Cell phone subscribership in the United States has grown dramatically in recent years, from 92,000 people in 1985 to more than 77 million in 1999. A recent National Highway Traffic Safety Administration survey reports that 44 percent of drivers have a cell phone with them while driving, a number that will only increase with the proliferation of phone ownership. We calculated that car accidents associated with phone use account for about 300 deaths per year. While small in comparison to the 41,000 annual deaths from car accidents, these deaths raise the question whether cell phone use while driving is justifiable. We think a ban is unwise at this time because vehicular cell phone use provides substantial personal and societal benefits, but does not contribute to a large number of serious accidents.
Cell phones in cars make life more convenient and safe: People can coordinate their errands and schedules with friends and family, parents can check up on their children, stranded motorists can call a tow truck or get help in an emergency, and motorists frequently use cell phones to report accidents and fires to police. Indeed, it is nearly impossible to quantify these everyday benefits. How does one determine the number of lives saved by the near‐instantaneous deployment of an ambulance to a crash scene, or the number of people who have been rushed to the hospital in time to prevent death or permanent injury?
A prudent regard for safety doesn’t imply cell phones should be banned. Americans are willing to tolerate some 41,000 annually deaths from car accidents. If we wish to decisively curtail automobile deaths, the national speed limit should be set at 10 miles per hour and vigorously enforced–yet we’re not willing to do that, because that inconvenience outweighs the pleasure and efficiency of being able to get places quickly.
And a ban on drivers’ use of cell phones might not make much difference anyway. Our best estimates of accident and fatality reductions do not take into account how drivers would alter their behavior in response to regulation; if police tried to enforce a ban, drivers might simply switch to other dangerous activities. Thus, the net reduction in accidents and fatalities is likely to be overstated, so the benefits of regulatory intervention could be quite small. Also, government regulations are notoriously slow to be enacted. Wireless technology is advancing so fast that regulation could soon be unnecessary.
Indeed, it is likely that the market will more effectively address cell phone risks than will government intervention. If the cell phone problem becomes serious enough, car insurance companies can classify drivers who use cell phones in higher‐risk groups and charge them commensurately higher premiums. Because an insurance company bears the burden of reimbursing injured parties for their losses, a company may decide to charge drivers who use cell phones higher premiums, to compensate for the increased risk that cell phones force the company to assume.
Instead of regulating now, the government should carefully monitor the problem and improve the information base for making regulatory decisions. For example, more research on the effectiveness of hands‐free devices in countries and municipalities that have required their use could help determine whether those devices should be encouraged. Also, if all states included a statistic in their accident reports describing whether the driver was using a cellular phone at the time of a collision, the resulting national data would be much more reliable.
The broader lesson is that the mere existence of a problem does not mean government action is needed. Drivers’ cell phone usage does lead to an increase in accidents and fatalities, but it is not obvious that new regulations would significantly reduce the problem. Moreover, a strong case needs to be made that the likely benefits of a ban exceed the costs by a significant amount. The case has yet to be demonstrated for prohibiting drivers from using cell phones, and we find just the opposite — a ban would flunk a cost‐benefit test.