Small Wars Journal

The Problem with Proxies: Ideology is No Substitute for Operational Control

Sun, 09/27/2015 - 8:43am

The Problem with Proxies: Ideology is No Substitute for Operational Control

Yelena Biberman and Orr Genish

The US strategy of selecting reliable proxies against the Islamic State of Iraq and Syria (ISIS) based on their ideological leanings is proving to be disastrous. The latest headlines about US-trained Syrian rebels handing over military equipment to al-Qaeda-linked fighters highlight the serious and enduring problems of outsourcing war.[1] Recruitment of proxies based on their ideology has taken more time and resources than anticipated – it is hard to measure one’s commitment to a set of ideas when the bullets are flying. The approach has also shown to be misguided. Without effective operational control, the “moderate” rebels are unlikely to bring victory against ISIS.  

The United States began the program of arming Syrian proxies at a base in Jordan in April 2013. It later expanded it to Saudi Arabia. In February 2015, Washington signed an agreement with Ankara to train and equip “moderate” Syrian rebels.

While the Obama administration debated whether to get involved in Syria, the CIA conducted several studies of the effectiveness of its past operations. Many of the findings remain classified, but the New York Times reported that one of the conclusions was that many of the state-sponsored militias “had a minimal impact on the long-term outcome of a conflict,” especially when “no Americans worked on the ground with the foreign forces in the conflict zones.”[2]

Using proxies in Syria was supposed to help the United States avoid placing the regular forces in harm’s way while benefitting from the former’s local knowledge. However, without the regulars on the ground, the few US-sponsored rebels who are actually engaged in combat are proving ineffective and even counterproductive. They are also becoming a target for groups against which they were not originally intended.[3]

Delegating the fight against ISIS in Iraq to Iran-backed militias, without direct involvement, has been no less precarious. In a March 2015 interview, former commander of US troops in Iraq Gen. David H. Petraeus identified the Shia militias as “the most serious threat” to Iraq’s long-term stability. He cautioned that they “could emerge as the preeminent power in the country, one that is outside the control of the government and instead answerable to Tehran.”[4] The Human Rights Watch linked the militias, along with the regular Iraqi forces, to looting and even destroying entire villages after ISIS had passed through them.[5]

The “Birdcage Dilemma” of Effective Operational Control

In 1982, Communist Party elder Chen Yun compared the relationship between China’s economy and central planning to that between a bird and its cage: “You mustn’t hold the bird in your hands too tightly or it would be strangled. You have to turn it loose, but only within the confines of a cage. Otherwise it would fly away.”[6] Chen’s birdcage analogy can be applied to the delicate and dangerous relationship between states and their proxies. In order for the proxies to be effective, states must keep them under control while simultaneously turning them loose. They must construct a kind of a “birdcage” that would allow proxies to be both useful and manageable.

The violence states typically outsource to proxies ranges from mostly defensive (e.g. guarding installations) to offensive (e.g. targeted killings). Proxies have been widely viewed as a Cold War relic: the United States and the Soviet Union used them to wage wars across the Third World without succumbing to direct, apocalyptic confrontation. Ultimately, the victory of the former was facilitated by US-supplied rebels waging an “anti-Soviet jihad” in Afghanistan during the 1980s.[7]

The practice of employing third parties for war-making is millennia-old. Egypt used mercenaries, including the Nubians, for seven centuries starting in 1479 BCE. So did the Israelites around 1250 BCE “so that local citizens would be free to maintain the economic output necessary to support both the kingdom and the army.”[8] In the 18th century, all the major European armies relied heavily on foreign fighters, while privateering played an important role in naval warfare.[9]

State outsourcing of war has survived and thrived in the post-Cold War era, not least due to the increased reliance on private contractors for combat-related activities.[10] Many states continue to use nonstate partners – from farmers to warlords – to counter as well as to foment insurgencies.[11]

The effective use of proxies requires the state sponsor to strike a delicate balance between freedom and control. Both too little and too much of either renders the proxy useless, and even dangerous, to its sponsor. Moreover, states that fail to use their nonstate partners effectively face the threat of other state and nonstate actors poaching them (and their equipment) or introducing their own, as did Iran. Rather than holding the proxy too tightly or letting it act freely, the state must construct a sort of a “birdcage” – a strategy for effective operational control. By “operational control” we mean the exercise of command authority over the irregulars: designating their operational objectives and providing authoritative direction for accomplishing those objectives through the tactical application of force.

