Diego was 23 years old, a poor Colombian living in a poor section of Cali, when his girlfriend had the baby. He was broke—everybody was broke—but his grandmother knew where he could earn some money: He could go work the coca plantations in the hinterlands like she had. She could get him a job working for FARC (Fuerzas Armadas Revolucionarias de Colombia, Colombia's Marxist-Leninist guerrilla army), which was better than working for the right-wing paramilitaries.

Diego is not his real name, and he's currently living in a different Latin American country—otherwise, he said, he wouldn't be talking to me.

The Colombian cocaine trade has been tangled up in the country's guerrilla wars for decades: FARC fights the government, the government fights FARC, and the right-wing paramilitaries fight FARC on the government's behalf with an extra measure of savagery. FARC and the paramilitaries constantly jockey for territory—to gain not only political power, but access to rivers, coastal mangrove swamps, and other secret routes where they can smuggle weapons and cocaine to fund their decades-old war.

The paramilitaries, according to Diego and his family, are the worst of the bunch. FARC at least tends to pay its workers, they say. The paramilitaries, on the other hand, have a reputation for stealing coca leaves from farmers at gunpoint and making their cocaine workers labor for months for little or no pay. Sometimes, Diego said, they pay you in lead—a bullet to the brain.

"You can't talk about FARC or the paramilitaries ever, because you don't know who's working for who or who they've recruited," Diego said. "Even people you know are on both sides, so you can't be drunk at a bar and bring up anything—your cousin's girlfriend might be working for the other side. People disappear and get killed all the time. It's very dangerous to talk, to do anything."

But Diego had a family to feed and his grandmother's assurance, so off he went.

His grandmother told him to go to Buenaventura, a port city on Colombia's Pacific coast and one of the most dangerous cities in the western hemisphere. Buenaventura has the highest murder rate in Colombia, the New York Times reported in 2007, and when Diego was there, around 2000, Colombia had the highest murder rate in the world. Buenaventura, Diego said, is "hell on earth." With no tourism to speak of, the government has limited incentive (or ability) to crack down on the leftist guerrillas, right-wing paramilitaries, and narcos who walk the streets with their guns out in the open.

Diego remembers getting on a ferry that left Buenaventura at 5:00 p.m. After a several- hour nighttime journey with 70 or 80 other people (plus rice, beans, beer, and other staples bound for the camp), they arrived at a dock. He was then loaded onto a speedboat that, he said, went 90 miles an hour for seven hours. He had no idea where he was or where they were going, which was the point. The speedboat traveled through swamps and up rivers, stopping only to refuel.

Some of Diego's fellow travelers were laborers, like him. Others were delivering goods (the beans and rice and beer) to the temporary village that serviced the camp—like an old Alaskan gold rush village, except in a tropical jungle patrolled by guerrilla soldiers with machine guns.

When the speedboat finally arrived at the village, it was greeted by guerrillas asking each new arrival for information: Who are you? What job are you here to do? Who told you about the camp? "You can't just not know anything," Diego said. "Some people died at the entrance because they did not have the right answers, so they got killed right there." Diego saw laborers—who said they'd heard they could board the boat in Buenaventura to make some money—murdered at the dock. This particular camp was known as the Black-Bag Camp, so called because the guerrillas would put a black plastic bag over your head before executing you.

And this was FARC—these were the good guys, according to Diego's family.

The coca camps move around to make it harder for police to find them. "It's very, very hard to get to them—it cannot be underestimated that guerrillas know the countryside and the police don't," Diego said.

The camp, Diego said, was maybe six football fields big. (We conducted our interviews over Skype, and his current girlfriend—not the one he had when he joined the coca trade—laughed in the background: "That is your measurement for everything!") He was set to work picking coca leaves, which are thick and sharp. The guerrillas don't issue gloves, and Diego was a city boy with soft hands: "It was really funny to everybody else but hard for me," he said. "It was like ripping palm trees apart. You can still see the scars on my hands."

The conditions weren't very good. He'd wake up at 5:00 a.m. on a dirt floor with 20 other people, every man under his own mosquito net. "After 5:00 p.m., it was just mosquitoes. It was boiling hot, but you had to wear long sleeves and pants, and still a lot of people got yellow fever and dengue fever." He'd work for three hours, from 5:00 a.m. to 8:00 a.m., filling a 50-kilo (110-pound) sack with coca leaves, and take a break for breakfast. Then he'd be back at it until lunchtime at noon. "We ate rice and fish and lots of bananas," he said. "Sometimes they had juice. But you had to drink a lot of water, because it was really hot."

