Good Hunting: An American Spymaster's Story

by Jack Devine

In 1992, Roeber led one of the most controversial estimates on heroin production. Many of the law enforcement agencies and the military wanted to downplay the spike in poppy production worldwide, since a high number would mean they might be forced to commit more of their stretched resources. Again, it is worth remembering that at this time counternarcotics remained a top national security priority. The world was at peace, and terrorism was not the overarching problem it would become in the late 1990s.

  Roeber was a strong leader in any meeting and knew how to press hard on people who came to the table with shoddy data and analysis. This made him an extremely valuable ally for me but did not endear him to those who felt the sting of his penetrating questions. The 1992 CNC data clearly showed a substantial increase in heroin production, but the rest of the intelligence community wanted no part of this conclusion. Some of the participating agencies became so vexed by the numbers that the level of dissent had reached DCI Gates. He was, in turn, pressing John Helgerson, the CIA’s top analyst, to see what all the fuss was about. Helgerson rarely became involved in counternarcotics matters, but on this occasion he set up a special meeting to tell me that many in the intelligence community were accusing Roeber of the mortal sin of the “politicization of intelligence.” The charge made no sense. I explained the situation to Helgerson and assured him that Roeber and the CNC were just doing their jobs.

  In the end, the estimate produced by the intelligence community went against our views, which were voted down by representatives from other agencies. The assigned non-CIA writer drafted a vanilla product. To his credit, Gates signed off on a CNC dissent of the key judgments in the heroin estimate. We were proud in the CNC that we had dissented. It epitomized what I believe to be key to good analysis: independence of thought. As it turned out, heroin continued to grow as a worldwide problem and remains a major problem today in Afghanistan, where it has greatly complicated our efforts to combat the Taliban there.

  It was around this time that Pablo Escobar “escaped” from prison. Government authorities tried to move him to a more rigid prison, but he was tipped off and escaped under highly suspicious circumstances. Thus, the hunt for Escobar began. He was hounded from two sides. On one side, we intercepted his communications. And four days after his escape, U.S. military forces arrived in Colombia to train the Colombian police and military in his capture. On the other front, Escobar’s rivals in the Cali Cartel financed the creation of “Los Pepes” (Los Perseguidos por Pablo Escobar), led by the infamous right-wing paramilitary officer Carlos Castaño. He led the group of vigilantes who helped hunt down Escobar and were thought to have ties to the Colombian National Police. The CIA didn’t run Los Pepes and didn’t give it any support. The Cali Cartel obviously benefited the most from the activities of Los Pepes, but the Colombian military and police did, too. One colleague said, “Thank goodness for Los Pepes. We didn’t support them, but they provided a service that we couldn’t do. These people had the Medellín Cartel scared to death.” The U.S. ambassador to Colombia at the time, Morris Busby, said that “Los Pepes ended up being an excuse for any and all killings,” and we made it clear to President Gaviria that we would pull our support if we learned the Colombian government was behind them. Jay Brant, an experienced Latin America case officer, said, “As far as I know, no U.S. government organization had any contact with Los Pepes. They were a bunch of terrorists.”

  There was, in fact, a standing rule that if it was discovered that any of our assets were involved in a breach of human rights or a violation of U.S. law, the Agency would have to cut off contact and/or turn them over to Justice. Brant remembered a highly legalized environment in the CIA at the time and stressed that there was great sensitivity about human rights. The Agency’s lawyers were all over it. There was some concern at the State Department, the Pentagon, and even in the Agency’s Directorate of Intelligence that some of the intercepts and other technical information we were passing to our liaison partners in Colombia were ending up in the hands of Los Pepes. When information is passed to legitimate liaison counterparts, there are never guarantees that some of it will not be siphoned off. All you can do is put down markers with liaisons and try to keep an independent check on things. To the best of my knowledge, we were never aware of information ending up in the wrong hands in Colombia, but I can’t really prove it. Los Pepes was quite effective in countering the violence of the traffickers at first. But as time went by, it, too, became corrupt and indiscriminately violent. It certainly was not our creation or our tool.

