Thus, beginning in 2005–2006, the officers in charge of the region launched a counterinsurgency program that is interesting both for its theoretical approach and its results.17 If we follow David Kilcullen, in the face of a politically fragmented opposition and in a tribal region, the strategy adopted was to use the construction of a road between Jalalabad and Asadabad as a means of signing agreements with local tribes and marginalizing the most aggressive (non-local) fighters, who came from Pakistan. Kilcullen explains that the remarkable successes of this strategy can be applied elsewhere. Certain provinces, notably Kunar and Khost, received a massive infusion of American credits through Provincial Reconstruction Teams (PRTs). Those two provinces had the largest PRT budgets in 2008–2009: $80–90 million apiece.18 Numerous development projects have been put in place in the Kunar Valley since the 1980s, particularly irrigation projects and roadwork (embankments built by PRTs). The PRT of Gardez, the first established in Afghanistan, spends tens of millions of dollars each year on infrastructure projects.19
What is the impact on security? In reality, there is no proven correlation or simple causality between aid levels and security.20 Experience shows that in the most stable places, there are regions that have not received aid, or very little, and that regions that have received the most aid are often very unstable. As a general rule, the “proposal of aid/establishment of shuras/resolution of conflicts/support for the government” model does not work. In the case of Kunar, despite Kilcullen’s misleading assessment, the situation actually deteriorated rapidly after 2008. With no possible control over the border with Pakistan, the infiltration of militants continued freely. In those circumstances, it was hard for the tribes to dissociate themselves from the insurgency, should they have wished to do so.21 In reality, the effect of aid on security can only be marginal because it does not significantly strengthen state structures and most of the district capitals are practically without means.
That’s a pretty hefty quote from Gilles Dorronsoro’s latest paper from Afghanistan, but he’s describing something important: the abysmal failures of COIN “gurus” like David Kilcullen. This is something I’ve harped on for years, on Kilcullen in particular for his insistence things in Kunar worked when they were really an absolute disaster. In a paper I have coming out soon, I present a literature review on aid in counterinsurgency, and there is almost universal agreement that development work has no affect on COIN. It just doesn’t apply. Yet that was the heart of Kilcullen’s COIN theories, which he managed to enact in Afghanistan for years.
Yet within the DC-military complex, you fail upward. This paper won’t change a thing. But it’s important to have it out there anyway.