Working Paper: NBER ID: w20333
Authors: Don Fullerton; Li Gan; Miwa Hattori
Abstract: Using three years of data from the 47 prefectures of Japan, we estimate behavior of households who simultaneously make discrete decisions about vehicle ownership and continuous decisions about driving distance. We use the estimated parameters to calculate elasticities and to simulate the effects of alternative pollution control policies such as taxes on gasoline, on distance, or on particular cars. Given choices about cars and distance, we also calculate emissions. Since we model simultaneous choices, both the chosen distance and the chosen car can be affected either by a tax on distance or by a tax on car characteristics. We find expected signs for coefficients on price and income. Car choices are relatively inelastic, however, either to taxes on cars or to taxes on gas or distance. Thus emissions are more affected by taxes on gasoline than by taxes on particular vehicles. \n \nYet taxes on cars have lower costs on consumers and thus lower marginal cost of abatement. Given that the existing gas tax already achieves some abatement, mostly through driving reduction, this analysis suggests that further abatement from the use of distance-reducing taxes is more costly than achieving some marginal abatement from induced changes in car choices. The option with the lowest cost is to tax each car at a rate proportional to its emission rate.
Keywords: vehicle emissions; tax policy; Japan; household behavior; environmental economics
JEL Codes: H23; Q52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Gasoline tax (H29) | Driving distance (R48) |
| Driving distance (R48) | Emissions (Q52) |
| Gasoline tax (H29) | Emissions (Q52) |
| Emission-based taxes (H23) | Marginal costs of abatement (Q52) |
| Vehicle characteristics tax (R48) | Vehicle choice (L92) |
| Vehicle choice (L62) | Emissions (Q52) |