As expected, my recent post on arming Syria’s rebels (see here) generated a bit of debate among Notes on the Periphery readers. Danny, among others, made an important conceptual point: in assessing the “success” of four past cases of proxy arming—Afghanistan, Angola, Chad, and Nicaragua—it helps to define what “success” means.
Point taken. My effort to correct this lapse will still be unsatisfying to many: measuring “success,” of course, is inherently subjective, and not all parties involved in the decision making have the same idea of success (even within the U.S. government). Yet maybe it is appropriate to address, at least in broad strokes, a few possible metrics: recognizing that no judgment is perfectly scientific, how could one define success?
The possibilities differ depending on the objective of the “donor” country (though in many cases, even the goal is unclear or disputed). Accordingly, I have identified five potential definitions of “success” (and there are certainly more):
- Did arming play a prominent role in expelling or defeating an adversary?
- Did arming protect civilians?
- Did arming help the recipient gain power?
- Did arming help end the conflict?
- Did arming help bring about a negotiated settlement?
|Expel or defeat an adversary?||Somewhat|
|Help recipient gain power?||N/A|
|Facilitate end to conflict?||No|
|Bring about negotiated settlement?||Somewhat|
Did proxy arming play a prominent role in expelling or defeating an adversary?
On one hand, Pakistan CIA station chief Milton Bearden called the Agency’s covert distribution of precision anti-aircraft Stinger missiles in 1986 to the Afghan mujahideen the “most significant battlefield development” in the 10-year war against Soviet occupation (Coll 2004: 150). On the other hand, as Dr. Artemy Kalinovsky explains, while the Stingers were “an important antiaircraft tool,” they “hardly changed the course of the war” (Kalinovsky 2011: 43). Soviet air forces simply changed tactics, replacing low-altitude precision strikes with higher-altitude indiscriminate bombing. Yet on a strategic level, with Soviet appetite for the long occupation already growing thin, the CIA signaled to Moscow that the United States was serious about expelling their Communist adversaries. This may have been the nail in the coffin that drove the Soviets to sign the Geneva Accords in 1988, paving the way for withdrawal. In this sense, proxy arming had moderate success.
Did proxy arming protect civilians?
Using this definition of success, it is hard to argue arming the Afghan mujahideen met its objective. While the Reagan administration had lamented the extraordinary human cost of the war, propping up an anti-Soviet resistance force probably generated more violence and loss of life, not less. In fact, estimates indicate that 100,000-200,000 civilians were killed from 1980-1983 (Amstutz 1994: 163); six years later (and after the CIA escalation), the death toll had exceeded 1.25 million (a full 9 percent of the Afghan population was killed ino the war; Khalidi 1991: 101). Unless one concedes that the assistance may have saved lives by precipitating an earlier end to the Soviet occupation, proxy arming failed.
Did proxy arming help the recipient gain power?
Until 1988, the United States had one goal in Afghanistan: expel the Soviets. What happened next—whether the Soviet-backed Najibullah government hung on or an Islamist government took power—was rarely the subject of discussion. It was not until the Soviets announced their withdrawal that policymakers started to think seriously about a potential post-Soviet state—and even then, the CIA was highly skeptical: with the Soviets gone, did the United States have any responsibility to help install a new government?
U.S. strategy after 1989 was frenetic, muddled, and ever-changing. While objectionable to the State Department, CIA support was guided by a marriage of convenience with the Pakistani ISI and its mujahideen proxies, led by Gulbuddin Hekmatyar. It was Ahmad Shah Massoud, Hekmatyar’s Tajik rival, however, that played a decisive role in ousting Najibullah and sacking Kabul in 1992. While the fragile but more representative government that followed better suited U.S. interests, for the question at hand, the primary recipients of U.S. military assistance during the Soviet war (Hekmatyar and other factions favorable to Pakistan) failed to take power. (Of course, four years later, the Taliban succeeded in capturing Kabul, but by this time, the CIA contribution had been reduced to a trickle.) But since it is unclear whether a Hekmatyar victory was ever the U.S. objective, let’s classify this as N/A—not applicable.
Did proxy arming help end the conflict?
