Kilcullen on Afghanistan: "It's Still Winnable, But Only Just."

November 14, 2008
New Yorker Magazine
Article by George Packer
Kilcullen on Afghanistan: "It's Still Winnable, But Only Just."

AFGANISTAN.JPGI wrote about David Kilcullen two years ago, in a piece called "Knowing the Enemy." Few experts understand counterinsurgency and counterterrorism better than this former Australian army officer and anthropology Ph.D, who has advised the American, British, and Australian governments, was one of General Petraeus's strategic whizzes at the start of the surge, in early 2007, and writes so well that you'd never imagine he's spent his whole career in government, the military, and academia. Kilcullen is now a senior fellow at the Center for a New American Security, which has provided Obama with foreign-policy advisers and advice.

This week, Kilcullen agreed to do an e-mail Q. & A. on Afghanistan and Pakistan, where he's spent a lot of time, and where the most pressing foreign crisis awaits the new Administration. Though Kilcullen is still an adviser to the State Department, he emphasized that his views are his own. And they are characteristically blunt.

The White House briefed both campaigns on Afghanistan before the election. Apparently that's how little time we have to turn things around. So how bad is it?

It's bad: violence is way up, Taliban influence has spread at the local level, and popular confidence in the government and the international community is waning fast. It's still winnable, but only just, and to turn this thing around will take an extremely major effort starting with local-level governance, political strategy, giving the Afghan people a well-founded feeling of security, and dealing with the active sanctuary in Pakistan. A normal U.S. government transition takes six to nine months, by the time new political appointees are confirmed, briefed, and in position. But nine months out from now will be the height of the Afghan fighting season, and less than a month out from critical Presidential elections in Afghanistan. If we do this the "normal" way, it will be too late for the Obama Administration to grip it up. I think this is shaping up to be one of the smoothest transitions on record, with the current Administration going out of its way to assist and facilitate. That said, the incoming Administration has a steep learning curve, and has inherited a dire situation—so whatever we do, it's not going to be easy.

It sounds like you're proposing classic counterinsurgency strategy: a combination of offensive and defensive military operations, political and economic development, and diplomacy. Isn't that what we've been doing these past seven years? Have we just not been doing enough of all these? Or do we need to change strategy to something fundamentally new?

Well, we need to be more effective in what we are doing, but we also need to do some different things, as well, with the focus on security and governance. The classical counterinsurgency theorist Bernard Fall wrote, in 1965, that a government which is losing to an insurgency isn't being out-fought, it's being out-governed. In our case, we are being both out-fought and out-governed for four basic reasons:

(1) We have failed to secure the Afghan people. That is, we have failed to deliver them a well-founded feeling of security. Our failing lies as much in providing human security—economic and social wellbeing, law and order, trust in institutions and hope for the future—as in protection from the Taliban, narco-traffickers, and terrorists. In particular, we have spent too much effort chasing and attacking an elusive enemy who has nothing he needs to defend—and so can always run away to fight another day—and too little effort in securing the people where they sleep. (And doing this would not take nearly as many extra troops as some people think, but rather a different focus of operations).

(2) We have failed to deal with the Pakistani sanctuary that forms the political base and operational support system for the Taliban, and which creates a protective cocoon (abetted by the fecklessness or complicity of some elements in Pakistan) around senior al Qaeda and Taliban leaders.

(3) The Afghan government has not delivered legitimate, good governance to Afghans at the local level—with the emphasis on good governance. In some areas, we have left a vacuum that the Taliban has filled, in other areas some of the Afghan government's own representatives have been seen as inefficient, corrupt, or exploitative.

(4) Neither we nor the Afghans are organized, staffed, or resourced to do these three things (secure the people, deal with the safe haven, and govern legitimately and well at the local level)—partly because of poor coalition management, partly because of the strategic distraction and resource scarcity caused by Iraq, and partly because, to date, we have given only episodic attention to the war.

So, bottom line—we need to do better, but we also need a rethink in some key areas starting with security and governance.

