Lawrence Freedman: This war reveals the limitations of American military strategy

'Air power works best in conjuction with ground forces – and that means finding reliable allies'

As a result of the past couple of months, the Pentagon is now confident there are few enemies that cannot be battered into submission through the application of carefully targeted but also overwhelming air power. They largely accept, after the Kosovo and Afghanistan experiences, that air power works best when used in conjunction with ground forces – to oblige the enemy to occupy open positions, to identify targets and to follow through after the bombing. Because they are reluctant to put their own troops in harm's way they need local allies. In this case they were fortunate in the availability of the Northern Alliance.

As a result the Americans are emerging from yet another major conflict with few casualties. Of the seven US personnel killed, more were caught by friendly fire than by the enemy. The Western media suffered greater losses. Afghan civilian casualties were substantial but precise numbers are hard to find. Reports from places such as Kandahar confirm that generally the American bombing was accurately targeted.although the power of their bombs are such that, when they do hit the wrong target, the effect is horrendous.

It is almost certainly the case that the combination of the US Air Force and anti-Taliban warlords, backed by American and British special forces, meant the fighting was far less bloody than would otherwise have been the case. This was largely because of satellite phones, which enabled intense bargaining among the Afghans so that Taliban commanders felt able to agree to defect or disperse, and occasionally to surrender. Underestimating the importance of these deals led so many commentators, including myself, to overestimate the capacity and readiness of the Taliban to resist.

If the Americans had been acting on their own, not only would they have taken far longer to get their forces in place for any ground offensives but, since their cultural disposition is to demand unconditional surrender, there would have been few deals. In addition, they did not attract the anti-foreigner sentiment to themselves: instead it became focused on the al-Qa'ida contingent.

The Americans may have been relieved by the speed of the Taliban surrender but they did not always appreciate its conditional quality, and we have yet to see the full impact on efforts to bring the Taliban and al-Qa'ida leadership to book for their past misdeeds, let alone on the future governance of the country.

Remarkably few Taliban fighters appear to have been disarmed and many appear to have drifted back, still armed, to their villages or into banditry. Despite the Defence Secretary Donald Rumsfeld's insistence that members of al-Qa'ida must be captured or killed, a number seem to have been ransomed and are now out of the country. When the Americans objected to attempts to organise conditional deals in the final battle for Tora Bora, in part because they bought time for those attempting to escape, the obvious answer was: if you want them so badly, you go and get them.

The heart has certainly been taken out of al-Qa'ida and its operational capacity severely degraded, if not fully eliminated, but Washington still needs Osama bin Laden "dead or alive". Mullah Mohammed Omar also appears to have vanished. It was the failure to finish off Saddam Hussein that took the shine off George Bush snr's victory over Iraq in 1991.

Furthermore, whatever the quality of the government being established in Kabul, the immediate consequence of the war will have been to localise power and encourage lawlessness. The Pentagon at times gives the impression that its obligations to any country ends once it has rooted out the bad guys, but the State Department appears to understand that the military achievement will be diminished if a degree of real stability is not now brought to Afghanistan.

Another reason for not rushing to quick judgements about the lessons of a particular war is that what works in one set of circumstances may not work in another. Last week Mr Bush spoke enthusiastically about the combination of "real-time intelligence, local allied forces, special forces, and precision air power", adding that this conflict "has taught us more about the future of our military than a decade of blue-ribbon panels and think-tank symposiums".

Yet there are many reasons to question this approach. Your campaign is only going to be as good as your local allies. They may not always be available or reliable. If the US does decide to return to Somalia, it will have to do better with the local warlords than it did in 1992-3. Evaluations of possible operations against Iraq depend on whether the Shia and Kurdish oppositions to Saddam's regime can ever amount to much. If your allies flounder you may end up without any sensible options, while an excess of enthusiasm may bring with it guilt through association with massacres and plunder.

A strong presence on the ground is necessary to sustain political influence and to prevent a country that has been brutalised through decades of warfare collapsing back into anarchy. Handing over political influence to whatever group may be prepared to work with you, an action probably based on opportunism rather than adherence to Western political norms, may well cause problems for otherwise-friendly local neighbours. Furthermore, by giving prominence to their low tolerance of casualties, the American create an incentive for their enemies to target any forces that are accessible.

This presents a sharp contrast with the view of the British Government that properly applied Western armed forces can serve as a force for good, so long as they are prepared to establish a robust presence on the ground. Tony Blair's readiness to authorise British deployments into Kabul almost as soon as the city fell to the Northern Alliance reflected this belief, and also his view that unless the West engages fully in the political and economic life of states such as Afghanistan then they will continue to fail and continue to cause trouble for the rest of the world.

The Americans may accept that they dare not ignore the more distressed and tumultuous regions, but they are still disinclined to volunteer for humanitarian operations and "nation-building". They prefer to provide funding, occasional diplomatic muscle and logistical support.

Their dependence on suitable local allies will impose limits on what the Americans can achieve militarily in the future. It may be that a sense of limits is no bad for thing for a superpower, and recognising that somebody's ground troops are needed is at least an advance on the assumption that air power can achieve all strategic objectives by itself.

lawrence.freedman@kcl.ac.uk

The author is Professor of War Studies at King's College, London