Measuring Effectiveness in Counterinsurgency

​HOW MUCH DOES the ability to move around a conflict zone matter to the counterinsurgency (COIN) effort? That question is a difficult one to answer, according to a recent report out of RAND Corp.

Freedom of movement (FoM) has rarely been precisely defined in literature, according to the report Assessing Freedom of Movement for Counterinsurgency Campaigns, which specifically looks at how FoM relates to the Afghanistan COIN campaign. There are different definitions of FoM for various groups, but it essentially means the ability of certain actors to move freely in an area. The report authors advocate looking at five different types of FoM: FoM for civilians, FoM for Afghan National Security Forces and the International Security Assistance Force, FoM for government officials, FoM for nongovernmental organizations, and FoM for insurgents.

One of the most important findings was that FoM cannot necessarily be used as a benchmark for COIN campaign success. That’s because different types of FoM (or lack thereof) in different locations indicate various things. Report coauthor Ben Connable, RAND international policy analyst and retired Marine Corps intelligence and foreign area officer, points out that more movement does not necessarily mean COIN is working.

For example, more FoM among local citizens could also indicate that insurgents might be able to move around just as much as other citizens.

Connable also points out that even though there may be movement in an area, it may not be true FoM. He uses as an example a person who needs to bring his goods to market who may drive on a road that he suspects is filled with IEDs because he needs to sell the items to support his family. It is not true FoM or even perception of FoM that puts him on that road, but rather necessity. “That’s a terrible inconsistency for somebody that’s trying to assess freedom of movement as it ties into government legitimacy,” says Connable.

David Kilcullen, a counterinsurgency expert and CEO of Caerus Associates, says that evaluating patterns of movement can be more helpful than looking at general FoM.

For example, when he was studying Indonesia, he found that people were heading into the city at night to sleep, rather than a more normal pattern of heading into the city in the morning to work. This indicated that things outside of the city were not safe at night.

“Of course it’s very complicated to interpret that... There’s a lot of stuff about it that gives you some indication,” explains Kilcullen.

One metric that Kilcullen advocates using instead of FoM is the price of goods. “Say you’re growing bananas down the road, and you want to sell them in the market in my town. If the roads are safe, you have to pay your driver 10 cents to drive them to the market. If the roads are dangerous you’re going to have to pay some kind of a danger premium where you’re going to have to pay insurance or that kind of thing on top of that.”

That price increase will be reflected in the price of the banana. The price reflects FoM because the more difficult it is to move about freely, the more expensive the banana.

It is very difficult to come up with a list of metrics to assess COIN success. Kilcullen says that one of the biggest mistakes that groups make when generating assessments is doing incident counts. He says it can be a misleading measurement, because if you have more troops, they will experience more incidents, and also, the level of violence doesn’t necessarily indicate whether the counterinsurgency is winning or losing.

Everyone loves metrics, but Connable warns that there is no way to simplify and quantify COIN assessment. “Understanding the progress or lack of progress in a counterinsurgency campaign requires some depth of understanding, and it requires reading some detailed narrative, and it requires understanding some things on the local level in context,” says Connable.

Kilcullen also advocates the local approach. He says that analysts need to figure out what measures work for each specific area. In Afghanistan, for example, he says that COIN assessors built models of districts and used those models to figure out metrics specific to the district. He emphasizes that there is no standard set of metrics—analysts must go out and figure out what’s normal for each environment and derive metrics from the patterns discovered.