05 March 2015

Review of Bruce Riedel What We Won: America's Secret War in Afghanistan Brookings Institution Press (2014)

What We Won
This book has been very well-received in most serious circles.  The author, who advised President Barack Obama on Afghanistan policy from the beginning1, has a startling message: Western intelligence agencies, including the American CIA, Britain's MI5, and others, did help defeat the USSR by supplying the mujahidden rebels in Afghanistan; but they were certainly not responsible for the appearance of groups like the Taliban or al-Qaeda.  The main error of the USA in supporting the mujahidden was to abandon them after February 1990.  These are the main points, and I would like to explain why they are implausible.

A few points of clarification before I begin: one, I am NOT a member of any intelligence organization.  I have no security clearance; I'm only reviewing a book based on publicly available information.  Two, the ISI is the Directorate of  Inter-Services Intelligence  of the Pakistani armed forces, which was the principal interlocutor of the CIA during the secret war in Afghanistan.2  Three, the USA has been a major supplier of foreign aid to Pakistan since independence (with the exception of a brief interlude, which I'll mention below)3; in addition, or separately, the CIA funneled money directly to the mujahidden4). So when writers speak of the US "abandoning" Af-Pak" after the Soviet withdrawal, there's reason to question this narrative.  I'll discuss that below also.


It is frequently alleged, especially in Pakistan, that the USG abandoned Pakistan after the Soviet-Afghan War.  In some senses, this is a fair allegation: in October 1990, after six years of looking the other way, the White House "noticed" Pakistan's nuclear weapons program and drastically all forms of aid. Diplomatically, the US was suddenly antagonistic to Pakistan; even as it shifted from military to civilian rule.  The reason was that Pakistan was determined to use its alliance with the USA in a war against India, which would have been utterly horrific. 

Aid continued to flow to the mujahidden after 1990; it continued until April 1992.5 The USG was faced with an insoluble dilemma: the ISI was essentially running Pakistani foreign policy through fait acompli, while the Bhutto (and subs. Sharif) governments were locked in a political death-match with Pres. Ishaq Khan and each other.  The intelligence services (remember, Pakistan has several agencies) were implicated in the  political turmoil,6 and it became apparent that aid to the government was sure to result in a political crisis for Pakistan. The rival factions of mujahidden were fighting each other; the comparatively benevolent ones, like Ahmed Shah Massood, were not accessible to the CIA.

For this reason, "sticking with the job" was not feasible. All of the parties in Pakistan were seeking to avoid public association with the USA anyway.

On p.149, in his list of recommendations for future covert wars, Bruce Riedel warns against "mission creep," which in this case meant going from the Carter plan of harassing the Soviets, to the Reagan [era] plan of defeating the Kabul government.  This would have meant an even more "egregious" and arbitrary-seeming abandonment of Pakistan than what actually happened.7 But at least it's original.


As is now, I think, universally known, on 3 July 1979 Carter issued a presidential "finding" for assistance to the mujahidden; writing for The Nation, Eric Alterman writes "First revealed by former Director of Central Intelligence Robert Gates in his 1996 memoir From the Shadows, the $500 million in nonlethal aid..." Alterman's article is about as shoddy as it's possible to get: Gates' memoir cites a figure of $500 thousand, while cumulative US aid to the mujahidden did not reach the $500M level until 1985. He also ignores the fact that the DRA government was a junta with massive Soviet support, which had accelerated during its troubled 18-month history.; The Soviets had not yet invaded, but they had definitely taken over (analogous to South Vietnam, prior to October 1963--in this case, with the USG).8 Gates hints that more "non-lethal aid" followed, but doesn't specify how soon or how much.In any event, the inference drawn by Alterman and Blum is totally false: the USSR was committed to propping up a client state despite it swerving out of control.  There were already ample Soviet/KGB assets in Afghanistan, and had been since the Saur (April) "Revolution" of 1978.

