The CIA received very good information on at least the initial stage of the Soviet strategic debate. An intelligence source code-named Veil provided access to Russian general staff material and other data. Renegade CIA officer Aldrich Ames began to spy for the KGB that summer. Source Veil dried up.

  It is not clear how much the Soviet plan should be attributed to Gorbachev. Morton I. Abramowitz, then head of the Bureau of Intelligence and Research (INR) at the State Department, believes the Russian change of course was more like Jack Kennedy’s Bay of Pigs—already in motion before the leader came onto the scene. Most interestingly, the CIA had already anticipated this Soviet strategy—in a major analysis by the DI in May. This five-year retrospective discussed the war’s cost and Russian prospects. While Moscow had incurred substantial costs, the CIA felt Russia saw the price of remaining as more than the international costs of getting out. Rebel casualties (40,000) were estimated as far less than those of the Soviets and DRA (25,000 Russians, including 8,000 dead, plus 67,000 losses for the Kabul regime). In a guerrilla war, casualties to the rebel side usually greatly outnumber those for the counterinsurgents.

  Langley’s experts were not universally admired. But in this important survey the CIA had it exactly right: the agency did not believe Moscow foresaw early victory and the Russians were unlikely to make important progress during the next two years. The CIA expected Soviet reinforcements—five thousand to ten thousand (the actual would be closer to twenty thousand), primarily specialized troops—exactly what the Russians did. The American analysts did not think this could alter the outcome.

  If the CIA erred in this analysis it was on the pessimistic side. The report noted, “We cannot rule out a more serious deterioration of the Soviet position in Afghanistan than we have estimated,” and if that happened it “would result in a move not toward a political settlement but toward an expanded Soviet military commitment and a wider war.”

  An increase of 50,000 troops (three divisions) would require Soviet mobilization and an expanded logistical base and take many months to accomplish. It would also utilize the kinds of troops the Russians had found the least effective. An increase of 100,000 or 150,000 soldiers might enable the Russians to clear the cities and hold substantial portions of the countryside, while one of 200,000 to 400,000 offered a possibility of inroads against the rebellion—but the CIA doubted that that could be sustained. Moreover, “We believe this view underestimates insurgent morale and military performance . . . and exaggerates Moscow’s effort to Sovietize the country.”

  The DI followed up in October with a study of the rebels, concluding that the divisions among the tribes were such that no near-term victory should be expected. Thus the CIA projected a stalemate.

  The major mujahedeen initiative in 1985 came in southeast Afghanistan—an effort to capture the city of Khost. The campaign required coordination among the warlords, but personal and religious differences made cooperation impossible. Brigadier Youssaf attempted to knock heads together but failed. The Soviets, aware of the threat, mounted two major offensives. That led to the battle of Zhawar, a key receiving area for supplies from Quetta. The muj had developed it into a major base and training center, using bulldozers and explosives to open tunnels. Zhawar became a showcase for the rebels, with a hotel in one cave and others with a mosque, a communications center, repair shops and garages, an infirmary, and arms depots. The base had a permanently assigned unit, perhaps the first such formation of the resistance, equipped even with a captured Russian artillery piece and a couple of tanks manned by DRA defectors. During offensives the unit was hard-pressed. At the time of the second attack many rebel leaders were absent, on the Hadj, and one key commander died in the collapse of a tunnel. The Russians plus DRA forces captured most of Zhawar but could not hold it. The base remained in rebel hands and years later passed into those of the Taliban, until 2001 when U.S. troops in the war on terrorism finally took the place.

  Supplies of rifles and missiles would not be the only sore points. Afghan task force paramilitary specialist Michael Vickers emphasized the mix of weapons and looked to where the resistance should be in a few years. Whether the rifles were Lee-Enfields or AK-47s seemed less vital than whether the warlords could supplement their firepower with heavy machine guns, mortars, and other weapons. Similarly a mixture of machine guns with the Oerlikons and SAMs provided the right set of capabilities. In fact the Russians would lose more aircraft and helicopters to machine guns than to missiles. The object was to drive the aircraft to high altitudes, robbing them of effectiveness, not to down planes. That was accomplished.