  As I approached my second anniversary at the CNC with Escobar still on the loose, Tom Twetten, my old friend and mentor, who had risen to deputy director for operations, asked me if I would like to accompany him on a trip to India, He thought it would be worth asking the Indians to cooperate with us on counternarcotics. George Crile, in Charlie Wilson’s War, notes that Gust Avrakotos did not hold Twetten in high regard. Avrakotos not only missed Twetten’s essence but greatly underestimated his intellect, operational acumen, and tough-mindedness. While Twetten may have appeared mild-mannered, he had that exceptional blend of political acumen and sharp operational skills. I never sat in on a meeting with him and foreign dignitaries or U.S. government officials when he wasn’t constantly assessing the situation or analyzing the next move on the board. He also was not afraid to ask the tough question: “Will you work for me as a spy?” We had worked closely together going all the way back to Iran-Contra and Afghanistan.

  On the trip to India, Twetten, an old India hand, took me to “Old Delhi” to see how the other half lived in the poorer sections of the city. He convinced two rickshaw drivers to ride us through the rougher part of town. At six five, I felt rather silly being pulled along by a bony, one-hundred-pound young man, but it was the only way to do it. On foot would have been impossible. Twetten particularly wanted to show me a “factory” that bound books with homemade paper liners and gold gilt. He was an avid collector of antique books, and after retiring he would study bookbinding and set up an antiquarian book company in Vermont. When we got to the site, I was amazed to see what passed for a “factory”: it consisted of a dozen Indians sitting on the floor, hammering out the artwork. I felt I was somehow intruding, even though they seemed immensely proud of their work. It reminded me in no uncertain terms what a privileged life most of us in America live. I later sent half a dozen personal books to them for rebinding, which I cherish to this day.

  When we settled back in New Delhi, we relaxed over a few beers and very spicy food. Inevitably, we returned to talking about the business, and Twetten said that he wanted to run by me a couple of positions that would be coming open soon, to see if I was interested in either of them. The two jobs he was looking to fill were chief of the Latin America Division and chief of the Counterintelligence Center staff. Which one would appeal more to me? He said he couldn’t tell me the whole story but made it clear that there was something really significant happening inside the Counterintelligence Center, something he described as “critically important.” He knew what I didn’t—that the mole hunters were closing in on their prey. It was like the moment, fourteen years earlier, when I was offered the choice by Ray Warren between being a base chief or a station chief in Latin America. This time, it took me about as long to decide.

  TEN

  The Rooster and the Train

  Washington/Haiti, 1992–94

  Latin America was home to me, the part of the world where I’d served five overseas tours—three as chief of station—and overseen the drug war as head of the Counter Narcotics Center. When Tom Twetten gave me the choice of running either the Counterintelligence Center or the Latin America Division, I chose Latin America in an instant, with a sense of pride. Ever since I was a young operations officer in Chile, I had wanted to be division chief. And Latin America was the obvious choice; after all, I’d spent the first half of my career there. After Chile, I had traveled frequently on assignment to Mexico and other countries and then served as station chief in Argentina befo
re returning to headquarters’ Near East Division.

  The job that Twetten was offering me was the rough equivalent of a three-star general in the military. It was a position that required a new set of skills and offered challenges that tested my mettle and political agility. There are only two jobs above division chief in the Directorate of Operations: the deputy director for operations and the associate deputy director for operations. The chief of the Latin America Division could influence Agency policy and operations in the region and have a say at the DO board level. And two of the men whose leadership I most admired in the Agency, Ray Warren and Nestor Sanchez, had both held the position with distinction. Warren was station chief during my first tour in Chile and remained a mentor throughout my career, counseling me, encouraging me, promoting me. Some of us referred to him as the “Gray Fox,” for his wavy gray hair. He was six four and could have easily passed for an ambassador or Fortune 500 executive. He was measured in his demeanor and thoughtful and balanced in his deliberations. When, as chief of the Latin America Division, he sent me off on my first assignment as station chief, he gave me only a few words of caution: “Remember, most people in your position break their pick on the ambassador. Watch that.” In other words, don’t tangle with the ambassador on trivial issues and, even more important, develop a close collaborative relationship with him or her. Ultimately, you are expected to work out problems locally and not bring them back to headquarters for adjudication, which rarely works to a chief of station’s long-term advantage. It was invaluable wisdom that applied well beyond the embassy, to relations with liaison partners in foreign intelligence services, Congress, and other U.S. agencies. Warren was a man of impeccable judgment.

