This article on the American failure to listen to the will of Afghanistan falls into the category “the plural of anecdote is not data” — which basically means that what you and your circle see/believe/experience is not absolutely going to be representative of a wide population. I am sure some people in Afghanistan were happy with the US occupation. Some were probably happy with the Russian occupation a few decades back too. Does that mean the majority of Afghan citizens want US troops there? No way for us to know. And, even if there were accurate public opinion surveys available … would it matter?

Since the second George Bush, I’ve thought it’s a bad idea to start saying “well, a simple majority of the population doesn’t like the government, that means it’s a-OK for us to invade and depose that government”. Because I’m pretty sure GW2’s approval ratings were well under fifty percent at many points in his presidency. That mean any other “well meaning” country could invade and liberate us from our unwanted government?

I don’t even think the intel community was seriously surprised by the post-withdrawal results. There’s a meta component to publicized intel analysis — what we say about a situation can influence the situation. Could we realistically publish a document saying “OK, we blew a trillion or two over here and spent a decade or two training their military … it’s all gonna fall apart within two weeks of us leaving”?! Of course not — that becomes a self-fulfilling prophecy. We’ve got to communicate confidence in that government and military. A week later? We have to act surprised. I’m sure there were position papers that included the not unlikely scenario of a complete collapse.