Sounds good until you hear the details. In the case of Afghanistan, it doesn't matter if local politicians and bureaucrats want to develop western-style institutions; they are the elite and are by definition corrupted by their exposure to the west. The new theory is that we need to send in teams of sociologists and anthropologists to study indigenous power structures and determine how to transfer authority to them. It took the West 300 years to evolve democratic systems, the argument goes, so there should be no attempt to impose a quicker time frame on countries in conflict. We should work with the war lords and the Taliban to rebuild the country, and not fuss about corruption: our Western disdain for third world corruption is a prejudice born in ignorance of the way other cultures get things done.
In a way I'm predisposed to agree with this argument. My work in Africa led me to believe that much of Western aid does a lot more harm than good. But the idea of teams of Western academics imposing their latest ideologically-based theories on vulnerable populations is a far, far scarier notion. And any argument that dismisses the importance of the middle class in creating a stable society is just plain crazy. A grand experiment that has the West funding warlords to try to bolster feudal institutions makes me shudder.
I can certainly see the need for sensitivity to local power structures. And in some cases, like perhaps the more moderate sections of the Taliban, it may be necessary to do business with the enemy. I don't see that we have to be so morally pluralistic that we can't agree on any norms at all, and must accept whatever our researchers say the local population believes - especially when the researchers exclude the educated, working population from their study. I have no doubt they will over-emphasize the views of women and groups that look particularly ethnic, after indoctrinating those groups with their own, western, politically correct notions.