Gideon Rose. Foreign Affairs. Volume 92, Issue 2. March/April 2013.

In July 2010, General Stanley McChrystal retired from the U.S. Army after almost three and a half decades in uniform. Soon after graduating from West Point, McChrystal had joined the U.S. Special Forces, and he eventually led the Rangers, the Joint Special Operations Command, and all U.S. and international forces in Afghanistan.

A knowledgeable author wrote in a recent issue of this magazine that “as head of the U.S. Joint Special Operations Command … , McChrystal oversaw the development of a precision-killing machine unprecedented in the history of modern warfare,” one whose “scope and genius” will be fully appreciated only “in later decades, once the veil of secrecy has been removed.” What did he mean?

I was part of a [special operations] effort that we can call Task Force 714. When the counterterrorist effort against al Qaeda started, it was narrowly focused and centralized; you only did occasional operations with a high degree of intelligence and a tremendous amount of secrecy. That worked well for the pre- 9/11 environment, but in the post- 9/11 environment-particularly the post-March 2003 environment in Iraq- the breadth of al Qaeda and associated movements exploded. This gave us an enemy network that you couldn’t just react to but actually had to dismantle. It also gave us a very complex battlefield-not just terrorism but also social problems, an insurgency, and sectarian violence.

So the first thing we did when I took over in late 2003 was realize that we needed to understand the problem much better. To do that, we had to become a network ourselves-to be connected across all parts of the battlefield, so that every time something occurred and we gathered intelligence or experience from it, information flowed very, very quickly.

The network had a tremendous amount of geographical spread. At one point, we were in 27 countries simultaneously. Inside Iraq, we were in 20 and 30 places simultaneously—all connected using modern technology but also personal relationships. This gave us the ability to learn about the constantly evolving challenge.

People hear most about the targeting cycle, which we called F3EA-“find, fix, finish, exploit, and analyze.” You understand who or what is a target, you locate it, you capture or kill it, you take what intelligence you can from people or equipment or documents, you analyze that, and then you go back and do the cycle again, smarter.

When we first started, those five steps were performed by different parts of our organization or different security agencies. And as a consequence, each time you passed information from one to another, it would be like a game of telephone, so that by the time information got to the end, it would be not only slow but also corrupted. We learned we had to reduce the number of steps in the process.

In 2003, in many cases we’d go after someone, we might locate them and capture or kill them, and it would be weeks until we took the intelligence we learned from that and were able to turn it into another operation. Within about two years, we could turn that cycle three times in a night. We could capture someone, gain intelligence from the experience, go after someone else, and do three of those in a row, the second two involving people we didn’t even know existed at the beginning of the night.

In August 2004, in all of Iraq, our task force did 18 raids. And we thought that was breakneck speed. I mean, we really thought we had the pedal to the metal. These were great raids, very precise, a high percentage of success. But as great as those 18 raids were, they couldn’t make a dent in the exploding insurgency. Two years later, in August 2006, we were up to 300 raids a month-ten a night. This meant the network now had to operate at a speed that was not even considered before, not in our wildest dreams. It had to have decentralized decision-making, because you can’t centralize ten raids a night. You have to understand them all, but you have to allow your subordinate elements to operate very quickly.

But then, we had to be able to take all of that and make it mean something- because it’s not just about capturing and killing people; it’s about synchronizing into the wider theater campaign. And that took us longer. We really didn’t mesh completely into the conventional war effort [in Iraq] until 2006, 2007.

So that was the revolution. I didn’t do it. The organization I was part of became this learning organization. If I take any credit, it is for loosening the reins and yelling “Giddyup!” a lot. I allowed, encouraged, required the team to push forward. And they just rose to the occasion.

Did the tactics of the special operators under your command change in any way?

The operational change and the mental change was by far the more significant part of it. However, tactically there were some things that changed, and part of that was technological. We started with well-trained commandos. We had always had those. They shoot well, they move well, they think brilliantly. But three things changed.

The first was global positioning systems. These allowed you to be exactly where you wanted to be without fits and starts. Navigating from point A to point B wasn’t a big part of the task anymore. People take that for granted now, but as I grew up in the military, half of doing something was getting there.