One point about the Nouvelle Observateur article: it's been translated twice (English to French, French to English).  The title of the article declares that ‘Oui, la CIA est Entrée en Afghanistan avant les Russes...’ ("Yes, the CIA was in Afghanistan before the Russians") which is emphatically false.  The Soviet government had a long-standing relationship with PM Muhammad Daoud Khan, long before he ousted his cousin, King Zahir, in the 1973 coup, and dating back to 1965. This isn't necessarily a bad thing, but clearly Brzezinski was not listening to his interlocutor and assumed "intervention" meant "invasion."  Even if one acknowledges (as the CIA eventually did, internally) that the Soviets were surprised by the PDPA coup in 1978, they nevertheless acted quickly to shore it up militarily.  In any event, declassified documents do not support any of Brzezinski's remarks. For one thing, the Soviets were demonstrably not concerned about the rural uprising because they did not expect it would be hard to suppress.  They were angry with the renegade Hafizullah Amin, who had murdered their client Nur Taraki.9

On the other hand, there  the case of Gen. Naseerullah Babar, former interior minister of Pakistan.  In 1973, as mentioned above, the monarchy had been overthrown by Daoud Khan. The new '73 coup was, like the subsequent 1978 "Saur Revolution," pro-Soviet and comparatively left-leaning; more urgently, from the point of view of Gen. Babar (and Pres. Zulfikar Ali Bhutto), the new regime favored a major border revision. Gen. Babar successfully recruited Gulbuddin Hekmatyar, Burhanuddin Rabbani, and Ahmed Shah Massood to launch attacks on the Afghani army from the Panjshir Valley.  This insurgency failed to oust the government but may have extracted concessions from it; at the time, Pakistan's relationship with the USA was in suspended animation.10


Riedel claims many times in his book that the CIA "had very little direct contact" with the fighters or their leaders (p.41, quoting Robert Gates) as well as the Arab volunteers (p.81, preceding a detailed account of UbL's and Abdullah Azzam's CIA un-influenced experience in Afghanistan.  Steve Coll corroborates this claim, which surprised me. For example, as late as 1989, Milton Bearden, CIA station chief in Islamabad (of the sort typically described as "colorful")  describes running into a Gulf Arab (?) volunteer intent on killing "infidels" in Afghanistan.  At first I expected to  debunk this claim with documents available since 2004, but there really isn't a lot of evidence that the USG/CIA had any interaction with foreign  volunteers.  Sure, Prince Bandar told Larry King that he had been thanked by UbL for the latter's efforts to bring "the Americans, our [Usama's and Bandar's, presumably] friends, to help us against the atheists" (In case it isn't obvious, I don't think this ever happened).  But for the most part, encounters between either the Wahhabi volunteers or the higher ranking figures of the mujahiddin leadership were with Saudi entourages.

Interviews with US station personnel in Pakistan validates this thesis.  Much of intelligence work done in Pakistan during the mid-late 1980s was to determine what the ISI was doing with US money and weapons.  A key opponent was the embassy and CIA staff in Islamabad, who dreaded leakage of internal criticism to their Pakistani hosts.  But gradually, the message reached Washington by 1988 that the ISI was totally devoted to the most extreme Wahhabist factions, rather than ones that were effective or likely to win respect in Afghanistan.  Pakistan's hoped-for transition to democracy was a descent into anarchy and potential nuclear war with India; amid the endless repining about abandoning the region in the 1990s (along with Leonard Shelby-levels of forgetfulness about why "we" did so), the political establishment in the USA was unwilling to fully disavow the ideology and goals of the Taliban.  Members of the US Congress certainly would not have liked to live in Taliban-controlled Afghanistan, but they clearly preferred the Talib for the people of Afghanistan.


