Our government says we're winning the war on drugs. At a press conference to release results of the government's major annual drug use survey Sept. 6, both White House drug czar John Walters and Secretary of Health and Human Services Mike Leavitt said so, with Walters touting "fewer teens using drugs today."

Not quite.

When you cut through the spin and look at the actual numbers, it's clear that Walters is again trying to fool the public - much as President Richard Nixon did back in 1972, when he first claimed we were "winning" the war on drugs.

While drug-use rates reported in the just-released 2006 National Survey on Drug Use and Health are essentially unchanged from 2005, Walters and Leavitt touted declines in current teen use of illicit drugs since 2002, from 11.6 to 9.8 percent, and a parallel decline in current marijuana use from 8.2 to 6.7 percent.

That sounds impressive - until you look at the long-term trends. If you go back another 10 years, to 1992, the rate of current teen use of illicit drugs was just 5.3 percent, and current marijuana use was at 3.4 percent. So while it edged down a bit in the last five years, teen drug use is actually nearly double what it was 15 years ago.

Walters and Co. has an explanation for this, of course. They say that the methodology of the survey was changed in 2002, so you can't compare earlier figures with recent ones. But that claim is shaky, at best.

First, not all experts agree that the changes in the survey were enough to drastically alter the results. Second, another government-funded survey of teen drug use that hasn't changed its methodology, called "Monitoring the Future," has documented strikingly similar trends.

In the 2006 "Monitoring the Future" survey, released last December, 16.8 percent of 10th-graders reported using at least one illicit drug - a drop from 20.8 percent in 2002, but a substantial increase over the 11 percent rate in 1992. For marijuana, current use among 10th-graders soared from 8.1 percent in 1992 and 14.2 percent in 2006.

None of this stopped Leavitt from claiming, "The trends in general are very encouraging." Do these people not read their own data?

The fact is that Walters and colleagues have squandered well over a billion of our tax dollars on a failed ad campaign, mostly aimed at demonizing marijuana, and are desperate to show some results. So they cherry pick a few numbers that seem to make their case, and ignore the rest.

And before you buy Walters' frequent claim that "we took our eye off the ball" fighting drug abuse in the '90s, don't forget that between 1991 and 2000, marijuana arrests skyrocketed from 282,000 to 734,497.

But buried in the new national survey on drug use results are some fascinating and sometimes disturbing tidbits. The percentage of Americans who reported using illicit drugs in the past year or past month edged up slightly, and this increase was driven by jumps in use of some of the most dangerous drugs: cocaine, narcotic pain drugs, and stimulants (a category that includes methamphetamine).

While most of the changes were small and not statistically significant, those that were significant are alarming. For example, among 14- to 15-year-olds, past-month use of deadly inhalants (glues, spray paints and solvents) rose significantly, as did past-month use of sedatives. This raises the disturbing possibility that scare campaigns focused on marijuana are driving kids to try drugs that are far more dangerous.

The drug czar will never admit it, but the long-term picture is clear: Our drug policies don't work. The government's bizarre overemphasis on marijuana - a drug that is safer than such legal drugs as alcohol and tobacco - has had little effect on marijuana use, but may well be making our hard-drug problem worse. It's long past time we had policy based on facts, not spin.

 To view the 2006 survey of drug use and health, go to www.oas.samhsa.gov/nsduhLatest.htm. To view the "monitoring the Future" survey, go to www.monitoringthefuture.org/pubs/monographs/overview2006.pdf