The second thing was the use of night-vision goggles and night-vision equipment on aircraft and other things. These allowed you to have superiority in what you can see and do in the dark. Our entire force operated with night vision, so at night we used no visible lights. We had laser-aiming lights on our weapons and infrared illumination if it was too dark for the night vision. And as a consequence, we just dominated night firefights and night operations dramatically. That was a big deal.

The third was the use of things like the Predators-unmanned aerial vehicles-and some manned aircraft. The big breakthrough was that we could put these up and send the downlink or the video feed down to the command or the force on the ground in real time. That doesn’t give you complete situational awareness, but it lets you see a bird’s-eye view of the battlefield, even though you’re standing on the ground in the mud or dirt.

Traditionally, if we did a raid and we thought we were going to need 20 commandos to actually be on the target, we might take 120, because we had to put security around the site to protect it from enemy reinforcements, and we might have to put a support section and a command-and-control section there, because you need all those things to account for the unexpected. But when you have very good situational awareness and good communications, you only send the 20, because your security comes from being able to see, and then you can maneuver forces if you need them. So suddenly, the 120 commandos aren’t doing one raid; they’re doing six raids simultaneously, and you start to get the ability to do 300 raids a month.

And that’s important, because if you’re going at an enemy network, you’re trying to paralyze its nervous system. If you just hit it periodically, say, every other night, it not only heals itself; some would argue it gets stronger because it gets used to doing that. But if you can hit it in enough places simultaneously, it has a very difficult time regenerating. And that’s when we started to have decisive effects.

In your book, you take a somewhat contrarian view about the changeover from the Casey era to the Petraeus era in Iraq. How decisive was the change in military leadership and the “surge,” and how much of it was just a natural evolution of what had been going on in Iraq for years already?

People tend to simplify things. They try to say, “It was all screwed up here and then it got all good there,” or, “This decision was decisive.” I never found anything that clear. I found the move towards counterinsurgency to be one that was more gradual than sudden. It started under General Casey; he pushed it.

I will say that when the president made the decision to surge more forces, it intersected with some things which were happening. [Iraqi] Sunnis had grown disenchanted with al Qaeda, for good reason. I think the Sunnis also came to the conclusion that they were fighting the coalition, and we were beating on them pretty badly. And so I think they said to themselves, “We had better not fight the wrong war.” People were exhausted, were not sure what was going on. And then suddenly, President Bush effectively says, “OK, we are going to double down.” Even though people knew it couldn’t be permanent, I think there was a sense that this pushes it past the tilting point.

And then, of course, there was General Petraeus, who brought a level of energy and a commitment to the counterinsurgency campaign. All of these together produced a pretty amazing result.

Counterinsurgency typically requires three things to work: a long time, a lot of troops, and a very sensitive, low-impact, politically aware mindset. Given that the American public doesn’t like long wars, and given that large numbers of forces and a politically aware, sensitive approach to the use of violence seem to be at odds, when, if ever, is counterinsurgency going to be something that the United States should actually embark on?

That’s a valid and difficult question, because there are also two other factors which ought to be thrown in. Successful insurgencies usually need an outside safe haven and access to the war zone. Pakistan gave the insurgents in Afghanistan that, and so they had something that we really needed to take away. But we couldn’t seal the border.

And a successful counterinsurgency needs a legitimate government. You need to offer to the people an alternative to what the insurgent is offering. The Taliban don’t offer a very compelling narrative or popular government, but the government of Afghanistan has huge problems with its popular legitimacy as well.

So in reality, what we had is a situation where we had been there a long time, the coalition was tired, the people of Afghanistan were scared, the insurgency was growing in confidence, the insurgency had a safe haven, and the government of Afghanistan was weak and somewhat conflicted about the war. So there were a lot of factors against it. And that is a very valid argument on why the success of the endeavor is certainly not assured.

What lessons did you learn in your Iraq and Afghanistan tours?

In Iraq, when we first started, the question was, “Where is the enemy?” That was the intelligence question. As we got smarter, we started to ask, “Who is the enemy?” And we thought we were pretty clever. And then we realized that wasn’t the right question, and we asked, “What’s the enemy doing or trying to do?” And it wasn’t until we got further along that we said, “Why are they the enemy?”

