The iconic photos of the last helicopter leaving the roof of the Saigon embassy vividly encapsulated the collapse of the US effort in Vietnam. And there has been no end of commentary comparing those earlier, fruitless US efforts in Vietnam with the current, troubled effort in Afghanistan. Building Afghanistan is seen at best as a race against time, but often as a simple refusal to acknowledge the inevitable. The Taliban smell victory, Pakistan hedges its bets, Afghan President Karzai has lost faith in the US ability to defeat the Taliban, and so have many Americans. There are visions of another helicopter on a roof, of Taliban marching into Kabul as the North Vietnamese did into Saigon, or even as Castro did in Havana. The furies are loose.

Analysis of past insurgencies and how they end identifies a wide variety of pertinent factors such as outside support, troop ratios, and sanctuaries. But there are so many common factors and such a wide range of relevance, that each insurgency is essentially unique. Few of them provide helpful parallels for Afghanistan, which is a fragmented nation with several major ethnic regions and dozens of tribal groups. Bosnia provides one of the few examples of multi-ethnic conflict and it was essentially a three-sided civil war with multiple outside supporters. The only really useful model for Afghanistan is Afghanistan and the insurgent campaign which eventually drove the Soviets from the country. But the Najibullah government they left behind survived for over two years, only collapsing when a major supporter (Rashid Dostum) switched sides. Afghanistan then devolved into violent internecine warfare. So the Soviets were the clear losers, but there was no clear winner. The Taliban eventually came to power, but never did succeed in controlling the entire country.

Presently, the prospects of some kind of total collapse of the Afghan state are minimal, regardless of the Taliban resurgence.

• The Taliban ability to seize control of the state in 1995 was due to a total lack of US involvement and a concerted Pakistani effort to support a new regime beholden to Islamabad, as well as noninvolvement by other regional powers, particularly the Soviet Union which by then had had enough of Afghanistan. That will not happen again. The United States will certainly continue to play a role, if only as a military spoiler and supporter of local commanders in areas not under any Taliban control. Pakistan is restrained, though hardly benign. Russia, India, China and even Iran are all hostile to the Taliban and unlikely to simply acquiesce in a new Taliban regime in Kabul.
• During the Taliban rule in Kabul, the north was never totally pacified. Even with minimal outside support, local commanders led determined resistance to Taliban control. Mazar-i-Sharif did not fall to the Taliban until 1998, and it was the first major city recaptured by the Northern Alliance when the United States initiated hostilities in 2001. Any major Taliban triumphs in the south or east of Afghanistan would certainly lead to significant outside support of the non-Pashtun groups in the north and west.
• Kabul undoubtedly has Taliban sympathizers, and the Taliban have demonstrated a capability to disrupt activities in the city. But while Taliban forces, mostly Pashtun tribal elements, might be able to seize control in some of the fundamentalist Pashtun areas in the south, there are no waiting manpower resources to support columns pushing into Kabul. Unhappiness there with the government does not necessarily translate into support for the Taliban, whose repressive rule is still fresh in memory. A collapse of government control in areas of the south or east would be unlikely to quickly translate into a loss of control in Kabul, or in other major cities.
• Rather than melting away in the face of some Taliban onslaught, regional leaders, who fought the Taliban before, would almost certainly do so again. Anti-Taliban sentiment remains strong in wide areas of the country; resilient local officials can draw significant support from the populace. This is particularly true since many of them would undoubtedly receive outside support. In fact, the economic development which has already taken place provides strong incentive for regions to tenaciously resist any Taliban control.

So what would a Taliban "victory" look like?

For starters, it would certainly include loss of government control in areas of the south and east, possibly including Kandahar. It would almost certainly be a rolling victory, a spread of Taliban control from the most fundamentalist areas toward more developed areas, meeting a steadily increasing resistance. Taliban authorities would certainly have their hands full consolidating control over areas they might dominate. Recognizing this, they have embarked on a broad national effort to enhance their organizational structure and address the concerns of Afghans who are unhappy with their tactics. In an unusual move, the brother of Taliban leader Mullah Omar has criticized the burning of hundreds of schools, complaining that it merely increases illiteracy. Mullah Omar even has called on Afghans to respect the country's different religions. In a ten-page statement issued in November 2009 he also pledged that a future Taliban regime would bring peace and noninterference from outside forces, and would pose no threat to neighboring countries - implying that al-Qaeda would not be returning to Afghanistan along with the Taliban. While this certainly raises credibility challenges, the Taliban would get no obvious benefits from a continuing al Qaeda association and are well aware that it was precisely this association which brought about the destruction of the earlier government. At any rate, the Afghan Taliban appear much more focused on local issues than their counterparts in Pakistan.

Although the Taliban certainly have a capability to disrupt activities in Kabul, as well as the north and west, they have no demonstrated ability to take control of these areas, particularly in the face of continuing outside support to anti-Taliban groups. Once again, local commanders and tribal leaders would actively oppose Taliban inroads, this time with significant outside support, including continuing development support. Obviously the extent to which continued struggle would be violent would also depend on what sort of negotiated settlements or cease fires might be arranged. But the entire country is clearly weary of fighting and the Taliban would have every incentive to minimize disruption of their own consolidation efforts, so some kind of power sharing agreements are highly likely.

It is certainly not clear what a Taliban success would mean, but it is clear what it would not mean - a unified Afghanistan under Taliban control. Rather there would be some kind of fragmented Afghanistan. A residual central government would certainly have a greatly diminished authority; regional authorities - almost all non-Pashtun - would work hard to consolidate their own positions. The West would have every incentive to provide assistance, including development assistance, to buttress their own local support. In fact, significant economic development in non-Taliban areas would greatly stress Taliban control and serve as a force for moderation. Afghanistan would revert to tribal domains and could remain fragmented for many years.

A diminished central government and a fragmented nation would not pose any direct strategic threat to the United States; an al Qaeda presence is of course possible, though an established base of operations in the Pakistan tribal areas would seem to serve them better - particularly since the Afghan Taliban seem to be so locally oriented. At any rate, the only significant al Qaeda threat is some kind of weapon of mass destruction (WMD); Afghanistan offers no prospects of supporting such an effort.

The most significant impact could be demonstrating the United States as an unreliable ally, already a serious problem with Pakistan. This impact would certainly be smaller if US withdrawals were part of agreements negotiated by the central Afghan government, especially if there were continued and effective US assistance to non-Taliban areas. A Taliban triumph in Afghanistan would certainly exhilarate the Pakistani Taliban and increase tension in that already unstable country. But the extent of coordination between the Pakistani intelligence service (ISI) and the Taliban remains murky. So an area of de facto Pashtun control in the Afghan border areas might well be matched with Pakistan once again ceding control of its border areas to local elements there, allowing for stabilization and development in the rest of the country.

The future of course remains unknowable. But the past gives little reason to expect some kind of total control of Afghanistan by the Taliban. The underlying strategic challenge to the United States is not retaining close control over all of Afghanistan, but demonstrating to the world in general and the Muslim world in particular support of those who wish to better their lives. So military efforts may draw down in the year or so ahead, but development commitments need to be for the long term. They need to emphasize grass roots development and transparency of efforts. Afghanistan is a focal point of that effort, but it has to extend more broadly to the entire Muslim world.