The latest analysis from what used to be known as the Senlis Council says that 80% of the territory of Afghanistan currently experiences "heavy" insurgent activity. 17% experiences what they call "substantial" insurgent activity. And a mere 3% of the territory, in a region called Sari Pul where the dominant language is Dari Persian and the dominant ethnicity Uzbek, experiences only "light" insurgent activity. The number of insurgents, as estimated by the US, has risen from 7,000 in 2006 to about 25,000 today, which slightly more than the total number of insurgents reported killed.

The figure offered by the US seems likely to be a sizeable underestimate. This 25,000 or so insurgents are supposed to be ranged against almost 65,000 ISAF troops, 45,000 non-ISAF American troops, 9,000 British troops and purportedly 100,000 members of the Afghan National Army (most of whose troops are probably working for the ruling pro-US warlords). The implication is that a combined army of over 200k troops armed to the teeth and with godlike aerial power to back them up can't thwart an insurgency of an eight of the size with comparatively poor weapons and no air force. There must be a substantially larger hardcore of insurgents, and a very large periphery in the supporting population. This is what is so illogical about the continued pretense by US-led forces that their foes are an unpopular rump. They may once have been, but evidently now command the loyalty of broad social layers, perhaps comprising a majority in places such as Helmand. Still, if the figures nonetheless correctly identify a trend, then the insurgency has more than tripled in size since 2006.

Not only are the insurgents growing in number, the sophistication of their attacks is increasing. For example, a recent attack on a military outpost in Nuristan killed eight American soldiers. Another attack on a UK base in the Helmand province killed a British soldier. These are just samples from the dozens of weekly attacks that strike occupation forces. Now, Obama - anxious to justify that Nobel prize, no doubt - is looking at the idea of buying off a section of the insurgency, just as Bush was able to do with a layer of the Iraqi resistance. The alternative is the McChrystal plan of sending up to 60,000 more troops, which is known to divide the Democrats and will force Obama to rely on GOP support if he wants to push it through. The assumption behind the idea of paying insurgents to fight on the American side, though, is that the majority of those fighting the US take up arms because it pays well. Perhaps that's true of some, but the reality is that what has escalated the insurgency from being a relative nonentity into a force that could (so military leaders predict) defeat the combined occupying forces is the mode of rule and repression that the US has developed. The client-state of warlords, the air war, the selective 'war on drugs' are all mainstays of the occupation, and can't easily be dispensed with. Moreover, the success of this strategy in the 'Sunni triangle' depended on the occupiers' ability to coopt the leadership of some of the disarticulated networks of military resistance that characterised the Iraqi insurgency. The leadership of the insurgency is nowhere near as divided in Afghanistan, and the 'neo-Taliban' are waging a smarter war than those fragmented groups that have been fighting in Iraq. The only realistic option for those still committed to this war is escalation. However, that then raises the question of whether America's allies are prepared to throw in more troops and money - an issue over which NATO has divided before.