Abstract
We analyse the effects of distortionary company car taxation through increased household car consumption for the Netherlands. We use several identification strategies and demonstrate that for about 20 % of households company car possession increases car ownership. The annual welfare loss of distortionary company taxation through increased car ownership is generally rather small, maximally €120 per company car, and much less than the welfare loss through increased expenditure on the company car. However, for policies that exempt households from paying tax on their company car, the annual deadweight loss is likely higher. Our first-best tax policy recommendation is to increase company car tax rates. However, our current results suggest that a second-best policy, which keeps average company car taxation constant but which reduces the marginal tax on cheaper cars and increases the marginal tax on expensive cars, would be welfare improving as overconsumption of company cars will be reduced.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Company cars are usually new, used for a period of about 3 to 4 years, and then sold on the secondhand market. In Europe, about 15 to 20% of two-adult households have access to a company car (Gutiérrez-i-Puigarnau and Van Ommeren 2011).
To give a concrete example, in the Netherlands maximally 24% of the company car’s purchase price is added to annual taxable income. However, the annual cost of such a car is about 40% of its purchase price. In the literature, it is common to distinguish between the price of owning and the price of using a car. This difference is not essential here, because employers usually pay for the possession of the car (usually through a car lease) including insurance and taxes, but also for its use including fuel consumption and repairs (Wuyts 2009).
However, even when a worker uses a company car for business purposes then this does not necessarily imply that the worker is more productive, because most workers use a car for commuting. So when not offered a company car, they may use a private car for business purposes (and receive reimbursement from their employer for the marginal costs related to business trips).
Empirical car ownership studies, which aim to explain the number of cars in the household are abundant in the transportation literature (see e.g. Holtzclaw 1994; Dargay and Gateley 1999; Kumar and Rao 2006). However, these studies ignore company cars, which is a major omission when applied to European countries.
We focus on the long run, because it is then plausible that households fully adapt to changes in prices, similar to households that have a company car and are confronted with different prices.
It is suggested in the literature that most of the reported car price elasticities of ownership are downward biased, because they omit car quality measures, so the price elasticities are plausible somewhat larger.
For example, in the UK, until recently, the tax level decreased strongly with business mileage.
The results by Zax (1988) suggest that monopsony has little effect on the welfare effects of fringe benefits taxation, so this issue will be ignored. The car market is described by the partial-equilibrium perfectly competitive model, where costs and pre-tax prices are not affected by the tax distortion. We abstract from market power by car suppliers (see Berry et al. 1995; Verboven 1996).
In general, it is preferred to model the company car ownership decision as a discrete choice problem, see, in particular, De Borger and Wuyts (2011). However, as explained later on in Sect. 3.2, we will estimate linear probability models, rather than discrete choice models. Hence, in order to relate theory to empirical application, we use a model which relies on the assumption that the demand for the second car is a continuous good.
When the type of car received as a company car is not the preferred choice of the household then the company car is an imperfect substitute. This was common in the 90s when the lease market for cars was not well developed in Europe, so company cars were usually owned by the employer without considering the preferences of households.
We ignore here that a car is a discrete good. Assuming a discrete demand function, welfare analysis is possible using the methodology proposed by Small and Rosen (1981).
This selection is not problematic, because merely 6 % of households do not own a car and these households seldom belong to the social group of households who receive a company car offer.
For most variables, information is almost complete. The main exception is household income which is missing for about one third of the observations. Including these observations, and using a control dummy variable to control for missing income, generates almost identical effects.
The control variables have the expected effects (e.g. income has a strong positive effect; the implied income elasticity is about 0.64).
Households who receive a company car may not immediately sell their private car for several reasons. First, it takes time to find a buyer; second, some households may be uncertain of the expected employment duration and company car possession may be considered temporary; third, some households might be uncertain of the benefits of having more than one car and keep the private car for a short period to experience the advantage of having two cars.
In this specification we ignore that commuting distance is potentially endogenous. We are not able to control for other private kilometres travelled as it is unobserved in this dataset. However, this is less relevant as private travel (excluding commuting) is roughly the same for households with and without a company car (Gutiérrez-i-Puigarnau and Van Ommeren 2011).
In the dataset, we do not have specific information about the type of industry, but we have information about the names of the pension fund of the current employer. In this way, we are able to derive the type of industry in an unusual way. So, we are able to distinguish between employees that work in public organisations (government, universities) or private organisations. We are also able to distinguish between broad categories of private organisations. This information is later on used in the sensitivity analysis.
It also appears that our instruments have no direct impact on car ownership conditional on including company car possession suggesting that any effect of sector is through company car possession.
As an informal test, we exclude company car in the OLS analysis, and find a strong effect of type of private sector which also implies that the effect of type of private sector is entirely through company car possession, supporting our assumption that public sector is a valid instrument.
One objection to our analysis may be that employers that lease a large number of cars receive bulk discounts due to a reduction in transaction costs. Bulk discounts are typically less than 10 %, but precise figures are unknown to us. If we assume that employers provide company cars at an average reduction of 5 % (and that these reductions are not due to market power), then the welfare loss due to the tax distortion is overestimated by 5 %.