Of course, the answer to this question depends on what one calls the “end” to conflict. The Soviet withdrawal in 1989? The exit of the USSR proxy—Najibullah—in 1992? The Taliban takeover in 1996? These events usually help separate the Afghan war into three distinct phases—war against Soviet occupation (1979-89), civil war part I (1989-92), and civil war part II (1992-96). As the paragraph immediately above notes, U.S. strategy (if there was a coherent one) differed during each of these three phases.
Is even this delineation of three separate “conflicts” acceptable? If yes, then follow the narrative outlined in the first paragraph (expelling an adversary) above—arming played a moderately important role in changing the Soviet calculus. But as my friend and Notes on the Periphery reader Beth Goldberg noted, it is hard to separate “conflict” and “post-conflict.” By this logic, the campaign to dislodge Najibullah (with CIA assistance) post-1989 was simply an extension of the original war to expel the USSR. Many could argue that the war even intensified, rather than slowed, after the Soviet withdrawal. By this metric, the arming campaign failed to halt conflict—in essence, it fed a 3-year bloody stalemate.
Did proxy arming help bring about a negotiated settlement?
Again, one could argue the answer here depends on where one draws the mark: the Soviet withdrawal in 1989, or the last-ditch negotiations that pushed Najibullah out of power in 1992? In both cases, however, I would argue proxy arming helped bring about a settlement. With the CIA-backed mujahideen proving more durable than expected in 1986-7, Moscow finally accepted the open invitation to Geneva for a negotiated withdrawal in 1988. In this sense, political scientists are likely to characterize this as a classic case of using military force (in Afghanistan, a blank check to the mujahideen) to leverage negotiations.
After the Soviets’ departure, the CIA continued to support Hekmatyar’s faction (and to a lesser degree, Massoud’s), in hopes of achieving a military victory that ousted Najibullah. Though they failed to deliver in the 1989 Battle of Jalalabad, sustained pressure helped persuade Najibullah to accept a UN transition plan in April 1992. While this also happened to coincide with the collapse of the Soviet Union—Najibullah’s most reliable financier—it is fair to say that arming played a moderately important role in producing the power-sharing agreement (Peshawar Accord) that followed.
Of course, these descriptions do not take into account the unintended consequences of the CIA’s arming campaign: the development of what would become future U.S. adversaries, Taliban and al Qaeda. Accordingly, one could add a sixth category: did arming serve U.S. long-term interests? In hindsight, the answer was clearly no—but one has to keep in mind that the United States at the time was far more concerned with the spread of Communism than the rise of Islamic extremism.
This makes the Syria debate all the more difficult to assess. Few working on AfPak policy in the 1989 could foresee that Osama bin Laden, based in Afghanistan, would 12 years later launch an attack that killed nearly 3,000 Americans. It is equally impossible to predict what Syria might look like 12 years from now—and what our “interests” will look like in hindsight.
Even still, Syria analysts today are increasingly harping on President Obama to provide a clearer description of at least U.S. short- to medium-term objectives. The Economist puts it well:
“The problem here, as Robert Malley of the International Crisis Group argues, is that America still needs to figure out what the goal of an intervention is. Do we want to safeguard civilians? Then we should establish a no-fly zone and humanitarian safe areas with sufficient military strength to ensure Mr Assad cannot overrun them. Do we want to topple Mr Assad’s dictatorial government? Then we should arm the most effective jihadi rebel groups. Do we wish to prevent the spread of anti-American jihadi terrorist groups? Then we should arm the more moderate groups. Do we want to enforce the international ban on the use of chemical weapons? Then we should attack Mr Assad’s forces directly. Do we wish to simply shorten the civil war and end the killing? But what if the fastest route to a stable, unified Syria were actually a rapid government victory?” (Economist, Democracy in America blog, 6/14/13)
This reaction is understandable–as those observing the civil war in Syria feel they deserve an explanation from the President on whether the means (arming) are commensurate with the ends (Obama: “a stable, non-sectarian, representative Syrian government that is addressing the needs of its people through political processes and peaceful processes.”)
History suggests that American attempts to leverage a military opposition through arming are often ad hoc and produce mixed results. No historical analogy is perfect, but can the United States produce a better outcome in Syria than it did during the 1980s in Afghanistan?