Will Afghanistan Collapse When the US Withdraws?

On February 8, 1988, then-President Mikhail Gorbachev announced that the Soviet Union would withdraw its military forces from Afghanistan. That withdrawal began on May 15 and was completed on February 15, 1989. In its wake, the Soviet Union left a government under the leadership of Dr. Mohammad Najibullah Ahmadzai. At the time, as scholar Thomas Barfield wrote, “It was universally believed that once the Soviets left, the mujahedin would overrun government outposts, take the regional cities, and march on Kabul in triumph in a matter of months, if not weeks.” And yet, President Najibullah held his government together for three years after the Soviet withdrawal, its collapse finally occurring in April 1992, after the dissolution of the Soviet Union and cessation of Soviet aid in December 1991.

On April 14, President Joe Biden announced that the United States would withdraw its military forces from Afghanistan. That withdrawal began on May 1 and is to be completed no later than September 11. In its wake, the United States will leave a government under the leadership of Dr. Ashraf Ghani. At the time of Biden’s announcement, there were a flurry of predictions that the Taliban would move quickly to overrun government outposts, seize regional cities, and ignite a broader civil war or march on Kabul to topple the government. But will these predictions turn out to be accurate? Or will they, like those of Najibullah’s quick demise, prove to be overly dire and based on faulty assumptions? What should we expect to happen after the U.S. withdraws from Afghanistan?

Answering those questions requires examining the Soviet withdrawal in more detail, and comparing the situation that existed in 1989 to the situation today. As we will see, there are both striking similarities and distinct differences between the two, which should inform our thinking about the future of Afghanistan.

The Soviet Withdrawal

Writing as a RAND analyst in 1991, Zalmay Khalilzad – today the United States’ special representative for Afghanistan reconciliation – identified three assumptions that underpinned predictions of Najibullah’s hasty overthrow: his regime was brittle and likely to fragment once the Soviets left; the departure of Soviet advisers would decisively shift the military balance to favor the mujahideen; and the Soviets cared only about securing a “decent interval” between their departure and the government’s collapse, and would therefore only provide limited assistance to Najibullah for a short time. These assumptions, of course, were incorrect. But it’s instructive to observe just how incorrect they were.

First, Najibullah’s government proved more flexible and resilient than many observers predicted. As author Tamim Ansary recalls:

Najib tried, he really did. He tried everything he could. He changed the name of KhAD [Khadamat-e Aetela’at-e Dawlati] to WAD [Wizarat-i Amaniyyat-i Dawlati], but everyone knew it was the same old dreaded secret police. He renamed the [People's Democratic Party of Afghanistan] the “Fatherland Party,” but no one started singing patriotic anthems. He had a new constitution written declaring Afghanistan an Islamic republic and guaranteeing the freedom of all citizens, but no one believed him. He started building mosques and religious schools. He called for national reconciliation. He offered cabinet positions to selected Mujahideen leaders. He even offered to step down if certain conditions were met.

As Ansary states, Najibullah’s attempts to reform the state were not particularly successful. But his self-reformation was more so. As the Soviets withdrew, the communist “Dr. Najib” became the Muslim “Najibullah,” who used his eloquence in Pashto to recite passages from the Quran at prominent mosques on Fridays. He established a new center for research in Islamic Studies and put scores of mullahs on the government payroll.

Patronage was a key part of Najibullah’s strategy, which combined powerful security forces with purchased support, just as the British-funded emirs of Afghanistan’s past had done. In a workforce of 3 million people, he put 700,000 on the government’s dime. He increasingly relied on tribes – and the militias they could raise – to provide security and contest the mujahideen in areas beyond the reach of his forces. He allowed militia commanders to autonomously run their areas of control (with financial assistance from the state), if they agreed not to fight his government. By 1991, Najibullah could count around 170,000 militia forces aligned to his cause, a number larger than his formal security forces could muster. The strongest groups, however, never aligned with him – despite his attempts at inducements up to and including senior positions in his government – and many of his deals proved fleeting and unstable. Najibullah faced numerous coup attempts, including one by his own chief of general staff. But the combination of continuous personal and governmental reforms, a time-tested strategy, and constant political dealing showed Najibullah to be considerably more dexterous and resilient than many had assumed.

