The Insurgent Sanctuary in Pakistan

CSIS Briefs


The Issue

Despite tough talk from Washington, Taliban leaders continue to enjoy a sanctuary in Pakistan and support from Pakistan security agencies. A review of insurgencies since World War II suggests that groups like the Taliban, which retain a sanctuary in neighboring states, either win insurgencies or successfully drag them out. If Washington is serious about ending the war in Afghanistan—including through a peace settlement—it needs to put significant pressure on the Taliban and Haqqani Network in Pakistan, not just in Afghanistan.

This month’s visit to Pakistan by U.S. Secretary of State Mike Pompeo and Chairman of the Joint Chiefs of Staff Joseph Dunford highlights a continuing challenge for the United States: how to coerce or cajole Pakistan to curb its support to proxies and support a peace process. Since late 2001, when individuals like Taliban leader Mullah Muhammad Omar slipped across the Pakistan-Afghanistan border, Pakistan has been a sanctuary for Afghan insurgent groups. U.S. officials have taken a tough public stance. In July 2017, President Trump remarked that “Pakistan often gives safe haven to agents of chaos, violence, and terror.”1 On New Year’s Day in 2018, President Trump went further, tweeting that the United States has “foolishly given Pakistan more than 33 billion dollars in aid over the last 15 years” with little in return, promising “No more!”2 Pakistan officials quickly returned fire. Foreign Minister Khawaja Asif said that Mr. Trump is likely “disappointed at the U.S. defeat in Afghanistan and that is the only reason he is flinging accusations at Pakistan.” Mr. Asif also warned that Pakistan wouldn’t budge. “We have already told the U.S. that we will not do more,” he said, “so Trump’s ‘no more’ does not hold any importance.”3

Just days before Secretary of State Pompeo’s September 2018 trip to Islamabad, the United States added pressure by announcing that it was cancelling $300 million in military aid, a move that was part of a broader suspension of aid announced in January. But despite these statements and efforts, the Trump administration has not been successful in changing Pakistan’s behavior—much like the Bush and Obama administrations before it. While there are multiple factors that have caused the current war in Afghanistan to persist for nearly two decades—including Afghan governance failures, a host of U.S. and broader NATO mistakes, and the Taliban’s access to resources from foreign donors and the drug trade—sanctuary in Pakistan has been critical. The Taliban’s senior shura (also called the Rahbari shura, or leadership council) continues to reside in Pakistan, as do the Taliban’s regional shuras that support the Afghan war. In addition, Pakistan’s Directorate of Inter-Services Intelligence (ISI) still provides sanctuary and aid to groups like the Afghan Taliban and Haqqani Network that are fighting the United States and its allies in Afghanistan.

A continuing failure to curb the Taliban and Haqqani Network’s sanctuary in Pakistan will undermine Washington’s ability to achieve even limited goals in Afghanistan—including a peace settlement. The historical record is bleak. Insurgent groups like the Taliban that retain a sanctuary in neighboring states, particularly a safe haven for their leadership, either win insurgencies or successfully drag them out over years and even decades. If the United States is serious about reaching a negotiated settlement with the Taliban—or even the unlikely goal of defeating the Taliban on the battlefield—it needs to quickly change its strategy and put pressure on the Taliban leadership in Pakistan. At the moment, the Taliban likely believes it is winning. A review of past peace negotiations suggests that insurgent groups usually negotiate a settlement when they believe they can’t win on the battlefield. Until Taliban leaders in Pakistan begin to feel some heat, the war will likely persist.

A continuing failure to curb the Taliban and Haqqani Network’s sanctuary in Pakistan will undermine Washington’s ability to achieve even limited goals in Afghanistan—including a peace settlement.

A Draw in Afghanistan

The war in Afghanistan is one of nearly 200 insurgencies since World War II. An insurgency is a political and military campaign by a nonstate group (or groups) to overthrow a regime or secede from a country. Insurgencies can be understood, in part, as processes of alternative state-building. Groups like the Taliban tax populations in areas they control, establish justice systems, and attempt to provide other services—much like governments. Consequently, insurgent groups need to mobilize local populations to secure recruits, intelligence, logistics, finances, and legitimacy. Based on both the Taliban and Afghan government’s need to mobilize the local population and govern territory, there are several indicators that provide a useful gauge of the war today.

