How America Lost the War on Drugs

After Thirty-Five Years and $500 Billion, Drugs Are as Cheap and Plentiful as Ever: An Anatomy of a Failure.

Ben Wallace-Wells

1. AFTER PABLO

On the day of his death, December 2nd, 1993, the Colombian billionaire drug kingpin Pablo Escobar was on the run and living in a small, tiled-roof house in a middle-class neighborhood of Medellín, close to the soccer stadium. He died, theatrically, ­ridiculously, gunned down by a Colombian police manhunt squad while he tried to flee across the barrio's rooftops, a fat, bearded man who had kicked off his flip-flops to try to outrun the bullets. The first thing the American drug agents who arrived on the scene wanted to do was to make sure that the corpse was actually Escobar's. The second thing was to check his house.

The last time Escobar had hastily fled one of his residences - la Catedral, the luxurious private prison he built for himself to avoid extradition to the United States - he had left behind bizarre, enchanting ­detritus, the raw stuff of what would ­become his own myth: the photos of ­himself dressed up as a Capone-era gangster with a Tommy gun, the odd collection of novels ranging from Graham Greene to the Austrian modernist Stefan Zweig. Agents from the Drug Enforcement Administration, arriving after the kingpin had fled, found neat shelves lined with loose-leaf binders, carefully organized by content. They were, says John Coleman, then the DEA's assistant administrator for operations, "filled with DEA reports" - internal documents that laid out, in extraordinary detail, the agency's repeated attempts to capture Escobar.

"He had shelves and shelves and shelves of these things," Coleman tells me. "It was stunning. A lot of the informants we had, he'd figured out who they were. All the agents we had chasing him - who we trusted in the Colombian police - it was right there. He knew so much more about what we were doing than we knew about what he was doing."

Coleman and other agents began to work deductively, backward. "We had always wondered why his guys, when we caught them, would always go to trial and risk lots of jail time, even when they would have saved themselves a lot of time if they'd just plead guilty," he says. "What we realized when we saw those binders was that they were doing a job. Their job was to stay on trial and have their lawyers use discovery to get all the information on DEA operations they could. Then they'd send copies back to Medellín, and Escobar would put it all together and figure out who we had tracking him."

The loose-leaf binders crammed in Escobar's office on the ground floor gave Coleman and his agents a sense of triumph: The whole mysterious drug trade had an organization, a structure and a brain, and they'd just removed it. In the thrill of the moment, clinking champagne glasses with officials from the Colombian police and taking congratulatory calls from Washington, the agents in Medellín believed the War on Drugs could finally be won. "We had an endgame," Coleman says. "We were literally making the greatest plans."

At the headquarters of the Office of National Drug Control Policy in Washington, staffers tacked up a poster with photographs of sixteen of its most wanted men, cartel leaders from across the Andes. Solemnly, ceremoniously, a staffer took a red magic marker and drew an X over Escobar's portrait. "We felt like it was one down, fifteen to go," recalls John Carnevale, the longtime budget director of the drug-control ­office. "There was this feeling that if we got all sixteen, it's not like the whole thing would be over, but that was a big part of how we would go about winning the War on Drugs."

Man by man, sixteen red X's eventually went up over the faces of the cartel leaders: KILLED. EXTRADITED. KILLED. José Santacruz Londoño, a leading drug trafficker, was gunned down by Colombian police in a shootout. The Rodríguez Orejuela brothers, the heads of the Cali cartel, were extradited after they got greedy and tried to keep running their organization from prison. Some U.S. drug warriors believed that the busts were largely public-relations events, a showy way for the Colombian government to look tough on the drug trade, but most were less cynical. The crack epidemic was over. Drug-related murders were in decline. Winning the War on Drugs didn't seem such a quixotic and open-ended mission, like the War on Poverty, but rather something tangible, a fat guy with a big organization and binders full of internal DEA reports, sixteen faces on a poster, a piñata you could reach out and smack. Richard Cañas, a veteran DEA official who headed counternarcotics efforts on the National Security Council under both George H.W. Bush and Bill Clinton, can still recall the euphoria of those days. "We were moving," he says, "from success to success."

This is the story of how that momentary success turned into one of the most sustained and costly defeats the United States has ever suffered. It is the story of how the most powerful country on Earth, sensing a piñata, swung to hit it and missed.

2. THE MAKING OF A TRAGEDY

For Cañas and other drug warriors, the death of Escobar had the feel of a real pivot, the end of one kind of battle against drugs and the beginning of another. The war itself had begun during the Nixon administration, when the White House began to get reports that a generation of soldiers was about to come back from Vietnam stoned, with habits weaned on the cheap marijuana and heroin of Southeast Asia and hothoused in the twitchy-fingered freakout of a jungle guerrilla war. For those in Washington, the problem of drugs was still so strange and new in the early Seventies that Nixon officials grappled with ideas that, by the standards of the later debate among politicians, were unthinkably radical: They appointed a panel that recommended the decriminalization of casual marijuana use and even considered buying up the world's entire supply of opium to prevent it from being converted into heroin. But Nixon was a law-and-order politician, an operator who understood very well the panic many Americans felt about the cities, the hippies and crime. Calling narcotics "public enemy number one in the United States," he used the issue to escalate the culture war that pitted Middle Americans against the radicals and the hippies, strengthening penalties for drug dealers and devoting federal funds to bolster prosecutions. In 1973, Nixon gave the job of policing these get-tough laws to the newly formed Drug Enforcement Administration.

By the mid-1980s, as crack leeched out from New York, Miami and Los Angeles into the American interior, the devastations inflicted by the drug were becoming more vivid and frightening. The Reagan White House seemed to capture the current of the moment: Nancy Reagan's plaintive urging to "just say no," and her husband's decision to hand police and prosecutors even greater powers to lock up street dealers, and to devote more resources to stop cocaine's production at the source, in the Andes. In 1986, trying to cope with crack's corrosive effects, Congress adopted mandatory-minimum laws, which hit inner-city crack users with penalties as severe as those levied on Wall Street brokers possessing 100 times more powder cocaine. Over the next two decades, hundreds of thousands of Americans would be locked up for drug offenses.

The War on Drugs became an actual war during the first Bush administration, when the bombastic conservative intellectual Bill Bennett was appointed drug czar. "Two words sum up my entire approach," Bennett declared, "consequences and confrontation." Bush and Bennett doubled annual spending on the drug war to $12 billion, devoting much of the money to expensive weaponry: fighter jets to take on the Colombian trafficking cartels, Navy submarines to chase cocaine-smuggling boats in the Caribbean. If narcotics were the enemy, America would vanquish its foe with torpedoes and F-16s - and throw an entire generation of drug users in jail.

Though many on the left suspected that things had gone seriously awry, drug policy under Reagan and Bush was largely conducted in a fog of ignorance. The kinds of long-term studies that policy-makers needed - those that would show what measures would actually reduce drug use and dampen its consequences - did not yet exist. When it came to research, there was "absolutely nothing" that examined "how each program was or wasn't working," says Peter Reuter, a drug scholar who founded the Drug Policy Research Center at the RAND Corp.

But after Escobar was killed in 1993 - and after U.S. drug agents began systematically busting up the Colombian cartels - doubt was replaced with hard data. Thanks to new research, U.S. policy-makers knew with increasing certainty what would work and what wouldn't. The tragedy of the War on Drugs is that this knowledge hasn't been heeded. We continue to treat marijuana as a major threat to public health, even though we know it isn't. We continue to lock up generations of teenage drug dealers, even though we know imprisonment does little to reduce the amount of drugs sold on the street. And we continue to spend billions to fight drugs abroad, even though we know that military efforts are an ineffective way to cut the supply of narcotics in America or raise the price.

All told, the United States has spent an estimated $500 billion to fight drugs - with very little to show for it. Cocaine is now as cheap as it was when Escobar died and more heavily used. Methamphetamine, barely a presence in 1993, is now used by 1.5 million Americans and may be more addictive than crack. We have nearly 500,000 people behind bars for drug crimes - a twelvefold increase since 1980 - with no discernible effect on the drug traffic. Virtually the only success the government can claim is the decline in the number of Americans who smoke marijuana - and even on that count, it is not clear that federal prevention programs are responsible. In the course of fighting this war, we have allowed our military to become pawns in a civil war in Colombia and our drug agents to be used by the cartels for their own ends. Those we are paying to wage the drug war have been accused of ­human-rights abuses in Peru, Bolivia and Colombia. In Mexico, we are now ­repeating many of the same mistakes we have made in the Andes.

"What we learned was that in drug work, nothing ever stands still," says Coleman, the former DEA official and current president of Drug Watch International, a law-and-order advocacy group. For every move the drug warriors made, the traffickers adapted. "The other guys were learning just as we were learning," Coleman says. "We had this hubris."

3. BRAINIACS AND COLD WARRIORS

"At the beginning of the Clinton administration," Cañas tells me, "the War on Drugs was like the War on Terror is now." It was, he means, an orienting fight, the next in a sequence of abstract, generational struggles that the country launched itself into after finding no one willing to actually square up and face it on a battlefield. After the Cold War, in the flush and optimism of victory, it felt to drug warriors and the American public that abstractions could be beaten. "It was really a pivot point," recalls Rand Beers, who served on the National Security Council for four different presidents. "We started to look carefully at our drug policies and ask if everything we were doing really made sense." The man Clinton appointed to manage this new era was Lee Brown.