They'd pick coca leaves until dark, hoping for rain to cool them down and to make their leaves heavier, meaning they had to pick fewer to fill their sacks. (He typically filled the sack three times in a day—150 kilos total.) Sometimes, when it didn't rain, they'd piss in their sacks to make them heavier.

"Not one person I met out there used cocaine," Diego said. "We would chew on the leaves—to kill the hunger, the fatigue, to stop the pain of the work. You'd get bit by spiders and scorpions, mosquitoes or snakes, and you'd chew so you wouldn't feel the pain. Some people believed the coca leaves would stop the poison and save your life."

At his peak earning period, Diego was making the equivalent of US$600 per month. "Back then, that was good cash and it was fast," he said. It was incredibly dangerous, too: People would go back home with wads of money, usually to bring to their families, and get robbed and killed along the way, sometimes by the same people they'd been working with in the fields for four or five months. When Diego traveled, he always went with an entourage of uncles or cousins.

After a while, Diego graduated from the fields to the "factory," which was more like a shed, where he helped turn the raw leaves into cocaine paste. "Making the paste is gnarly," his girlfriend said. "That's where the real scars come in."

"There's a big pool with all the leaves, a big wood tub," Diego explained. Workers would pour leaves into the tub, stomp them down, and then add gasoline to extract the cocaine alkaloids. "That's the easiest way for the government to find the camps," Diego said. "Gasoline is expensive, and most farmers don't use that much—sugarcane and bananas are all farmed by hand—so you find whoever's buying vats and vats of gasoline."

They'd lay a tarp over the tub for 24 hours, with someone stirring the gasoline-coca stew every four or five hours. Then they'd taste the brew to see if it was strong enough. If it numbed the tongue, it was good. If not, it needed more chemicals. Diego doesn't remember exactly which chemicals they used: "There were a lot of chemicals." Eventually, they'd pull the plug on the tubs, collect the cocainized gasoline, add ammonia and sulfuric acid, and chemically reduce the brew into a paste that was taken elsewhere to be turned into cocaine hydrochloride—powder.

Diego and the other workers were encouraged to spit into the tubs holding gas and coca, and even pass spit-mugs around the camp, under the premise that saliva helped the extraction process. Workers were also encouraged to ash their cigarettes into the vats, perhaps because traditional coca chewers sometimes added a dab of quicklime or the ash of burned quinoa plants to their wad of coca. (Diego said he wasn't sure why.)

Then there were the unauthorized additives. "Sometimes we'd piss or shit in the vats, just to be fuckers," Diego said. "Only the rich use cocaine, and we thought it was funny."

The territory where Diego was working—he still isn't 100 percent sure where they were—was a death zone. The paramilitaries and guerrillas were fighting upriver, and he said that sometimes when he went to fetch water, he'd see dead bodies or severed limbs floating past. "I often heard people say things like 'Yesterday I saw four bodies going down,'" Diego said. "All the time, people were talking about bodies. A lot of times they were tied together, big groups of people."

Stories circulated from the guerrillas to the workers about small bands of soldiers who "went to make some business away from the group" and were savaged by larger groups of paramilitaries: "First they cut their hands off, then they cut their legs, and then they'd kill them." The threat of infighting and defection was constant, as workers felt the tempting urge to abscond with packages of paste and sell them on their own.

Then, of course, there was the guerrilla war. In nearby villages, guerrillas and paramilitaries enforced curfews, telling villagers to "go to bed early, because if we see anyone walking around after 10:00 p.m., we'll kill them." They also went from village to village, Diego said, conscripting boys for their armies, sometimes leaving notes under people's doors ordering them to bring their sons to the town center (the church, the plaza) at a certain time. If anyone refused, the soldiers might return to murder the entire family.

One weekend, Diego and some of the workers had a Sunday off, so they went to the little village that serviced the camp to drink beer, loaf around, and look at the women. (Most of the women in the camps, he said, were either cooks or prostitutes, some of them very young local girls.) "We were playing billiards and chess, and all of a sudden people got really quiet. Then people freaked out and were running because the paramilitaries showed up." The paramilitaries forced everybody inside a classroom, then said that government soldiers were on their way and not to tell the guerrillas. "A lot of people were shot," he said. "Then they left. That was it—people were shot because the paramilitaries wanted to show something or prove something, to make a point."

During his time working in the cocaine kingdom—several months-long shifts—Diego saw a parade of corpses, guns, bags of cocaine paste, and barrels full of U.S. dollars that mysteriously appeared and disappeared. One of his cousins and some of his uncles were killed. (He declined to say how or why.) The breaking point came one day when he went to get drinking water, saw yet another corpse floating past, and snapped. "And I was like, Oh my God, I can't do this anymore. No, no, no, no, this is not the kind of life I want for me."

