Last week Austin Police Chief Art Acevedo advocated creating a new criminal offense: "driving while ability impaired." The problem with the current Texas law prohibiting "driving while intoxicated" (DWI), Acevedo explained, is that it doesn't allow him to arrest a driver whose blood-alcohol content (BAC) is below 0.08 percent without additional evidence of impairment.
"People sometimes focus on how many drinks they can have before they'll go to jail," Acevedo told the Austin-American Statesman. "It varies….A person may be intoxicated at 0.05, and you don't want them out driving." Acevedo wants to be able to arrest people with BAC levels as low as 0.05 percent, and he may have support for that idea in the state legislature. John Whitmire (D-Houston), chairman of the state Senate's Criminal Justice Committee, told the Statesman Acevedo's plan "might be one way to go," adding, "Some people shouldn't be driving after one drink—probably below the 0.08 limit—and this could address that."
Bill Lewis, head of the Texas chapter of Mothers Against Drunk Driving, agreed. "I don't see how it would hurt," he told the paper. "The level of 0.08 is where we know most people are good and drunk...and there are people who are driving at less than the limit who probably should not be. It might keep some people from driving [drunk] again."
Acevedo, Whitmore, and Lewis are right, although probably not in the way they intended. People do react to alcohol differently. For many people one drink may well be too many, while experienced drinkers can function relatively normally with a BAC at or above the legal threshold for presuming intoxication. A person's impairment may also depend on variables such as the medications he is taking and the amount of sleep he got the night before. Acevedo et al.'s objections to the legal definition of intoxication highlight the absurdity of drawing an arbitrary, breathalyzer-based line between sobriety and criminal intoxication.
The right solution, however, is not to push the artificial line back farther. Instead we should get rid of it entirely by repealing drunk driving laws.
Consider the 2000 federal law that pressured states to lower their BAC standards to 0.08 from 0.10. At the time, the average BAC in alcohol-related fatal accidents was 0.17, and two-thirds of such accidents involved drivers with BACs of 0.14 or higher. In fact, drivers with BACs between 0.01 and 0.03 were involved in more fatal accidents than drivers with BACs between 0.08 and 0.10. (The federal government classifies a fatal accident as "alcohol-related" if it involved a driver, a biker, or a pedestrian with a BAC of 0.01 or more, whether or not drinking actually contributed to the accident.) In 1995 the National Highway Traffic Safety Administration studied traffic data in 30 safety categories from the first five states to adopt the new DWI standard. In 21 of the 30 categories, those states were either no different from or less safe than the rest of the country.
Once the 0.08 standard took effect nationwide in 2000, a curious thing happened: Alcohol-related traffic fatalities increased, following a 20-year decline. Critics of the 0.08 standard predicted this would happen. The problem is that most people with a BAC between 0.08 and 0.10 don't drive erratically enough to be noticed by police officers in patrol cars. So police began setting up roadblocks to catch them. But every cop manning a roadblock aimed at catching motorists violating the new law is a cop not on the highways looking for more seriously impaired motorists. By 2004 alcohol-related fatalities went down again, but only because the decrease in states that don't use roadblocks compensated for a slight but continuing increase in the states that use them.
The roadblocks are also constitutionally problematic. In the 1990 decision Michigan v. Sitz ,the Supreme Court acknowledged that stops at sobriety checkpoints constitute "seizures" under the Fourth Amendment but ruled that the public safety threat posed by drunk driving made them "reasonable." In the years since, the checkpoints have become little more than revenue generators for local governments. When local newspapers inquire about specific roadblocks after the fact, they inevitably find lots of citations for seat belt offenses, broken headlights, driving with an expired license, and other minor infractions. But the checkpoints rarely catch seriously impaired drivers. In 2009, according to a recent study by researchers at the University of California at Berkeley, 1,600 sobriety checkpoints in California generated $40 million in fines, $30 million in overtime pay for cops, 24,000 vehicle confiscations, and just 3,200 arrests for drunk driving. A typical checkpoint would consist of 20 or more cops, yield a dozen or more vehicle confiscations, but around three drunk driving arrests.
Checkpoints are only the beginning of what California DWI attorney Lawrence Taylor calls "the drunk driving exception to the Constitution." The Fifth Amendment right against self-incrimination has been turned upside down by state laws that instantly suspend the licenses of drivers who refuse to take roadside breath tests. Those breath tests are also fraught with problems. Most manufacturers of breath test machines have refused to turn over their source code, meaning DWI defendants can't assess the machines' margin of error, which can be a significant factor in a case where the difference between 0.80 and 0.79 for a first offense can be $1,000 or more in fines, mandatory alcohol awareness classes, and loss of driving privileges for up to a year.
Blood tests are far more accurate, but by the time a driver is pulled over, questioned, taken to the nearest hospital, and had his blood drawn, his BAC may be significantly different from what it was when he was driving. Perversely, the time lapse can have the effect of protecting guiltier motorists. Imagine a driver pulled over or stopped at a checkpoint after having "one for the road," knowing his house is a short drive away and the last drink won't kick in until he's sitting on his couch. At the time he is stopped, he is under the legal limit. But his BAC is rising, and it tops 0.08 by the time his blood is drawn at the hospital. By contrast, a driver who is impaired when he's pulled over, but who stopped drinking an hour or so before, benefits from the delay, since his BAC is falling by the time he arrives at the hospital.
Many states have tried to solve this problem by claiming another invasive power: They are now allowing police to forcibly take a blood sample on the side of the road.
These ever-expanding enforcement powers miss the point: The threat posed by drunk driving comes not from drinking per se but from the impairment drinking can cause. That fact has been lost in the rush to demonize people who have even a single drink before getting behind the wheel (exemplified by the shift in the government's message from "Don't Drive Drunk" to "Don't Drink and Drive"). Several studies have found that talking on a cell phone, even with a hands-free device, causes more driver impairment than a 0.08 BAC. A 2001 American Automobile Association study found several other in-car distractions that also caused more impairment, including eating, adjusting a radio or CD player, and having kids in the backseat (for more on such studies, see the 2005 paper I wrote on alcohol policy for the Cato Institute).
If our ultimate goals are to reduce driver impairment and maximize highway safety, we should be punishing reckless driving. It shouldn't matter if it's caused by alcohol, sleep deprivation, prescription medication, text messaging, or road rage. If lawmakers want to stick it to dangerous drivers who threaten everyone else on the road, they can dial up the civil and criminal liability for reckless driving, especially in cases that result in injury or property damage.
Doing away with the specific charge of drunk driving sounds radical at first blush, but it would put the focus back on impairment, where it belongs. It might repair some of the civil-liberties damage done by the invasive powers the government says it needs to catch and convict drunk drivers. If the offense were reckless driving rather than drunk driving, for example, repeated swerving over the median line would be enough to justify the charge. There would be no need for a cop to jam a needle in your arm alongside a busy highway.
Scrapping the DWI offense in favor of better enforcement of reckless driving laws would also bring some logical consistency to our laws, which treat a driver with a BAC of 0.08 much more harshly than, say, a driver distracted by his kids or a cell phone call, despite similar levels of impairment. The punishable act should be violating road rules or causing an accident, not the factors that led to those offenses. Singling out alcohol impairment for extra punishment isn't about making the roads safer. It's about a lingering hostility toward demon rum.
Radley Balko is a senior editor at Reason magazine.