Brown had been a cop for almost thirty years when Clinton tapped him to be the nation's drug czar in 1993. He had started out working narcotics in San Jose, California, just as the Sixties began to swell, and ended up leading the New York Police Department when the city was the symbolic center of the crack epidemic, with kids being killed by stray bullets that barreled through locked doors. A big, shy man in his fifties, Brown had made his reputation with a simple insight: Cops can't do much without the trust of people in their communities, who are needed to turn in offenders and serve as witnesses at trial. Being a good cop meant understanding the everyday act of police work not as chasing crooks but as meeting people and making allies.

"When I worked as an undercover narcotics officer, I was living the life of an addict so I could make buys and make busts of the dealers," Brown tells me. "When you're in that position, you see very quickly that you can't arrest your way out of this. You see the cycle over and over again of people using drugs, getting into trouble, going to prison, getting out and getting into drugs again. At some point I stepped back and asked myself, 'What impact is all of this having on the drug problem? There has to be a better way.' "

In the aftermath of the Rodney King beating, this philosophy - known as community policing - had made Brown a national phenomenon. The Clinton administration asked him to take the drug-czar post, and though Brown was skeptical, he agreed on the condition that the White House make it a Cabinet-level position. Brown stacked his small office with liberals who had spent the long Democratic exile doing drug-policy work for Congress and swearing they would improve things when they retook power. "There were basic assumptions that Republicans had been making for fifteen years that had never been challenged," says Carol Bergman, a congressional staffer who became Brown's legislative liaison. "The way Lee Brown looked at it, the drug war was focused on locking kids up for increasing amounts of time, and there wasn't enough emphasis on treatment. He really wanted to take a different tactic."

Brown's staff became intrigued by a new study on drug policy from the RAND Corp., the Strangelove-esque think tank that during the Cold War had employed mathematicians to crank out analyses for the Pentagon. Like Lockheed Martin, the jet manufacturer that had turned to managing welfare reform after the Cold War ended, RAND was scouting for other government projects that might need its brains. It found the drug war. The think tank assigned Susan Everingham, a young expert in mathematical modeling, to help run the group's signature project: dividing up the federal government's annual drug budget of $13 billion into its component parts and deciding what worked and what didn't when it came to fighting cocaine.

Everingham and her team sorted the drug war into two categories. There were supply-side programs, like the radar and ships in the Caribbean and the efforts to arrest traffickers in Colombia and Mexico, which were designed to make it more expensive for traffickers to bring their product to market. There were also demand-side programs, like drug treatment, which were designed to reduce the market for drugs in the United States. To evaluate the cost-effectiveness of each approach, the mathematicians set up a series of formulas to calculate precisely how much additional money would have to be spent on supply programs and demand programs to reduce cocaine consumption by one percent nationwide.

"If you had asked me at the outset," Everingham says, "my guess would have been that the best use of taxpayer money was in the source countries in South America" - that it would be possible to stop cocaine before it reached the U.S. But what the study found surprised her. Overseas military efforts were the least effective way to decrease drug use, and imprisoning addicts was prohibitively expensive. The only cost-effective way to put a dent in the market, it turned out, was drug treatment. "It's not a magic bullet," says Reuter, the RAND scholar who helped supervise the study, "but it works." The study ultimately ushered RAND, this vaguely creepy Cold War relic, into a position as the permanent, pragmatic left wing of American drug policy, the most consistent force for innovating and reinventing our national conception of the War on Drugs.

When Everingham's team looked more closely at drug treatment, they found that thirteen percent of hardcore cocaine users who receive help substantially reduced their use or kicked the habit completely. They also found that a larger and larger portion of illegal drugs in the U.S. were being used by a comparatively small group of hardcore addicts. There was, the study concluded, a fundamental imbalance: The crack epidemic was basically a domestic problem, but we had been fighting it more aggressively overseas. "What we began to realize," says Jonathan Caulkins, a professor at Carnegie Mellon University who studied drug policy for RAND, "was that even if you only get a percentage of this small group of heavy drug users to abstain forever, it's still a really great deal."

Thirteen years later, the study remains the gold standard on drug policy. "It's still the consensus recommendation supplied by the scholarship," says Reuter. "Yet as well as it's stood up, it's never really been tried."

To Brown, RAND's conclusions seemed exactly right. "I saw how little we were doing to help addicts, and I thought, 'This is crazy,' " he recalls. " 'This is how we should be breaking the cycle of addiction and crime, and we're just doing nothing.' "

The federal budget that Brown's office submitted in 1994 remains a kind of fetish object for certain liberals in the field, the moment when their own ideas came close to making it into law. The budget sought to cut overseas interdiction, beef up community policing, funnel low-level drug criminals into treatment programs instead of prison, and devote $355 million to treating hardcore addicts, the drug users responsible for much of the illegal-drug market and most of the crime associated with it. White House political handlers, wary of appearing soft on crime, were skeptical of even this limited commitment, but Brown persuaded the president to offer his support, and the plan stayed.

Still, the politics of the issue were difficult. Convincing Congress to dramatically alter the direction of America's drug war required a brilliant sales job. "And Lee Brown," says Bergman, his former legislative liaison, "was not an effective salesman." With a kind of loving earnestness, the drug czar arranged tours of treatment centers for congressmen to show them the kinds of programs whose funding his bill would increase. Few legislators came. Most politicians were skeptical about such a radical departure from the mainstream consensus on crime. Congress rewrote the budget, slashing the $355 million for treatment programs by more than eighty percent. "There were too many of us who had a strong law-and-order focus," says Sen. Chuck Grassley, a Republican who ­opposed the reform bill and serves as co-chair of the Senate's drug-policy caucus.

For some veteran drug warriors, Brown's tenure as drug czar still lingers as the last moment when federal drug policy really made sense. "Lee Brown came the closest of anyone to really getting it," says Carnevale, the longtime budget director of the drug-control office. "But the bottom line was, the drug issue and Lee Brown were largely ignored by the Clinton administration." When Brown tried to repeat his treatment-centered initiative in 1995, it was poorly timed: Newt Gingrich and the Republicans had seized control of the House after portraying Clinton as soft on crime. The authority to oversee the War on Drugs passed from Rep. John Conyers, the Detroit liberal, to a retired wrestling coach from Illinois who was tired of drugs in the schools ? a rising Republican star named Dennis Hastert. Reeling from the defeat at the polls, Clinton decided to give up on drug reform and get tough on crime. "The feeling was that the drug czar's office was one of the weak areas when it came to the administration's efforts to confront crime," recalls Leon Panetta, then Clinton's chief of staff.

4. THE YOUNG GUNS

The administration was not doing much better in its efforts to stop the flow of drugs at the source. Before Clinton had even taken office, Cañas - who headed drug policy at the National Security Council - had been summoned to brief the new president's choice for national security adviser, Anthony Lake, on the nation's narcotics policy in Latin America. "I figured, what the hell, I'm going back to DEA anyway, I'll tell him what I really think," Cañas recalls.

The Bush administration, he told Lake, had been sending the military after the wrong target. In the 1970s, drugs were run up to the United States through the Caribbean by a bunch of "swashbuckling entrepreneurs" with small planes - "guys who wouldn't have looked out of place at a Jimmy Buffett concert." In 1989, in the nationwide panic over crack, Defense Secretary Dick Cheney had managed to secure a budget of $450 million to chase these Caribbean smugglers. (Years later, when a longtime drug official asked Defense Secretary Donald Rumsfeld why Cheney had pushed the program, Rumsfeld grinned and said, "Cheney thought he was running for president.") The U.S. military loved the new mission, because it gave them a reason to ask for more equipment in the wake of the Cold War. And the Bush White House loved the idea of sending the military after the drug traffickers for its symbolism and swagger and the way it proved that the administration was taking drugs seriously.

The problem, Cañas told Lake, was that the cocaine traffic had professionalized and was now moving its product through Mexico. With Caribbean smugglers out of the game, the military program no longer made sense. The new national security adviser grinned at Cañas, pleased. "That's what we think as well," Lake said. "How would you like to stay on and help make that happen?"

Taking a new approach, the Clinton administration shifted most military assets out of the Caribbean and into the Andes, where the coca leaf was being grown and processed. "Our idea was, Stop messing around in the transit countries and go to the source," Cañas tells me. The administration spent millions of extra dollars to equip police in Bolivia and Colombia to bust the crop's growers and processors. The cops were not polite - Human Rights Watch condemned the murders of?Bolivian farmers, blaming "the heavy hand of U.S. drug enforcement" - but they were ­effective, and by 1996, coca production in Bolivia had begun a dramatic decline.

After Escobar fell, the American drug agents who had been chasing him did not expect the cocaine industry to dry up overnight - they had girded for the fallout from the drug lord's death. What they had not expected was the ways in which the unintended consequences of his downfall would permanently change the drug traffic. "What ended up happening - and maybe we should have predicted this would happen - was that the whole structure shattered into these smaller groups," says Coleman, the veteran DEA agent. "You suddenly had all these new guys controlling a small aspect of the traffic."

