Policy commentary

Pay-As-You-Drive Auto Insurance - Weekly Policy Commentary - December 15, 2008

Dec 15, 2008 | Jason Bordoff, P Noel

Welcome to the RFF Weekly Policy Commentary, which is meant to provide an easy way to learn about important policy issues related to environmental, natural resource, energy, urban, and public health problems.

Automobile use in the United States is underpriced as motorists do not pay for the full costs of pollution, congestion, and traffic accidents when deciding how much to drive. However, there is intense opposition to gasoline taxes, peak-period congestion fees, and other policies that would reduce these externalities by raising the costs of driving. This week, Jason Bordoff and Pascal Noel discuss pay-as-you-drive insurance, which offers a novel approach for reducing automobile use, without raising the private costs of vehicle ownership and use for the majority of drivers. 

Under the current lump-sum pricing structure for auto insurance, drivers who are similar in other respects—age, gender, location, driving safety record—pay nearly the same premiums if they drive 5,000 or 50,000 miles a year, even though the likelihood of being in­volved in a collision increases with each mile driven. Hardly an efficient approach, to put it mildly.

Just as an all-you-can-eat restaurant encourages more eating, all-you-can-drive insurance pricing encourages more driving because drivers don’t face the marginal insurance cost for each mile driven. The extra driving that results imposes significant costs on society: more accidents, congestion, carbon emissions, local pollution, and dependence on oil.

Moreover, the current structure is inequitable. It forces low-mileage drivers to subsidize the accident cost of high-mileage drivers in each risk class, even though the former are responsible for fewer accidents. This problem is particularly disturbing given that low-income people tend to drive less on average.

 A simple alternative, known as pay-as-you-drive (PAYD) auto insurance, avoids the problems of the current system. With PAYD, the price of auto insurance would be tied to the number of miles driven. Other rat­ing factors such as location, age, vehicle type, and driving record still would be incorporated into this price, so higher-risk drivers would pay more per mile than lower-risk drivers.


Switching to PAYD could yield substantial benefits, according to our recent findings, which are based on data from the 2001 National Household Transportation Survey. The average driver would face a per-mile insurance premium of 6.6 cents per mile, instead of a yearly lump-sum cost of about $800. Because drivers could save money by driving less, we estimate driving (miles travelled) would fall by about eight percent.

Achieving a reduction on this scale would yield social benefits of about $60 billion a year, mostly from reduced accidents and congestion, but also from reduced carbon emissions, local pollution, and oil dependence.

And PAYD could achieve these gains while actually reducing the cost of driving for most drivers. Almost two thirds of households would save money under PAYD, with average savings (for those households that save) totaling $270 per vehicle. Most of the savings result from the elimination of the current subsidy from low-mileage to high-mileage drivers. The high proportion of drivers that would pay less reflects the fact that a minority of high-mileage drivers is responsible for a majority of driving within each risk class. In fact, we find that the top 20 percent of drivers are responsible for 45 percent of all miles driven.

Our research also shows that low-income families would especially benefit from PAYD, because low-income people tend to drive fewer miles. Every household income group making less than $52,500 (in 2001) would save money on average. Further, the savings for low-income groups are significant as a share of their total income, whereas any losses by high-income groups are not significant.


Despite the large social benefits from PAYD, there are currently barriers to its widespread adoption. For one, insurance regulations in many states prohibit or pose significant obstacles to pricing insurance by the mile. Since regulations were always written with yearly premiums in mind, per-mile premiums are sometimes technically illegal even if that was never the intention of the regulators. California, for example, just acted to address this issue and make it easier for firms to offer PAYD.

A second problem is that, even where it is legal, certain costs reduce the likelihood that firms will independently offer PAYD insurance. In order to price insurance per mile, firms or their customers would need to incur the cost of verifying mileage, either through odometer checks or devices that fit in each vehicle. While odometer readings could be inexpensive procedures if done on a widespread basis, there currently is no infrastructure of certified providers that insurance firms can use. And technological devices that automatically monitor and transmit mileage to insurance companies can be expensive, costing as much as $100 to install. Moreover, to institute PAYD, firms must develop new billing and administrative infrastructures, retool their advertising, and develop new actuarial models to determine appropriate risk-adjusted per-mile prices.

