« My Solutions to "Inte… | Home | I Bleed Blue and Gold… »

How to Fix "Intelligence is Evidence" in Counter-Insurgencies

With all my writing on emotions, gratitude theory and cultural understanding, I suppose I have been unclear about my thoughts on counter-insurgency. Here is all you need to know to beat an insurgency:

Don’t kill civilians. Killing civilians is the worst thing that can happen in an irregular war.

First, let’s tackle the counters to this thesis. The first comes from the “war-is-war”-iors who say, “when we go to war, we go to win and that means defeating the enemy’s will. If the population supports the enemy, than they are the enemy.” Rarely does a “war-is-war”-ior make this full statement. Instead, they generally let the second part fall as an unstated assumption. So they either actively support targeting civilians--any blog/forum that advocates nuking or carpet bombing Afghanistan falls in this category--or just don’t sympathize with a people who let “terrorists live within their ranks.” Wayne Resnick summed this up on a syndicated radio program last September, which we wrote about here.

The second counter is much more common and widespread, though equally damaging to a counter-insurgent, “Well, in war, civilians can expect to pay the price.” In other words, once a war starts, the military/invading army can’t be expected to keep civilians alive. This is more a resigned neglect; the “s*** happens” excuse if you will. I have seen this in two different comment threads, and this Lew Rockwell article describes people with this sentiment.
(The middle ground is: if you don’t want to be killed in a war, you shouldn’t allow insurgencies to happen in your country. This is vaguely the “you are with us or against, you’re either with those who love freedom, or those who hate innocent life.)

As I have written before, I completely disagree. As a counter-insurgent, or “foreign internal defender”, the U.S. military is naturally perceived as an outsider. Killing innocents, or perceived innocents, will only set back our efforts. There is a fine line between killing the bad guys to encourage them to switch sides, and killing innocents which discourages anyone from supporting the government. (More on this in our next round on “Gratitude Theory” and emotions in warfare.)

So, from the tactical perspective, killing civilians hinders the counter-insurgent. (I know that I ignored the ethics of killing civilians. As a “just war”-ior, I hold protecting non-combatants as the highest ideal. I just think I can make my argument in this case without resorting to ethics.)

As such, here are my recommendations to help our Army (and Marine Corps) wage counter-insurgencies better. They all center around the idea I put forward on Monday: the more accurate our forces are during offensive operations (“targeting”), the more likely they are to win.

1. Change the metrics to value accuracy, not totals. The most important number for a battalion should not be number of enemies killed or detained, but the accuracy of those targeting efforts. For instance, the JSOC “kill-capture” program often releases detainees within weeks of their initial detention. Senior commanders emphasize the total number of enemy captured, when they should measure the accuracy of our detentions and lethal operations. This report by the Afghanistan Analysts Network shows how commanders emphasize initial captures, and neglect to mention how many detainees get released.

2. Encourage the intelligence community to study law enforcement techniques. Studying law enforcement and the thoroughness they (mostly) use to put together cases should encourage intelligence professionals to do the same.

3. Start the position of “Devil’s Advocate”. I wrote about this when I explained “Why Intel Goes Bad: The Devil’s Advocate”. The military is a top-down organization, and a lot of commanders don’t want their opinions second-guessed. They should. An official “devil’s advocate” would help.

4. Vet potential targets with more than one source of intelligence. I’ll use The Wire, the best open source example of this. (I would use examples from my intelligence work in Iraq or infantry time in Afghanistan, but that comes dangerously close to revealing classified information.) On the show, the detectives used signal intelligence (the wire tap), human intelligence (confidential informants) and imagery intelligence (pictures from roof tops) to take down the biggest drug dealer in B-Mo. Way too many targeting efforts simply rely on a single source of intel to launch a mission in Afghanistan and Iraq (during the war). Of course, we are only as good as the intelligence we collect which leads to...

5. Get rid of all the analysts at higher headquarters. Every counter-insurgency guide says that higher headquarters should shrink in favor of bolstering lower headquarters. The U.S. military took a middle ground: it made headquarters bigger everywhere. So the division headquarters had literally hundreds of military intelligence analysts and intelligence contractors. JSOC--according to Top Secret America--runs a brigade-sized intelligence staff of 3,000 people. The battalions got some more, but still had less than ten people in an S2 shop. Yet, the battalions often produced better intelligence products for a variety of reasons, including location and ability to patrol. My advice is simple: get rid of gigantic division and corps staffs.

These ideas aren’t terribly original, and plenty of units do some or all of them as it is, but not all do, especially the more “special” people. Further, these ideas need to be enshrined in whatever doctrine emerges in the post FM 3-24 world. Like many things, I’m not optimistic.

nine comments

“Don’t Kill Civilians.” I’d suggest expanding this to how to handle civilians on the battlefield- something that we hand wave and only pay lip service to.

In combat, people are going to die. Perhaps, a better solution is to adopt a do no harm mentality.

For further expanding this out, look at the current effects in and around Iraq

1. Internally and Externally displaced people.
2. Poverty levels
3. Homelessness
4. of civilians mentally/emotionally/spiritually wounded
6. of women who lost a husband.
8. Unemployment or underemployment

During this reconstruction period (or next phase of Civil War) the group that mobilizes to address these issues will ultimately win the power of the people. Currently, Maliki is only exasperating these problems by his strongman approach and seeking revenge and hate on his perceived opponents.