Diego left the trade and eventually the country. As of our last conversation—we had multiple hours-long conversations over two months—he does not intend to return. At one point, he called his family to clarify a few details he couldn't remember. A few days later, he said his family was very upset that he'd been talking to me and upset that he'd asked them about the trade, and that he couldn't talk to me anymore. That was the last I heard from him.

Diego never worked on the hydrochloride processing, the most chemically advanced and sensitive part of the process—the stage when the paste is turned into powder and, presumably, the stage at which levamisole, a cattle-deworming medicine that's been showing up in the world's cocaine supply, is added. As covered in the first part of this series—"The Mystery of the Tainted Cocaine," August 19—the DEA reported finding levamisole in 73.2 percent of cocaine seized in the United States in 2009, up from a paltry 1.9 percent in 2005. The DEA has also found levamisole-tainted cocaine in busts in Colombia and even in the plastic laminate on glossy calendars shipped into the U.S.—the laminate itself is impregnated with cocaine. (The DEA has agents working across Latin America.) This indicates that levamisole is being added at the source, in the labs where the paste is turned to powder.

When levamisole is ingested by humans, it can trigger a catastrophic immune-system crash called agranulocytosis; it has led to an unknown number of hospitalizations and multiple deaths among cocaine users in the past two years. Physicians in Seattle have reported seeing the same cocaine users land in the hospital with agranulocytosis on multiple occasions.

Levamisole is an unusual—and unprecedented—cutting agent because it's more expensive than other cuts, it makes some customers sick, and it's being cut into the cocaine before it hits the United States. Smugglers typically prefer to move pure product, which is less bulky and results in less chance of detection. "The Mystery of the Tainted Cocaine" offered a few theories about why South American drug manufacturers (mostly Colombian) are cutting their cocaine with levamisole.

A quick review of those theories:

1. Levamisole might produce a cocainelike stimulant effect either on its own or in conjunction with cocaine (in 2004, racehorses treated with levamisole were found to metabolize the deworming drug into an amphetamine-like stimulant called aminorex), meaning the product could produce a more substantial high with less pure cocaine.

2. Levamisole, unlike other cutting agents, retains the iridescent, fish-scale sheen of pure cocaine, making it easier to visually pass off levamisole-tainted cocaine as pure.

3. Levamisole passes the "bleach test," a quick street test that reveals cuts like sugar or lidocaine (but, because of a chemical anomaly, not levamisole).

4. Levamisole is a bulking agent for crack. Making crack involves purifying cocaine and washing out the cutting agents, but levamisole molecules slip through this process—meaning a dealer can produce more volume of crack with less pure cocaine.

5. All of the above.

If levamisole can do several of these things, it becomes (in evolutionary terms) an advantageous genetic mutation, a cut that may have started as an accident but showed beneficial properties, so was passed on from one batch of cocaine to the next, from one generation to another, like a new gene.

The evolutionary, all-of-the-above theory is especially compelling because Colombia's cocaine-production market is so fractured and decentralized, it's unlikely that one mind is calling all the shots about cocaine's manufacture and distribution. After the breakup of the Medellín and Cali cartels in the mid-1990s, hundreds (maybe even thousands) of independent cocaine producers leaped into the void. While FARC and the paramilitaries control much of the cocaine trade—such as the camps where Diego worked—they don't control and centralize production technology to the degree that the Medellín and Cali cartels did.

Meaning: Those hundreds (and maybe thousands) of independent producers must have independently decided to use the same cutting agent. And even if levamisole looks like cocaine, behaves in bleach tests like cocaine, and can bulk up crack in a way that other cutting agents can't, why not wait to cut it once it has crossed the border?

That's still a mystery.

First thing: The theory that levamisole is being cut into cocaine because it's a stimulant might be bunk. Dr. Geoff Baird, the director of clinical chemistry at Harborview Medical Center, has been giving patients who test cocaine-positive (or admit to having recently used cocaine) a comprehensive drug screening that detects levamisole as well as aminorex and other drugs. Two-thirds of his cocaine-positive patients are testing positive for levamisole—which is roughly consistent with the DEA estimates of the rate of cocaine contamination—but none tested positive for the stimulant aminorex.

That might kill the first theory. Racehorses metabolize levamisole into aminorex, but humans don't, at least not in any significant amount. But even if levamisole isn't a stimulant on its own, it might have a synergistic effect with cocaine that boosts the high. Another doctor at Harborview, Dr. Mike Clark, is currently testing rodents to see whether levamisole plus cocaine does something that neither levamisole nor cocaine can do on its own. Results on his experiments are pending.

