RAND on COIN

I know. The tide of war is receding. The Army has been told that it won't be needed any more. But the world goes on despite soothing words from the White House. So RAND updated their study of counter-insurgency practices.

RAND couldn't say what ratios of troops are necessary because their data isn't good enough to know the ratios. So the 2% rule (you need security forces equal to 2% of the population) is a guideline rather than something based on hard data.

But local self defense forces and police have a role in defeating insurgents. So whether COIN forces need 1% or 4%, don't make the mistake so many made during Iraq and Afghanistan of assuming only US forces counted in adding up the percentage. Throughout the Iraq War, I defended the troop numbers as sufficient to win as the fight dragged on.

One factor that is more easy to recognize and one that I repeatedly mentioned was the need to "atomize" (my term--not RANDs) the enemy so that they cannot fight as conventional forces. If government forces are strong enough to compel the insurgents to fight as guerillas, the COIN side is more likely to win.

More important is their conclusion that using harsh violence to brutalize a population into submission is highly correlated with defeat for COIN forces. Already outnumbered by the anti-war side, I felt outnumbered by the pro-victory side's proponents who wanted us to "take the gloves off" and use maximum violence.

I always wanted focused violence on actual enemies and noted that the "crush them" strategy can only work if you are willing to commit actual or near genocide on the population that supports the insurgency. Aside from the moral problems of doing that, practically speaking we couldn't do that as a nation even if such a strategy was optimal.

Nor did I want to ask our troops to fight that way. It was not "supporting the troops" to make them into war criminals who'd have to live with what they did when they came home.

So this study reinforces the old wisdom that maximum indiscriminate use of force does not work well.

An interesting part is that foreign conventional support for COIN forces is not associated with defeat for COIN forces. So although we still want to turn over COIN duties to locals simply because we'd rather they fight their war than our people, our participation does not make it less likely for the COIN side to win. That is, we don't "taint" the government side, as our left likes to assert.

Further, external conventional support for an insurgency is a war winner for insurgents (again, no "taint" as I've long argued) unless the COIN forces have external conventional support, too. This could be support for providing air support to rebels in Syria, although external conventional support is not a requirement for insurgency victory. But if we did provide air support, Iran and Russia might understand that they must provide conventional support, too, if they hope Assad can survive our escalation.

Anyway, regardless of what we intend, it might be useful to keep this knowledge handy. Enemies tend not to cooperate with our ideas of what wars are acceptable to us.