The following case studies represent two different scenarios in which the state exercised too little and too much operational control for the proxy to be useful. Pakistan provides us with multiple cases of proxy warfare that include significant success in Afghanistan against the USSR as well as failures in Kashmir, India.[12] We can learn from Pakistan’s mistakes in proxy use by analyzing the dynamics of militia control through the lens of the bird cage dilemma.

The “Free Bird” Scenario: Pakistan’s 1947 Covert Operation in Kashmir 

The Muslim-majority princely state of Jammu and Kashmir (henceforth referred to as “Kashmir”) was ruled by a Hindu maharaja, who sought independence following the British withdrawal from India in August 1947. Neither Pakistan nor India shared this vision. In October 1947, Pakistan covertly armed and transported Pashtun tribesmen to take control of Kashmir. The panicked maharaja turned to Delhi for assistance, but was told first to sign an Instrument of Accession to India. He had little choice but to oblige. The tribal invasion turned into the first India-Pakistan war. Since then, Kashmir has been a major source of rivalry between the two regional powers. When both countries acquired nuclear weapons in 1998, the Kashmir conflict made South Asia, as then-US president Bill Clinton put it, “the most dangerous place on earth.”[13]

Pakistan’s ineffective proxy strategy in Kashmir in 1947 was indeed very costly. This section examines why the lashkars (tribal raiding parties) failed to serve their purpose. It shows that the problem lay in the birdcage dilemma: the Pakistani sponsors gave the lashkars too much freedom.

Neither the currently dominant narrative of uncontrollable lashkars nor the original British estimation of them being “under highly competent leadership”[14] are entirely correct. The tribesmen enjoyed high, but not total, autonomy. Pakistani Prime Minister Liaquat Ali Khan signed off on the plan, North-West Frontier Province authorities mobilized the Pashtun tribesmen in their region, and the Pakistani army equipped and transported the invaders. Some Pakistani soldiers turned a blind eye when boxes of .303 ammunition mysteriously disappeared from armories. A number of regulars took leave, or became technically “deserters,” to join the fray. Several of them took it upon themselves to assume senior responsibilities. However, they were not fully in control of the tribesmen, especially once the latter arrived in Kashmir.

Having to swear on the Koran “to refrain from killing any neutrals and from individual looting”[15] did not prevent many of the tribesmen from “massacre arson and loot to the very gates of Srinagar.”[16] Srinagar was the summer capital of the princely state and the site of an airport, which the tribesmen failed to secure by the time the Indian troops arrived on October 27.

Journalist Andrew Whitehead traces the story of a convent ransacked by tribesmen on their way to Srinagar. It is a story of Muslims who greeted the tribesmen as liberators from the unpopular maharaja, only to be violated on par with their non-Muslim counterparts; and of a Christian mission assaulted and terrorized.

Upon arrival, the raiders pillaged the convent, shot hospital patients as well as a British officer and his wife, and raped and killed nuns. Those who survived the initial onslaught were ordered by the raiders to line up. They were told that they would be shot. However, as the nuns were prepared to be killed, a Pakistani army officer suddenly appeared. One of the nuns recounted him saying something to the tribesmen in their language, and they put down their arms.[17]

Striking about the Pakistani officer is both the conspicuous presence and absence of his authority. On the one hand, he was able to stop the tribesmen from executing innocent civilians. On the other, he was unable to redirect them toward Srinagar, where the first Indian troops were landing that very day. The tribesmen stayed at the convent for over a week, looting and assaulting. When nothing was left to take, many of them headed back home. Had the lashkars been under tighter operational command, they would not have been preoccupied with the type of gratuitous violence they displayed at the convent. The Indian troops could then have been prevented from landing in Kashmir and, thus, mounting a rapid offensive. The outcome of the ensuing war between India and Pakistan may have been very different.