Regardless, levamisole is certainly out there and is still making people sick. "We've seen levamisole in rock [cocaine] for sure," Dr. Baird said. He has found it not only in the patients who use cocaine but directly in crack rocks that patients have given him for analysis.

(Please note, drug users: Dr. Baird says he and his fellow physicians have no interest in reporting drug users to the police. You can be honest about your habit without fear of prosecution—and you should be honest, both for your own sake and for the sake of research about drug-related health issues like the levamisole question. "It is not our job to arrest patients or police them or enforce those sorts of laws," Dr. Baird said. "If someone has a crack rock, we'll analyze it and that's it—we won't give it back to you, but unless you say you're going to feed it to a little girl, we don't have to talk to the authorities.")

Second thing: While reporting the first part of this series, several drug-trade experts suggested to me that cartels were only cutting levamisole into their cocaine because they happened to have piles of the stuff lying around, maybe even as part of agricultural subsidies from the United States.

Peter Reuter, a professor and drug- economy expert at University of Maryland's School of Public Policy and Department of Criminology and Criminal Justice, took this line, saying "the presence of levamisole doesn't seem to be any great mystery to me," since it's a common agricultural medicine in the cocaine-producing regions of South America. Ethan Nadelmann, the executive director of the Drug Policy Alliance in New York City (he is sometimes described as "the godfather of the drug-reform movement") echoed Professor Reuter, saying by e-mail that "it could be the case that South American cocaine producers have access to cheap (or no-cost) levamisole, making it inexpensive and convenient."

And new intelligence from the DEA shows that Colombian cocaine producers are actively trying to keep their access to levamisole. A source familiar with an ongoing DEA investigation, who agreed to be interviewed on condition of anonymity, said: "When we first began testing, we found pharmaceutical- grade, highly processed levamisole that probably came from us [the United States]. New intelligence says they probably have illicit labs making levamisole in South America—it's a concerted effort. They also have evidence of container ships from China with metric tons of levamisole. The DEA is hoping to use levamisole in interdiction efforts. They can tell from the impurities, from the actual processing of coca leaves into powder form, which batch it came from, which region. Now they're adding the levamisole angle to their investigations as a way to follow where things came from and where they're going. The addition of levamisole is definitely intentional—it makes you think it's got to be something. And the DEA has no doubt it's being cut in in Colombia."

Most of the cocaine in the U.S. comes from Colombia (overland via Mexican gangs or on ships and narco-submarines sneaking up the Pacific coast), and most of the cocaine in Europe comes from Peru and Bolivia, via Brazil. (Brazil is also experiencing a cocaine- consumption boom, not only among the wealthy but also among street kids who find discarded 50-gallon drums that contained cocaine paste. They reportedly scrape the insides for the residue and smoke it as a kind of crude crack.)

As the anonymous source said, the DEA has a Cocaine Signature Program (CSP) that, according to a DEA report from January of this year, can "give evidence of how and where coca leaf was processed to cocaine base (geographical origin), and how cocaine base was converted into cocaine hydrochloride (processing method)." Traditionally, the report says, the processing from paste to powder can be broken down into a Colombian Method, a Peruvian Method (which involves acetone), and a Bolivian Method (which involves ether and acetone).

The report goes on to claim that the CSP can even detect which valley the cocaine has come from: In 2009, for example, 0.4 percent of the cocaine in the U.S. came from Peru's Cuzco/Apurímac valleys and 1.3 percent came from Peru's Huallaga/Ucayali valleys, a couple hundred miles away.

During the fourth quarter of 2009, 97 percent of the cocaine in the U.S. came from Colombia and 1.5 percent from Peru. (The remaining 1.5 percent came from undetermined sources.) And of all the cocaine bricks analyzed by the CSP (which averaged 65 to 85 percent purity), 67 percent were cut with levamisole, 9 percent were cut with diltiazem (more on that in a minute), and 20 percent were uncut.

So, you'd think levamisole was simply a Colombia/U.S. problem. The vast majority of our cocaine comes from Colombia, the vast majority of our cocaine is tainted with levamisole, and the DEA has independent confirmation that cocaine producers in Colombia are scrambling to keep ahold of their levamisole supply.

