What is the value of high value targeting?

I do not usually make it to think tank events in DC, but I made the time today to sit in at the start of the Brookings Institute's Defense Challenges and Future Opportunities confab put together by the awesome Pete Singer, he of Wired for War fame. Dave Kilcullen moderated the first panel of the morning on irregular war, and the first presentation of that panel was a keeper.

Veteran intelligence analyst Matt Frankel, on leave from his service in the intelligence community as a federal executive fellow at Brookings, gave a compelling presentation on high value targeting (HVT) campaigns and their utility. His findings:

1. HVT campaigns are more effective against centralized opposition -- but decentralization is the trend.
2. HVT campaigns do not work in a vaccum. They have to be connected to a broader CT or COIN strategy.
3. Indigenous attacking forces have the edge in HVT campaigns, mainly due to local knowledge.
4. Along the same line, third party HVT campaigns are less likely to succeed, and in order for them to do so, the objectives of the host nation must be alligned with those of the third party.
5. Capture when you can, kill when you must. Obviously, the intelligence yield is better in the case of the former. Dead mean tell no tales.
6. Understanding enemy organizational dynamics is vital. What will the effect on an enemy organization be? And -- and this is my concern -- are we killing the people we might need to do a deal with later?

Frankel's presentation matches up with a lot of what I have often argued, which for the most part has been based either on my personal experiences (I learned the danger of ignoring Lesson #2 in Iraq in 2003, for example) or case study analysis (fun fact: Hizballah's HVT campaign against the SLA in southern Lebanon was more successful than the IDF's HVT campaign against Hizballah, lending support to Lessons #3 and #4). I am pretty sure Frankel's analysis supports a lot of the concerns Dave and I have had about the drone program in Afghanistan and Pakistan (namely, that it ignores Lessons #1, #2 and #6), but I would want to see Frankel's presentation in an article supported by footnotes and with methodology laid out in greater detail. Regardless, Frankel's presentation was a great one, and I was sorry that an obligation at CNAS kept me from hearing the discussion that followed it.