What are the risks of counterinsurgency?

Some of the pushback against Gates' Defense proposal is that he's shifting too far from conventional warfare to counter-insurgency.  This is sometimes described as fighting the last war despite the fact that the wars are still ongoing and the programs being cut were made for the Cold War and thus even further out of date.  Regardless of what we need for counter-insurgency, we're overly invested in conventional warfare capability.  There is a risk in trying to make a force that can do both, better to split capabilities and allow specialization.

However, Michael Cohen does get to some of the real problems with counter-insurgency:

It is simply incorrect to say that only the Army can perform post-conflict reconstruction and I'm utterly unconvinced that its proper for the US military to be expanding its skill set to include aid and development functions. Isn't this why we have a civilian agency dedicated to aid and development?

Now as some of my friends at the Pentagon often remind me AID and State, as currently formulated, are not as well positioned as they should be to play these roles. But the solution is not to outsource this stuff to the military, it's to build up capacity at civilian agencies so they are better able to play their assigned roles! One of the reasons the military has taken on responsibilities that used to be restricted to civilian agencies is that they were given the responsibility at the outset of the Iraq War - and under the Bush Administration the capacity of our civilian agencies was allowed to diminish.

In the end, this is perhaps the greatest problem I have with counter-insurgency doctrine, and the most intractable divide between myself and COIN-danistas: embedding COIN in military doctrine is not a benign exercise. It risks shifting power dramatically and perhaps irreversibly toward the military and away from civilian agencies - and it provides a rationale for ever-expanding military budgets. Considering that the greatest security challenges facing the US in the future will come from non-state actors and transnational threats - and thus best confronted by the non-military elements of our national security toolbox -- the result could be a US national security and foreign policy apparatus that is ill-prepared and badly positioned to confront them.

So, ultimately, is the Gates budget spending too much on counter-insurgency?  The Sec. Def. estimated that 50% of the budget is conventional, 40% dual use, and 10% straight counter-insurgency.  That means between $50B and $70B that's completely counter-insurgency focused.  Given that we do have two wars going, that doesn't seem crazy.  However, it's also comparable to what we're spending overall on civilian foreign policy agencies.  That could give a rough rule of thumb, military counter-insurgency spending should not exceed civilian spending.

How do we get to wear a large scale civilian shift is possible?  I don't know.  However, until we find that political approach, I'm going to try to be careful not favor any interventions that, even if legitimate, the U.S. lacks the capability to successfully implement.