Among them was a hired gun known as Don Berna, who had served as a bodyguard for Escobar. Double-crossed by his boss, Berna broke with the Medellín cartel and struck out on his own. For him, the disruption caused by the new front in America's drug war presented a business opportunity. But with the DEA's shift from the Caribbean into Bolivia and Colombia, Berna and other new traffickers had a production problem. So some of the "microcartels," as they became known, decided to move their operations someplace where they could control it: They opened negotiations with the FARC, a down-at-the-heels rebel army based in the jungles of Colombia. In return for cash, the FARC agreed to put coca production under its protection and keep the Colombian army away from the coca crop.

Berna and the younger kingpins also had a transportation problem: Mexican traffickers, who had been paid a set fee by the cartels to smuggle product across the U.S. border, wanted a larger piece of the business. The Mexican upstarts had a certain economic logic on their side. A kilo of cocaine produced in Colombia is worth about $2,500. In Mexico, a kilo gets $5,000. But smuggle that kilo across the border and the price goes up to $17,500. "What the Mexican groups started saying was, 'Why are we working for these guys? Why don't we just buy it from the Colombians directly and keep the profits ourselves?' " says Tony Ayala, a retired DEA agent and former Mexico country attache.

The remaining leaders of the weakened Cali cartel, DEA agents say, traveled up to Guadalajara for a series of meetings with Mexican traffickers. By 1996, the Colombians had decided to hand over more control of the cocaine trade to the Mexicans. The Cali cartel would now ship cocaine to Guadalajara, sell the drugs to the Mexican groups and then be done with it. "This wasn't just happenstance," says Jerome McArdle, then a DEA assistant agent for special operations. "This was the Colombians saying they were willing to reduce their profits in exchange for reducing their risk and exposure, and handing it over to the Mexicans. The whole nature of the supply chain changed."

Around the same time, DEA agents found themselves picking up Mexican distributors, rather than Colombians, on the streets of New York. Immigration and customs officials on the border were meanwhile overwhelmed by the sheer number of tractor-trailers - many of them loaded with drugs - suddenly pouring across the Mexican border as a consequence of NAFTA, which had been enacted in 1994. "A thousand trucks coming across in a four-hour ­period," says Steve Robertson, a DEA special agent assigned to southern ­Texas at the time. "There's no way we're going to catch everything."

Power followed the money, and Mexican traffickers soon had a style, and reach, that had previously belonged only to the Colombians. In the border town of Ciudad Juárez, the cocaine trafficker Amado Carrillo Fuentes developed a new kind of smuggling operation. "He brought in middle-class people for the first time - lawyers, accountants - and he developed a transportation division, an acquisitions division, even a human-resources operation, just like a modern corporation," says Tony Payan, a political scientist at the University of Texas-El Paso who has studied the drug trade on the border. Before long, Carrillo Fuentes had a fleet of Boeing 727s, which he used to fly cocaine, up to fifteen tons at a time, up from Colombia to Mexico. The newspapers called him El Señor de los Cielos, the Lord of the Skies.

The Mexican cartels were also getting more imaginative. "Think of it like a business, which is how these guys thought of it," says Guy Hargreaves, a top DEA agent during the 1990s. "Why pay for the widgets when you can make the widgets yourselves?" Since the climate and geography of Mexico aren't right for making cocaine, the cartels did the logical thing: They introduced a new product. As Hargreaves recalls, the Mexicans slipped the new drug into their cocaine shipments in Southern California and told coke dealers, "Here, try some of this stuff - it's a similar effect."

The product the Mexican cartels came up with, the new widget they could make themselves, was methamphetamine. The man who mastered the market was a midlevel cocaine trafficker, then in his late twenties, named Jesús Amezcua. In 1994, when U.S. Customs officials at the Dallas airport seized an airplane filled with barrels of ephedrine, a chemical precursor for meth, and traced it back to Amezcua, the startling new shift in the drug traffic became clear to a handful of insiders. "Cartels were no longer production organizations, whose business is wrapped up in a single drug," says Tony Ayala, the senior DEA agent in Mexico at the time. "They became trafficking organizations - and they will smuggle whatever they can make the most profit from."

5. THE LOBBYISTS & THE MAD PROFESSOR

It is only in retrospect that these moments - the barrels of ephedrine seized in Dallas, the quiet suggestion that meth had worked its way into the cocaine supply chain - take on a looming character, the historic weight of a change made manifest. Up until methamphetamine, the War on Drugs had targeted three enemies. First there were the hippie drugs - marijuana, LSD - that posed little threat to the general public. Then there was heroin, a horrible drug but one that was largely concentrated in New York City. And, finally, there was crack. What meth proved was that even if the DEA could wipe out every last millionaire cocaine goon in Colombia, burn every coca field in Bolivia and Peru, and build an impenetrable wall along the entire length of the Mexican border - even then, we wouldn't have won the War on Drugs, because there would still be methamphetamine, and after that, something else.

Gene Haislip, who served for years as one of the DEA's top-ranking administrators, believes there was a moment when meth could have been shut down, long before it spiraled into a nationwide epidemic. Haislip, who spent nearly two decades leading a small group at the agency dedicated to chemical control, is his own kind of legend; he is still known around the DEA as the man who beat quaaludes, perhaps the only drug that the U.S. has ever been able to declare total victory over. He did it with gumshoe methodicalness: by identifying every country in the world that produced the drug's active ingredient, a prescription medication called methaqualone, and convincing them to tighten regulations. Haislip believes he was present the moment when the United States lost the war on methamphetamine, way back in 1986, when meth was still a crude biker drug confined to a few valleys in Northern California - a decade before the Mexican drug lords turned it into the most problematic drug in America. "The thing is, methamphetamine should never have gotten to that point," Haislip says. And it never would have, he believes, if it hadn't been for the lobbyists.

Haislip was known around the DEA as precise-minded and verbal. His impulse, in combatting meth, was the same one that had pushed the drug warriors after Escobar: the quixotic faith that if you could just stop the stuff at the source, you could get rid of all the social problems at once. Assembling a coalition of legislators, Haislip convinced them that the small, growing population of speed freaks in Northern California was enough of a concern that Congress should pass a law to regulate the drug's precursor chemicals, ephedrine and pseudoephedrine, legal drugs that were used in cold medicine and produced in fewer than a dozen factories in the world. "We were starting to get reports of hijacking of ephedrine, armed robbery of ephedrine, things that had never happened before," Haislip tells me. "You could see we were on the verge of something if we didn't get a handle on it."

All that was left was to convince the Reagan administration. One day in late 1986, Haislip went to meet with top officials in the Indian Treaty Room, a vast, imposing space in the Eisenhower Executive Office Building: arches, tiled floors, the kind of room designed to house history being made. Haislip noticed several men in suits sitting quietly in the back of the room. They were lobbyists from the pharmaceutical industry, but Haislip didn't pay them much attention. "I wasn't concerned with them," he recalls.

When Haislip launched into his presentation, an official from the Commerce Department cut him off. "Look, you're way ahead of us," the official said. "We don't have anything to suggest or add." Haislip left the meeting thinking he had won: The bill he proposed was submitted to Congress, requiring companies to keep records on the import and sale of ephedrine and pseudoephedrine.

But what Haislip didn't know was that the men in suits had already gone to work to rig the bill in their favor. "Quite frankly," Allan Rexinger, one of the lobbyists present at the meeting later told reporters, "we appealed to a higher authority." The pharmaceutical industry needed pseudoephedrine to make profitable cold medications. The result, to Haislip's dismay, was a new law that monitored sales of ephedrine and pseudoephedrine in bulk powder but created an exemption for selling the chemicals in tablet form - a loophole that protected the pharmaceutical industry's profits.

The law, drug agents say, sparked two changes in the market for illegal meth. First, the supply of ephedrine simply moved overseas: The Mexican cartels, quick to recognize an emerging market, evaded the restrictions by importing powder from China, India and Europe and then smuggling it across the border to the biker groups that had traditionally distributed the drug. "We actually had meetings where we planned for a turf war between the Mexicans and the Hells Angels over methamphetamine," says retired DEA agent Mike Heald, who headed the San Francisco meth task force, "but it turned out they realized they'd make more money by working together." Second, responding to a dramatic uptick in demand from the illegal market, chemical-supply companies began moving huge amounts of ephedrine and pseudoephedrine out to the West Coast in the form of pills, which were then converted into meth. Rather than stemming the tide of meth before it started, the Reagan administration had unwittingly helped accelerate a new epidemic: Between 1992 and 1994, the number of meth addicts entering rehab facilities doubled, and the drug's purity on the street rose by twenty-seven percent.

Haislip resolved to have another go at Congress, but the issue ended up in a dispiriting cycle. The resistance, he says bitterly, "was always coming from the same lobbying group." In 1993, when he persuaded lawmakers to regulate the sale of ephedrine in tablet form, the pharmaceutical industry won an exception for pseudoephedrine. Drug agents began to intercept shipments of pseudoephedrine pills in barrels. Three years later, when lawmakers finally regulated tablets of pseudoephedrine, they created an exception for pills sold in blister packs. "Congress thought there was no way that meth freaks would buy this stuff and pop the pills out of blister packs, one by one," says Heald. "But we're not dealing with normal people - we're dealing with meth freaks. They'll stay up all night picking their toes."