Points 5,6,7 didn’t come out right.

5. Percentage of civilians physically wounded.
6. Percentage of civilians mentally/emotionall/spiritually wounded.
7. Percentage of women who lost a husband
8. Percentage of children without a father

Okay I am all about expanding metrics, and you are right. Basically, “don’t kill anyone” means, don’t displace them or wound them and understand the effects of your actions. I need to write more about metrics, but I have a personal theory that any single metric can be gamed, but if you have a range it is much harder. So, for example, if you want decrease crime, you could simply incarcerate more people. The better option is to find policies that decrease the prison population and decrease the crime rate. More to follow.

As for the “do no harm” I feel like you could add the word, “unnecessarily” to my theory. So don’t kill anyone means don’t do actions that are guaranteed to kill civilians unless you absolutely have to. This will impart more risk to the counter-inusrgent/us, but war is risky.

The theory I am wrestling with right now is the “Sovereignty Solution.” I feel that my thoughts tend to be diametrically opposed to that viewpoint. What do you think?

I don’t think that you are too far off from the thoughts behind the “Sovereignty Solution.” Remember, you’re talking military force. They are talking policy. See also,

American Force: Dangers, Delusions, and Dilemmas in National Security
Richard K. Betts

Betts describes himself as a Cold War hawk who became a post–Cold War dove. In this collection of essays, he addresses all the central issues of recent U.S. strategy: the maintenance of primacy and the prospective rise of China, humanitarian intervention and the struggles in Afghanistan and Iraq, and the problems posed by weapons of mass destruction, terrorism, and the possibility of a link between the two. This is not mainstream international relations scholarship. Betts combines serious thought, common sense, and deep historical knowledge, rather than simply applying abstract theories, and his conclusions are expressed in plain English, rather than with mathematical models. His judgments are therefore contingent, but they are always considered and often incisive. Betts is not opposed to the occasional use of force for the right purposes, and he explains why it is difficult to get strategic policy right. But he deplores the persistent American tendency toward military activism, especially in pursuit of what he describes as a “liberal empire.” As he himself recognizes, he is by no means a lone voice arguing for American restraint, but he is certainly among the most articulate.

I’d just like to point out that the “middle ground” point of view is overly simplified. It fails to account for the nature of insurgency. The occupied people can be complacent, yes. However, there are also circumstances where compliance is forced at gun point. The local population may be forced to live with insurgency under threat of death. The American equivalent is the mafia and the protection racquet. You don’t report the mafia to the authorities because they’ll take your livelihood.


I just read the American Interest paper on the Sovereignty Solution. I am severely under-impressed. Their main proposal is abandoning the internationalist system we currently have for a state based one. Well, the state based system gave the world more and more destructive wars than the internationalist system. By far. It is a backward approach to a over-inflated problem. (And a reaction to another problem, nation building, when the solution is get involved in less over seas wars.)


I read it differently. It’s not abandoning the international system; it’s an evolution. For the past twenty years, all action has been forced military by states. There are many, many other ways.

I have a suggestion that may seem a minor semantic point, but I think it makes a real difference.

You are absolutely right that killing civilians is the worst thing a small warrior can do. But if you say “Don’t Kill Civilians” a lot of people are going to stop listening immediately after “Don’t Kill…”, get all upset and start saying things like “It’s a war!” etc etc. Their emotions will turn on and their thinking off.

I suggest saying something like “Kill the Right People” and after that develop the explanation as to the grave disadvantages of killing the wrong people. Nobody, no matter how pugnacious a posturer they are, can disagree with the statement Kill the Right People, and once they nod their heads in agreement they will be more inclined to listen why civilians shouldn’t be killed.

There is one thing you missed when you referred to The Wire. One of the reasons cops try so hard to make sure they go after the correct target is because there will be absolute hell to pay if they strike the wrong target, especially if somebody is injured or killed. They may escape going to jail, but before the whole thing is over they may wish they had pled out just to get some peace. That hell will be paid by not only the officers involved but by the big bosses in the dept.

I don’t know if the same kind of incentive to get it right exists in the military. I know that nothing much happened after the military got a number of the wrong people in the Takhar attack as detailed by Kate Clark of the Afghan Analyst Network.

Cops don’t try hard to get confirmation just because they are noble but because they have extremely strong incentive not to get it wrong. Until the military has a similarly strong motivation, human nature will preclude them trying as hard as American cops do.

Kill the right people might be a better way to say it. The only thing I worry about is giving carte blanche to expand “the right people” to include all sorts of insurgent service/support personnel who can be dissuaded without killing. (For instance, detention or imprisonment as opposed to killing; you shouldn’t kill unless you have to).

I would also argue that in The Wire-type situations in America (urban centers with high crime rates) you could argue there is a situation of arresting the wrong people. Basically, it could be argued that the arrest is lawful, but it might not be necessary. This is the “broken windows” theory as described by David Kennedy on NPR. Basically, if every black male (or a super majority) has an arrest record, will they still trust the government?