Population Control or Influence: The first indicator is changes over time in population control or influence. Data on territorial control—including control of districts—is less helpful, since it can’t distinguish between unpopulated mountain ranges or deserts and heavily-populated urban areas. As Figure 1 highlights, there has been a slight increase in Taliban and other insurgent control or influence of Afghanistan’s population—from 9 percent in August 2016 to roughly 14 percent in May 2018. This translates into roughly 4.8 million Afghans under Taliban control or influence.6 There has also been a slight decrease in Afghan government control—from 66 percent in May 2016 to 56 percent in May 2018. This translates into approximately 19 million people under Afghan government control or influence. In addition, just over a quarter of the population—or 10 million Afghans—live in contested areas where neither side has significant control or influence.

There has been a slight increase in Taliban and other insurgent control or influence of Afghanistan’s population—from 9 percent in August 2016 to roughly 14 percent in May 2018.

Figure 1: Trends in Control or Influence of Population, 2016-20187

 Yet the data also show that Taliban gains have been almost entirely in rural areas of the country, where it enjoys some support among Afghans that have become disillusioned with the Afghan government, endorse the Taliban’s religious zealotry, need a job, or support a tribe or community allied with the Taliban. The Taliban controls no major urban areas—at least not yet. While the Islamic State swept through Iraq in 2014, seizing key cities like Mosul, Fallujah, and Ramadi, the Taliban has done nothing of the sort in Afghanistan.

The data show that Taliban gains have been almost entirely in rural areas of the country.

After briefly seizing the northern city of Kunduz in September 2015, the Taliban quickly lost control of it within days. In 2017, the Taliban failed to mount a sustained threat against any provincial capital and instead engaged in high-profile attacks in Kabul and other populated areas. Even in Helmand Province, where the Taliban have made advances in rural areas, local commanders have repeatedly failed to seize and hold such cities as Lashkar Gah. In 2018, the Taliban temporarily seized the eastern city of Ghazni, though again failed to hold it. Still, Taliban and other insurgent groups have succeeded in overrunning Afghan checkpoints, destroying military bases, and—at least temporarily—seizing district centers. On September 9, 2018, for example, Taliban and allied insurgents destroyed an Afghan army base in Baghlan Province, assaulted the headquarters of Day Mirdad District in Wardak Province, and attacked a police post in the Obe Distrct of Herat Province.

Local Support: A second set of indicators includes analyzing changes over time in local support since both the government and Taliban need to hold and expand territorial control. The Taliban’s ideology may be amenable to some Afghans, such as those living in conservative rural pockets of the south and east. But it is generally too extreme for many Afghans who adhere to a much less conservative form of Islam that permits most modern technology, sports, elections, and some women’s rights. The Taliban and its ideology are deeply unpopular, even compared to the current government and it security forces. Data from the Asia Foundation indicates that sympathy for the Taliban has declined since 2009, including those with “a lot of sympathy” and “a little sympathy” for the Taliban.,8 Most support for the Taliban is concentrated in parts of southern, eastern, and western Afghanistan—including such provinces as Wardak, Nuristan, and Zabul.

Levels of Violence: data is not a particularly useful outcome measure since it doesn’t tell us how—if at all—violence translates into control or influence. Indeed, low levels of violence in some areas may indicate Afghan government or Taliban control, while high levels may indicate contested areas where the government and insurgents are fighting to control territory. But since violence impacts the local population, data over time is still useful to track. Figure 2 shows the average daily security incidents recorded by the United Nations. Recent trends indicate a slight decline in violence, with violence levels particularly high in the east. Armed clashes comprised most incidents (64 percent), followed by improvised explosive devices (15 percent). There was also an increase in targeted assassinations (35 percent) and suicide attacks (78 percent) from 2017 levels.9

Figure 2: Average Daily Security Incidents in Afghanistan, 2015-201810

The Afghan war is, at best, a draw today. The Afghan government controls or influences a majority of the population, but the percentage of Taliban control or influence has increased. The Taliban has sympathy and support from locals, though it appears to have declined somewhat since 2009. And average violence levels have dipped. Yet the Taliban is still able to conduct complex attacks in urban areas—including Kabul—and threaten cities like Kunduz and Ghazni by controlling or influencing rural terrain in surrounding districts.