Let's take these one at a time. Has there been too much emphasis on offensive operations, especially air strikes? We read a lot recently about civilian deaths and growing Afghan anger. Should we cut back on the use of air power and put in more ground troops, as Obama has said he will? Or is this not a matter of managing numbers and assets so much as changing the focus of our tactics?

It's both. There has been an emphasis on fighting the Taliban, which has led us into operations (both air and ground-based) that do a lot of damage but do not make people feel safer. Similarly, we have a lot of troops in rural areas—small outposts—positioned there because it's easier to bring firepower to bear on the enemy out in these areas. Meanwhile, the population in major towns and villages is vulnerable because we are off elsewhere chasing the enemy main-force guerrillas, allowing terrorist and insurgent cells based in the populated areas to intimidate people where they live. As an example, eighty per cent of people in the southern half of Afghanistan live in one of two places: Kandahar city, or Lashkar Gah city. If we were to focus on living amongst these people and protecting them, on an intimate basis 24/7, just in those two areas, we would not need markedly more ground troops than we have now (in fact, we could probably do it with current force levels). We could use Afghan National Army and police, with mentors and support from us, as well as Special Forces teams, to secure the other major population centers. That, rather than chasing the enemy, is the key.

Police are another main issue. We have built the Afghan police into a less well-armed, less well-trained version of the Army and launched them into operations against the insurgents. Meanwhile, nobody is doing the job of actual policing—rule of law, keeping the population safe from all comers (including friendly fire and coalition operations), providing justice and dispute resolution, and civil and criminal law enforcement. As a consequence, the Taliban have stepped into this gap; they currently run thirteen law courts across the south, and ninety-five per cent of the work of these courts is civil law, property disputes, criminal matters, water and grazing disuptes, inheritances etc.—basic governance things that the police and judiciary ought to be doing, but instead they're out in the countryside chasing bad guys. Where governance does exist, it is seen as corrupt or exploitative, in many cases, whereas the people remember the Taliban as cruel but not as corrupt. They remember they felt safer back then. The Taliban are doing the things we ought to be doing because we are off chasing them instead of keeping our eye on the prize—securing and governing the people in a way that meets their needs.

So, on the military side, three additional brigades isn't the answer? Or isn't the only answer?

That's right. The first thing we have to do is to "triage" the environment: figure out the smallest number of Afghan population centers that accounts for the greatest percentage of the population. Once we understand that lay-down (e.g., in the South, it's two towns that account for eighty per cent of the population, but the east is more rural, so it's a different calculation there), then we tailor a security plan for each major cluster of population, and for the key communications—roads, essentially—that link them together. Then we will have an idea of the extra troops we need, if any. But we can start right away with the troops we have.

Also, there are assets beyond (or, at a pinch, instead of) combat troops that would make a huge difference, without "breaking the bank" for combat troops elsewhere. These include construction engineers, aid and development personnel, aid project money, intelligence analysts, helicopters, trainers and advisers, mentors for local mayors and district officials, surveillance assets and so on—so it's not necessarily a straight zero-sum between having combat troops pull out of Iraq so we can send them to Afghanistan. (In any case, if you accept the argument that a key part of our grand-strategic problem is that we are over-committed in Iraq—and I do accept that argument—then it makes no sense to pull troops out of Iraq just so we can go and re-commit them somewhere else. We need to be reducing overall force commitment everywhere, not just moving troops from Iraq to Afghanistan. That would be tantamount to un-bogging ourselves from Iraq just so we can re-bog ourselves in Afghanistan).

On the Pakistani sanctuary, this seems to be the cancer in the bones of Afghanistan, and no one has a good answer. Both air power and special-forces incursions have drawn the wrath of the Pakistani government and people, but their efforts, as you say, have been weak at best and two-faced at worst. Our diplomats and development workers are being systematically targeted, and there's a question how well we can spend $750 million in the northwest. Is there a way to clear out this sanctuary, that doesn't cause the problem to metastasize?

