All Commentary
Friday, September 1, 2006

Privatize the DMV

John Semmens ([email protected]) is a transportation policy analyst at the Laissez Faire Institute in Arizona.

For most Americans, driving is the most dangerous activity they undertake on a regular basis. We ought to try to make the roads as safe as is humanly possible. Unfortunately, government ownership of the roads and the vehicle-registration/driver-licensing process undermine safety. In “Highways: Public Problems and Private Solutions” (The Freeman, March 1985), I wrote about privatizing the roads. Now I will examine the possibility of privatizing registration and licensing.

One approach governments have adopted to deal with the hazards of driving is mandatory insurance. In 44 states plus the District of Columbia, liability insurance is a requirement for vehicle registration. Still, many vehicles lack insurance. The Insurance Research Council has estimated that about 14 percent of motorists in the nation operate uninsured vehicles.

Uninsured vehicles, of course, are only half the problem. Other vehicles are grossly underinsured. Most states that require insurance allow absurdly low minimum amounts of coverage. Arizona, for example, permits vehicles to be operated with liability coverage as low as $15,000 for bodily injury to a single victim, $30,000 for multiple victims, and $10,000 for property damage. It doesn’t take much of a wreck to “total” a car. Replacing one “totaled” car could easily cost more than $10,000. Considering that, according to the National Highway Traffic Safety Administration, about 85 percent of the traffic accidents in the United States involve more than one vehicle, it should be readily apparent that many crashes will produce property damage in excess of the minimum mandated coverage. As it is, the average cost of a “property damage only” accident is in the thousands of dollars. Minor-injury accidents result in costs averaging nearly $18,000 per accident. Major-injury accidents generate an average cost of $55,000. Fatal accidents produce damages in the million-dollar range. Next to this, the typical $15,000 to $30,000 coverage for injuring or killing someone is hopelessly inadequate.

So “mandatory” insurance creates a false sense of security.

There are two models useful in analyzing this situation. On the one hand, we could view roads through the “ballpark” model. In the ballpark customers are warned that the management assumes no responsibility for any injuries or damages in the normal course of the game. If this model were applied to roads, anyone who purchased the necessary licenses and vehicle registrations would have access. Drivers venturing onto the roads would do so at their own risk and with the explicit warning that they might be harmed by others who were unable to pay compensation. Road users would determine whether to buy insurance or not.

The chief advantage of the “ballpark” model is that it would remove the ambiguity regarding who should bear the responsibility for insurance. No one could claim that the government has guaranteed that other drivers are insured. Knowing there are no insurance requirements would inspire those who want to be indemnified against damages to purchase their own adequate levels of coverage.

The chief disadvantage of the “ballpark” model is that driving might become more financially risky. Those without insurance could easily sustain losses that could bankrupt them. Those who do purchase insurance might well have to buy substantially larger amounts of coverage since there may be more uninsured drivers. So even though the total number of vehicle crashes and the social cost of traffic accidents would likely be lower, the increased incidence of bankruptcy among the uninsured and the subsequent redistribution of financial burdens to those who are risk-averse might be viewed as undesirable.

An alternative to the “ballpark” model is the “Disneyland” model. In Disneyland, customers are covered by the business’s liability insurance. Consequently, the management sets its own risk-reducing restrictions, such as barring some customers from some rides and attractions. Since the business is held strictly liable for any damages suffered by those entering the park, management will strictly enforce its rules.

Like the “ballpark” model, the “Disneyland” model applied to roads would also reduce the ambiguity concerning who will be responsible for damages. Enforcement of the insurance requirement would be best achieved by having insurers issue the licenses and vehicle registrations. This differs from the current system in which the mandatory insurance is sold by private vendors, but is enforced by a state agency. The government charged with enforcing mandatory-insurance laws assumes no liability for damage done by drivers who fail to comply. Thus there is no significant financial consequence for not enforcing the mandate.