Once the Indian troops landed, the lashkars were ill-prepared and organized for confronting a regular army. The Pakistani planners erroneously assumed that what their proxies lacked in operational command they would make up for in enthusiasm. The lashkars were at their best in hill battles. Usually relying on ambush and surprise attacks, they avoided set-piece battles and exposed positions on the plains. The execution was frequently haphazard. The tribal warriors were also highly prone to turning back after acquiring sufficient booty or when they saw the raid running into trouble. Even their fervent supporter, Pakistan’s Brigadier-in-Charge in Kashmir Akbar Khan, was baffled by the lashkars’ response when Indian troops began coming out of Srinagar: “The withdrawal of the tribesmen had not been a step by step falling back, but a breaking away and a total disappearance. A spectacular advance coming to such an abrupt end was most bewildering.”[18] Whitehead explains: “The glory of taking part in the raid was much greater motivation than territorial conquest. Often fighters would simply head home without any attempt at coordination within the lashkar.”[19]

While Whitehead attributes the ineffectiveness of the lashkars to their culture, Akbar Khan identifies the strategic flaws in Pakistan’s proxy plan. In his memoir, he laments that not even “ex-servicemen” of the Pakistani army were there to support and command the lashkars upon the arrival of the Indian troops. He points out that the lashkars’ ineffectiveness was primarily the result of the inadequate preparation and backup by the Pakistani regulars:

They [the tribesmen] had, of their own free will, agreed to come and fight in Kashmir but only against the [Kashmiri] State Army. In this they had done more than expected of them. But no one had arranged with them to fight also against the regular Indian Army with artillery, tanks and aircraft… they had naturally expected that, in the changed circumstances, the Pakistan Army would be coming up to support them. And soon they had been shocked to find that no troops, no artillery and no aircraft were coming to help them. Indeed, not even the most elementary requirements of something like a secure base behind them as being provided by them.[20]

Pakistan’s 1947 covert operation in Kashmir offers two important lessons about the state-proxy relationship. First, proxies require on-the-ground support and command. Moral hazard and lack of professionalism may otherwise prevent them from carrying out their mission in an efficient and timely manner. When Mao Tse-tung advised that “there must be no excessive interference” with the activities of guerrilla units, he did not mean that there should be absolutely no interference. All guerrilla units, he observed, require “political and military leadership” to discipline them: guerrilla warfare is not about “banditry and anarchism,” but about “severe discipline.”[21]

The second lesson of Pakistan’s 1947 covert operation is one which also appears in Mao’s guerrilla warfare manual: nonstate proxies may slow down or weaken the enemy, but they are insufficient for achieving “ultimate victory.” Regular troops are necessary. As Mao put it, “while we must promote guerrilla warfare as a necessary strategical auxiliary to orthodox operations, we must neither assign it the primary position in our war strategy nor substitute it for mobile and positional warfare as conducted by orthodox forces.”[22] Had the lashkars been better supported and commanded by Pakistani regulars, they would have unlikely exhibited high desertion rates and committed as many atrocities, thereby alienating the local population, much of which was initially sympathetic to their cause. They not only failed to prevent the arrival of the Indian troops, but also facilitated India’s recruitment of local collaborators.  

The “Smothered Bird” Scenario: Pakistan’s 1965 Covert Operation in Kashmir

Operation Gibraltar represented Pakistan’s second attempt to take Kashmir through a combination of proxies and regular forces. This time, however, the traumatic experience of 1947 led the military planners to keep their proxies too constrained to be effective. What resulted was, again, a botched operation and an inter-state war, as well as the re-freezing of the Kashmir conflict.