But now levamisole is turning up in Europe, in samples seized by police in France, the Netherlands, Spain, and the UK. Furthermore, in 2008, the DEA arrested Colombians turning paste into powder in Bolivia. According to the CSP report, "This 'technology transfer' for cocaine HCl [hydrochloride] processing was verified by the CSP from the samples seized at the clandestine laboratory site in Bolivia. The solvent profiles of those samples"—the geography-revealing "impurities" mentioned by the anonymous source above—"were 'Colombian HCl Process'... Distinguishing the processing origin of cocaine HCl is no longer possible based on the solvent profile of a seized cocaine HCl exhibit... It should be noted that the type of HCl conversion process has no effect on the 'base origin' determination of cocaine."

In plainer words, the DEA can still tell where the cocaine was grown and turned into paste. But it can no longer tell where the paste was turned into powder.

Which may explain why the DEA is so interested in levamisole and whether it's being cooked in illicit labs or coming over on container ships from China. With the aforementioned "technology transfer," the DEA is losing the ability to pinpoint the location of paste-to-powder labs (the easiest part of the process to conceal—certainly easier than hiding the coca-leaf farms). Could the ubiquity of levamisole serve as a smoke screen for the same purpose, to confuse the DEA? It seems unlikely, given how fractured the cocaine market is (or is supposed to be). But who knows?

Third thing: Nicholas Reuter, a senior analyst at the federal Substance Abuse and Mental Health Services Administration (SAMHSA), says that levamisole contamination in the U.S. started to peak as other cutting agents started to drop off—specifically diltiazem, a drug used to treat arrhythmia.

There are competing theories (this is a shadowy business, there are always competing theories) about why diltiazem was being cut into the cocaine: One theory, Reuter said, is that its antiarrhythmic properties might reduce the rates of overdose. The other is that it had been shown to reduce drug cravings in cocaine-addicted rats—in humans, diltiazem might prolong the cocaine high, meaning you could produce more cocaine-pleasure with less pure cocaine. Whatever you say about the cocaine producers, they have a serious interest in R&D. "They are," Reuter says, "very sophisticated." (The DEA's Microgram Bulletin, however, dismissed the calculated use of diltiazem, saying it was "most likely... an adulterant of convenience"—just something they had lying around.)

So perhaps levamisole is a replacement for diltiazem because it does have some kind of synergistic effect that boosts the cocaine high (even though Dr. Baird at Harborview has shown that patients who test positive for cocaine and levamisole aren't testing positive for the aminorex). "None of these theories are confirmed," Reuter said. "Another is that the levamisole might affect drug-detecting canines' ability to detect the drug as it's being shipped."

The answer to the levamisole puzzle is getting murkier and murkier. We know levamisole is very prevalent in the cocaine supply, and we know it's there for a reason. But we still don't know why.

Once the cocaine crosses the U.S./Mexico border, dealers cut it with agents other than levamisole: flour, baby powder, Epsom salts, laxatives, lidocaine, chalk. The street lore among cocaine and crack dealers in Seattle is that we get some of the worst product in the country—it gets cut and stepped on, mostly by Latino narcos, at every stop between San Diego and Seattle. It's rumored that high-end cocaine clients here, who want purer product and are willing to pay a premium for it, pay freelance smugglers who don't work with large narco-organizations to drive to Mexico, pick up cocaine, and drive it straight to Seattle.

Vancouver, Canada—a two-and-a-half-hour drive north—is rumored to have far superior cocaine than what is commonly available in Seattle. Vancouver is Canada's San Diego, the stories go: Asian gangs import product to Canada through Vancouver and step on it as it moves eastward across the continent—making Winnipeg cocaine roughly as crappy as Seattle cocaine. The Latino and Asian gangs are said to have a gentleman's agreement that the U.S./Canada border is also a gangland border. Latinos won't sell north and Asians won't sell south. As one dealer put it, "If we see any of those Asian guys down here, they're fucking dead."

That could be a just-so story told by street dealers, but Norm Stamper, who was the chief of police in Seattle from 1994 to 2000, confirmed this description about the way cocaine moves across the U.S. and Canada, saying that it was "consistent with what I was given to understand while I was chief of police." Since leaving the SPD, Stamper has joined LEAP (Law Enforcement Against Prohibition) and become a vigorous activist for the legalization and control of drugs—all drugs.

"The more sinister the drug, the greater the justification for their regulation instead of their prohibition," he said in an interview. "A regulatory model would give the government, imperfect as it is, the first opportunity since the beginning of the last century to exercise some control over the nation's drug consumption. As it is, it's just horrific chaos."

In the 1980s, 80 to 90 percent of the cocaine bound for the United States came in boats across the Caribbean Sea—it was the era of Miami Vice and cocaine cowboys and Pablo Escobar. Now, according to Dominic Corva, a geographer at the University of Washington who has done extensive research on the coca-cocaine commodity chain, 80 to 90 percent of U.S. cocaine comes overland through Mexico or up the Pacific coast in boats and narco-submarines.

