Here's some more from Derrick Crowe at FireDogLake; just another excerpt:

Note that all of these statements deal with the importance not just of the protection of civilians from killings by counterinsurgents, but the protection of the people in general. Counterinsurgency doctrine says that people aren’t going to switch to your side if they think they’ll get killed for it, no matter how low you drop the rate at which you cause civilian deaths. In other words, a drop in casualties caused by U.S. and allied groups is not sufficient for the hoped-for dynamic to take hold, according to COIN doctrine. It must be paired with an increase in security from insurgent violence as well. And that’s a problem for Spencer’s interpretation of counterinsurgency doctrine and his assessments of progress in Afghanistan, especially since the data he cites shows that in 2009:

"The escalation and spread of armed conflict resulted in the highest number of civilian casualties recorded since the fall of the Taliban regime in 2001 and in the further erosion of humanitarian space."

So, even if we just went with the information that was available early yesterday, which said that McChrystal and Co. were killing fewer civilians, they still hadn’t managed to increase security for civilians in Afghanistan as measured by the total civilian deaths caused by the parties to the conflict. The Afghans, especially those in Kandahar, know it. The elders who live in the area targeted for the next big offensive told Karzai and McChrystal they didn’t want an operation in their area and specifically cited the increased risk to the civilian population from insurgent IEDs. Even if McChrystal proved he could drive down civilian casualties when he puts his mind to it, he’s also managed to prove over the last year that he can’t protect the population.

People who claim to actually believe in the efficacy of and the necessity for actual counterinsurgency in Afghanistan need to start screaming, right now, about what’s going on in Afghanistan under General McChrystal because their credibility is now unambiguously on the line. To his credit, Spencer notes in today’s post that, "By McChrystal’s own reckoning…the system is blinking red and new measures have to be put in place…" The problem is, though, that in an honest reading of counterinsurgency doctrine should have indicated that the system was already blinking red in 2009, but for whatever reason people continued to sing the praises of Saint Stanley McChrystal and took up gross distortions of COIN doctrine to do so. Numerous prerequisites for success as articulated by COIN doctrine remained absent and/or further degraded over 2009, including host nation government legitimacy and security for the local population, yet many writers focused on one particular statistic (casualties caused by pro-government forces) because it let them tell the story they wanted to tell.

The facts are these: Not only are we not protecting the population generally, but we’re demolishing progress made on decreasing civilian deaths attributed to us and our allies. We’ve doubled the number of special forces in the country, forces responsible for some of the most outrageous, alienating incidents of the war. We don’t have a legitimate local partner or a legitimate host nation government. And after paying lip-service to getting local buy-in for a Kandahar operation, McChrystal’s people now inform us that we plan to go ahead whether the people of Kandahar like it or not. McChrystal is letting the COIN pretensions fall away as the reality of the Afghanistan war reveals them as the hypocritical bullshit they always were. What’s left is the uncompromising and ugly truth: we are fighting a brutal war in Afghanistan, it’s going badly and we don’t have a credible prospect for a turnaround.

Spencer is right, the system is blinking red. It’s been blinking for years.