While, in 1947, Pakistan’s proxies consisted of tribesmen with martial experience, in 1965 they comprised ordinary civilians from Azad Kashmir. Azad Kashmir was the roughly one-third of the princely state that was awarded to Pakistan following the 1947-48 war, with the rest given to India. The roughly 7,000 civilian recruits had the local knowledge that their 1947 counterparts lacked, but they were far less militarily skilled. Their training period was only three to seven weeks, with the main focus being guerrilla fieldcraft. Prisoner of war interrogation reports revealed that although the Azad Kashmir recruits were supposed to have been volunteers, “an element of coercion” was involved in their recruitment.[23] Some appeared to have known very little about their mission. They were told that they would be involved in “an ordinary hit and run exercise in the enemy territory.”[24]

In 1947, it was not uncommon for the Pashtun units to contain only one regular. In 1965, a unit (“Company”) typically contained an officer, several junior commissioned and non-commissioned officers, key personnel from the Azad Kashmiri battalions or units of the Northern Scouts, and several other ranks. The degree to which the proxies were constrained from exercising any significant role in the operation is illustrated by a story recorded in a secret 1966 report by the Criminal Investigation Department of Jammu and Kashmir.[25]

On August 4, 1965, a group of grazers were tending to their cattle on a pasture located in the high mountains of Baramulla District. Suddenly, they encountered armed men coming from the mountain passes. The commanding officer of the column of the infiltrators collected and addressed the herdsmen. He told them that his force had come from Pakistan to liberate the Kashmiri Muslims from the yoke of Indian imperialism, and that this was the beginning of a “Jihad” in which every Muslim had to participate. The herdsmen were sworn to secrecy and given money. Some of them were directed to go down and collect provisions for the infiltrators, while others were questioned about the disposition of the Indian Army units and other strategic information.

Towards the evening, the Pakistani infiltrators permitted several of the herdsmen to go down to their village, called Darakasi, to run some errands. One of the herdsmen, a young man named Mohammad Din, insisted that the villagers find a way to alert the Indian authorities. In order not to arouse suspicion of the infiltrators, he told his companions to carry on as usual, while he himself set out on a marathon run to the nearest Indian defense post.

What is particularly telling about this story is that it was the commanding officer, not the Azad Kashmiri guerrilla, who addressed the herdsmen. Had it been the latter, the story may have had a different ending since the guerrillas shared the same regional identity as the locals. The Pakistani forces faced a similar outcome in the rest of the region – rather than collaborating with them, the locals reported the infiltration to the Indian authorities, which soon led to the second India-Pakistan war.  

While, in 1947, Pakistan exercised too little control over its proxies, in 1965, the Azad Kashmiri guerrillas were far too inhibited to be effective. They lacked not only the necessary military skills and information about their mission, but also the freedom to deeply engage with the local population. The benefits of their local knowledge were, consequently, squandered, and the mission failed.


Proxy warfare requires effective operational control. Ideology is insufficient for predicting how proxies will behave during important military (and civilian) engagements. The advantages of outsourcing war cannot be realized without striking a delicate balance between freedom and control. If effective operational control is not an option, neither should be the fight.

End Notes

[1] Karen DeYoung, “U.S.-Trained Fighters in Syria Gave Equipment to al-Qaeda Affiliate,” Washington Post, September 25, 2015.

[2] Mark Mazzetti, “C.I.A. Study of Covert Aid Fueled Skepticism About Helping Syrian Rebels,” New York Times, October 14, 2014.

[3] Anne Barnard and Eric Schmitt, “Rival Insurgents Surprise Syrians Supported by U.S.,” New York Times, August 1, 2015, p. 1.

[4] Liz Sly, “Petraeus: The Islamic State isn’t our biggest problem in Iraq,” Washington Post, March 20, 2015.

[5] Human Rights Watch, “After Liberation Came Destruction: Iraqi Militias and the Aftermath of Amerli,” March 18, 2015.

[6] Daniel Yergin and Joseph Stanislaw, The Commanding Heights (New York: Simon and Schuster, 2002), p. 192.

[7] Steve Coll, Ghost Wars (New York: Penguin, 2004), p. 16.

[8] Molly Dunigan, Victory for Hire: Private Security Companies’ Impact on Military Effectiveness (Stanford: Stanford University Press, 2011), p. 6.

[9] Janice E. Thomson, Mercenaries, Pirates, and Sovereigns: State-Building and Extraterritorial Violence in Early Modern Europe (Princeton: Princeton University Press, 1994), p. 11.