By the time Haislip retired, in 1997, the methamphetamine problem was really two problems. There were the mom-and-pop cooks, who were punching pills out of blister packs and making small batches of drugs for themselves. Then there were the industrial-scale Mexican cartels, which were responsible for eighty percent of the meth in the United States. It took until 2005 for Congress to finally regulate over-the-counter blister packs, which caused the number of labs to plummet. But once again, the Mexican groups were a step ahead of the law. In October 2006, police in Guadalajara arrested an American chemist named Frederick Wells, who had moved to Mexico after losing his job at Idaho State University. An academic troublemaker who drove around campus with signs on the back of his pickup truck raging at the college administration, Wells had allegedly used his university lab to investigate new ways that Mexican traffickers could use completely legal reagents to engineer meth precursors from scratch. "Very complicated numerical modeling," says his academic colleague Jeff Rosentreter. By the time Wells was arrested, the State Department had only just succeeded at pressuring Mexico to restrict the flow of pseudoephedrine, even though Wells had apparently been hard at work for years creating alternatives to that chemical. The lobbying by the pharmaceutical industry, Haislip says, "cost us eight or nine years."

For some in the drug war, it was a lesson that even the most promising efforts to restrict the supply of drugs at the source - those that rely on legal methods to regulate legally produced drugs - remained nearly impossible, outflanked by both drug traffickers and industry lobbyists. The tragedy of the fight against methamphetamine is that it repeated the ways in which the government tried to fight the cocaine problem, and failed - racing from source to source, trying to eliminate a coca field or an ephedrine manufacturer and then racing to the next one. "We used to call it the Pillsbury Doughboy - stick your finger in one part of the problem, and the Doughboy's stomach just pops out somewhere else," says Rand Beers. "The lesson of U.S. drug policy is that this world runs on unintended consequences. No matter how noble your intentions, there's a good chance that in solving one problem, you'll screw something else up."

6. THE GENERAL & THE ADMAN

Within the Clinton White House, the reform effort spearheaded by Lee Brown had created a political dilemma. Republicans, having taken control of Congress in 1994, were attacking the administration for being soft on drugs, and the White House decided that it was time to look tougher. "A lot of people didn't think Brown was a strong leader," Panetta tells me. As senior figures within the administration cast about for a replacement, they started by thinking about who would be the opposite of Brown. "We wanted to get someone who was much stronger, much tougher, and could come across that way symbolically," Panetta says.

During the planning for a possible invasion of Haiti, Panetta and others had discovered a rising star at the Pentagon, a charismatic, bullying four-star general named Barry McCaffrey, who had annoyed many in the Pentagon's establishment. In 1996, halfway into his State of the Union address, Clinton looked up at McCaffrey, a lean, stern-seeming military man in the balcony, and informed the nation that the general would be his next drug czar. "To succeed, he needs a force far larger than he has ever commanded before," Clinton said. "He needs all of us. Every one of us has a role to play on this team." McCaffrey, the bars on his epaulets shimmering, saluted. It was one of the president's biggest applause lines of the night.

For the drug warriors in McCaffrey's office, "the General" was everything the languid, considered, academic Lee Brown had not been. "It was clear from the outset that here was a guy who would take advantage of the bully pulpit and who, unlike Brown, would probably be able to get things done," says Bergman, Brown's former liaison. "One thing that surprised us all was how thoughtful he was - he wasn't a knee-jerk, law-enforcement guy. He understood there needed to be money for treatment. He prided himself on being very sensitive to the racial issues, and he was sensitive to the impact of sentencing laws on African-American men." McCaffrey imported his own staff from the Southern Command - mostly men, all military. They lent the White House's drug operation - previously a slow place - the kinetic energy of a forward operating base. "We went to a twenty-four-hour clock, so we'd schedule meetings for 1500," one longtime staffer recalls. "His people sat down with senior staff and told us what size paper the General wanted his memos on, this kind of report would have green tabs, this would have blue tabs."

The General's genius was for publicity. "He was great at getting visibility," Carnevale says. McCaffrey held grandstanding events everywhere from Mexico to Maine, telling reporters that the decades-long narrative of impending doom around the drug war was out of date - and that if Congress would really dedicate itself to the mission, the country had a winnable fight on its hands. Drug-use numbers were edging downward; even cocaine seemed to be declining in popularity. "We are in an optimistic situation," McCaffrey declared.

For the first time ever, McCaffrey had the drug czar's office develop a strategy for an endgame to the drug war, a plan for finishing the whole thing. The federal government needed to reduce the amount of money it was spending on law enforcement and interdiction. But McCaffrey believed this was only possible once it could guarantee that drug use would continue to decline. "The data suggested very strongly that those who never tried any drugs before they were eighteen were very likely to remain abstinent for their whole lives, but that those who even smoked marijuana when they were teenagers had much worse outcomes," says McCaffrey's deputy Don Vereen. So the General decided to focus the government's attention on keeping kids from trying pot.

The "gateway theory," as it became known, had a natural appeal. Because most people who used hard drugs had also smoked marijuana, and because kids often tried marijuana several years before they started trying harder drugs, it seemed that keeping them off pot might prevent them from ever getting to cocaine and heroin. The only trouble is, the theory is wrong. When McCaffrey's office commissioned the Institute of Medicine to study the idea, researchers concluded that marijuana "does not appear to be a gateway drug." RAND, after examining a decade of data, also found that the gateway theory is "not the best explanation" of the link between marijuana use and hard drugs. But McCaffrey continued to devote more and more of the government's resources to going after kids. "We have already clearly committed ourselves," he declared, "to a number-one focus on youth."

"That decision," Bergman says, "was where you could see McCaffrey begin to lose credibility."

In 1996, less than a year into his term, the new drug czar met Jim Burke, a smooth-talking, silver-haired executive who chaired the Partnership for a Drug-Free America - the advertising organization best known for the slogan "This is your brain on drugs." "Burke personally was very hard to resist," one of his former colleagues tells me. "I've seen him sell many conservative members of Congress and also liberals like Mario Cuomo."

Burke told McCaffrey a simple story. In the late 1980s, he said, the major television networks had voluntarily given airtime to the Partnership to run anti-drug ads aimed at teenagers. The number of teenagers who used drugs - especially marijuana - declined during that period. But in the early 1990s, Burke said, the rise of cable TV cut into the profits of the networks, which became stingier with the time they dedicated to anti-drug advertising. The result, the adman told the General, was that the number of teenagers who used drugs was climbing sharply - to the outrage of Dennis Hastert and other conservative members of Congress. As a clincher, Burke handed McCaffrey a graph that showed the declining amount of airtime dedicated to anti-drug advertising on one axis and the declining perception among teenagers of the risks associated with drugs on the other. "I'm ninety-nine percent sure," one staffer at the Partnership tells me, "that it was that conversation that sold McCaffrey."

The General mobilized his office, lobbying Congress to allocate enough money to put anti-drug advertising on the air whenever teenagers watched television. His staff was skeptical. For all of McCaffrey's conviction and charisma, he didn't have much in the way of facts. "That was all we had - no data, just this one chart - and we had to go and sell Congress," Carnevale recalls. But Congress proved to be a pushover. Conservatives, who held a majority, were thrilled that soft-on-pot liberals in the Clinton administration finally wanted to do something about the drug problem. "At some point, you have to draw a line and say that some things are right and some things are wrong," says Sen. Grassley, explaining his support of the measure. "And using any drugs is just flat-out wrong." To the Partnership's delight, Congress allocated $1 billion to buy network time for anti-drug spots aimed at teenagers.

The General was also starting to make friends beyond the Clinton administration. The drug czar had found a natural ally in Hastert, who had become the GOP's de facto leader on drug policy. The former wrestling coach struck few as charismatic - his joyless and drudging style, his form like settled gelatin - but his experiences in high schools had left him with the feeling that the drug issue, in the words of his longtime aide Bobby Charles, "had become extremely poignant." Hastert wasn't quite Lee Brown; he believed that the prime focus of the drug war should be to increase funding for military operations in Colombia. But he and his staff had grown frustrated with the exclusively punitive character of drug policy and wanted the Republicans to take a more compassionate stance. His staff had studied the RAND reports and largely agreed with their conclusions. "We felt if you didn't get at the nub of the problem, which was prevention and treatment, you weren't going to do any good," says John Bridgeland, a congressional aide who helped coordinate Republican drug policy. Hastert eventually won $450 million to be used, in part, to expand a faith-based program discovered by Bridgeland: Developed by a former evangelical minister, it brought together preachers, parents and drug counselors to fight the problem of "apathy" through "parent training" and "messages from the pulpit."

But with McCaffrey's emphasis on kids came another, almost fanatical focus: going after citizens who used pot for medical purposes. If he was fighting marijuana, the General was going to fight it everywhere, in all its forms. He threatened to have doctors who prescribed pot brought up on federal charges, and dismissed the science behind medical marijuana as a "Cheech and Chong show." In 1997, voters in Oregon introduced an initiative to legalize medical marijuana in the state. "I'll never forget the senior-staff meeting the morning after the Oregon initiative was announced," Bergman says. "McCaffrey was furious. It was like this personal affront to him. He couldn't believe they'd gotten away with it. He wanted to have this research done on the groups behind it and completely trash them in the press." As the General traveled to the initiative states, stumping against medical marijuana, his aides sneered that the initiatives were "all being mostly bankrolled by one man, George Soros," the billionaire investor who favored decriminalizing drugs.