Second, the military balance of power did not shift decisively in favor of the mujahideen after the Soviet withdrawal. One reason for this was that Afghanistan’s security forces proved more capable than the Soviets believed they would be. Analyst Olga Oliker noted that this was due in part to the Soviet tendency to assess Afghan forces against Soviet standards – and to thus conclude that Afghan forces were incompetent – as opposed to using metrics attuned to the way Afghans preferred to fight. While there was significant deterioration of these forces over time (many police defected to militias and the army saw large numbers of defections and desertions), the dreaded KhAD remained largely operable during Najibullah’s tenure. After the mujahideen’s disastrous performance in its now infamous attack on Jalalabad, the morale of the Afghan army increased considerably and Najibullah found himself in a notably enhanced position, with many more local commanders willing to cut deals and align their militias with him.

Another reason was the mujahideen were not as politically and militarily astute – nor as cohesive – as many believed them to be. As stated in a Human Rights Watch report from 1991, “The belief that the Afghan resistance organizations were Western-style, hierarchically-organized political parties whose ‘leaders’ in Peshawar could decide matters for their ‘followers’ in Afghanistan was always an illusion; this became even clearer after the Soviet withdrawal.” There were 23 major mujahideen commanders in Afghanistan, most of whom acted independently and assumed victory would come naturally after the Soviets left. Neither they, nor their supporters in Pakistan, created a strategic plan or unified structure for the post-Soviet era. This, combined with the loss of a unifying principle for war – jihad against the godless Soviets – left the mujahideen fractious and squabbling. As Barfield stated, “Attracted by offers of arms and money, 20 percent of former mujahedin groups defected and joined Najibullah’s militia system, while another 40 percent agreed to ceasefires. They saw little reason to fight if their own needs were met and the Soviets were gone.”

The mujahideen also had difficulty adjusting their tactics from guerrilla attacks designed to bleed the Soviets and convince them that Afghanistan was ungovernable, to the massed, conventional assaults required to overtake and capture the country’s cities. The mujahideen’s first attempt at this – Jalalabad – was a hastily planned catastrophe that resulted in the deaths of thousands of fighters. The combination of these factors led to sharp reductions in mujahideen manpower, from an estimated 85,000 fighters in 1989 to 55,000 by 1990.

Third, contrary to expectations, the Soviets provided Najibullah with substantial support after their withdrawal. This was part of a well-crafted Soviet exit strategy that jettisoned idealistic objectives and instead aligned a de minimus goal – the survival of the Afghan government – with resources along four lines of effort: political governance, military and security assistance, economic and humanitarian aid, and diplomacy and regional engagement. Despite withdrawing the vast majority of their forces, the Soviets provided ammunition, fuel, weaponry (including thousands of SCUD-B missiles), salaries for government and militia fighters, and seven tons each of rice and kerosene for the families of fallen soldiers. The Soviets invested considerable resources and effort in creating a commando force (a Spetsnaz-type “Special Guard”) equipped with modern weaponry, intelligence, and reconnaissance assets, to serve as an on-call “shock force.” They also sought to maintain a large Afghan army, capable of deterring massed attacks on major cities, or at least stalling them until the Special Guard and airpower could arrive. They provided new aircraft (Su-22 bombers, MiG-21 fighters, An-12 transport planes, and Mi-35 attack helicopters) and maintenance, as well as training for Afghan pilots in the Soviet Union.

All told, Soviet assistance to Najibullah totaled something like $300 million per month, or roughly $3.6 billion per year ($7.7 billion in today’s dollars). According to Steve Coll, this was “at least twice the amount of aid being supplied by the CIA and Saudi intelligence to the mujahedin.” And since the Soviet patron and its Afghan client had the same goal of regime survival, the Soviets trusted Najibullah to use their aid to influence audiences to support (or at least not undermine) his government and gave him unprecedented latitude to do so.

Ultimately, when the Soviet Union dissolved in December 1991, Najibullah lost his patron. In the sudden absence of Soviet financial and material aid, his strategy of patronage and appeasement collapsed and he was forced to resign and seek sanctuary in the United Nations’ Kabul compound, where he lived under effective house arrest until the Taliban brutally killed him in 1996. As this discussion shows, however, the common assumptions that Najibullah’s government would be fragile, decisively outgunned, marginally supported – and would collapse within a year – turned out to be false. What can this example teach us about the future of Afghanistan today?

Afghanistan Today

Many of those making dire predictions about the future of Afghanistan cite the downfall of Najibullah after the cutoff of Soviet aid as a key point of reference, implying similar assumptions today as those stated earlier by Khalilzad. But are those assumptions valid? Testing them will make clear that the past and present eras are less directly analogous than many commentators assume.

