Quote Originally Posted by TheCurmudgeon View Post
You are correct. I wanted to take the lowest number I could. I used this reference: A Historical Basis for Force Requirements in Counterinsurgency by STEVEN M. GOODE.

There is also: Troop Levels in Stability Operations: What We Don’t Know by Peter J. P. Krause. It is an MIT pub and is where I took the more robust 20 per 1,000. I would think that this number is more logical if you had to do everything including military and civilian security and providing basic services. Krause includes military and civilian security forces in his number (based on Iraq):



Notibly, the numbers do not include contractor support or civilians helping with the Iraqi Government.

The 3 per 1,000 number is for a U.S. style police department, which is interesting if for no other reason than most people in the U.S. are not likely to challenge police authority or think that the local government and police are illegitimate ... mostly. It would not be a planning factor I would use for an invasion and occupation.
A few years ago we had a pretty extensive discussion on this very subject at the SWC. One of the things I uncovered was this article by Lidell Hart from Military review 1960. T.E. Lawrence had wanted Lidell Hart to study this very subject. Here is a link to the article. Not sure how relavent it is to COIN but it does have a lot to say about so called conventional forces.

http://cgsc.contentdm.oclc.org/cdm/s...1/id/705/rec/1

Having seen first hand how gangs take over a neighborhood in America I am very dubious of any magic ratio. It is usually a combination of factors that come together at the right time and place.