Afghanistan’s central government also poses a big question mark for the country. The government almost surely will be weak—a consequence in large measure of President Karzai’s two terms in office. His government has been undermined by corruption, familial and Pashtun nepotism, and a failure to engage consistently with the wider Kabul polity. At the provincial level and below, Karzai’s political situation is not much better. Lack of effective government and the Taliban challenge have undermined his standing, and his support among Pashtuns in the South has declined precipitately.

There are surface parallels between Karzai’s attempt to function as a national leader and the leadership of Mohammed Najibullah, head of the Soviet successor state in Kabul. Najibullah also sought to bring the nation together through his national reconciliation and pacification program of the late 1980s. But Najibullah was a far more effective national leader who understood and engaged with the wider societal and political issues of the day in a manner that Karzai has not been able to do. His People’s Democratic Party of Afghanistan was a political party with national reach; the Karzai years have not spawned any such party organization with comparable influence. The 2011 formation of the Truth and Justice Party, which seeks to represent a broad range of ethnic groups and ideological positions, is a belated attempt to put this right. But however rapidly the party develops, it isn’t likely to challenge successfully the well-established local and regional power brokers or take on the Taliban in the South and East.

Still, the events that preceded Najibullah’s fall in 1992 have a depressing contemporary resonance. His government fell after the collapse of the Soviet Union led to the end of Soviet support for the Kabul regime. But even before that, the country’s extended crisis contained elements of corruption, financial collapse, scarcity of resources and chronic overdependence on foreign support. In 1988, 75 percent of Afghan state revenue derived from projects dependent on Soviet support. And in 1991, internally generated revenue provided just 30 percent of a declining GDP. The inability of today’s Afghanistan to generate the revenue to meet the financial burden of maintaining an expanded national army and police force is eerily reminiscent of the immediate post-Soviet era, after the abrupt collapse of Soviet support left the country economically on its own.

A joint Afghan and World Bank report issued in November 2011 stated that, assuming effective development of Afghan minerals and national economic growth of 5–6 percent a year for a decade, expenditures would still exceed GDP by some 25 percent, or $7.2 billion a year. Even when the cost of maintaining the security forces is removed, spending exceeds income by 11 percent. And this is based on the assumption that security will improve sufficiently to allow for the exploitation of mineral resources. That may not be a realistic assumption. The Senate Foreign Relations Committee issued a similarly pessimistic report. The international community may make available greater support than the USSR was able to give Najibullah, but it still may not be enough to offset the weakness of any national government following Western disengagement, not to mention the existence and durability of the Taliban. A more ideologically coherent opponent than the mujahideen of the late 1980s and early 1990s, today’s Taliban is the dominant indigenous Afghan political influence in the South and East.

Then there are the twin issues of personality and deeply ingrained behavioral patterns at the national level. One might think that the civil war of the 1990s and the subsequent Taliban government, followed by the Taliban overthrow, would have brought new governmental and political players to the fore. But the Afghan national polity seems to be dominated by the same people as before, and it is striking how difficult it is for outsiders to break into it, even in the face of these major traumas. Many of the key figures have been major players in Afghan national politics far longer than Karzai, which may account for some of his difficulties. With the exception of Ahmed Shah Massoud and Burhanuddin Rabbani, both killed by the Taliban, almost all the key players of the 1990s remain active today.

But another important reality is that none of these men has obvious credentials as a potential national leader. They are distinctly ethnic or regional players. The result is that there has been a dearth of alternative potential national leaders. This reflects, in part, the ethnic and local nature of Afghan society and politics. But deliberate policy comes into play as well. Karzai raised concerns among American policy makers in 2010 when he sacked two top governmental officials—Interior Minister Mohammad Hanif Atmar and intelligence chief Amrullah Saleh—after they failed to prevent an attack on a Kabul peace council. Such actions belie any idea of an orderly political succession. Thus, the collective behavior of the Kabul polity is likely to revert to that of the early 1990s—jockeying for individual and ethnic advantage as well as the formation of unstable, shifting alliances susceptible to external exploitation and military pressure.