  Nestor Sanchez succeeded him as chief of the division. He was yet another Agency legend. Born in New Mexico and fluent in Spanish, Sanchez, as a young case officer, ran a Cuban agent named Rolando Cubela, point man in the CIA’s plan to assassinate Fidel Castro. Sanchez and Desmond FitzGerald, one of the founders of the Agency’s Clandestine Service, met with Cubela in Paris in early November 1963 to talk about the latest Castro assassination plan. At the meeting, Sanchez gave Cubela a ballpoint pen made by the Technical Services Division that was actually a hypodermic needle filled with Black Leaf 40, a poison. As the meeting broke up, Sanchez and FitzGerald were informed that President Kennedy had been assassinated in Dallas. As a consequence, the operation was aborted.

  Nestor, a consummate operator but also good at Agency politics, rose steadily through the Agency. He became the top-ranking Hispanic officer in the DO at the time and had a great feel for Latin American political and operational life, which he combined with passion and humor. He was an especially effective recruiter, exuded leadership skills, and was broadly liked in the division. With almost no introduction in 1980, he asked me to go abroad. The assignment was a plum, and there were many people in line in front of me. Not only that, I was only a GS-14 in the civil service system, and the work I would be doing was at a GS-16 level, roughly the equivalent of a one-star general in the military. At the time, there was a rule that you couldn’t jump two grades for a new assignment, so Nestor said he was going to downgrade the post temporarily so I could do the work—not a common occurrence and no doubt the cause of some hard feelings among those waiting in line for a senior posting.

  I assiduously observed Warren’s counsel on maintaining harmonious relations with the ambassador, and this cordiality was broken only once, when Sanchez visited. I brought him down to pay a courtesy call on the ambassador, and in a flash, both of them were agitated and practically yelling at each other. I wasn’t sure what exactly had set it off, but I quickly hustled Nestor out and back to my office. But for the next hour, I received calls from the ambassador telling me in no uncertain terms what he thought Nestor needed to know, and Nestor in turn instructed me on what to tell the ambassador. I was amazed that two exceptionally bright and talented professionals could push each other’s buttons so easily and dramatically. The ambassador, to his credit, did not hold this against me. It did, though, earn me some points with Nestor, for dealing with what he perceived to be a very difficult ambassador. The truth, of course, is that it took no especially adroit effort.

  I saw Nestor again when I was back at Langley for a visit. I had lunch with him and a colleague who was running another station in Latin America, and he told us both that we should be putting in for the station chief’s job in Tegucigalpa, Honduras. We both laughed. Tegucigalpa was a three- or four-man station; to say you were going to Tegucigalpa was like saying you were being sent to Siberia. Sanchez was visibly agitated by our reaction. “You know how many times I’ve been upstairs to talk about your stations in the past week? None,” he said. “You know how many times I’ve been up there to talk about Tegucigalpa? Ten times. It’s going to become the biggest station in Latin America.”

  That was my first introduction to the brewing operation to arm and train the Contras. Nestor’s skepticism about certain aspects of that operation seemingly cost him his job at the CIA. Casey is said to have told others that Nestor wasn’t “bold” enough for him. The operator who, as a young officer, ran the Cuban agents the CIA was counting on to assassinate Castro wasn’t a big enough risk taker? No, the real problem was that Nestor apparently did not fully share Casey’s enthusiasm for the Contra program. Here, again, we see the line between good covert action and bad covert action. Nestor reportedly thought using exiled members of the deposed Nicaraguan dictator’s National Guard was not the way to go. He questioned having other Latins training the Contras in Honduras. It likely was because of his resistance to parts of the operation that Nestor was eased out of the CIA, to become assistant secretary of defense for Latin America, a position he filled admirably.