First, is Ghani’s government fragile? In 1989, the war in Afghanistan had both intra-Afghan and extra-regional components. Most notably, the Soviet Union – whose invasion of Afghanistan had been internationally condemned – was backing Najibullah’s government (which it had installed), while a coalition of the United States, Pakistan, Saudi Arabia, and Israel were backing the highly fractious mujahideen. These states were pouring billions of dollars a year into their proxy war in Afghanistan. There were no diplomatic relations between the Soviets and the mujahideen. Rather, the Soviets were firmly aligned behind their client Najibullah in his attempt to survive. Despite Najibullah’s efforts to expand his appeal, his government remained widely despised and he was never able to generate a popular mandate.

Today, the war continues to have intra-Afghan as well as extra-regional components. The latter involves support to the government from the United States, its NATO allies, and India, mixed support to both sides from Russia and Iran, and support to the Taliban from Pakistan and several Gulf states. However, while all of these countries are pursuing their interests in Afghanistan, the war is not purely one of proxy conflict between them, and many of these countries – most notably the U.S. – have diplomatic relations with both the government and the Taliban. This has created a complex set of patron-client relations, which sometimes result in discrepancies between patron and client goals. The most notable example is the gulf between the United States’ goal of a power sharing peace deal that includes the Taliban and may involve substantial governmental restructuring, and Ghani’s goal of the survival of the Islamic Republic, as well as that of himself and his presidency. In addition, the most recent Asia Foundation polling numbers for the performance of Ghani’s government (from 2019) showed 65 percent of respondents expressing their satisfaction with it. That said, at least 88 percent of respondents have called corruption a problem in their daily lives every year of Ghani’s presidency – the highest numbers since 2006. The most recent survey of Afghans’ views of the economy shows an overwhelming majority see worsening situations all across the country.

Thus, while Ghani’s government will start the post-U.S. withdrawal period with an arguably stronger popular mandate than did Najibullah (though still with a fair degree of dissatisfaction), Ghani has less firm backing from his international patrons and has found himself increasingly isolated and politically threatened, as the U.S. has prioritized ending the conflict over the preservation of his government.

Second, will the Taliban have a decisive military advantage post-U.S. withdrawal? By 1989, Afghanistan’s armed forces totaled around 342,000 individuals, while the mujahideen were estimated to have about 85,000 fighters (and possibly as many as 180,000 total supporters). Both the government and the mujahideen were heavily armed: The Soviets had provided the government large numbers of airplanes, tanks and armored personnel carriers, and copious amounts of weaponry and ammunition. The CIA provided the mujahideen (via Pakistan’s intelligence service) Iraqi tanks and armored vehicles, heavy artillery, and most famously, thousands of Stinger anti-aircraft missiles. The mujahideen also had aircraft of their own. When the Soviets withdrew, the government was estimated to control only about 23 percent of Afghanistan (mostly cities and major roads). Even before the Soviet withdrawal, the mujahideen had seized the areas around Kunduz and they “controlled all posts along the contested frontier with Pakistan, as well as the provincial capitals of Kunar, Paktiya, Bamiyan, Takhar, and Laghman.”

Similar to the post-Soviet era, Afghanistan’s security forces currently number about 308,000 individuals, while the Taliban are believed to have around 60,000 core fighters, with manpower totals (including militias and supporters) possibly as high as 200,000. The country’s security forces again include a top-notch Commando force, a capable air force with a mix of over 100 fixed- and rotary-wing aircraft, a large army with modern equipment, and a capable state intelligence and security service. At present, the government is estimated to control only about 30 percent of the country’s districts.

Unlike the mujahideen, however, the Taliban lack notable conventional armaments, such as aircraft, tanks, and anti-air missiles, and are thus less capable of massed assaults on cities and completely vulnerable to attacks by air. On the other hand, the Taliban are a more cohesive organization than the mujahideen, as a result of years of focus on solidifying the structure of their forces. These differences account for why a recent net assessment of Afghanistan’s security forces versus the Taliban concluded that in the absence of U.S. advisers, the Taliban would have a slight military advantage, but not a decisive one. Additionally, while the fraction of the country under government control today is similar to what it was in 1989, Ghani’s government controls all of Afghanistan’s major cities and the Taliban control only about 20 percent of the country (the other 50 percent is considered “contested”). Thus, militarily, Ghani’s government will start its post-withdrawal situation in a slightly better position than did Najibullah.