There are numerous reasons for the challenges in militarily defeating the Taliban or reaching a negotiated settlement. They include such factors as the failure to integrate the Taliban into Afghan society after their overthrow in 2001, when Taliban leaders were hunted down instead of co-opted; weak and ineffective Afghan governance, including significant corruption; a mistaken U.S. and Western focus on building a top-down government in Kabul, rather than also working at the grass-roots level and supporting local communities and tribes; and a U.S. and Western mistaken decision to try to win the war for Afghans by deploying large numbers of Western military forces.

But outside assistance—including sanctuary—has also been a significant factor.

The Benefits of Safe Haven and State Support

The history of insurgencies demonstrates that external sanctuaries can be extremely beneficial for groups. A group can plot, recruit, proselytize, contact supporters around the world, raise money, resupply, and—perhaps most important—enjoy a respite from the government’s counterinsurgent efforts in a sanctuary.11 This enables operatives to escape from the constant stress of life underground. Ideally, a sanctuary should be on foreign territory outside the reach of government forces, yet close enough—such as in a neighboring country—so that it is relatively easy to transit. In addition, some research suggests that sanctuaries increase the likelihood of ferocious conflict and cause higher casualties than would otherwise be the case.12

There are numerous cases where outside sanctuary has been helpful to insurgents. In South Africa, for example, the African National Congress had access to safe havens in Mozambique, where militants could train, rest, and plan future operations.13 In Algeria, the National Liberation Front (FLN) benefited from sanctuaries in Egypt, Libya, Tunisia, and Morocco. Approximately 20,000 Algerian guerrillas were concentrated in Tunis alone. As Walter Laqueur argues, it would have been politically infeasible for the French to target FLN guerrillas in these sanctuaries: “Even a minor air attack against an FLN base on the Tunisian side of the border (Sakiet Sidi Yusef) provoked a major international scandal; a massive attack was altogether unthinkable since the French government felt it could not commit such an affront to world public opinion.”14 There are other examples of insurgents effectively using sanctuary to defeat a government, such as the Vietcong in North Vietnam and—somewhat ironically—the Afghan mujahideen in Pakistan in the 1980s.

In addition to safe haven, outside state support can help increase the probability of insurgent victory—or, alternatively, allow groups to survive until they reach a negotiated settlement.15 One study, which examined 286 insurgencies fought between 1800 and 2005, found that 70 percent of all insurgent groups that received external support either won or negotiated a settlement to the conflict. Groups without any external support won or negotiated a settlement only 28 percent of the time.16

The Sanctuary in Pakistan

The Taliban and Haqqani Network’s sanctuary in Pakistan and state support from organizations like ISI have been essential to their war effort, and the U.S. failure to undermine this safe haven may be Washington’s most significant mistake of the 17-year-old war. To conduct political and military operations, the Taliban has tried to establish a somewhat centralized organization. At the top is the Taliban’s senior leadership shura, which is divided into committees that oversee finances, military operations, propaganda, religious affairs, and other tasks. As in many other insurgencies, the Taliban’s chief decisionmaking body is a political, rather than a military, one. Below the senior shura are several regional shuras—such as in Peshawar, North Waziristan, and Quetta (all in Pakistan)—whose job is to coordinate operations in nearby Afghan provinces. Below the regional shuras, the Taliban typically appoints a shadow governor and a military commander for Afghan provinces and districts.

The U.S. failure to undermine the Pakistan safe haven may be Washington’s most significant mistake of the 17-year-old war.