Even for those who shared McCaffrey's philosophy, the theatrics seemed strange: There he was, on evening newscasts, effectively insisting that grandmothers dying of cancer were corrupting America's youth. His office pushed arguments that, at best, stretched the available research: Marijuana is a gateway drug that leads inexorably to the abuse of harder drugs; marijuana is thirty times more potent now than it was a generation ago. "It didn't track with the conclusions our researchers came to," says Bergman. "It felt like he was trying to manipulate the data."

McCaffrey had taken the drug war in a new direction, one that had little obvious connection with preventing drug abuse. For the first time, the full force of the federal government was being brought to bear on patients dying from terminal diseases. Even the General's allies in Congress were appalled. "I can't tell you how many times I went to the Hill with him and sat in on closed-doors meetings," Bergman recalls. "Members said to him, 'What in the world are you doing? We have real drug problems in the country with meth and cocaine. What the hell are you doing with medical marijuana? We get no calls from our constituents about that. Nobody cares about that.' McCaffrey was just mystified by their response, because he truly believed marijuana was a gateway drug. He truly believed in what he was doing."

7. THE HARVARD MAN

For the cops on the front lines of the War on Drugs, the federal government's fixation with marijuana was deeply perplexing. As they saw it, the problem wasn't pot but the drug-related violence that accompanied cocaine and other hard drugs. After the crack epidemic in the late 1980s, police commissioners around the country, like Lee Brown in Houston, began adding more officers and developing computer mapping to target neighborhoods where crime was on the rise. The crime rate dropped. But by the mid-1990s, police in some cities were beginning to realize there was a certain level that they couldn't get crime below. Mass jailings weren't doing the trick: Only fifteen percent of those convicted of federal drug crimes were actual traffickers; the rest were nothing but street-level dealers and mules, who could always be replaced.

Police in Boston, concerned about violence between youth drug gangs, turned for assistance to a group of academics. Among them was a Harvard criminologist named David Kennedy. Working together, the academics and members of the department's anti-gang unit came up with what Kennedy calls a "quirky" strategy and convinced senior police commanders to give it a try. The result, which began in 1995, was the Boston Gun Project, a collaborative effort among ministers and community leaders and the police to try to break the link between the drug trade and violent crime. First, the project tracked a particular drug-dealing gang, mapping out its membership and operations in detail. Then, in an effort called Operation Ceasefire, the dealers were called into a meeting with preachers and parents and social-service providers, and offered a deal: Stop the violence, or the police will crack down with a vengeance. "We know the seventeen guys you run with," the gangbangers were told. "If anyone in your group shoots somebody, we'll arrest every last one of you." The project also extended drug treatment and other assistance to anyone who wanted it.

The effort worked: The rates of homicide and violence among young men in Boston dropped by two-thirds. Drug dealing didn't stop - "people continued what they were doing," Kennedy concedes, "but they put their guns down."

As Kennedy reflected on the success of the Boston project, which ran for five years, he wondered if he had discovered a deeper truth about drug-related violence. If the murders weren't a necessary component of the drug trade - if it was possible to separate the two - perhaps cities could find a way to reduce the violence, even if they could do nothing about the drugs.

In 2001, Kennedy got a call from the mayor of San Francisco that gave him a chance to examine his theories in a new setting. The city had experienced a recent spike in its murder rate, much of it caused by an ongoing feud between two drug-dealing gangs - Big Block and West Mob - that had resulted in dozens of murders over the years. Could Kennedy, the mayor asked, help police figure out how to stop the killings?

Kennedy flew out to San Francisco and met with police. But as he researched the history of the violence, it seemed to confirm his findings in Boston. Though both Big Block and West Mob were involved in dealing drugs, the shootings were not really drug-related - the two groups occupied different territories and were not battling over turf. "The feud had started over who would perform next at a neighborhood rap event," says Kennedy, now a professor at John Jay College of Criminal Justice. "They had been killing each other ever since."

Such evidence suggested that drug enforcement needed to focus more narrowly on those responsible for the violence. "Seventy percent of the violence in these hot neighborhoods comes back to drugs," Kennedy says. "But one of the profound myths is that these homicides are about the drug trade. The violence is driven by these crews - but they're not killing each other over business." The real spark igniting the murders, he realized, was peer pressure, a kind of primordial male goad that drove young gang members to kill each other even in instances when they weren't sure they wanted to.

Given that police departments had already locked up every drug dealer in sight and were still having problems with violence, Kennedy thought a new approach was worth a try. "There's a difference between saying, 'I'm watching this, and you should stop,' and putting someone in federal lockup," he says. "The violence is not about the drug business - but that's a very hard thing for people to understand."

But in the early days of the Bush administration, police departments were in no hurry to experiment with an approach that focused on drug-related murders and mostly ignored users who weren't committing violence. Kennedy's efforts proved to be yet another missed opportunity in the War on Drugs - an experience that made clear how difficult it is for science to influence the nation's drug policy.

"If ten years ago the medical community had figured out a way to reduce the deaths from breast cancer by two-thirds, every cancer clinic in the country would have been using those techniques a year later," Kennedy says. "But when it comes to drugs and violence, there's been nothing like that."

8. HELICOPTERS AND COCA

Instead of pursuing the Boston Gun Project and other innovative approaches to fighting drug violence, the federal government decided to escalate its military response in Colombia. For the past decade and a half, cooperation from officials in Bogotá had been halfhearted, sporadic and deeply corrupt. But by 1999, the country, it seemed, was on the verge of collapsing into civil war. The drug money that had flowed into Colombia had found its way into the hands of the rebel militia - the FARC - which had been laying siege to the Colombian government. The Clinton foreign-policy team, having spent the previous few years dealing with the consequences of failed states in Somalia and the Balkans, was deeply concerned about the possibility of a failed narco­state in America's own back yard.

One afternoon in June 1999, a dozen senior Clinton officials filed into the National Security Council's situation room, summoned by Sandy Berger, the president's national security adviser. Even though Bogotá had ceded control of vast swaths of the country to the left-wing rebels, they were told, recent peace talks had collapsed. "The FARC had basically always been jungle campesinos - they were a pretty austere bunch," says Brian Sheridan, who was in charge of the Pentagon's counternarcotics effort at the time and attended the meeting. "All of a sudden, they were leveling these attacks that had gotten more and more audacious." When FARC rebels had emerged from the jungle for a round of peace talks the previous fall, they had brandished brand-new AK-47s and Dragunovs, as if on military parade. One U.S. official observed at the time that the weaponry was "far beyond" what the Colombian army had - in a pitched battle, the Clinton administration worried, the ­Colombian government could plausibly collapse.

The White House advisers weren't the only officials in Washington concerned about Colombia. Earlier that day, two men who attended the briefing - Rand Beers of the State Department and Charlie Wilhelm of the Defense Department - had gotten a call from the Republican caucus on the Hill. Dennis Hastert, who had been elevated to Speaker of the House six months earlier, wanted to see them right away. "It was kind of unusual," Beers recalls - but when Hastert called, you came.

When Beers and Wilhelm arrived, Rep. Porter Goss, then the chairman of the House Intelligence Committee, handed them a piece of paper. It was a copy of a supplemental spending authorization that the Republicans planned to offer immediately. Crafted by Bobby Charles, Hastert's longtime aide, the bill would have more than doubled military aid to Colombia to take on the rebels and narcotraffickers -to a staggering $1.2 billion a year. But it was the politics of the situation that worried Beers as much as the money. "It occurred to me that if the administration was going to do anything on Colombia, it better do it soon," he says now, "or the Republicans would once again outflank what they perceived as the I-never-inhaled Clinton administration." Beers told the Republicans he would take a look, and then hurried to Berger's meeting.

Throughout much of the Clinton administration, the hope had been that the United States would be able to reduce its military aid to the Andes as the cocaine epidemic waned. Now, as Berger's group heard from intelligence agents, that hope seemed to be fading. Narcotraffickers were paying off the FARC so they could grow coca in the jungles of Colombia. The FARC were then turning around and using the money to buy weapons to stage attacks on the Colombian government.

Berger decided to act. Rather than oppose the Republican plan, he agreed to negotiate on an assistance package to bail out the Colombian government. The result was Plan Colombia - nearly $1.6 billion to escalate the War on Drugs in the Andes. The new program would arm the military and police in their fight against the FARC, launch an ambitious effort to spray herbicide on coca crops from the air and provide economic assistance to poor farmers in rural villages. The initial aid, officials decided, would be heavily concentrated in Putumayo, a rebel-run province in the jungle.

No one is sure what convinced President Clinton to approve such an ambitious escalation in the War on Drugs. But some observers at the time speculated that the critical factor was a conversation with Sen. Christopher Dodd, the Connecticut Democrat, whose state is home to the helicopter manufacturer Sikorsky Aircraft. In early 2000, Clinton unveiled Plan Colombia - and Sikorksy promptly received an order for eighteen of its Blackhawk helicopters at a cost of $15 million each. "Much has been made of the notion that this was Dodd looking to sell Blackhawks to Colombia," Beers tells me. He pauses before adding, "I am not in a position to tell you it didn't happen."