Third, will support from the United States and the international community be insufficient to prevent collapse? The Soviets crafted a well-synchronized and resourced exit strategy in support of a capable client with a free hand to use that support as he saw fit. Today, the U.S. has indicated that while its troops will depart, its commitment to Afghanistan will continue in the form of security, economic, and humanitarian assistance. Of course, White House words are not always backed by congressional checks. But by the time the U.S. leaves Afghanistan, Congress will likely have appropriated another $4 billion or so for the country’s security forces, via the Afghanistan Security Forces Fund. This is a two-year appropriation, meaning the U.S. will have effectively pledged security assistance support through fiscal year 2024. For its part, NATO has pledged support through 2024 as well.

Those resource commitments notwithstanding, what is not clear is whether the United States has a well-crafted and synchronized post-withdrawal strategy. Nor is the U.S. likely to give Ghani a free hand to employ that assistance purely in support of his goals of governmental and political survival. Due to the extreme levels of corruption in Ghani’s government, the U.S. will instead likely establish a Security Cooperation Office inside its embassy and continue providing the bulk of its assistance “off budget” (i.e., not directly to the government) with oversight by U.S. officials and various inspectors general. Thus, Ghani will enter the post-withdrawal period with a high degree of confidence for sustained funding for at least two more years, but his ability to use that funding in direct support of his own political strategy will be highly circumscribed.

What the Future Holds

What does all of this mean for Afghanistan post-U.S. withdrawal? Two observations are clear.

First, as was the case with Najibullah’s government, those predicting the quick demise of the current government are likely to be wrong. Ghani will enter the post-withdrawal period in a slightly better military position than Najibullah had, though it is likely that this position will continue to erode in rural areas and those surrounding the cities, as a result of the Taliban’s slight military advantage in the absence of U.S. and NATO forces. The Taliban – many of whom believe they have already won – are likely to press their military advantage in 2022, to see how far they can get against Afghan security forces in the absence of U.S. advisers and air support. That said, the Taliban likely do not have the manpower or the conventional military resources to seize and hold more than a few provincial capitals – and any massed attacks on the cities are likely to incur the same vulnerability to airstrikes that the mujahideen faced in their assault on Jalalabad. They are also likely to struggle with at least some splintering of their forces once the unifying theme of driving out foreign infidels is gone. In other words, 2022 is likely to be another violent year for Afghanistan, but the Taliban are more likely to increasingly lay siege to the country’s cities in order to generate political concessions in Doha than they are to “empty the madrassas” in Pakistan for a Tet-style gambit to seize the country by force.

Second, Ghani’s fortunes are unlikely to be decided militarily. If the Taliban are successful in strangling (and possibly capturing one or two) major cities, Ghani’s demise may be accelerated. Conversely, if Afghanistan’s security forces can independently withstand the Taliban’s pressure and protect the cities, Ghani’s hand may be strengthened and his political fortunes may improve. Before that, though, he will need to survive the political fight that is unfolding today. Afghan warlords are already organizing forces as part of a new “national resistance.” Meanwhile, Ghani’s counter-proposal to the United States’ suggested interim government and his insistence that the only path to a new government is through elections has been roundly dismissed as unworkable. In short, it is unlikely that Ghani’s government will be overthrown purely militarily in the next couple of years. Rather, political dynamics will decide his future and that of his government.

Given these observations, it seems that predictions of the imminent demise of Afghanistan’s government are as precariously reasoned as those of Najibullah’s time. However, this is not to suggest that such an outcome is impossible. There are several key indicators to watch over the next year or so.

Militarily, Afghanistan’s security forces will enter the post-withdrawal period with certain advantages over the Taliban. The most notable of these are its air force, Commandos, size, and heavy weapons. The attrition rate of aircraft and heavy weapons will be key metrics to follow, as will trends in Afghan security force recruitment and retention. The extent to which the U.S. and its NATO allies continue to provide security assistance funding and replacements for combat attrition of aircraft and weapons, as well as their ability to provide or entice contracted support, will also be key to watch. For the Taliban’s part, the extent to which they choose to directly threaten or attack provincial capitals – and the extent to which they are successful if they do – as opposed to laying siege to them, will be critical to gauging the relative dominance of their political and military wings. Most important will be what happens to Ghani’s political coalition. The extent to which political elites currently aligned with or tacitly supportive of him stay that way, or realign to other political blocs (including the Taliban), will likely be the most important indicator as to whether the government will survive or collapse after the U.S. withdraws.