Twitter inspired a conversation today with Starbuck about Eliot Cohen, who wrote a very important book on civil-military relation, Supreme Command.  It features case studies where civilian leaders over-rode military officers, focusing on Lincoln, Churchill, Clemenceau and Ben-Gurion.  Clemenceau had said that war was too important to be left to the generals, and I agree.  I found Cohen’s book very persuasive despite:

• the extreme selection bias: no cases of civilians making bad decisions and over-riding good military advice.  What would do the trick? Including an Iraq chapter where Rummy ignores his mil experts on size of force going into Iraq and micro-manages them down.
• Using Churchill as a model.  Mr. Gallipoli.  Mr. Balkans Are the Soft Underbelly of Europe.  

The bigger problem is that Cohen was a Neo-Con who became a big advocate of the Iraq war, serving on the Defense Policy Board Advisory Committee and defended the Bush administration when it came to ignoring the advice of the military.  The Army Chief of Staff General Eric Shinseki said, when pushed by the Congress, that the US would need a couple of hundred thousand troops on the ground to occupy Iraq.  He based this on his experience in Bosnia and on other US experiences.  Rummy disagreed, implicitly following Cohen’s advice about civilian supremacy.

So, the question is: how do we think of Cohen’s book now, in retrospect?  Well, when I started doing work on NATO and Afghanistan, I started with the local case of Canada.  I learned that the Canadians, at least lately, have had little micromanagement from civilians, that the new generation of officers were able to reduce restrictions and adapt to the requirements of the Kandahar mission.  I found the absence of civilian interference to be a good thing, but that rankled me given that I believe that war is too important to be left to the generals.

The Steve and Dave book on NATO and Afghanistan, which focuses on a variety of ways in which countries control their contingents that they deploy on multilateral missions, tends to be critical of caveats.  I do think that civilians should have control over military missions, but I also think that caveats are a crappy way to do it, as they are too blunt.  Commanders in the field need to be able to react to events as they will have a greater understanding of the situation.  So, control should be asserted through picking the most suitable commanders (ones who understand the intent of the civilians the best), provide significant oversight, and reward folks who produce success.

The key is this: micromanagement is not really about oversight but about discretion.  The civilians need to be informed about how things are being done but rarely should tell the military about how to operate at the tactical level and only occasionally at the operational level and frequently at the strategic level.  Which front to attack (Normandy or Balkans) should be something that the civilians decide.  The number of military police battalions should be sent to Baghdad should be decided by the military. 

So, I do think that civilians need to remains supreme, that they are the ultimate bosses of militaries in democracies, but they need to listen to the military since the military folks are the experts in the deployment of violence.  Listening does not mean agreeing with or always respecting, but taking into account the various arguments, and making informed decisions.

When I think about the defense policies of Rummy and the rest, I am reminded of a saying that Mike Lombardi, former general manager of various football teams, often repeats: hope is not a plan.  Rummy et al were not focused on what history and experience and study had taught us, but what they hoped would happen.  Which makes one realize that the wide-eyed optimists were Rummy, Cheney, Wolfowitz, Feith, and Cohen.

So, Cohen was very, very wrong in 2001-2005.  That does not mean all of his work is wrong, but we will presume he is wrong unless we can be convinced otherwise.  He has lost the presumption of wisdom because he was so very, very wrong.  That is the price one pays for having access and media attention*–your mistakes make news and change opinions–including opinions of your work.

*  I have been wrong in the past–most notably predicting on Canadian TV in 2003 that there were not be suicide terrorism in Iraq once the war happened.  Why is this not so problematic?  Because I do not publish on suicide terrorism.  Of course, I should not have answered a question so far out of my area of expertise.  Lesson learned.  Sort of.