Plan Colombia would be the Clinton administration's primary and most ­costly contribution to the War on Drugs, the major counternarcotics program it bequeathed to the Bush administration. But as with so many other aspects of American drug policy, the plan had an unintended consequence: As it evolved, the emphasis on supplying arms to the Colombian government ended up having less to do with drugs and more to do with helping Bogotá fight its enemies. Colombia used the military aid to target the left-wing FARC - even though many believed that right-wing paramilitaries, who were allies of the government, were more directly involved in narcotrafficking. "It wasn't really first and foremost a counternarcotics program at all," says a senior Pentagon official involved in the creation of Plan Colombia. "It was mostly a political stabilization program."

9. THE TEMPLE OF HOPE

In July of 1999, Gov. George W. Bush of Texas traveled to Cincinnati to visit Hope Temple, a former crack house that had been turned into a church. It was an almost unbearably hot day. Bush was on a tour through the Midwest during which he was testing out his philosophy of compassionate conservatism, trying to see if its rhetoric and principles could sustain a winning presidential run. "The American dream is vivid," Bush told audiences, "but too many feel, 'This dream is not meant for me.' " John Bridgeland, the congres­sional aide who had helped steer federal funding to Hope Temple, says Bush was "overwhelmed" by his visit to the church that day, and stayed the whole afternoon. That evening, Bush spoke about the fervent ­religiosity of the place and the rough joys of the addict's redemptions. "These," he said, "are the armies of compassion."

This was a strange moment in the politics of the drug war: Just as the Clinton administration was toughening its rhetoric, influential Republicans were going all soft and gentle. John DiIulio, a political scientist from the University of Pennsylvania who would become a key Bush adviser, was disgusted by the "perverse consequences" of harsh sentencing laws that had put millions of young Americans in prison, disbelieved the "sweeping scientific claims" made about the dangers of medical marijuana and wanted to expand "meaningful drug-treatment opportunities in urban areas." DiIulio and his contemporaries were troubled, too, by the racial imbalances of the War on Drugs: Blacks, who comprised only fourteen percent of drug users, made up seventy-four percent of those in prison for drug possession. It was not as if the Republican Party had suddenly taken up a position on the far left of the drug war. But it did seem, for a moment during the 2000 campaign, as if some moderation were possible.

Three months later, when the Bush campaign released its drug policy, even the most experienced drug warriors were impressed. The platform balanced spending between demand- and ­supply-side programs, stressed treatment and doubled the number of community anti-drug coalitions. When Bush won the White House and DiIulio became the director of the Office of Faith-Based Programs, they raided the team of compassionate conservatives surrounding Hastert: Bridgeland became director of the White House Domestic Policy Council, and Charles became assistant secretary of state for narcotics control. The new administration, DiIulio believed, would take the lead in "reforming drug-related sentencing policies that ­research had shown were having perverse consequences."

"If you look back at that campaign document, it really is pretty impressive," says Carnevale, who ended up heading the drug office's transition team for the Bush administration. "Which is kind of remarkable, given what happened next. They've appointed a drug czar who ran like hell from a very sensible policy."

It took Bush nearly a year to pick his drug czar, and almost no one felt encouraged by his choice: John Walters, a laconic Midwesterner who had served as Bill Bennett's chief of staff during the administration of George H.W. Bush. "We all knew who Walters was," one longtime drug warrior tells me, "but he wasn't what you would call an inspiring figure, even to conservatives." When Walters submitted his first National Drug Control Strategy to Bush in February 2002, it became clear that the administration's focus had narrowed: Walters was devoted to Plan Colombia and to a prevention campaign that would keep kids from trying drugs for the first time, aimed particularly at marijuana - even though the number of first-time pot smokers had been flat for half a decade. Longtime drug warriors like Carnevale were stunned. "We were going back to an Eighties-style drug policy," he says - one that emphasized the kind of military and law-and-order programs that had been proven not to work, while ignoring programs, particularly treatment, that did.

Walters also had a complaint with the ads that the Partnership for a Drug-Free America had created for the drug czar's office under McCaffrey. They were, he said, too soft. He had a point. The ads, which ran under the slogan "The Anti-Drug," had been designed by a committee of academics who apparently believed that kids needed to be shown that not doing drugs could be fun too. In one characteristic spot, a pen draws an animated landscape, with a cartoon boy avoiding the advances of cartoon dealers before driving off into the distance with a cartoon dragon on a cartoon motorcycle. "My name is Brandon, and drawing is my anti-drug," the narrator says sweetly. The commercials made abstinence seem so lame they could have been designed by the cartels. "A lot of the ads that were produced were really boring," admits Philip Palmgreen, a University of Kentucky communications professor who served on the ad committee. Walters not only wanted harder-hitting messages - he also wanted the focus "to narrow around marijuana," according to one staffer at the Partnership who asked not to be identified. "Very candidly, the Partnership pushed back against that because the problems associated with marijuana are not very dire." But Walters disagreed, the staffer adds, "and we lost."

Walters refused to be interviewed for this story, but his office did make available one of his top advisers, David Murray. I asked him why his boss had narrowed the focus to marijuana, even though studies had disproved the causal link between marijuana and hard drugs. "If you're going to have a national office of drug-control policy, you look at the most prevalent drug in the society that's readily available - you don't go after meth first thing," he says. "You think about it like an epidemiologist, and you go for the vector that's most likely to spread, and that's teen marijuana users."

The new ads took a counterintuitive approach. "We wanted to make sure we were getting through to the thrill-seekers - those teenagers who are much more likely to use drugs - and convince them that it was more exciting not to do drugs," says Palmgreen. In a heralded spot called "Pete's Couch," the teenage narrator says, "I smoked weed and nobody died. I didn't get into a car accident. I didn't OD on heroin the next day. Nothing happened. We sat on Pete's couch for eleven hours." Then the camera shifts to show other teenagers, presumably those who haven't smoked weed, doing fun things - biking, playing basketball, flirting with girls. "You have a better shot at dying out in the real world," the narrator says, "but I'll take my chances out there." The advertising community was impressed with the spot: "Finally, an admission that smoking pot isn't calamitous," cheered Slate's advertising columnist, Seth Stevenson. Said Palmgreen: "Really good spots. The focus groups of thrill-seekers gave them great grades."

But the reality is that such ads - no matter how persuasive - do little if anything to prevent teens from trying pot. In 2005, a government-commissioned study designed to evaluate the prevention campaign over five years delivered its conclusions: Kids who had been ­exposed to the campaign ended up with rates of drug use that were roughly the same as those of the control group, who had not seen the ads. Murray loudly challenged the study's methodology, but when Congress asked federal analysts at the Government Accountability Office to assess the findings, the GAO upheld the report. The anti-drug campaign had not worked at all.

There was another problem with the Walters approach: Just as the federal government asserted the dangers of smoking pot, the states - first California, then three others - were permitting doctors to legally prescribe marijuana to relieve the chronic pain that came with cancer, polio and other debilitating long-term diseases. Attorney General John Ashcroft dispatched federal agents to begin raiding the suppliers and purchasers of medical marijuana in California - people who were operating completely within state law. The raids were even more surreal in their theatrics than the ones that had been launched by McCaffrey: In one particularly ludicrous incident, a forty-four-year-old post-polio sufferer named Suzanne Pfeil, who smoked prescription marijuana to relieve her pain, was hauled off to jail by DEA agents who pointed automatic rifles at her head and handcuffed her to her wheelchair. The rhetoric reached the level of crusade: Walters called citizens who plant and tend marijuana gardens "terrorists who wouldn't hesitate to help other terrorists get into the country with the aim of causing mass casualties."

What was striking to many veteran drug warriors was how fully the drug czar's office had bet on the youth marijuana initiative. For all Ashcroft's bluff talk about wanting to "escalate the War on Drugs," only a very small portion of it was being escalated. Funding for drug courts, which channel nonviolent drug offenders through treatment programs rather than prison, was zeroed out, and funding for local police was gutted. Carnevale, who quit his job after overseeing the transition in 2000, began to feel he was in a time warp. "This White House is walking away from prevention funding and treatment," he says now. "They haven't supported the community anti-drug coalitions, which actually work pretty well, and domestic law enforcement is flat or declining. To have a successful drug policy, you need all these elements, and what this administration has done is go crazy on exactly the element that doesn't work."

By the summer of 2005, the drug czar's failures were beginning to spill out into the open. For four years, while he focused obsessively on pot, Walters had done virtually nothing about meth, which was rapidly devastating the red states that had elected his boss. Walters struck a strangely discordant note on the growing epidemic, insisting that even as methamphetamine spread from the West Coast to the East, it remained a regional problem, not a national one, and therefore did not place high on his list of priorities. That September, the House's meth caucus asked Walters to come in for a meeting, to see if they could restore some element of dialogue and begin to rebalance the budget. The drug czar, once again downplaying the issue, sent Murray in his place. The congressmen, who had excluded the press to prevent grandstanding, went through the budget in detail and told the drug deputy what they wanted restored to fight meth. But, according to one staffer, Murray just sat there: "He didn't even bother to ask a question."

