ownership in any economy has several determinants which include income of the household, the gender of the household head, the costs of car ownership (having factored in import costs and duty), short term and long costs and use, the public transport system in terms of organization, fares and other socio-demographic characteristics of the households. Another important factor is the inclusion of government transport policy considerations, rate of infrastructure development and infrastructure maintenance policy. All these combined with GDP are important determinants of car ownership. Further, the data should be taken over a long period of time and not just within a year. For the research report to be valid and reliable, car ownership determinants including income of the household, the gender of the household head, the costs of car ownership-short term and long term costs, adequacy of the public transport system in terms of organization, fares and other socio-demographic characteristics of the households should have been considered. In this case, then, the model would have been. CARS = β0 + β1X1 + β2X2 + β3X3 + β4X4 + β5X5 +…+ βnXi Where βn are the coefficients of the explanatory variables while Xi is the explanatory variables (income of the household, the gender of the household head, the costs of car ownership-short term and long term costs, adequacy of the public transport system in terms of organization, fares and other socio-demographic characteristics of the households). Holding GDP constant, the model shows that car owners’ increases by 96 per 1000 while a constant 2000 US$ increase GDP leads to 11 more (more than 96) car ownership per 1000 persons. The coefficient for GDP is significant (β = 0.011, p = 0.001, plt.0.05). A model with car ownership as the dependent variable and GDP as the independent variable is significant. GDP explains car ownership at 77%. Holding transformed GDP constant, the model shows that car owners’ decreases by -2 cars per 1000 persons while a constant 2000 US$ increase GDP leads to 0.822*log (2000), 2.71 = 2 more car ownership per 1000 persons. The coefficient for GDP_Ln is significant (β = 0.822, p = 0.000, plt.0.05). Models with transformed car ownership as the dependent variable and transformed GDP as the independent variable is significant. Transformed GDP explains car ownership at 90.2%. In this case, the paper aimed at analyzing the relationship between car ownership and income in 15 different countries. The data was secondary and is internationally accepted in addition to being used by The World Bank (IBRD). Two models (CAR = 96.634 + 0.011GDP + up and Ln(CAR) = -2.272 + 0.822Ln (GDP) + up) were formulated from the data. One was based on the original data while the other was based on the transformed data. Further tests, shows that the data has an outlier, New Zealand. Omitting the values for the country and running regressions, linearity, normality of the residuals and serial correlations are eliminated. Running stem and leaf plots for the residuals and leverage values shows that the data is corrected by transforming using the natural log values (Dobson 1990. Alvesson and Skoldberg 2006. Alasuutari et al. 2008).