  Other acrimonious disputes also arose. One concerned mines—first the weapons themselves, later mine-detection equipment. Both rebels and Soviets laid many thousands of mines during the war, and both progressed to new-generation nonmetallic versions. Later, as the rebels began assaulting defended positions, mine clearing became an issue. Around 1983, prodded by the war, the Pentagon began developing a device called “Lightfoot” to detect nonmetallic mines so they could be disarmed. Elie Krakowski recalls that the design work was completed in just a few months, but the military then had a devil of a time getting CIA to accept the product. The agency demanded lengthy tests. Lightfoot reached the field only in late 1985. Agency officers say the detection devices never worked properly. Krakowski insists the method was so effective it was used by the United States in the Gulf War. Another issue would be anti-tank weapons. Quite well supplied with armor, the Russians relied on tanks in every battle. Until 1987 the muj had only rudimentary means to fight them. Then the CIA brought in Italian-made anti-tank missiles, a measure that probably cost the Russians more than the Stingers. There were so many “ideal” weapons introduced at various times that one CIA officer took to calling them silver bullets.

  Langley again revamped its high command for the secret war. John McMahon retired in February 1986, his post going to Robert Gates. Gust Avrakotos, the task force chief, an inside agitator for the victory camp, had been protected from Clair George by his connections with associate DDO Ed Juchniewicz and Congressman Wilson. But as CIA budgets increased radically, congressional backing became ever more important, so Director Casey built his own bridge to Wilson. Then Juchniewicz stepped down. Bert Dunn, his successor, had no dog in this race, and Avrakotos was banished to the Africa Division. In a significant breach of Langley’s tribal cultures, the new Afghanistan Operations Group chief, Jack Devine, came from the Latin America crowd by way of a brief stint running the Iran Branch of NE. The division itself got a new baron when Tom Twetten succeeded Dunn.

  Finally, the station chief at Islamabad moved on. At Langley one day in May 1986 Milton A. Bearden found himself in Clair George’s Seventh Floor office. He hardly needed the weekend to decide. Milt Bearden became CIA’s new field commander for the Afghan project. A Texan, Bearden had the advantage of coming from Charlie Wilson’s home state.

  About the time George appointed Bearden station chief, the Russians began their summer offensive. They concentrated on several provinces, tried to clear Khost again, and fought a new battle for Zhawar. The operations were standard, the results insignificant. The Russian military had failed. In Moscow Chairman Gorbachev came to the end of his patience. In a July speech he announced a limited withdrawal, and six Soviet regiments left the theater. Washington made an issue of claiming the Russians had offset the pullout by sending new reinforcements, but in fact the Soviet contingent in Afghanistan diminished by fifteen thousand troops and enough equipment for a mechanized brigade. Gorbachev had told mediator Diego Cordovez that the UN mission would ultimately be successful. Washington remained skeptical. In the summer of 1986 DRA officials prevented any progress in the negotiations, but Karmal, politically bankrupt, resigned. His successor Najibullah gradually softened. That December Najibullah went to Moscow where Gorbachev told him that Russia would be out of Afghanistan within two years. Deprived of source Veil, Washington needed that intelligence but did not get it.

  Milt Bearden arrived in July 1986, just as the ISI was beginning to turn out rebels trained to use the Stinger missiles. The CIA had doubts about the effectiveness of the SAMs in muj hands, but the matter settled itself. In early September Stingers were used for the first time—and gunners hit three of the four helicopters targeted. Bill Casey, ecstatic, wanted to fly to Pakistan right away, but instead Bob Gates made the trip to familiarize himself with the front line in this secret war. During that visit the Iran-Contra affair began, calling Casey’s entire enterprise into question. Through the controversy the Afghan project stood out as the jewel among Casey’s wars, but beneath the surface it had begun to wind down.

  “Uncle Milty,” as Bearden was known, carried on. One key issue with the Stinger had been “accountability,” which meant tracking the SAMs so they did not fall into the wrong hands. The CIA devised techniques for this, never perfect, but really the problem applied to every weapon down to the AK-47 rifle. Beginning in 1985 deliveries multiplied by a factor of ten. Sixty thousand tons of CIA weapons were floating around Pakistan. Bearden adopted a simple rule: if prices in the bazaars were much higher that what the CIA paid, that meant the market had not been flooded with rake-offs from agency shipments. He stayed cool.