You're right. Pakistan is extremely important; indeed, Pakistan (rather than either Afghanistan or Iraq) is the central front of world terrorism. The problem is time frame: it takes six to nine months to plan an attack of the scale of 9/11, so we need a "counter-sanctuary" strategy that delivers over that time frame, to prevent al Qaeda from using its Pakistan safe haven to mount another attack on the West. This means that building an effective nation-state in Pakistan, though an important and noble objective, cannot be our sole solution—nation-building in Pakistan is a twenty to thirty year project, minimum, if indeed it proves possible at all—i.e. nation-building doesn't deliver in the time frame we need. So we need a short-term counter-sanctuary program, a long-term nation-building program to ultimately resolve the problem, and a medium-term "bridging" strategy (five to ten years)—counterinsurgency, in essence—that gets us from here to there. That middle part is the weakest link right now. All of that boils down to a policy of:

(a) encouraging and supporting Pakistan to step up and effectively govern its entire territory including the FATA [Federally Administered Tribal Areas], and to resolve the current Baluch and Pashtun insurgency, while
(b) assisting wherever possible in the long-term process of state-building and governance, but
(c) reserving the right to strike, as a last resort, at al Qaeda-linked terrorist targets that threaten the international community, if (and only if) they are operating in areas that lie outside effective Pakistani sovereignty.

During the campaign, McCain talked about transferring the surge from Iraq
to Afghanistan. We've discussed the military side. On the political side, is there any possible counterpart to the Sunni Awakening in Afghanistan—perhaps local Taliban disenchanted with foreign influences on their leadership? Should part of our political strategy be to talk to Taliban leaders who might be prepared to negotiate with us?

Well, I doubt that an Anbar-style "awakening" is likely in Afghanistan. The enemy is very different from A.Q.I. and, in any case, Pashtun tribes have a very different makeup from Arab tribes. So even if an awakening happened it would likely play out differently from Iraq. Rather than talking about negotiations (which implies offering an undefeated Taliban a seat at the table, and is totally not in the cards) I would prefer the term "community engagement." The local communities (tribes, districts, villages) in some parts of Afghanistan have been alienated by poor governance and feel disenfranchised through the lack of district elections. This creates a vacuum, especially in terms of rule of law, dispute resolution, and mediation at the village level, that the Taliban have filled. Rather than negotiate directly with the Taliban, a program to reconcile with local communities who are tacitly supporting the Taliban by default (because of lack of an alternative) would bear more fruit. The Taliban movement itself is disunited and fissured with mutual suspicion—local tribal leaders have told me that ninety per cent of the people we call Taliban could be reconcilable under some circumstances, but that many are terrified of what the Quetta shura and other extremists associated with the old Taliban regime might do to them if they tried to reconcile. So, while an awakening may not happen, the basic principles we applied in Iraq—co-opt the reconcilables, make peace with anyone willing to give up the armed struggle, but simultaneously kill or capture all those who prove themselves to be irreconcilable—are probably very applicable.

You spoke of Iraq's effect in draining our energy and focus away from Afghanistan. President-elect Obama has made it clear that he plans to alter the balance significantly. But, as you say, he doesn't have much time. If you had his ear, what would be your basic advice?

Well, I don't have his ear, and I don't envy the pressure he must be under. But if I did have his ear, I think I would argue for the four major points we discussed above. First, the draw-down in Iraq needs to be conditions-based and needs to recognize how fragile our gains there have been, and our moral obligation to Iraqis who have trusted us. As I said, we don't want to un-bog ourselves from Iraq only to get bogged in Afghanistan while Iraq turns bad again. Second, our priorities in Afghanistan should be security, governance, and dealing with the Pakistan safe haven—and we may not necessarily need that many more combat troops to do so. Third, the Afghan elections of September 2009 are a key milestone—we can't just muddle through, and the key problem is political: delivering effective and legitimate governance that meets Afghans' needs. And finally, most importantly, this is a wartime transition and we can't afford the normal nine-month hiatus while we put the new Administration in place: the war in Afghanistan will be won or lost in the next fighting season, i.e. by the time of the September elections.

The situation in Afghanistan is dire. But the war is winnable. We need to focus our attention on the problem, and think before acting. But we need to think fast, and our actions need to involve a major change of direction, focussing on securing the population rather than chasing the enemy, and delivering effective legitimate governance to the people, bottom-up, at the local level. Do that, do it fast, and we stand an excellent chance of turning things around.