Not until you walk yourself along that intellectual path do you realize that’s what you have to understand, particularly in a counterinsurgency where the number of insurgents is completely independent of simple math. In World War II, the German army could produce x number of military-aged males. In an insurgency, the number of insurgents isn’t determined by the population, but by how many people want to be insurgents. And so figuring out why they want to be insurgents is crucial. And that’s something we had never practiced.

Second, it’s all about teams. Nobody wins the war alone. We had a culture in our force, and in many forces, of excellence. It was, “How good can I be at my task? How good can I be at flying an airplane, dropping a bomb, locating an enemy target?” But that’s not as important as how well those pieces mesh together. The real art is, if somebody builds a bridge, you have the people ready to drive over it and take advantage of that. It’s cooperating with civilian agencies, it’s cooperating with conventional forces, it’s tying the pieces together. That’s the art of war, and that’s the hard part.

It seems like the methods you pioneered in Iraq have been embraced by the U.S. government and the American public as a general approach to managing smallscale irregular warfare, and doing so in a way short of putting lots of boots on the ground or walking away entirely. Some would argue that this is the true legacy of Stan McChrystal-the creation of an approach to counterterrorism that is halfway between war and peace, at such a low cost and with such a light footprint that it’s politically viable for the long term in a way that war and disengagement are not. Do you disagree?

I question its universal validity. If you go back to the British tactics on the North-West Frontier, the “butcher and bolt” tactics, where they would burn an area and punish the people and say, “Don’t do that anymore,” and simultaneously offer a stipend to the leader while saying, “If you will remain friendly for a period of time, we’ll pay you”-that approach worked for a fair amount of time. It managed problems on their periphery. But it certainly didn’t solve the problems.

The tactics that we developed do work, but they don’t produce decisive effects absent other, complementary activities. We did an awful lot of capturing and killing in Iraq for several years before it started to have a real effect, and that came only when we were partnered with an effective counterinsurgency approach. Just the strike part of it can never do more than keep an enemy at bay. And although to the United States, a drone strike seems to have very little risk and very little pain, at the receiving end, it feels like war.

Americans have got to understand that. If we were to use our technological capabilities carelessly-I don’t think we do, but there’s always the danger that you will-then we should not be upset when someone responds with their equivalent, which is a suicide bomb in Central Park, because that’s what they can respond with.

So it’s incorrect for someone to say, “I like the Iraq Stan McChrystal of raids and drones and targeted strikes, but I don’t like the Afghanistan Stan McChrystal of clear, hold, and build and counterinsurgency; I want to deploy the first but avoid the costs and difficulties of the second”?

I would argue they should like all the Stan McChrystals. If you look at the role I had in Iraq, it is sexy, it is satisfying, it is manly, it scratches an itch in the American culture that people like. But I was doing that as part of a wider effort in Iraq, and it was that wider effort that I took control of in Afghanistan. And those wider efforts were about people. The whole point of war is to take care of people, not just to kill them. You have to have a positive reason that protects people, or it’s wrong. So while I did what I had to in Iraq, and did a lot of that in Afghanistan, too (because we had a significant effort along those lines there), the broader purpose is what’s important, and that’s what I think people need to be reminded of. The purpose is the Afghan kid. The purpose is the Afghan female. The purpose is the 50-year-old farmer who just wants to farm.

Did the success of your efforts in Iraq lead to an overemphasis on the use of direct action by Special Forces, raids and drone attacks and targeted killings, rather than indirect action, such as training and building local capacity?

My wife Annie and I are not golfers, but some years ago, we took part in a golf tournament in our unit. After having significant trouble, on one of the tees, Annie used a Kevlar driver. She hit this amazing drive straight down the fairway, and she was elated. For the rest of the afternoon, the only club she used was the Kevlar driver. She chipped with it. She putted with it. She used it for everything.

That’s the danger of special operating forces. You get this sense that it is satisfying, it’s clean, it’s low risk, it’s the cure for most ills. That’s why many new presidents are initially enamored with the Central Intelligence Agency, because they are offered a covert fix for a complex problem. But if you go back in history, I can’t find a covert fix that solved a problem long term. There were some necessary covert actions, but there’s no “easy button” for some of these problems. That’s the danger of interpreting what we did in Iraq as being the panacea for future war. It’s not.