If the current weakly enforced laws were to be replaced by a fully privatized system based on the “Disneyland” model, there would be a much stronger incentive to see that all vehicles had adequate insurance coverage. This would also help to keep the worst drivers off the roads. If insurance companies had to accept full liability for whomever they issued a driver’s license and vehicle registration, the problem of “underinsureds” would vanish. Bad drivers would not have the option of buying a woefully inadequate policy. They would be required to pay the full cost of their actuarial risk in order to obtain a driver’s license and vehicle registration. The insurance company would see to this as a matter of business survival.

Given the significant uncompensated costs inflicted on victims of underinsured and uninsured drivers, a coherent solution is needed. The contrasting “ballpark” and “Disneyland” models offer this. While each model is likely to make the roads safer, the “ballpark” model does this by imposing more of the burden on the cautious drivers. The “Disneyland” model would make the roads safer by removing more of the high-risk drivers.

Disneyland in Detail

Let’s look in detail at how this model might work. If insurers issued licenses and registrations it would be absolutely clear who was legally responsible for a particular vehicle being on the road. Insurers would have a strong incentive to make sure that every vehicle and driver they insured had adequate coverage. They would also have a strong incentive to keep uninsured vehicles and drivers off the roadways.

To clarify financial responsibility for potential damages, the privatization law would state that as long as a vehicle bore the license plate of an insurer, that insurer would be held liable for any damages caused by that vehicle. With this kind of provision it is likely that insurers would only issue plates after thoroughly testing the skills and investigating the driving records of customers. If an insurer was not satisfied that a prospective customer is truthful or a good risk, it could refuse to issue a policy. Under privatization, suspect applicants would not be able to use the roads legally until they found an insurer willing to issue a policy and its verifying license plates.

Since issuing an auto insurance policy and license plates would be simultaneous, the uninsured would be easier to spot than at present. Indeed, given the greater degree of responsibility placed on each insurer, it seems likely that steps to improve the visibility of license plates would be taken. For example, the much-tested, but little-deployed electronic license plate would likely be a widespread innovation, permitting plateless vehicles to be detected by automated means.

The dodge of buying insurance just to obtain a registration tag and then canceling the insurance once the tags were received would become much more difficult. Since the insurer would be responsible for damages, it would have a strong incentive to require a substantial payment or deposit refundable only when the plates are returned. Given the many insurance-sales locations and computer networks, obtaining or returning plates ought to be a lot more convenient than dealing with the DMV.

Individuals would be free to shop for the best license, registration, and insurance deal they could find. Each insurer would be free to establish its own criteria for issuing policies and license plates. Some insurers may cover only low-risk drivers. Others may cover high-risk drivers at correspondingly higher premiums. Some insurers may wish to give written and/or road tests to prospective customers. Some insurers may want to conduct regular safety inspections of their customers’ vehicles. Others may wish to encourage or require some or all the vehicles they insure to be equipped with safety-enhancing devices (for example, an ignition that can only be activated after the driver passes an automated, on-board breathalyzer test). Customers may choose to accept some limitations on driving in exchange for lower premiums (for example, driving only during daylight hours). Others may prefer to pay more to escape restrictions. In short, there is likely to be a wide variety of payment/license/registration options available. Nevertheless, those who could not meet the minimum requirements of any insurer would not be issued vehicle plates. The streets would be safer.

Potential Benefits

The potential benefits would occur in two areas. Privatizing vehicle registrations would shift costs from innocent victims of bad driving to those who cause the accidents. As the perpetrators of damage were forced to bear a larger share of the consequences of their actions, we could expect some modifications in their behavior. To remain profitable, insurers would have to be good at matching premiums to risk. This would motivate them to reduce risk. A price structure that accurately reflected risk would push drivers toward safer behavior. So over the long run, not only would uncompensated losses be shifted back to those at fault, total losses also would likely be reduced.

The other source of potential benefits would come from eliminating functions of the state departments of transportation. If we privatized the registration and driver’s license functions, we could probably reduce government spending by about $3 billion per year.

The potential for “one-stop-shopping” convenience under full privatization is substantial. One could buy a car and get it registered and insured all at one location. The insurers and auto dealers would have an incentive to make the process as expeditious as possible. In fact, competition among insurers and dealers would help promote efficiency and convenience.