There is a well-established pattern of the USG targeting modernizing foreign governments (sometimes with compelling reasons, sometimes not), ousting them at terrible cost, and leaving sectarian tyranny in place.  Bruce Riedel is very anxious to remind us that it was the Soviet invasion and occupation of Afghanistan that created the international jihadist movement, and it was General Zia ul-Haq who made the decision to launch the war of resistance.  Saudi-Pakistani motives were compatible, in Afghanistan, and in some respects Pakistan had opened itself up to internal development by the Saudis.11

This was, in effect, a vending-machine regime-change; the USG dropped a lot of coins into the Pakistani slot, prevented do-gooders from unplugging the machine, and after ten years there was a loud clonking in the dispenser; but the soda was a little fizzier than anticipated.  Who could have expected such a thing?  In case you think I'm attacking a straw man, here's Bruce Riedel:
The Muslims who came to Pakistan in the 1980s to fight with the mujahedin did not come because of the CIA, the ISI, or the GID*, they came because of the Soviet invasion. [...] To suggest there is some inevitable link between President Carter's and President Reagan's backing of the mujahidin and 9/11 is tortured and incorrect logic.  It also shows a lack of understanding of the U.S. role in the Afghan war.  CIA officers in Pakistan had no interaction with the Arab volunteers, just as they had little interaction with the mujahedin themselves.  The business of fighting the war was in the hands of the ISI, not the CIA. (p.81)
On p.151, Riedel briefly reprises this message.  One might point out that Riedel contradicts himself: if the ISI and the Saudi GIP were actually managing the fighting, and the CIA was merely providing money, weapons, and diplomatic cover, then he can't use the same argument to exonerate them as well as the CIA.  At the very least, the ISI and the GIP bear some of the blame for the extreme comparative success of Gulbuddin Hekmatyar and Burhanuddin Rabbani--who won because they let others fight the Soviets, while they fought the other mujahiddin. But even then, it's absurd to imagine that the CIA, or its political masters (snort) in Washington had no responsibility to notice the disturbing ideology of the fighting forces.  The huge increase in the number of madrassas in Pakistan, funded by Saudi Arabia, was clearly intended to remold public perceptions of what Islam means.  Pakistan was not a sequestered Shangri-La closed to Western eyes; its society was under intense scrutiny by official Washington. The notion that the State Department and the CIA were unable to predict the rise of a quasi-Islam targetting the USA, is impossible to take seriously.

Either the US endeavor in Afghanistan had no effect on the history of the region (unlikely), or it did.  The Soviet commitment to the Afghan War was small: a peak commitment of 100,000 troops, in a war on its very border.  In contrast, the US commitment to the War in Viet Nam was massive: a peak of 540,000 troops (both figures apply for a short time only).  The famously casualty-averse USA lost 58,000 Americans in the Vietnam War; the poorer, authoritarian USSR lost 15,000 troops in Afghanistan.  As a share of available resources, the Afghan War for the Soviets was about the same as the Vietnam War on the USA.12 Economically, I find it hard to take seriously the proposition that the Afghan War caused the collapse of the USSR, or even contributed significantly; it merely made thousands of Soviet conscripts miserable.  Liberating Afghanistan was naturally not a goal of the CIA; the ISI, as Riedel acknowledges, wanted to win control of Afghanistan for Pakistan.

On the other hand, insisting that the foreign jihadi volunteers who traveled to Afghanistan were not influenced by the CIA effort there is a bad faith argument.  The CIA was well aware that the ISI's goal was to get at India, not the USSR or Communism. The Saudi goal for Afghanistan was directly visible in the refugee camps of the Northwest Frontier Province.  Emotional antipathy towards the Americans was constantly on display in Pakistan; the ISI definitely felt the need to hide its cooperation with the Americans (but not the Saudis) from the Pakistani public.  In 1979, rumors that the US was to blame for a Wahhabist uprising in Mecca caused Pakistanis to riot at the US embassy in Islamabad, killing four inside.  This was, at the very least, a place that was (a) very poorly understood by the Americans, and (b) obviously receptive to the most explosive anti-American hatred.  This alone ought to have given the US authorities pause, but it did not.  Likewise, the ideological chasm between the GIP and American society could reasonably have been expected to forestall any joint black operation when the Americans were barred any chance to observe what was happening.   A genuine ally does not have an uncontrollable urge to kill you if he meets you; this was what GIP and ISI agents explicitly said was the motive for denying CIA officers access to mujaheddin in theater.  There is no way this could have failed to produce a massively deadly terrorist attack on the US and its allies, and of course it did.