  Bearden never solved the problem of the warlords. He could have had some impact there—Uncle Milty had originally impressed Casey as station chief in Nigeria, then the Sudan, both nations riven by tribal differences, and he should have been sensitive to this issue. Bearden made some moves toward creating a rebel central command, only to be undermined by the Pakistanis. He left it at that. Field officers pushed for aid to Massoud, whom all agreed had the best troops and the most finely honed political sense. But that meant tangling with the ISI. Of course, with the huge expansion of aid the CIA might have demanded a revised distribution formula, but Bearden shied away from that too. He felt his orders were to beat the Russians, not quibble over which warlords were the most fundamentalist. Citations of the percentages of weapons given to rebel groups from the early and late war periods indicate shipments to the more radical groups actually increased. Bearden focused on sharpening the rebels’ fighting edge. He had major differences with Brigadier Youssaf, who left ISI not long after the CIA man came to Islamabad.

  In one area, that of the “Arab Afghans,” Bearden had no control whatever. Early on the CIA had considered—and rejected—recruiting a corps from other Arab lands to fight in Afghanistan. Langley thought this unwise, as proved to be the case. But the secret warriors could not prevent Muslims getting the same idea, and it became a form of Arab radical chic to go fight the Russians in the mountains of Allah. There is no data on how many Arab Afghans took part in the war, but estimates hover around three thousand to five thousand. Fund-raising in Arab lands also became a source for the warlords, and by Soviet estimates brought in as much as $250 million annually by 1987. Much of this cash went to Peshawar and Quetta for medicine, food, and general assistance, and to fund the Arab Afghans themselves.

  The young Saudi engineer Osama bin Laden, one of the Arab Afghans, began building an organization in Peshawar at this time. There is conflicting evidence on whether bin Laden actually fought in the war—perhaps in one battle is the consensus. There is little reason to doubt CIA assertions that it never helped him during the war. The secret warriors had no identifiable role with the Arab Afghans, and bin Laden was no more than a minor character at the time.

  In the spring of 1987 Langley again changed its task force chief. Frank Anderson, a card-carrying Middle East warrior, became Bearden’s new boss. Anderson spoke Arabic and had served in Yemen, Beirut, and North Africa, most recently as station chief in Morocco. He tried to keep up momentum with new weapons every few months, starting with the Milan anti-tank missiles. Anderson often felt like a supply sergeant. Congressional overseers once asked him what Washington could expect in terms of support from a future muj government, to which Anderson replied, “gratitude in the Afghan’s dictionary is gonna be found somewhere after gimme and gotcha.”

  In April 1987 the muj mounted attacks into Russia. Moscow protested to Washington. In a maneuver worthy of CIA’s peregrinations on Central America, DDO Clair George phoned Bearden to certify that the agency had had no role or knowledge of the attacks. The station chief probably did not know that exactly this sort of operation had been advocated by the Pentagon’s Elie Krakowski in 1983 and by Bill Casey after that. Casey is said to have given ISI data on Soviet bases drawn from satellite photos. A year after the Russia attacks the KHAD or KGB struck back, blowing up a huge ammunition dump in Rawalpindi, Pakistan. Meanwhile a new director, Gen. Hamid Gul, took over ISI. The greatest change occurred in August 1988 when a C-130 crashed while carrying generals Zia and Akhtar as well as U.S. Ambassador Arnold L. Raphel. Pakistani elections then brought to power Benazir Bhutto, daughter of an earlier strongman. Once imprisoned by the ISI, Bhutto held them in little esteem. Her ascendancy increased prospects for negotiations, which began to move quickly. Diego Cordovez shuttled from New York to Geneva, Moscow, and Pakistan.