Consider a typical transaction with the current DMV. It’s your lunch hour. You have chosen this opportunity to take care of some business down at the department. Your first task is to find the nearest branch office. They’re not always conveniently located. They’re certainly scarcer than any other auto-related business location. As your search for the nearest office drags on, you pass numerous gas stations, a half-dozen auto-parts stores, several auto-insurance offices, and a few auto dealerships.

Finally, you locate the DMV office. You walk in and get in line. But your advance to the head of the line is slowed because many employees have chosen this time to take lunch. When your turn finally arrives you are greeted by an employee whose occupation has ranked last in a survey of civility. You think the service could be more convenient, expeditious, and courteous, but it isn’t and won’t be likely to get that way. The monopoly position of the state agency pretty much assures that it won’t. It is not as if you could take your “business” elsewhere.

Shifting the vehicle registrations and driver’s licenses to the private sector would require legislation. This proposal would likely be resisted by the DMV bureaucracy. After all, if the agency is no longer needed to register motor vehicles or issue driver’s licenses, we may well question whether it is needed at all. At the very least, we may be talking about a 50 percent cutback in its budget. Hundreds of people would see their government jobs eliminated. These prospective consequences would inspire objections.

We may also expect some initial opposition from the auto-insurance industry. At the outset the increased responsibility would provoke uncertainty. New means of coping with this uncertainty would have to be learned. However, once the insurers understood that they would be compensated by their customers for the costs of issuing registrations and licenses, and that privatization would not require them to provide subsidies to high-risk drivers, they should be more receptive to the idea. The opportunity to play a more direct role in controlling the risk of the roadway should be perceived as a means of reducing underwriting losses over the long term. As the roads become safer, insurers’ losses would fall. (Ultimately, premiums would be expected to fall as well, but not as rapidly as underwriting losses.) In addition, many of those currently evading the insurance mandate would become the reluctant customers of the insurance industry. Lastly, many of those currently underinsuring their vehicles would be required by insurers to buy adequate coverage. All this should improve insurer profitability.

The general public may be apprehensive about letting private-sector insurers decide who gets on the roads. But it seems odd to prefer having that determination made by a bureaucracy with no responsibility for its decisions. We have grown accustomed to allowing businesses to decide who obtains credit, and we expect the decisions to be made on rational criteria. And they are. Reflection on how the private sector has handled this vital segment of contemporary life should help to alleviate some of the public’s apprehension.

Of course, that segment of the population that is currently flouting the insurance mandate or exploiting it by underinsuring their vehicles would be expected to raise quite a fuss. While we should not be persuaded by the objections of those who wish to continue passing the burdens of their own risky driving on to others, we can envision some means of addressing their legitimate concerns.

The case most deserving of sympathy is that of the individual whose past driving has taught him a lesson. It is unfortunate for such individuals that many more proclaim to have learned lessons than actually have. Consequently, individuals with bad driving records would undoubtedly have trouble obtaining insurance and permission to use the roads. We should expect insurers to establish methods of serving this market niche. One method would be to require the vehicle of such a person to be equipped with devices that enhance safe operation. We already mentioned the possibility of a breathalyzer-ignition link. Other options could include vehicles that could only run during daylight hours (perhaps having a solar-collector connection to the engine or transmission) or vehicles whose maximum speed could not exceed a low setting (perhaps having “speed governors” placed on the engine). Insurers might want to require periodic safety inspections of the vehicle as a condition of issuing a registration. They might see fit to require regular driving tests for those whose driving behavior has been demonstrated to be more hazardous than average.

Some high-risk but repentant drivers could work toward a full reinstatement of driving privileges by demonstrating meritorious performance under limited driving privileges. For others, though, the outlook will be less sanguine. There are some people who should not be behind the wheel. Stopping them from driving not only helps to preserve the health and lives of others, but also may save the driving-deprived individual from injuring or killing himself. Incompetent drivers must find other means of meeting their transportation needs.

  • John Semmens is a research fellow at the Independent Institute and research project manager in the Arizona Department of Transportation Research Center.