Incensed, Rep. Mark Souder, a Republican who chairs the House Subcommittee on Drug Policy, walked out of the room and held an angry press conference. Murray's testimony, he said, had been "pathetic" and "an embarrassment," and Walters was not doing his job: "If he does not lead, we need a change of the drug czar." Sen. Grassley, the Iowa Republican, echoed Souder a few days later. "What I've never understood," he said, "is why they took marijuana so much more seriously than methamphetamine, when methamphetamine is a much more serious drug."

By virtually every objective measure, the White House had lost the War on Drugs. Last year, Walters boasted that drug use among teenagers has fallen since 2002 - ignoring the fact that overall drug use remains unchanged. The deeper problem is that the drug czar has stopped measuring anything other than drug use. During the 1990s, at the direction of Gen. McCaffrey, Carnevale had created a comprehensive system to measure whether we were winning the drug war. The system took into account drug price and availability in the United States, how difficult it was for drug smugglers to get their product into the country and the consequences of drug use on public health and crime. But Walters simply tossed out that system of evaluation - as well as the unflattering facts it highlighted. "Had we kept it," Carnevale tells me, "we would see that the Bush administration has not made a positive impact on any of the measures."

Most unexpectedly of all, crime - a problem that seemed to have been licked a decade ago - is beginning to creep back up. In October 2006, the Police Executive Research Forum released a report declaring that violent crime in the country was "accelerating at an alarming pace." Murders were up twenty-seven percent in Boston over the previous year, sixty percent in San Antonio and more than 300 percent in Orlando. Even in the cloistered world of policing, complaints began to build about the numbers and about the cuts in federal funding. "The reality is a lot of police officers are politically conservative folks," says Ron Brooks, the president of the National Narcotics Officers' Association. "But there's been a lack of leadership in this administration on this issue."

10. THE RETURN OF DON BERNA

While the drug czar was cracking down on medical marijuana, the Bush administration was also overseeing a dramatic escalation in its overseas front of the War on Drugs. From the start, the White House had trumpeted Plan Colombia as an essential weapon in its anti-drug arsenal, eliminating inconvenient rules that had gotten in the way of a full military commitment to the project. For "those in the drug business," Walters declared in January 2002, "now is the time to get out." But despite the billions the administration spent on the program, and the new impunity given to the Colombian military, nobody really knew whether it was working. In July 2006, Adam Isacson decided to see for himself.

Isacson, a scholar who runs the Colombia program at the Center for International Policy, flew down to the Andes to construct his own assessment of Plan Colombia. He decided to make two stops - in Medellín, to determine how much the country's security situation had improved, and in Putumayo, to determine the success of the plan to eradicate the drug traffic. Regular assessments compiled by the White House drug office suggested that the crop-eradication program had reduced the acreage under coca cultivation in Colombia, but Isacson was skeptical: The price of cocaine on the American street had not risen, and separate estimates by the United Nations undercut the Bush administration's findings.

The modern Medellín he found looked more like Miami than a front in the drug war. The government and its paramilitary allies had secured the city, and U.S. officials went out of their way to praise the cooperation they were getting from Colombian police and military units - which had been cleansed, they said, of corruption. When Isacson pressed people about why the violence had decreased so dramatically, he was told repeatedly that "the paramilitaries won" - that government-supported forces had simply driven off the left-wing guerrillas and ended civil war in the city.

The paradox for Americans was that paramilitary commanders, such as Don Berna, had also taken control of the cocaine trade and retained enough political clout, according to a study by a Colombian think tank, to alter the composition of the Colombian Senate. When Don Berna was arrested two years ago, the entire bus transportation system of Medellín shut down for a day. "The command came down from the prison phone," says Aldo Civico, a professor of international relations at Columbia University who has done ­extensive research on drug smugglers and the paramilitaries. Don Berna is now in a jail cell south of Medellín, from which he continues to control his trafficking organization. "It is a signal to everyone that Don Berna is the one who is in power in Medellín," Civico says.

In Putumayo, Isacson found tent cities buried in the thick jungle, migrants living underneath sheets of plastic. Though tens of millions of American dollars had been spent on trying to improve the local economy, the main road that farmers were supposed to use to ferry their legitimate products to market was still unpaved, and a factory American money had built in 2003 was already shut down. Putumayo had been the first target of Plan Colombia's spray-eradication efforts and the site of its initial success: Coca cultivation had been cut by ninety-three percent from 2000 to 2004. But the place Isacson saw only two years later was "depressed." With no real financial incentive to switch to legitimate crops, farmers in the region had once again begun planting coca: Cultivation doubled in 2005. "We didn't see anything to suggest the improvement was sustainable," Isacson tells me.

The problem was that coca had simply moved next door, to the rural province of Nariño, along the country's Pacific Coast. Traffickers were planting strains of coca that could grow from seed to harvest in just six months. "The spray planes eradicated Putumayo," Isacson says, "and then all of a sudden coca cultivation starts in Nariño, and you see the same pattern - coca money means all these nightclubs and stores go up in these nothing towns, the police start reporting a sharp increase in murders, and eventually the provincial government is overwhelmed." The traffickers hopscotched across the country - Putumayo to Nariño, Nariño to Antioquia - always one step ahead of the drug agents and soldiers.

"As a drug-control policy," Isacson says, "it's hard to come to any conclusion other than that Plan Colombia has failed." In June of this year, the CIA released an assessment that confirmed Isacson's conclusion. Admitting that it had previously been undercounting the coca crop, the agency issued revised numbers showing that six years of Plan Colombia, at nearly $1 billion a year, had not cut coca cultivation at all. The effort to stop cocaine at its source had not made a dent.

"We've been working in Colombia for thirty years, and we don't have a hell of a lot to show for it," says Myles Frechette, the American ambassador to Colombia during the Clinton administration. "This is like a cancer. Every year the lesion, if you took a snapshot, would be bigger."

11. THE WATER BALLOON

At night, the population of el Paso, Texas, is 700,000, and that of Ciudad Juárez, just across the border, is 1.4 million. During the day, those numbers shift, as Mexicans stream across the cobblestone bridge over the Rio Grande for legal work in the United States. Every twelve hours, the two cities pass 100,000 people back and forth, squeezing them from end to end like the contents of a water balloon. "Among them," says Tony Payan, the political scientist at the University of Texas-El Paso and an expert in the dynamics of the local drug trade, "you see the spotters, the lingerers, mostly young men who are just standing there, watching out for when the coast is clear or when an American border agent who's been paid off by the cartel comes on duty. Then they tell the people that need to know, so they can make their drug runs across the border into Texas." With the failure of Plan Colombia, a handful of bridges along the Mexican border have become the main front in the War on Drugs.

Cocaine trafficking in Mexico has its own prehistory. For generations, family networks of smugglers had moved marijuana and cheap, black-tar ­heroin across the border -veteran DEA agents were accustomed to arresting the grandsons of men they had arrested years earlier - and the whole drug traffic in Mexico was small enough, by the mid-1980s, that it was effectively controlled by one man, Miguel Angel Felix Gallardo, who ran a violent trafficking organization out of Tijuana. As Colombian groups, chased from the Caribbean by American interdiction efforts, began to look to the southwest border in the early 1990s, Felix Gallardo discovered he could no longer control the traffic himself from prison. "He had a meeting with his lieutenants and divided the Mexican border crossings up among them, creating the modern cartels," Payan says. "His nephews kept Tijuana, and one group got the Sinaloa-Arizona crossing, another got Laredo-Nuevo Laredo, and Amado Carrillo Fuentes got El Paso-Juárez."

Mexican officials along the border, whose PRI party had kept a lock on national power for seventy years, allowed traffickers to move their product in exchange for reduced violence. "In order to coexist, the government looked the other way as long as the cartels didn't wreak havoc in the country," says Armand Peschard-Sverdrup, director of the Mexico Project at the Center for Strategic and International Studies. "It became somewhat of a safety valve in terms of dealing with organized crime, as a way of mitigating the political instability." Though the U.S. government pushed Mexican officials to crack down on corruption, its pleas and threats went largely unheeded. By 1997, Carrillo Fuentes - the Lord of the Skies - was moving tons of cocaine across the border every year and had amassed a fortune worth $25 billion. But that same year, Carrillo Fuentes died on an operating table in Mexico City, where he had been undergoing plastic surgery to change his appearance and avoid detection: In the ghoulish post-mortem photographs, his face is speckled like a snake's skin, two shades of brown and one of pink. Juárez fell into a testy, three-way competition for control of the drug trade, and the murders took on a symbolic vocabulary of their own: Tortured victims piled in oil barrels filled with concrete and buried alive, members of opposing cartels murdered and left to rot in car trunks in their own neighborhoods, snitches killed and left on the side of the road. The violence between cartels is so pervasive, Payan says, "if you move into a home in Juárez, you will never know whether there's a body underneath the floor in your dining room."