  Frank Anderson felt his CIA project had no impact on the talks, but it was not in America’s interest to permit the Russians to withdraw from their quagmire intact. With victory the goal, CIA budgets redoubled. Close to $300 million in 1984, by 1986 they stood at $700 million—and that funded only the CIA contribution. Adding Saudi money made the totals enormous. Soviet intelligence toted up the numbers in 1987 and came up with almost $2 billion. Humanitarian aid added close to a billion dollars through the decade. Estimates of total CIA spending on the Afghan project through 1991 (the United States gave the rebels a golden parachute) range up to $9 billion.

  Robert Gates remarked at his 1991 confirmation hearings for CIA director that he would have grave doubts about the efficacy of covert operations, save for Afghanistan. That leads to the heart of the matter. Afghanistan supplanted Laos as the biggest paramilitary affair in CIA history up to then—and, more, an apparent success. By 1988 the Soviet military had stalled, its air effort neutralized by the mujahedeen’s SAM weapons. Russian troops could punch through but achieved nothing. In UN-mediated negotiations the Soviets agreed to a military withdrawal and carried it out. At 11:55 A.M. on February 15, 1989, Gen. Boris V. Gromov stepped back inside Russian territory, the last Soviet soldier to leave Afghanistan. Milt Bearden and Frank Anderson, breaking CIA standing orders not to cross the border, went to have their pictures taken at the same bridge where Gromov had exited.

  But the Afghan War continued. Najibullah’s Communist regime fought on, and both the CIA project and Russian aid halted only in 1992. The farewell gift of a generous allotment of arms to the factions—which CIA justified on grounds that the United States had a moral responsibility from mobilizing the tribes—served to fuel continuing civil war. A coalition government under a general of the former regime lasted about a year until the warlords fought again, over fundamentalist primacy in addition to political power. Ahmed Massoud and Gulbuddin Hekmatyar became enemies. Four battles for Kabul occurred, three of them after the Soviet withdrawal. Elections to be held in 1993 never took place. In the midst of the civil war a different, even more extreme fundamentalist force arose. The Taliban conquered Kabul and most of the country. Massoud—now allied with Najibullah’s former general Dost—fought on. Before long they got help from the CIA and the Russians. The details of their war spill over into the story of the war on terrorism, but the facts make a mockery of pious arguments that the United States intervened to support democracy in Afghanistan, and the evaporation of American aid after the Russian withdrawal shows Washington’s proclivity to discard pawns that outlive their usefulness.

  It is not possible to conclude, as some like to say, that the Afghan War brought about the fall of the Soviet Union. The war was increasingly unpopular and costly in both blood and treasure, but its burden cannot be separated from pressures of waste, inefficiency, and cultural ennui. The Afghan War paled in comparison to Soviet economic and nationality problems. Ironically the immediate catalyst of the Communist downfall came with Gorbachev’s attempt to craft a new union and the secession of the Baltic states, those theaters of long-ago CIA projects since terminated.

  The Afghan operation also cost the CIA and the United States. Those at the CIA and elsewhere who worried that Stinger missiles given to the muj would show up on world arms markets proved exactly correct. At the height of the Iran-Contra affair, intermediaries for Teheran told CIA operative George Cave how pleased they were that the United States had begun supplying Stingers since they now had ten of them. As early as 1988 U.S. forces in the Persian Gulf, inspecting a motorboat belonging to the Iranian Revolutionary Guards, found two Stingers. During the CIA operation a thousand went to the Afghan guerrillas. About 350 are reported to have been fired. No doubt some were expended in training and others defective, but by the spring of 1994 estimates were that Afghan warlords still had at least 400 SAMs. Indeed the Stinger became such a prestige item that the warlords took to parading them on national holidays. Appeals to return unused missiles were useless. In 1994 the CIA began a new $65 million covert effort to buy back the same missiles it had given the guerrillas, apparently an enhancement of a CIA buyback program already under way. American demand fueled a surge in prices, fraud, and further corruption. Missiles that cost the U.S. army $35,000 each were being sought for $50,000 or $100,000, then $150,000, with limited success. Agency proprietary aircraft flew to Middle Eastern cities to bring back the SAMs thus acquired. It is reported that the CIA bought back perhaps 200 Stingers. A weapon that had a positive effect in one paramilitary campaign became a long-term security and terrorism headache.