Hence, Bruce Riedel has got it exactly backwards: the intervention in Afghanistan's 1979-1992 "Civil War" produced, not freedom for Central Europe, but a large, well-seasoned, well-connected movement of eager terrorists, delighted to cleanse themselves of the stain of collaboration with the CIA.  Some even made it to New York City.

* Riedel probably meant "GIP," not "GID."  The GID is a Jordanian intelligence agency which sees Islamic fundamentalism as a major threat to the survival of Jordan.  The Saudi intelligence agency's English acronym is GIP.


Russian General Staff + Lester W. Grau & Michael A Gress (editors & translators), The Soviet-Afghan War: How a Superpower fought and Lost (PDF), University Press of Kansas (2002)

Steve Coll, Ghost Wars: The Secret History of the CIA, Afghanistan, and Bin Laden, from the Soviet Invasion to September 10, 2001,  Penguin Press (2004)

Frank Mazzetti, The Way of  the Knife: the CIA, a Secret Army, and a War at the Ends of the Earth (PDF), Penguin (2013)


  1. Bruce Riedel, a former CIA official, was selected by the Obama Administration to chair the White House review of policy in Afghanistan. See Julian E. Barnes, "Obama team works on overhaul of Afghanistan, Pakistan policy," Los Angeles Times (11 February 2009; accessed 2 March 2015)
  2. ISI is the best-known intelligence agency in Pakistan, but--like the American CIA--is neither the only intelligence agency in that country, nor entirely identical in interest and outlook to its government, military, or rival social classes.  See Peter Lyon, Conflict Between India and Pakistan: An Encyclopedia, ABC-CLIO (2008). While an analysis of the dynamics of the Pakistani intelligence/police services is not possible here, it is definitely relevant to events that followed.
  3. For data on US aid to Pakistan, see Claire Provost, "Sixty years of US aid to Pakistan: Get the data" The Guardian (11 July 2011).  This article includes links to source material and a database.  Data on foreign aid is classed by type; in no year since 1951 has there been no US financial aid to Pakistan, but military aid plummeted after the 1963 Indo-Pak War, and ceased after 1968. It increased after 1983, although no doubt much "direct" (i.e., non-USAID) economic assistance was actually secretly intended to strengthen the Pakistani armed forces (Riedel discusses this a bit, around pp.103-105).