  Like Warren and Sanchez, I was not intimidated by the challenge of running the Latin America Division, perhaps because I didn’t fully appreciate what lay ahead. The country was banking on a peace dividend, having won the Cold War, and looking to a new leader after the U.S. presidential election in November 1992. Most officers at the Agency do not wear their politics on their sleeves; for a civil servant, it is unprofessional. I suspect most were silently rooting for President George H. W. Bush to beat Democrat Bill Clinton that fall, not so much for Bush’s politics, but because he had been a popular DCI, from 1976 to 1977, and was respected as a serious public servant himself, with extensive international experience. But the CIA workforce took Clinton’s election in stride. There was no gnashing of teeth about what it would portend. We were more concerned about the general depreciation of CIA stock because of the end of the Cold War, and what the peace dividend would mean, with both parties in Congress deemphasizing intelligence.

  Clair George, the former operations director who had sent me to Rome, was convicted in the Iran-Contra scandal in December for reportedly misleading Congress six years earlier about what he knew about a cargo plane shot down in Nicaragua. Notwithstanding a Christmas Eve gift of a pardon of George from the lame-duck president, this was not, by any stretch of the imagination, a good time to be taking over the division, or any division. With the Cold War over, the focus had turned to transnational threats, such as counternarcotics and, to an extent, counterterrorism—with good reason. On January 25, 1993, a Pakistani national named Mir Aimal Kansi opened fire with an AK-47 on CIA employees stopped at a traffic light on Route 123 in Langley, just outside headquarters. While not initially considered in the context of the al-Qaeda–inspired terrorists attacks of today, the Kansi attack was particularly stunning to those of us working at Langley. Agency officers Lansing Bennett and Frank Darling were killed in their cars, and three others were wounded in the attack. That year the general assumption at headquarters was that Kansi was a lone deranged Pakistani. No one linked his singleton action in late January to the bigger plot behind the truck bomb detonated below the North Tower of the World Trade Center a month later, or to any fundamentalist Islamic terrorist organization. But like Ramzi Yousef, the mastermind of that 1993 bombi
ng in lower Manhattan, Kansi said in 1997, when captured by the FBI in Pakistan, that the Langley shootings were in retaliation for U.S. policies in the Middle East and for how Muslims were treated by the CIA in that region. Though Kansi could not be connected to any group in Pakistan or elsewhere, in hindsight, it’s clear that he, too, was a forerunner to Islamist terrorism à la al-Qaeda. After his capture, Kansi was rendered back to the United States, where he stood trial for the murders. He was found guilty and executed by lethal injection in 2002.

  * * *

  The Latin America Division was on the third floor of headquarters. I had the largest office of all the division chiefs, because it once had been the deputy director of operations’ office, before the DDO was moved up to the seventh floor. It even had its own bathroom, a novelty for an office in that building. There were numerous stations in the division reporting to me. I had a division staff of about twenty, including reports officers, counterintelligence officers, a lawyer, and a human resources officer, but all field operations were run out of the individual stations. I brought with me my deputy from the Counter Narcotics Center, Marty Roeber, and made him deputy of the division, because I needed someone I could rely on and trust to pay very close attention to the day-to-day operational and intelligence details. It had been natural for me to choose Roeber, a career Latin America analyst from the Directorate of Intelligence, as counternarcotics deputy, since the CNC resided in the DI. But my decision to bring him over was fraught with peril and risk—for both of us. It is hard for an outsider to appreciate how rare a move this was, given the closed nature of the DO. To many in the directorate who had waited years for this job to open up, it appeared that I was bringing in an outsider unfamiliar with operations. But Roeber was a great partner, the most knowledgeable person I knew on Latin America in the Agency. He knew Latin America in a way that was different and more substantial than I did, even though I had lived there, and since counternarcotics remained a huge account in the division, it was a natural transition for us. Indeed, the overlap between the CNC and the Latin America Division was great, given that much of the money being channeled to the region in the early 1990s was for the purpose of counternarcotics. Marty helped immeasurably with the continuity and with coordination. He was also a bulldog in the Latin America office, and we worked extremely well together. I normally came in at 7:00 a.m. and would leave at 7:00 p.m. Roeber was there when I arrived and when I left at the end of the day. It must have been the Dr Pepper, which he drank incessantly, that kept him awake.