click on LOGO to return to HOME page

"Men stumble over the truth from time to time, but most pick themselves up and hurry off as if nothing happened"---Winston Churchill



I got my renewed car registration in the mail the other day. With it was a notice about drunken driving. It included the following statement: "Nearly half of all fatal accidents involve a drunk driver."

This is a wild exaggeration. I thought I knew the source: Mothers Against Drunk Driving. MADD started out years ago as a grass-roots group with a noble goal, but it has turned into yet another bunch of Beltway lobbyists trying to fool people by spinning statistics.

I called the New Jersey Motor Vehicle Commission and asked to speak to the author of the notice. Sure enough, Derrick Stokes told me that he had gotten the information from MADD. He was kind enough to e-mail me the MADD literature he cited: "In 2002, 17,970 people were killed in crashes involving alcohol, representing 42 percent of the 42,850 people killed in all traffic crashes."

The above sentence might lead you to conclude what Stokes concluded: that drunken drivers are responsible for nearly half of highway fatalities. So might the following recent statement by MADD president Wendy Hamilton in a Reuters article just prior to the July 4 weekend: "Last year, 18,000 people were killed in drunk driving crashes."

This is not true, however. The MADD people admitted that when I phoned their Washington office.

"She may have been misquoted or she may have misspoken," said spokesperson Stephanie Manning when I asked her about Hamilton's statement. "The official line is to say 'alcohol-related deaths.'"

The "official line?" Let me make an observation here. By the time any grass-roots group acquires an official line, it is no longer a grass-roots group. I asked Hamilton if she knew exactly how many people were actually killed by drivers who were legally drunk in 2002. She said she didn't, but she invited me to call the National Highway Traffic and Safety Administration.

The NHTSA spokesman said he didn't know either. But he was kind enough to point out that the "alcohol-related" category employed by Mothers Against Drunk Driving includes a great number of accidents in which the participants were either not drunk or not driving. If a car hits you as you walk across the street after having a glass of wine with dinner, for example, then the .01 blood-alcohol content in your corpse will be sufficient for NHTSA to classify your untimely demise as alcohol-related-- even though the driver was totally sober.

As it happens, NHTSA also compiles statistics on true drunken driving. Among those, we find that about 21 percent of drivers in fatal accidents have a blood-alcohol content above 0.08. This is exactly half the percentage touted as "the official line" of MADD, you'll note. But that line was successful in getting Congress to pass legislation to lower the national standard on blood alcohol content for drunken driving to 0.08.

And that in turn caused a whole lot of drivers to be subject to drunken-driving arrests. A lot of these people got together in an online chat group. Out of it came a true grass-roots group called RIDL (Responsibility in DUI Laws). It is on the Web at

Its executive director is a computer scientist named Jeanne Pruett. Pruett is a numbers nerd - and I mean that as the highest possible compliment. She downloaded the federal data base on drunken-driving and began to rip it apart. "They're trying desperately to make the problem worse than it is," Pruett said of MADD. "The reason why they have to make such a big deal is they have to justify their existence."

The actual number of people killed annually in accidents involving drunken drivers is not 18,000 per year but 7,500, said Pruett. And most of those victims are the drunken drivers themselves. For example, almost a third of the drunken drivers involved in fatal crashes are on motorcycles. Many others kill only themselves when they run off the road late at night.

When you get to the statistic that should be central to the debate - the number of innocent people killed by drunken drivers - it is about 2,500 annually, said Pruett. That's still tragic, but to put it into perspective, NHTSA argues that we could save almost as many lives by simply bringing sport utility vehicles and other light trucks down to the same bumper height as cars. Roadway improvements, more seat-belt usage and better vehicle design also offer major gains in safety.

And these gains are predictable and achievable. But Pruett argues that the major gains in reducing drunken-driving deaths may already have been achieved. MADD deserves credit for this, she said, but perhaps it's time for the group to simply declare victory and disband.

"Their original start was fine but they are now becoming a detriment by continuing to overemphasize the connection of drunk driving to the total safety picture," she said. "They know that the real numbers for drunk- driving deaths have stabilized over the past 10 years. And these are about the best numbers you're ever going to get."

And when it comes to numbers, RIDL's add up. This is the nature of grass-roots groups. They are made up of concerned citizens who volunteer their time because they are concerned about a real problem. And since the problem is real, they are comfortable quoting statistics that are equally real.

If such a group stays in business long enough, however, the amateurs turn pro. They hire public relations people who know how to spin statistics so they appear to show something that they don't really show.

That's the level MADD is at. It may not be true, as that Motor Vehicle notice states, that nearly half of all fatal accidents involve a drunk driver. But it's a safe bet that nearly half of all bad statistics involve a pressure group that has gotten too comfortable inside the Beltway.

Paul Mulshine is a Star-Ledger columnist.

The Star Ledger Archive
COPYRIGHT The Star Ledger 2003
Reprinted with permission from Jeanne Pruett, president of RIDL. Visit their website at

"It is dangerous to be right
when the government is wrong.

Hit Counter