This shift in smuggling routes is, in part, why Mexico has become a meat grinder over the past few years—30,000 dead since 2006, hours-long firefights on city streets, journalists murdered for simply reporting the news, mayors assassinated in big cities and small towns, beheaded corpses and bodies dissolved in lye dumped in downtowns in the middle of the afternoon with messages for the government and rival narcos pinned to their chests, Mexico's entire state apparatus severely compromised by violence and fear. (Not to mention, Corva says, an estimated half million Mexicans are employed in the narco-trade.)

This September, El Diario of Ciudad Juárez—one of the few Mexican newspapers to push back a bit against narco-censorship, a newspaper whose editorial staffers have been murdered for their efforts—published a pleading editorial to the narcos, titled "What Do You Want from Us?"

We'd like you to know that we're communicators, not psychics. As such, as information workers, we ask that you explain what it is you want from us, what you'd intend for us to publish or to not publish, so that we know what is expected of us.

You are at this time the de facto authorities in this city because the legal authorities have not been able to stop our colleagues from falling, despite the fact that we've repeatedly demanded it from them... We don't want any more dead. We don't want any more injured or any more threats. It is impossible to exercise our function in these conditions. Indicate to us, therefore, what you expect of us as a news outlet.

A newspaper publicly groveling at the feet of the narcos—it makes the Miami Vice era of targeted assassinations and speedboat chases seem almost quaint.

Before Mexico took over cocaine smuggling, its narco-capitalists "specialized in poppy/heroin and marijuana production, while performing transit services for a percentage of the more lucrative cocaine flow," Corva writes in his dissertation. Now the narco-capitalists of Mexico have gained a much bigger stake in—perhaps even a stranglehold on—the American cocaine trade. Many factors have enabled this transition, from U.S. and Colombian fights with Colombian narcos to effective interdiction in the Caribbean to banking liberalization and the North American Free Trade Agreement (NAFTA).

Corva and other drug-trade experts hesitate to call the narco-capitalist organizations "cartels" because the word implies alien entities separate from the state (the way that, say, the Bloods or the Crips are separate from the state). Mexico's narcos are thoroughly integrated into the government—through both contemporary corruption and long-standing historical alliances that predate U.S. drug-prohibition policy.

Major Mexican landowners had been growing marijuana and opium poppies and selling them to the U.S. long before the Harrison Narcotics Act of 1914 (the first major federal drug prohibition—prior to that, even the Sears, Roebuck catalogue advertised a syringe and a dose of cocaine for $1.50). Those Mexican landowners were aligned with, or outright members of, the Mexican political establishment. One brief example: Colonel Esteban Cantú Jiménez, who had a personal army of 1,800 soldiers and political control of Baja California Norte, started taking a cut from Mexican opium traders as soon as the Harrison Act was passed. The Mexican army eventually flushed him out in 1920—with a force of 6,000—but Colonel Jiménez secured amnesty with the help of a former military colleague.

The overlap between Mexico's military officers, politicians, and drug barons goes all the way back to the beginning.

Mexico hit the big time as a destination for American vice on October 28, 1919, with the passage of the Volstead Act (alcohol prohibition—in the early 1900s, the U.S. was in a prohibiting mood). In 1919, only 14,130 American tourists formally requested to visit Mexico. The following year, after Prohibition set in, 30 times as many tourists—418,735—ran south of the border to visit its bars, brothels, and casinos. According to Mónica Serrano, a professor of politics at El Colegio de México and a research fellow at Oxford University, American celebrities Jean Harlow, Rita Hayworth, and Al Jolson first glamorized Tijuana by hanging around its famous racetrack. By this time, Mexico's marijuana- and opium-smuggling economy was booming. American officials, Serrano writes, "contended that drug trafficking was simply unstoppable. The unintended impact of tighter [domestic] drug-control policies on the rise of trafficking was not addressed."

The American refusal to acknowledge that illegal trafficking and its problems are a direct result of prohibitionist drug policy also goes all the way back to the beginning.

Eventually, over the course of years in the drug trade, the major landowners and marijuana and poppy growers of northern Mexico mutated into the so-called Sinaloa cartel. Their longtime political-establishment allies became the Partido Revolucionario Institucional (PRI), which held power in Mexico for over 70 years. Their relationship, relative to today's narco-chaos, was mostly stable and calm. It is not entirely accidental that the PRI lost its political hegemony at roughly the same time (the late '90s) that the Sinaloa lost its narco-hegemony—nor is it accidental that "higher levels of violence connected with drug trafficking in the 1990s were observed mostly in those states where the political opposition [to the PRI] had gained power," as historian Luis Astorga notes.