At the beginning of the Bush administration, it looked like Mexico might actually begin to bust corrupt cops who did business with drug smugglers. In 2000, when Vicente Fox, the reforming, conservative rancher and friend of George W. Bush, took power, he began prosecuting dirty police officers, throwing tens of thousands of them off the force. "There were unintended consequences," says Peter Andreas, a Brown University professor who has studied drug trafficking along the border. "Many of the corrupt cops went to work in the drug trade" - a shift in power that had the effect of professionalizing the violence. In addition, an estimated 90,000 Mexican soldiers deserted during the Fox administration, many of them signing up with the cartels.

In Juárez, the effect was devastating. Free to operate as they pleased, the cartels began to split, with capos challenging one another openly for control of the drug corridors. Local and state police killed each other over the right to protect the traffic. A new gang called the Zetas, made up of Mexican soldiers who had quit their day jobs to take over the drug trade, waged war in Juárez and killed 100 people in the corridor around Nuevo Laredo in the summer of 2005. The gaudy theatrics of the murders have only intensified as drug gangs seek to guarantee that their killings send a message by getting media attention: Last year, gunslingers wearing military uniforms walked into a popular nightclub in Uruapan and dumped the severed heads of five rivals on the dance floor, like soccer balls. Over the past year, drug-related murders in Mexico's border states have doubled, driven primarily by the booming trade. "What we're seeing is the Colombianization of Mexico," says Andreas.

For those who have studied American drug policy, the catastrophe along the border looks like a final reckoning for overseas interdiction. "It's like a balloon effect - we've never succeeded in cutting off the traffic, we've just pushed it around," says Payan. "We cut off supply in the Caribbean, and it came here. We cracked down on the Colombian traffickers, and it just meant the Mexicans traffickers got wealthier, and the violence came here." Like many DEA agents and border experts, Payan was consumed last summer by the story of Zhenli Ye Gon, a Chinese pharmaceutical executive whose house Mexican police raided, suspecting him of diverting meth components from China for illegal use. Inside they found $206 million in cash -final evidence of just how far the meth epidemic has spiraled out of control since pharmaceutical lobbyists prevented Gene Haislip from forestalling it with a simple federal regulation. Payan believes, as do many in the DEA, that Ye Gon is a harbinger of the next frontier in the War on Drugs.

"Even if somehow we could manage to get the drug trade away from the Mexican border, it will come through Asia next," he says. "Instead of fighting a border war, we'll be fighting it in containers. But unless we can reduce demand, it's a zero-sum game."

12. THE PRIVATEERS

Even by conservative estimates, the War on Drugs now costs the United States $50 billion each year and has overcrowded prisons to the breaking point - all with little discernible impact on the drug trade. A report by the Government Accountability Office released at the end of September estimated that ninety percent of the cocaine moving into the United States now arrives through Mexico, up from sixty-six percent in 2000. Even Walters acknowledges that for all of the efforts the Bush administration has devoted to overseas drug enforcement, the price of cocaine has dropped while its purity has risen. More than forty percent of Americans support legalizing marijuana, yet the government continues to target pot smokers. In October, the administration announced it was planning a new military offensive, dubbed Plan Mexico, with a price tag of $1.4 billion. Things look so bleak that Walters was recently moved to describe a momentary upward blip in drug prices as "historic progress."

There are a handful of battles in the War on Drugs that have actually been won, times when fresh thinking prevailed over politics - but they are not the kind of victories that the Bush administration is eager to trumpet. In the summer of 2003, the police department in High Point, North Carolina, held its annual command-staff retreat in a small conference center themed to look like the log cabins of the pioneers who settled the region. One topic dominated the conversation: an increase in violent crime that was concentrated in three drug-dealing neighborhoods in the city. "The place we were at was that all the traditional enforcement was making no difference," says the department's deputy chief, Marty Sumner. "We agreed we weren't going to be able to eliminate drug use. We weren't even going to try to go after drug use. We wanted to change the marketing of the drug."

Sumner's department called in the Harvard criminologist David Kennedy. The High Point police had worked with Kennedy before, adopting the Boston Gun Project's policy of trying to break the link between drugs and crime. Now the criminologist told them that he had a new kind of project to propose, one that went beyond the Boston experiment. Kennedy's pitch was simple: The trick, he said, wasn't to focus on eliminating drugs but rather to shut down the most "overt" drug markets, the ones operating so openly that they attracted prostitution and violent crime. "Instead of looking at it as a drug problem, we decided to think of it as a drug-market problem," Sumner says. "What the public really couldn't stand was the violence associated with public drug markets." Dealers operating in the open are targets for stickup men and other would-be robbers, and the public swagger and turf consciousness of street slingers can cradle violent, simmering beefs.

High Point police began in the West End neighborhood, one of the city's three overt drug markets. A team of officers staked out the site, videotaping hundreds of hand-to-hand sales and mapping out a complete anthropology of the West End drug market. They found it was strikingly small: Sumner had expected as many as fifty dealers working there, but it turned out there were only sixteen. Before long, the ­officers had enough evidence to put away each of the sixteen dealers for good. But they didn't. Instead, Sumner and Kennedy called them in for a meeting. They showed each of them the portfolio of evidence against them and said that unless they stopped dealing drugs, the whole file would be handed over to the prosecutors and they'd be in jail for years. Family members were brought in to urge the dealers to stop, and social-service providers pledged assistance with food, housing and job training.

"We didn't think it would work," Sumner tells me, "but the drug markets have disappeared."

For five years before the program went into effect, the number of drug-related murders in High Point had stayed steady, around fifteen a year. In 2007, in the program's fourth year, it has plummeted to two. Violent crime in the West End has declined by thirty-five percent. "The use of drugs isn't something we could affect," says Kennedy. "But the violence was." His logic has an appealing clarity for overworked police departments: There are now more than sixty cities in the United States that use some version of Kennedy's program, edging away from thirty-five years of punitive measures that have turned the United States into the world's leading jailer to a social-work model that encourages communities and cops to engage the problem on a more human level. The real radicals of the War on Drugs are not the legalization advocates, earnestly preaching from the fringes, but the bureaucrats -the cops and judges and federal agents who are forced into a growing acceptance that rendering a popular commodity illegal, and punishing those who sell it and use it, has simply overwhelmed the capacity of government.

In 2000, voters in California, whose prisons now hold nearly twice as many inmates as they were designed to incarcerate, passed a referendum called Proposition 36, which has since sent more than 150,000 nonviolent drug ­offenders to treatment instead of prison. The program is not perfect: Though the outcomes for those who make it through treatment are surprisingly strong, many convicts simply skip the sessions, and there are few enforcement mechanisms to compel them to attend. But the program, according to a study conducted by researchers at UCLA, still saves tax­payers $2.50 for every dollar put in. And a pilot program in Honolulu which requires near-constant drug tests of those on probation and provides incremental punishments for each extra failed test - suggests an effective model for treating hardcore addicts, says Angela Hawken of UCLA and Pepperdine University. "It offers the promise that we might ­really be able to solve this problem."

In recent years, there have been flickers of political progress that suggest America's drug policy is ready for a historic shift. Democrats in both the House and Senate have voted to cut proposed funding for Plan Colombia and have pushed for hearings on sentencing reform. As the politics of crime and drugs have lost their power to move votes, some conservatives, including Republican senators Jeff Sessions and Sam Brownback, have begun to question the logic of mandatory-minimum sentences. "There is a more promising environment for drug-policy reform than at any time since the Carter administration," says Ethan Nadelmann, executive director of the Drug Policy Alliance and one of the country's foremost critics of the drug war.

But despite their evident success, the most forward-looking programs remain buried at the fringes of drug policy, featured not in the president's budgets but in academic journals and water-cooler talk in cities like High Point. Experimentation at the community level is more imaginative than programs that are federally sanctioned. "We haven't had the kind of national leadership that blesses this and encourages it," says Caulkins, the RAND researcher from Carnegie ­Mellon. "So this kind of innovation stays below the radar." Thirty-five years after Richard Nixon launched the War on Drugs, the most promising ­programs continue to be shunted aside by Washington's unswerving emphasis on law and order.

The drug war, in the end, has been undone in no small part by the sweeping and inflexible nature of its own metaphor. At the beginning, in the days of Escobar, the campaign was a war as seen from the situation room, a complicated assault that spanned multiple fronts, but one which had identifiable enemies and a goal. Today, the government's anti-drug effort resembles a war as seen from the trenches, an eternal slog, where victory seems not only unattainable but somehow beside the point. For the drug agents and veterans who busted Escobar, the last decade and a half have been a slow, agonizing history of defeat after defeat, the enemy shifting but never retreating. "You get frustrated," Joe Toft, a former DEA country attache in Colombia, tells me. "We've never had a true effort where the U.S. as a whole says, 'We're never going to crack this problem without a real demand-reduction program.' That's something that's just never happened."

Toft, now a private security consultant, thinks back to the heady days after the fall of Escobar, the days when winning the War on Drugs seemed only a matter of dispatching more American helicopters to the Andes. "The first couple years, I had this very naive idea that I was really going to make a huge impact," he says. "But after a while, you start realizing that without a concerted effort to reduce demand, it's not going to happen. Over the years, I came to see my job as basically keeping the lid on the garbage can trying to sit on that lid and prevent that garbage can from overflowing. If you talk to a hundred agents, that's what almost all of them would say. We're just being realistic."