Figure 3: Example of Taliban Organizational Structure

Current and former Pakistan government officials have repeatedly stated that the government is not providing assistance to the Taliban or the Haqqani Network—and that these groups are not using Pakistan as a sanctuary. “The Taliban don’t need Pakistan for safe havens, Afghanistan is a safe haven,” remarked Lt. Gen. Amjad Shuaib, a retired Pakistan military officer during Secretary Pompeo’s September 2018 trip to Pakistan.

But this is disingenuous. Recent interviews with senior Western government officials indicate that the Pakistan government—especially the ISI—has provided several types of assistance. It has given money, intelligence, and strategic guidance to the Taliban and Haqqani Network and has helped provide medical care for Taliban fighters. The Pakistan government has also provided housing and logistics to some Taliban and Haqqani leaders. This includes allowing the Taliban to establish a sanctuary on its soil with freedom to operate in Pakistan. The Pakistan military has also conducted combat operations against militants in Pakistan, including in such areas as the Federally Administered Tribal Areas of Pakistan. But Pakistan military or police units have not conducted sustained operations against Taliban insurgents or their support network in areas like Baluchistan.

Indeed, the leadership of the Taliban and the Haqqani Network, which the United States and its allies are fighting in Afghanistan, are situated on the Pakistan side of the border. Examples include the Taliban’s leader, Haibatullah Akhunzada; his deputies, Sirajuddin Haqqani and Mohammad Yaqub; and a range of senior leaders like Abdul Qayyum Zakir, Ahmadullah Nanai, Abdul Latif Mansur, and Noor Mohammad Saqib. All reside in Pakistan, though the bulk of Taliban and Haqqani foot soldiers live in Afghanistan and fight a government they consider corrupt and incompetent.

Pakistan’s leaders have long been motivated to remain engaged in Afghanistan’s affairs—including through military engagement—to promote its national security interests. As Pakistan leader General Zia-ul-Haq remarked in 1979 to the head of the ISI, Lieutenant-General Akhtar Abdul Rehman Khan, “the water in Afghanistan must boil at the right temperature.”17 Part of Pakistan’s impetus is the continuing conflict with India and Islamabad’s fear—as one ISI official explained to me—of a “double squeeze” by New Delhi from both India and Afghanistan. India remains a major ally of Afghanistan and has provided significant assistance since 2001. The Indian-Afghan axis has left Pakistan isolated in South Asia. During the 1990s, Pakistan had a close relationship with the Taliban government in Afghanistan, which it had nurtured since the Soviet wars.

Today, Pakistan is surrounded by mostly hostile neighbors. Its relations with India and Afghanistan are frosty. The political scientist Hans Morgenthau once noted that balancing is the result of a natural struggle among states for power and influence. Without a state of equilibrium among them, he argued, “one element will gain ascendancy over the others” and “encroach upon their interests and rights.” Pakistan and India have long been involved in a balance-of-power struggle in South Asia. Both lay claim to the Kashmir region, and have fought three wars over Kashmir since 1947. Afghanistan is not the ultimate objective of either country, but rather an arena for competition in what has long been called the “great game.”18

Future Steps

The United States needs to re-think its strategy toward Pakistan. With U.S. and other outside aid to the Afghan government, the Afghan Taliban is likely too weak to overthrow the Kabul government or hold a major Afghan city. And the Afghan government is too weak to defeat the Taliban on the battlefield. As Figure 4 highlights, the war has impacted all areas of the country and caused significant suffering among the Afghan population.