As the Medellín and Cali cartels were being busted up in Colombia and the Mexican narco-political scene decentralized, political parties and narco-capitalists began seriously competing with each other—and killing each other in large numbers—for the first time. Meanwhile, other forces, especially the United States, inadvertently pushed cocaine smuggling from the Caribbean into Mexico.

The big problem for the narco-capitalists in Colombia, Corva explains, wasn't so much how to get the drugs north as how to get the cash back south. NAFTA was designed to dramatically increase the flow of goods, services, trucks, people, and money across North America's borders. It was also, in the words of Guilhem Fabre, professor at l'Institut Universitaire de France in Paris, "a true Trojan horse for the narco-traffickers."

The banking deregulation and privatization that came along with NAFTA made it much easier to launder money from the U.S. into Mexican banks, where it could disappear into South America. In an excellent story this June, Michael Smith of Bloomberg News detailed how Wells Fargo, Wachovia, American Express Bank, and Bank of America became funnels for sending narco-money back into Mexico. And (alleged) narco-capitalists like Carlos Cabal Peniche ran entire banks. "You don't have to pay a bank president (or a teller) to look the other way," Corva writes in his dissertation, "when you are already paying their salaries."

"The whole point of NAFTA was to free up the trade of commodities across the border," Corva says. "And that means all of them." Mexican narco-capitalists were now moving cocaine north and money south, the middlemen between Colombian producers and U.S. buyers.

Suddenly Mexico, in the midst of its own political and narco power shifts, found itself the central corridor for cocaine and cash—with much more product and money to fight over.

During the 1980s and '90s in South America, the United States invested money and political pressure in trying to strong-arm the Colombian government into fighting its cocaine capitalists. The U.S.'s position pinched a succession of Colombian leaders, as the country's ongoing civil strife with FARC and the paramilitaries kept domestic politics precarious. The Medellín and Cali organizations began assassinating politicians, cops, journalists, and judges—especially judges who supported the extradition of narco- capitalists for trial, at the insistence of the U.S. Some Colombian political leaders wondered if it was all worth it, and whether they should abandon the U.S.'s lead on drug policy. But defection from the U.S. meant "decertification" from participating in the war on drugs, which would lead to economic sanctions. It was like the exasperated wife of an alcoholic holding the liquor store hostage for her husband's drunkenness—the U.S. was making its own cocaine consumption Colombia's problem.

Ironically, Corva says, Colombia doesn't have a tradition of growing coca leaf. In the early days of cocaine, the plantations were mostly located in Peru and Bolivia, where farmers would make paste (like Diego did) and ship it to Colombia for processing into powder for smuggling through the Caribbean. The U.S. tried to combat this Peru-Bolivia-Colombia triangle with several tactics, including the Air Bridge Denial Program—in which Peruvian and Colombian military aircraft, with the help of the U.S. military and DEA, would identify and shoot down drug planes.

The program was briefly halted in 2001, after the Peruvian military screwed up and shot down a civilian plane full of Baptist missionaries, killing two American citizens. In 2003, George W. Bush approved the reinstatement of the program, though the U.S. Government Accountability Office (GAO) said its results were "mixed." Between 2003 and 2005, the Colombian Air Force "located only 48 aircraft out of about 390 suspicious tracks pursued; and the military or police took control of just 14 aircraft—four were already on the ground. Only one resulted in a drug seizure." During that time—again, according to the GAO—the U.S. had devoted $68 million to the program. And for that $68 million, they got one single drug seizure.

Imagine what that $68 million could have done for drug-treatment programs in the United States.

Aerial coca-eradication programs—spraying herbicide—also showed some results but poisoned big swaths of the countryside. "Where they've sprayed, nothing lives there, nothing grows there," Diego said during our conversations about his time on the cocaine farms. "The water, the rivers, they die. It's killing everything and the people who live there."

"From cancer?" I asked.

"They don't even live long enough to get cancer," he said. "They die that same year. It's really, really strong poison."

The results of spraying were uncertain at best, not least because international estimates of how much coca is being grown are, in Corva's words, "complete shit—they're all political." For example: The difference between U.S. and UN estimates of how many hectares of coca are under production in any given country can vary by as much as 50 percent.

