Despite every complexity of the counter-insurgency in Afghanistan, and each pained effort to counterbalance the two great strategic obstacles — the weakness of the Kabul government and the position of Pakistan — to its durable success, the final say on the war's outcome will be decided by one blindingly simple element: money. Since its inception as a sovereign entity, Afghanistan has always been a rentier state, economically dependent on revenues from foreign patrons paying for influence. Money has proved a far more effective tool in denying Afghan soil to rival powers than the military has ever been. For long periods between their first three Afghan wars the British managed to pacify Afghanistan through subsidy, and for as long as they were bankrolled by Russia Najibullah's Communist regime proved surprisingly resilient, even after the Soviet withdrawal. After 2014 it will be money again, rather than a political settlement, which will dictate Afghanistan's future.

There is no sign that the Afghan government will be ever able to fund its forces independently; to date the ANA is entirely American-financed. The US pays, equips and trains them at a level of expense that   exceeds its military assistance to Israel. The US Department of Defense has spent $20 billion on the ANA since 2002. A further $7.5 billion were requested for the 2011 fiscal year. But after the withdrawal of their troops, given the shaky state of the global economy, for how long and to what level will America and Nato governments be able to continue to finance their influence there?

Fear, the dynamic that has always attracted foreigners to Afghanistan, will govern the answer. Every Great Game player of the past two centuries, whether Russian, British, American, Pakistani or Indian, has been drawn to the Afghan board by fear of their opponent's presence there, and has been repelled only when the fear of the continued cost of that involvement outweighed the sense of its reward. In assessing how best to calibrate its own presence in Afghanistan after 2014, and thus decide Afghanistan's mid-term destiny, America has to work out how much it fears the ungoverned spaces that will result from a resurgent Taliban and civil war, then fund its post-2014 vision accordingly.

"Not losing" may prove the least unattractive option on the table, the only viable alternative to writing off the thousands of lives and billions of dollars so far invested there, and accepting civil war in Afghanistan as the consequence of a total withdrawal. In practical terms "not losing" will mean financing the army of a corrupt and vacillating government indefinitely to contain the conflict as best it can. Afghanistan will not see peace for many more years, another generation at least, and will remain on the lengthening list of nations around the world with unresolved conflicts.