While private firms and their customers would have to bear these costs, much of the benefit from reduced mileage would accrue to other insurance companies and to society as a whole. In our analysis, we find the social benefit to be about $250 per vehicle per year. This is a classic case of a positive externality, and in these cases the government has a clear role to play in promoting a better social outcome. To address the market failure around monitoring costs, the government could require that odometer readings be performed as part of required safety and emissions inspections or by certifying vehicle service businesses in other states to perform odometer readings.


  Jason BordoffJason Bordoff is policy director of the Hamilton Project, an economic policy initiative housed at The Brookings Institution committed to promoting more broadly shared prosperity. He has written on a broad range of economic policy matters, with a focus on climate and energy policy, income security and inequality, and tax policy. Bordoff is also a term member of the Council on Foreign Relations and serves on the board of the Association of Marshall Scholars. He graduated with honors from Harvard Law School, where he was treasurer and an editor of the Harvard Law Review.  

Pascal NoelPascal Noel is a research analyst at The Brookings Institution’s Hamilton Project. He has written on the economics of climate change policy, energy security, alternative energy, and a range of transportation issues. He holds a masters degree in economics from the London School of Economics and a BA from Yale University in economics and a joint degree in ethics, politics, and economics.

The government could also offer a tax credit for each new mileage-based policy that an insurance company writes. We recommend a $100 tax credit, which would cover the cost of most technological devices that could easily measure and transmit mileage data. The tax credit could be phased out, once roughly five million vehicles (two percent) are signed up, after which point PAYD is expected to take off on its own. To address the development costs, the government could increase the funding available to PAYD pilot programs.

While we believe that PAYD would be a significant improvement, it is not an adequate policy response to driving related harms all by itself. It does not force drivers to internalize the external social costs of the congestion, accidents, pollution, and oil dependence they cause. It simply corrects a failure with the way that auto insurance is priced today and the inefficient and inequitable consequences of that pricing structure.

Ideally, PAYD would be complemented with other policies, such as carbon pricing and a congestion charge, which directly target the driving-related social harms. But many of these other policies raise the cost of driving, which is politically challenging, especially in these tough economic times. The promise of PAYD is that it can achieve some of the benefits of these user fees by creating incentives to reduce driving without raising the cost of driving in aggregate, and indeed lowering it for the majority of drivers. What is good for drivers, in this case, is also good for society.


Views expressed are those of the author. RFF does not take institutional positions on legislative or policy questions.

To receive the Weekly Policy Commentary by email, or to submit comments and feedback, contact comments@rff.org.


Additional Resources:

Bordoff, Jason E. and Pascal J. Noel. 2008. Pay-As-You-Drive Auto Insurance: A Simple Way to Reduce Driving-Related Harms and Increase Equity. The Hamilton Project, the Brookings Institution. July.

Bordoff, Jason E. and Pascal J. Noel. 2008. The Impact of Pay-As-You-Drive Auto Insurance in California. The Hamilton Project, the Brookings Institution. July.

Edlin, Aaron D. 2003. Per Mile Premiums for Auto Insurance. In Economics for an Imperfect World: Essays in Honor of Joseph Stiglitz, ed. Richard Arnott, Bruce Greenwald, Ravi Kanbur, and Barry Nalebuff. Cambridge: MIT Press.

Parry, Ian W. H. 2005. Is Pay-as-You-Drive Auto Insurance a Better Way to Reduce Gasoline Than Gasoline Taxes? AEA Papers and Proceedings 96 (2): 287-93 (May). Related RFF Discussion Paper 05-15, available at http://www.rff.org/files/sharepoint/documents/RFF-DP-05-15.pdf

Harrington, Winston, Ian Parry, and Margaret Walls. 2007. Automobile Externalities and Policies. Journal of Economic Literature. Vol. XLV pp. 374-400. Related RFF Discussion Paper 06-26, available at http://www.rff.org/files/sharepoint/rff/Documents/RFF-DP-06-26-REV.pdf.