However, at the same time, the tax on company cars was slightly increased, which reduces the probability that a company car is offered by the firm, so the increase in welfare loss per company car may have been offset by a reduction in the number of company cars. Furthermore, employees have a strong incentive to cheat by declaring that the car is only used for commuting but not for other trips. Cheating households do not have an additional incentive to increase car ownership. According to the tax authorities, about 20 % of households cheat.
Tax advantage given to company cars may (partially) compensate for the presence of distortionary purchase taxes on personal cars that are substantial in Europe. Since in most other European countries purchase car taxes are lower than in the Netherlands, our conclusion that the tax system is distortionary regarding company cars can be generalised to other European countries. This conclusion may not hold in countries such as Denmark and Norway, where purchase taxes are 180 and 100 %, respectively, of the car’s purchase price. For these countries, it is possible that favourable company car taxation generates substantial welfare benefits as purchase taxes on personal cars can be argued to be too high in these countries, so that company car taxation corrects for the purchase tax distortion.
References
Arzaghi M, Henderson JV (2008) Networking off Madison Avenue. Rev Econ Stud 75(4):1011–1038
Bento AM, Goulder LH, Jacobsen MR, Von Haefen RH (2009) Distributional and efficiency impacts of increased us gasoline taxes. Am Econ Rev 99(3):667–699
Berry S, Levinsohn J, Pakes A (1995) Automobile prices in market equilibrium. Econometrica 63(4):841–90
Brons MRE, Nijkamp P, Pels AE, Rietveld P (2008) A meta-analysis of the price of elasticity of gasoline demand: a SUR approach. Energy Econ 30(5):2105–2122
Copenhagen Economics (2010) Company car taxation. Taxation Papers, working paper 22
Craft ED, Schmidt RM (2005) An analysis of the effects of vehicle property taxes on vehicle demand. Natl Tax J 55(4):697–720
Dargay JM, Vythoulkas PC (1999) Estimation of a dynamic car ownership model: a pseudo-panel approach. J Transp Econ Policy 33(3):287–302
Dargay JM, Gateley D (1999) Incomes effect on car and vehicle ownership, worldwide: 1960–2015. Transp Res A 33(2):101–138
De Borger B, Wuyts B (2011) The tax treatment of company cars, commuting and optimal congestion taxes. Transp Res B 45(10):1527–1544
De Palma A, Lindsey R, Proost S (2006) Research challenges in modelling urban road pricing: an overview. Transp Policy 13(2):97–105
Ehrenberg RG, Smith RS (2003) Modern labor economics: theory and public policy. Addison-Wesley, New York
Goodwin P, Dargay JM, Hanly M (2004) Elasticities of road traffic and fuel consumption with respect to price and income: a review. Transp Rev 24(3):275–292
Gutiérrez-i-Puigarnau E, Van Ommeren JN (2011) Welfare effects of distortionary fringe benefits taxation: the case of employer-provided cars. Int Econ Rev 52(4):1105–1122
Han B (2001) Analyzing car ownership and route choices using discrete choice models. Dissertation, Royal Institute of Technology, Stockholm
Holtzclaw J (1994) Using residential pattern and transit to decrease automobile dependence and costs. Natural Resources Defense Council, New York
Johansson O, Schipper L (1997) Measuring the long-run fuel demand for cars. J Transp Econ Policy 31(3):277–292
Kumar M, Rao KV (2006) A stated preference study for a car ownership model in the context of developing countries. Transp Plan Technol 29(5):409–425
Parry IWH, Small KA (2005) Does Britain or the United States have the right gasoline tax? Am Econ Re 95(4):1276–1289
Shoup DC (2005) The high cost of free-parking. Planners Press, Chicago
Small KA, Rosen HS (1981) Applied welfare economics with discrete choice models. Econometrica 49(1):105–130
Small KA, Verhoef ET (2007) The economics of urban transportation. Routledge, London
Tillema T, Van Wee B, Rouwendal J, Van Ommeren JN (2008) Firms: changes in trip patterns, production prices, locations and in the human resource policy due to road pricing. In: Verhoef E, Bliemer M, Steg L, Van Wee B (eds) Pricing in road transport: a multi-disciplinary perspective. Edward Elgar, Cheltenham, pp 106–127
Ubbels B (2006) Road pricing: effectiveness, acceptance and institutional aspects. Dissertation, VU University, Amsterdam
Van Ommeren JN, Wentink D (2012) The (hidden) costs of employer parking policies. Int Econ Rev 53(3):965–978
Verboven F (1996) International price discrimination in the European car market. RAND J Econ 27(2): 240–268
Vickrey WS (1963) Pricing in urban and suburban transport. Am Econ Rev 53(2):452–465
Whelan G, Fox K, Daly A (2001) Updating car ownership forecasts. Institute for Transport Studies, Leeds
Wuyts B (2009) Essays on congestion, transport taxes, and the labour market. Dissertation, Universiteit Antwerpen
Zax JS (1988) Fringe benefits, income tax exemptions and implicit subsidies. J Public Econ 37(2):171–183
Acknowledgments
This research has been supported financially by the Netherlands Organisation for Scientific Research (NWO).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
van Ommeren, J.N., Gutiérrez-i-Puigarnau, E. Distortionary company car taxation: deadweight losses through increased car ownership. Empir Econ 45, 1189–1204 (2013). https://doi.org/10.1007/s00181-012-0659-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-012-0659-0