It should be noted that, in Najibullah’s day, the expectation that Soviet withdrawal would precipitate a large-scale and successful mujahideen offensive effectively undercut Najibullah’s policy of national reconciliation. When the mujahideen failed to capture Kabul following a military defeat at Jalalabad, the idea of reconciliation temporarily gained renewed momentum. A similar pattern could emerge today as imminent Western withdrawal is sensed. But a major Taliban defeat in southern or eastern Afghanistan isn’t likely. Without active Western military partners, it is doubtful the Afghan National Army will prosecute a successful counterinsurgency campaign against the resilient and resourceful Taliban. Once the extent of Taliban political control over the hinterland becomes plain, the national army’s Pashtun soldiers could leave en masse. That would mean the struggle taking on an ethnic cast, as the residue of trained Tajiks (overrepresented within the officer corps), Panjshiris and Uzbeks assume the bulk of resisting any Taliban spillover from the Pashtun areas. Thus there is a strong possibility that the country will return to the politics and conflict of the 1990s, characterized by ethnic and geographic divisions and passions.

It is premature and perhaps unduly pessimistic to talk of a Taliban protostate in southern and eastern Afghanistan. But after 2015, the Pashtun South and East will almost inevitably come under increased Taliban influence. Taliban strength and resilience are based as much on a natural affinity with the population as on intimidation or the Kabul government’s weakness. Before 2010, each successive attempt to extend control and governance was followed by Taliban success in retaking that territory. Away from the areas of direct Western military control, Taliban “shadow governance” is far stronger than the writ of Kabul. It is true that the surge of American forces in southern Afghanistan has produced significant tactical gains, and Afghan forces, mentored by Western soldiers, have begun to perform more effectively. But once Western military forces are removed, Taliban influence and control will likely expand once again. The models of provincial governance imposed or attempted by the West are not sufficiently deep or rooted to endure in Pashtun-majority areas.

In Helmand, the residual British model, based as it is on an external technocrat, effectively relied upon one man, Governor Gulbuddin Mangal, for several years. Even without a Taliban challenge, in the absence of Western military forces, local rivals with genuine roots in Helmandi society such as Sher Muhammad Akhundzada would have rapidly engineered Mangal’s removal. These men draw their power and authority as much from business interests, including narcotics, as from any traditional tribal structures or patronage networks, which were substantially destroyed in the Soviet occupation. However, despite their ability to raise and arm militias, they are unlikely to be any more effective in resisting the Taliban’s political and religious appeal and military power in 2015 than they were in 1994–96. By 2011, the structures of governance underpinning Mangal were more resilient, but their viability in the absence of the security provided by Western soldiers remains questionable.

There are few reasons to anticipate durability in the U.S. model in the Pashtun areas of eastern Afghanistan. The practice was to install a strongman from outside the province as governor. In Nangarhar, Governor Gul Agha Sherzai established a credible level of security. But then bombings in a Jalalabad bazaar in 2010 demonstrated just how tenuous that security really was. The resurgence of Taliban influence was starkly illustrated by an attack on Kabul Bank in Jalalabad in early 2011. Residual tribal structures are stronger in the East than in the South, but there is little evidence that tribally based militias could resist a reversion to Taliban control. Western withdrawal would thus almost certainly be followed quickly in both the South and East by restored Taliban influence.

The Taliban will probably also increase its influence in areas of mixed ethnicity, such as Wardak and Logar, near Kabul. In the 1990s, these areas formed the initial boundary between Taliban and governmental forces. Since 2008, the Taliban, using the capabilities of the Haqqani network, has infiltrated suicide attackers through Wardak and Logar to targets in Kabul. The influence of the warlord Gulbuddin Hekmatyar and his militia, loosely affiliated with the Taliban, is also strong in the area.