    Military assistance to Pakistan ended in late 1990, as a result of the Pressler Amendment; it resumed in 2002, for obvious reasons, and is now immense.
  4. Military assistance to the mujahidden included cash and weapons bought with that cash--most notably, the Stinger missile system, authorized for use after 1985. The US Dept. of State reported (in 2011) a total outlay of $3 billion to the mujahidden; press reports cite a flow of $20M (in 1980) "rising to to about $300 million per year during FY1986-FY1990." (Kenneth Katzman, "Afghanistan: Post-Taliban Governance, Security, and U.S. Policy," Congressional Research Service, 24 Feb 2015,  p.4.).  Footnote 2, same page, mentions, " For FY1991, Congress reportedly cut covert aid appropriations to the  mujahedin  from $300 million the previous year  to $250 million, with half the aid withheld until the second half of the fiscal year. See “Country Fact Sheet:  Afghanistan” (PDF) in  U.S. Department of State Dispatch,  vol. 5, no. 23 (June 6, 1994), p. 377."
  5. Robert Gates, From the Shadows, (1996), p.432.; Robert Gates was DCI from Nov 1991 to January 1993. Gates later claimed otherwise, switching over to the narrative that the USG perfidiously abandoned the region after the USSR evacuated.
  6. For a brief account of this period see Husain Haqqani, Pakistan: Between Mosque and Military, Carnegie Endowment (2010), pp199ff; Zahid Hussain, Frontline Pakistan: The Path to Catastrophe & the Killing of Benazir Bhutto,  Penguin Books (2008), pp.23-26;  Edward A Gargan, "Pakistan Government Collapses; Elections Are Called," NY Times (19 July 1993); and Barbara Crossette, "Bhutto Facing New Misconduct Charges," NY Times (17 Oct 1990). Chapter 8 (pp.129-140) of What We Won is not bad either, and spells some of this out, but his claim that the CIA became aware in 1990 that Pakistan perhaps really did have a bomb, and Benazir Bhutto was notably more hawkish toward India than Zia ul-Haq was, cannot be taken seriously.
  7. Because in that case the CIA would have broken off collaboration with its ISI partner before its partner's objectives in that theater were finished, and while the personnel in Pakistan was largely the same.  After 1989, ties of loyalty to individuals in the Pakistani government became inherently conflicted because the political establishment was divided against itself. ALSO: p.132, Riedel objects to the post-1990 neglect of the mujahidden, Afghanistan, and Pakistan.  This is admittedly not the same thing as calling for more money to be supplied, but it seems most unlikely any options were then available to the CIA.

    Steve Coll (Ghost Wars, p.200) mentions that Frank Anderson, director of the Afghan Task Force at Ft. Langley (CIA HQ), was arguing for CIA withdrawal from the the Secret War well before the Soviet troop withdrawal (Feb 1989).
  8. Eric Alterman, "Blowback: the Prequel" (The Nation, 25 Oct 2001):  The legend of a massive US scheme to provoke a Soviet invasion of Afghanistan by "stirring up Moslems" [sic] was promoted by Zbigniew Brzezinski (Alterman tries to distance himself from William Blum, who undoubtedly drew attention to the interview Brzezinski gave to Le Nouvelle Observateur in 1998, but fails; see Killing Hope appendix, or Rogue State).   Steve Coll, Ghost Wars (2004), p.46 also cites Gate's 1996 memoir on the 3 July finding.
  9. For a sadly amusing account of the deadly triangle of Amin, Taraki, and the KGB, see Steve Coll, Ghost Wars, pp.48-49.
  10. Ibid., p.114.
  11. Saudi Arabia's massive foreign assistance program probably ought to be regarded as an ideological developmentalist regimen.  Western critics of the Saudi state usually make the mistake of imagining it to be self-consciously seeking to restore the middle ages to the Islamic world.  In fact, the Saudi state is quite visionary; its foreign assistance program does in fact pursue the economic development of its partners, albeit in an opaque and decentralized manner. Pakistan's ruling institutions certainly appreciated the aid received from Saudi Arabia, and evidently approved of its objectives for Pakistani society.
  12. There are a lot of pitfalls in my comparison, but Riedel says the Afghan War cost the Soviets 2% of their GDP in 1986 (p.31, citing the DIA estimate; admittedly, GDP estimates for the USSR were probably excessive); the Vietnam War cost the USA about 2.3% of its GDP in 1968 (Stephen Daggett, "Costs of Major U.S. Wars" (PDF), Congressional Research Service-29 June 2010, p.2).  Both estimates are in current accounting and ignore indirect costs.  Both wars lasted about the same time, with the Soviets and Americans getting involved indirectly before officially invading, and continuing to fund their respective sides for a couple of years after departing.  Note this comparison is not decisive; it can be argued that a socialist economy is less adversely affected by massive amounts of unproductive spending than is a market economy; on the other hand, market economies are more resilient, and can recover.



Post a Comment

<< Home