自動車利用とCO<sub>2</sub>排出量で評価する税制の影響

Abstract

In this paper, we study the effectiveness of tax incentives and the variation in vehicle user costs on the environmental impact of private car use across cities in OECD countries and in some non-OECD countries in Asia. We present a simple model of private car stocks, vehicle kilometres travelled (VKT) per capita and per car and CO2 emission in relation to acquisition, ownership and fuel taxes, the pre-tax user cost of cars, and income per capita. In contrast with earlier studies that have tended to focus on samples of revealed or stated preference data from specific highly developed countries (or small groups of countries), we estimate our model by means of data from 68 large cities, among which 49 from OECD countries and 19 from non-OECD countries in Asia. The implications for sustainable development of the structural differences between these taxes in terms of the model outcomes are identified. Our results indicate that acquisition and ownership taxes have little impact on behavior and environmental performance, but fuel and other use-related taxes are effective.Hence our econometric analysis suggests that taxes on variable costs have a greater impact on car ownership and use than taxes on fixed costs and that the former can be an effective way to reduce CO2 emission levels, despite the rebound effect. However, given the complexity of the transport sector and the wide range of behaviour responses that can take place in any given socio-economic and institutional setting, a ‘one size fits all’ conclusion should not be drawn from the rather low-dimensional analysis conducted in this study.A very clear finding is that our results confirm the importance of the fuel tax as an effective policy instrument in transportation. The fuel tax coefficient is highly significant in our econometric model. A 1 percent increase in the fuel tax per kilometre travelled reduces car ownership density, VKT per capita and CO2 emission all by about 0.6 percent. The results are consistent with the expectation that increased fuel taxes discourage further growth in car ownership, lead to switches to other transportation modes and also lead to manufacturer and driver responses that lower CO2 emission per capita. The impact on average annual vehicle kilometres travelled per car is much less, however. The elasticity is -0.1 (significant at the 5 percent level). This low elasticity is due to the so-called rebound effect : the increasing price of fuel has led manufactures to design and sell new cars with ever-more economical engines. Consequently, the disincentive of the higher fuel price is partially offset by the greater economy, leading to a lower overall impact on average VKT travelled per car per year. While the results suggest the greatest responsiveness of private car ownership and use to the fuel tax, they do not preclude that a combination of tax instruments would be more effective than a single one.JEL classification : O57, Q51, Q56, Q58, R48