[10] P. W. Singer, Corporate Warriors: The Rise of the Privatized Military Industry (Ithaca, NY: Cornell University Press, 2007); Allison Stanger, One Nation Under Contract: The Outsourcing of American Power and the Future of Foreign Policy (New Haven: Yale University Press, 2009). 

[11] Yelena Biberman, Gambling with Violence: Why States Outsource the Use of Force to Domestic Nonstate Actors (PhD diss., Brown University, 2014).

[12] Author’s interviews with military officials, Islamabad, Pakistan, December 2014.

[13] Lowell Dittmer, “South Asia’s Security Dilemma,” Asian Survey 41, no. 6 (November/December 2001), p. 897.

[14] Inward Telegram from India (Govt) to Commonwealth Relations Office, November 1, 1947; War Staff, India Office, “Kashmir” (Top Secret); 7, L/WS/1/1139; L/WS War Staff Papers, 1921-51, Asian and African Studies Collection, British Library, London, United Kingdom.

[15] Top Secret Cypher Telegram from U.S. High Commissioner in India, New Delhi, to Ministry of Defence, London, November 9, 1947; War Staff, India Office, “Kashmir” (Top Secret); 7, L/WS/1/1139; L/WS War Staff Papers, 1921-51, Asian and African Studies Collection, British Library, London, United Kingdom.

[16] Inward Telegram from India (Govt) to Commonwealth Relations Office, November 1, 1947; War Staff, India Office, “Kashmir” (Top Secret); 7, L/WS/1/1139; L/WS War Staff Papers, 1921-51, Asian and African Studies Collection, British Library, London, United Kingdom.

[17] Andrew Whitehead, A Mission in Kashmir (New Delhi, Penguin/Viking, 2007), p. 4.

[18] Akbar Khan, Raiders in Kashmir (Islamabad: National Book Foundation, 1975), p. 47.

[19] Whitehead, A Mission in Kashmir, p. 45.

[20] Khan, Raiders in Kashmir, p. 56.

[21] Mao Tse-tung, Guerrilla Warfare (1937), “What Is Guerrilla Warfare?”

[22] Ibid., “The Relation of Guerrilla Hostilities to Regular Operations.”

[23] Criminal Investigation Department, Report on Pakistani Organized Subversion, Sabotage and Infiltration in Jammu and Kashmir (Jammu and Kashmir, India, 1966), p. 31.

[24] Ibid., pp. 36-37.

[25] Ibid., p. 43.


About the Author(s)

Orr Genish is a Government major at Skidmore College.

Dr. Yelena Biberman is an Assistant Professor of Government at Skidmore College, Saratoga Springs, NY. Her research focuses on unconventional warfare and militias in South Asia. She is currently writing a book on the outsourcing of war.


This article focuses entirely on how sponsoring states can control and manipulate proxies to meet their needs, treating the proxies as passive birds that the sponsoring state can cage or release at will. That's a serious deficiency: proxies are in fact active players who have as much interest in manipulating the sponsoring state as the sponsoring state has in manipulating them. Of course proxies have their own agendas, which may or may not be in alignment with those of the sponsoring state. Their task is to squeeze as much support as they can out of their sponsor and divert it to their own agendas. It is impossible to effectively use proxies without viewing the equation realistically from their side. Treating them as passive tools is a sure road to failure.

Looking at Syria, it is easy to assume that proxies would be more effective with American forces in place, but if it were politically acceptable to have American forces in place the proxies wouldn't be necessary, so that observation, while likely true, is hardly useful. The whole point of using proxies is not to have state sponsor forces in place.

The US got into the current Syria mess by adopting goals driven by domestic policy, not by practical, achievable purpose on the ground. The goal seems to be to appear engaged enough that we can't be accused of ignoring ISIS, while avoiding commitment of ground forces. That is not a strategy for Syria, it's a strategy for domestic image management.

Of course what's missing from much of the ranting over Syria is any hint of a viable long term or even medium term objective. "Crush ISIS" is not enough. It's not enough to know what you're against, you have to know what you're for; what the desired end state is and what the exit strategy is. Anyone who seeks intervention in Syria but cannot supply those goals is talking through the wrong end.