Figure 4: Taliban Attacks 2015-2017

Most insurgent groups that agree to a peace settlement assess they can’t win on the battlefield, a situation that some have referred to as a “hurting stalemate.”20 In many of these situations—such as with the Provisional Irish Republican Army (IRA) in Northern Ireland, Farabundo Martí National Liberation Front (FMLN) in El Salvador, and South West African People’s Organization (SWAPO) in Namibia—government forces successfully infiltrated, targeted, and weakened insurgent groups. Without effective targeting of groups—including in their sanctuaries—most prefer to keep fighting. After all, only a quarter of insurgencies end with a settlement. Nearly three quarters end on the battlefield. Since World War II, insurgent groups successfully overthrew a government or gained independence in 35 percent of insurgencies, and governments defeated insurgents on the battlefield another 36 percent of the time.21

If Pakistan continues to harbor Taliban and Haqqani leaders and fails to support an Afghan peace process, the United States should consider several steps on an escalatory ladder:

  • Provide more public transparency about Pakistan activities. This could include, for example, publicly disclosing the names of senior Taliban and Haqqani leaders residing in Pakistan. It might also include declassifying intelligence—including satellite imagery—of Taliban locations in Pakistan.
  • Commit to aggressively pursue U.S. enemies wherever it finds them. The United States should be prepared to target the Taliban, al-Qaeda, Islamic State, and other groups wherever it finds them—including in Pakistan. This might include the United States directly targeting the Taliban and Haqqani Network in Pakistan, or—perhaps more likely—providing direct or indirect support to surrogates that do it.
  • Continue to freeze or terminate most military aid to Pakistan.
  • Continue to develop alternative routes to bring material to U.S. forces in Afghanistan, including through countries like Uzbekistan situated along Afghanistan’s northern distribution lines.
  • Consider suspending or terminating Pakistan’s status as a non-NATO ally. Non-NATO ally status offers military and financial advantages that generally are not available to non-NATO countries.
  • Consider making it more difficult for Islamabad to get access to multilateral financial lenders.
  • Consider placing Pakistan on the U.S. State Department’s list of state sponsors of terrorism. After all, U.S. and other Western intelligence agencies have collected an abundance of information about Pakistan ties to terrorist groups operating in Afghanistan and India, from Lashkar-e-Taiba (or Jamaat-ud-Dawa) to the Afghan Taliban and Haqqani Network.

Washington should be prepared to carefully escalate if there is not an improvement in cooperation. Ultimately, the United States cannot accept a situation where Islamabad covertly supports insurgents—some of which are targeting U.S. forces. This situation is a far cry from the 1980s, where Islamabad and Washington worked together to run a covert campaign to support the mujahideen in Afghanistan against the Soviet Union. Today, Pakistan is running a not-so-covert campaign in Afghanistan, and the United States has done little to stop it. Pakistan officials warn that U.S. actions are driving Islamabad toward China. But this development is not new. Islamabad and Beijing have increasingly established close political, security, and economic relations. The China-Pakistan Economic Corridor (CPEC) has the potential to bring valuable infrastructure and economic activity to Pakistan, and China is already Pakistan’s number one supplier of armaments and defense technology.

If Pakistan continues to harbor Taliban and Haqqani leaders and fails to support an Afghan peace process, the United States should consider several steps on an escalatory ladder.

Pakistan has long-term interests in a safe and stable Afghanistan, and Pakistan’s relationship with the Afghan Taliban makes it an important player in peace negotiations. Peace efforts need to continue, and Pakistan will be essential in reaching a political resolution in Afghanistan if it happens. With the Trump administration’s decision to keep U.S. forces in Afghanistan for the foreseeable future, Washington should continue to emphasize that it is in Islamabad’s interest to work toward a peace settlement and curb its sanctuary for senior Taliban and Haqqani leaders. After all, the Afghanistan and Pakistan people have suffered far too long.

Seth G. Jones holds the Harold Brown Chair, is director of the Transnational Threats Project, and is a senior adviser to the International Security Program at the Center for Strategic and International Studies in Washington, D.C. He is the author, most recently, of “A Covert Action: Reagan, the CIA, and the Cold War Struggle in Poland” (W.W. Norton).

The author gives special thanks to Nicholas Harrington and Danika Newlee for their comments and assistance in collecting and analyzing data.

CSIS Briefs are produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).

© 2018 by the Center for Strategic and International Studies. All rights reserved.