Aerial eradication and the Air Bridge Denial Program met with "mixed" success—or abject failure, depending on whom you ask. But other factors conspired to shove cocaine smuggling out of the Caribbean and into Mexico: successful DEA and Coast Guard campaigns to interdict drugs in the Caribbean, the breakup of the Medellín and Cali cartels in Colombia (with help from the CIA and the U.S. military's elite and secretive Delta Force), and the subsequent fracturing of the cocaine-production networks in South America. The big cartels specialized in cocaine-hydrochloride production and smuggling. Once they had been broken, they left behind small producers who could make powder, but nobody who had the resources to pull off major international smuggling operations besides the narcos in Mexico.

Then there was NAFTA. Just as the deck was being reshuffled for narcos and politicos and Mexico and Colombia, the North American Free Trade Agreement showed up and suddenly made everything so much easier.

Ulysses S. Grant was a famous cocaine addict. He took daily doses to kill the pain of his throat cancer so he could complete his Personal Memoirs—Mark Twain's new-at-the-time publishing company expected the memoirs to be best sellers. Twain desperately wanted Grant to finish, and some have speculated that he urged Grant to take cocaine so he could work through his pain. (In one of Twain's memoirs, he writes of "a longing" to go to the Amazon and "open up a trade in coca with all the world"—he wanted to be the first cocaine baron to the United States, but ran out of money in New Orleans before he could make the trip south.)

One hundred and ten years later, three men sat down to sign the North American Free Trade Agreement: Canadian prime minister Brian Mulroney, United States president George H. W. Bush (it wouldn't go into effect until Bill Clinton was in office), and Mexican president Carlos Salinas (whose brother would later be accused of being heavily involved in narco-capitalism and convicted of orchestrating a hit against their brother-in-law).

A few policymakers—including Ross Perot and Phil Jordan, one of the DEA's experts on Mexico—sounded the alarm about how much NAFTA was going to help the drug trade. But the Clinton administration, which struggled with organized labor, business, and nearly everybody else over NAFTA, gave Jordan a gag order when it came to how free trade might help the narcos. "We were prohibited from discussing the effects of NAFTA as it related to narcotics trafficking, yes," Jordan said in an interview with ABC News in 1997. "For the godfathers of the drug trade in Colombia and Mexico, this was a deal made in narco heaven."

Peniche, one of the (alleged) narco- capitalists who started buying up banks after NAFTA to make laundering easier, is another example (like old Colonel Esteban Cantú Jiménez) of how narcos and Latin American governments are fully integrated. Peniche contributed tens of millions to the PRI party, then had to flee Mexico after the government seized his Union Bank and interest in the Del Monte Produce chain.

After NAFTA, the Mexican narcos controlled both ends of the trade, allowing them to impose steep taxes on the Colombians. (The chaos and violence in Mexico does not make smuggling any cheaper.) Cutting cocaine with levamisole, which behaves like cocaine (its color, the way it responds to a bleach test), could be a strategy from Colombian producers to recover the profits they've lost to Mexican smugglers. It's possible that the Mexican narcos think they're smuggling pure product and have no idea (or don't care) that levamisole even exists.

More importantly, NAFTA revealed the fundamental contradiction of the drug war: The goal of NAFTA (and the U.S. banking deregulation that came with it) was to free the market and minimize state control. The goal of the drug war is to restrict the market and maximize state control. The result? Over 30,000 murdered in four years in Mexico alone, with all the grief and suffering that follows for each of those 30,000 families.

"The history of the last 20 years of the cocaine and heroin trade shows how much mobility there is in cultivation and trafficking," Professor Reuter told Gil Kerlikowske, director of the Office of National Drug Control Policy (and former chief of police in Seattle), earlier this year in front of a subcommittee of the House of Representatives. "What we do has a predictable effect. When we pushed down on trafficking in Florida, that led to increases in Mexico. The evidence is striking that all we are doing is moving the trade."

While Kerlikowske worked in Seattle, he was moderately progressive on drug issues—but as the nation's drug czar, he is legally bound (by the Office of National Drug Control Policy Reauthorization Act of 1998) to "oppose any attempt to legalize." Even if the nation's drug czar believes that prohibition has been a failure—and it has—it's against the law for him to say so.

This state of affairs cannot last. It is fundamentally immoral—if you consider the protection and betterment of human life a moral good. U.S. drug policy has maximized the murder rate in Latin America while defaulting on the problem of addiction and treatment, opting instead for incarceration, in the United States.

By embracing a contradiction, we have created the worst of both worlds.

Levamisole test kits— so you can test any substance for the presence of levamisole—will be available next week from the U-District needle exchange (the People’s Harm Reduction Alliance), The Stranger, DanceSafe, and other partner organizations. Look for more information on how to obtain a kit and how to use it in next week’s issue of The Stranger.

This article has been updated since its original publication.