Car fleet policy evaluation: The case of bonus-malus schemes in Sweden

ABSTRACT

In this paper, we evaluate the bonus-malus schemes proposed by governmental investigation in Sweden in 2014. The objective was to reach the target of maximum 95 g/km of average ${\text{CO}}_2$ emissions for new car sales by 2020. Two bonus-malus schemes along with several other policies were introduced in different scenarios as measures to reach the target. The scenarios differ in design regarding bonus-malus schemes, vehicle circulation tax, clean car premiums, company car benefits tax, and fuel tax. Both private and company car segments are targeted in these scenarios. We use a nested logit model for car type choice to predict the effects of the proposed policy scenarios on Swedish new car sales. Moreover, we introduce a methodology to predict the future of car supply. Our model results show that none of the three proposed policy scenarios is successful enough to meet the average ${\text{CO}}_2$ emissions target. Furthermore, although the number of electric, plug-in hybrid, and alternative fuel cars will increase, new car sales will still be dominated by fossil-fueled cars in all scenarios. The average ${\text{CO}}_2$ emissions in the scenarios containing bonus-malus schemes are not lower than that of the business-as-usual scenario. However, introducing bonus-malus schemes on its own would reduce ${\text{CO}}_2$ emissions showing that interacting with other non-${\text{CO}}_2$-differentiated policies counteracts their effects. Moreover, bonus-malus schemes are predicted to give a budget surplus effect.

KEYWORDS:

• Bonus-malus
• car supply prediction methodology
• car type choice models
• nested logit model
• policy interaction
• taxation

Acknowledgments

This research was supported by the Centre for Transport Studies in Stockholm, which provided the resources for this study. We would like to acknowledge Ronny Svensson, who generously put the Ynnor⁷ database on vehicle attributes at our disposal. We also thank Björn Hugosson, City of Stockholm, for comments and suggestions.

Notes

¹ In this paper, we use vehicle and car interchangeably and by both words we mean personal cars.

² Employees that have cars for private use provided by the employer are benefit taxed. The benefit value is based on the prices of the new cars and interest rate for each year, which is calculated by the tax authorities and taxed as income. For detailed information refer to Appendix A.

³ Since the vehicle circulation tax is the same in scenarios B and C, the difference in results of these two scenarios is due to the difference in bonus-malus schemes. Therefore, only checking the bonus-malus scheme of scenario B would be sufficient to illustrate the effects also for scenario C.

⁴ ${\beta }_{\text{purchase price}}\text{(SEK}\text{1000}\text{)}=-0.009473$ and ${\beta }_{\text{operating cost}}=-0.000291$ (Transek, 2006).

⁵ ${\beta }_{\text{taxable benefit}}\text{(SEK}\text{1000}\text{)}/{\beta }_{\text{operating cost}}=-0.095515/-0.000354=0.27$ for company cars with lease and ${\beta }_{\text{taxable benefit}}\text{(SEK}\text{1000)}/{\beta }_{\text{operating cost}}=-0.083094/-0.000364=0.23$ for company cars without lease.

⁶ Source: www.bilsweden.se

⁷ www.ynnor.se

^a We assume the same value for the average curb weight for the following years as well.

Additional information

Funding

This work was supported by the Center of Transport Studies, Stockholm.