Af-Pak strategy - what works and what doesn’t.

A bunch of smart analysts over at CNAS (that’s pronounced CEE-EN-AY-S and don’t you dare call it “C-NAS”) have published a paper on the US’s next twelve months in Af-Pak this week in advance of the annual CNAS’ conference.

They offer three suggestions I agree with, and one that I don’t.

This paper makes four operational recommendations – two on each side of the Durand line – which allow the new strategy articulated by the White House a better chance of success. In Afghanistan, we recommend that protecting the population take precedence over all other considerations for the time being. At the same time, however, any “civilian surge” must be used to increase the legitimacy of the Afghan government in the eyes of the Afghan population. In Pakistan, meanwhile, the U.S. government should place a moratorium on drone strikes on non-al Qaeda targets in the Federally Administered Tribal Areas and the Northwest Frontier Province until such strikes can be incorporated into a coherent strategy for separating the population of these areas from al Qaeda. And the United States should refocus its train and equip mission in Pakistan to place a greater emphasis on the police – the only Pakistani security service focused entirely on domestic security. Especial emphasis should be placed on the security services in those areas where Pakistani authority is strongest, such as in Punjab and Sindh.

I agree with everything but the moratorium on non-AQ targets.  I think they should be extraordinarily limited, not stopped.  Though I understand that drone strikes often have unintended consequences - killing innocent civilians - that sticks a finger in the eye of counter-insurgency strategy, they’re still a valuable tool in certain, rare circumstances.

Instead, I suggest the following criteria for prone strikes:

1. Target must be of the highest value, deemed irreconcilable from cooperating with the Pakistanis, and aligned with al Qaeda, the Taliban, or other organization set on destabilizing a local or national government.

2. Target’s location must be verified by multiple, credible sources from differing collection platforms.  SIGINT, HUMINT, IMINT, or another “INT” must corroborate one another.

3. Due to the targets’ mobility, the corroborated intelligence must not be more than two hours old and there must not be a reasonable expectation that the target has moved in the interim.

4. The main problem with drone strikes is that they accidentally kill civilians, but it’s unrealistic to say that drone won’t fire on population centers because then the targets would just hide in plain sight.  However, the US must carefully weigh the chance of civilian casualties and seek to avoid them - by using smaller missiles, modifying times of the strikes, etc. - at all costs.

5. The Pakistani government should engage in a massive public relations campaign that seeks to employ more credible voices (tribal elders, trusted local governors, and the few national level politicians who aren’t despised) to educate the public on the drones’ goals.

Hopefully these steps would limit the civilian casualities while maintaining a valuable tool in the local fight against regional destabilizers.