1Donald Trump, “Remarks by President Trump on the Strategy in Afghanistan and South Asia,” Fort Myer, Arlington, VA, August 21, 2017, available at:

2The quote came from President Trump’s Twitter account, @realDonaldTrump, on January 1, 2018. The full tweet was: “The United States has foolishly given Pakistan more than 33 billion dollars in aid over the last 15 years, and they have given us nothing but lies & deceit, thinking of our leaders as fools. They give safe haven to the terrorists we hunt in Afghanistan, with little help. No more!

3“Trump’s ‘No More’ Holds No Importance, Says Pakistan,” Daily Times (Pakistan), January 2, 2018, available at:

4On the number of insurgencies see Seth G. Jones, Waging Insurgent Warfare: Lessons from the Vietcong to the Islamic State (New York: Oxford University Press, 2017).

5I use “control or influence” since neither the government nor insurgents controls populations 24 hours a day, 7 days a week in all areas. In some cases, they rely on allies to coerce or co-opt locals, which is closer to “influencing” a village or city than “controlling” it.

6The estimate assumes a currently Afghan population of 34,124,811. See the CIA Factbook’s estimate at:

7Special Inspector General for Afghanistan Reconstruction (SIGAR), Quarterly Report to the United States Congress (Washington, DC: Special Inspector General for Afghanistan Reconstruction, July 30, 2018), p. 68, available at:

8Asia Foundation, A Survey of the Afghan People: Afghanistan in 2017 (Kabul: Asia Foundation, 2017), p. 58, available at:

9United Nations, Report of the Secretary General, “The Situation in Afghanistan and Its Implications for International Peace and Security,” A/72/888-S / 2018/539, June 6, 2018, p. 5, available at:

10SIGAR, Quarterly Report to the United States Congress, p. 74, available at:

11Daniel Byman, Deadly Connections: States That Sponsor Terrorism (New York: Cambridge University Press, 2005), pp. 53-78.

12Ariel E. Levite, Bruce W. Jentleson, and Larry Berman, eds., Foreign Military Intervention: The Dynamics of Protracted Conflict (New York: Columbia University Press, 1992); Patrick M. Regan, Civil Wars and Foreign Powers: Outside Intervention in Intrastate Conflict (Ann Arbor, MI: University of Michigan Press, 2000).

13Byman, Daniel, et al., Trends in Outside Support for Insurgent Movements (Santa Monica, CA: RAND Corporation, 2001), p. 10, available at:

14Walter Laqueur, Guerrilla Warfare: A Historical and Critical Study (New Brunswick, NJ: Transaction Publishers, 1998), p. 295.

15Ben Connable and Martin C. Libicki, How Insurgencies End (Santa Monica, CA: RAND Corporation, 2010), p. 62, available at:; Jason Lyall and Isaiah Wilson, “Rage Against the Machines: Explaining Outcomes in Counterinsurgency Wars,” International Organization, Vol. 63, No. 1 (Winter 2009), pp. 67-106.

16Lyall, “Rage Against the Machines,” pp. 67-106. External support is defined as the insurgent group having an international patron that provided material support, a rear base across international borders that could serve as a sanctuary in which to organize and train fighters and evade government countermeasures, or both.

17Quoted in Brigadier Mohammad Yousaf and Mark Adkin, Afghanistan – The Bear Trap: The Defeat of a Superpower (Havertown, PA: Casemate, 1992), p. 20.

18See, for example, Larry P. Goodson, “The New Great Game: Pakistan’s Approach to Afghanistan after 2014,” Asia Policy, No. 17, January 2014, pp. 33-39.

19Data is from the Global Terrorism Database at the University of Maryland, available at:

20See, for example, I. William Zartman, ed., Elusive Peace: Negotiating an End to Civil Wars (Washington, DC: Brookings Institution Press, 1995). Also see Monica Duffy Toft, Securing the Peace: The Durable Settlement of Civil Wars (Princeton, NJ: Princeton University Press, 2009).

21The data come from Seth G. Jones, Waging Insurgent Warfare: Lessons from the Vietcong to the Islamic State (New York: Oxford University Press, 2017).

Seth G. Jones
Senior Vice President; Harold Brown Chair; and Director, International Security Program