Irregular Warfare Podcast - Does Building Partner Military Capacity Work?

Episode Date: June 19, 2020

This episode of the Irregular Warfare Podcast features a conversation with two guests about an important question: Does building partner military capacity work? Dr. Stephen Biddle is a professor of international and public affairs at Columbia University and served as an advisor to Gen. David Petraeus and Gen. Stanley McChrystal during the wars in Iraq and Afghanistan. Matt Cancian is a doctoral candidate in political science at the Massachusetts Institute of Technology and a former Marine officer who has researched Western efforts to build partner capacity among the Kurdish Pershmerga during the fight against ISIS. Intro music: "Unsilenced" by Ketsa Outro music: "Launch" by Ketsa CC BY-NC-ND 4.0

I hear that in some meetings repeated like a mantra that's going to ward off evil, by, with, and through. We're by, with, and through. We're empowering the local government. And in fact, that's not what you want. Take a more political understanding of the phenomenon, that implies you need conditionality. Conditionality is all about understanding the interest divergences between yourself and the ally that you're working with. Welcome back to the Irregular Warfare podcast. In this episode, we ask the question, does building partner military capacity work?

We begin by asking why building partner capacity matters in the current national security environment. Security force assistance programs where U.S. forces train, equip, and advise indigenous partner forces have played an important role in the U.S. approach to address insurgency and terrorist threats around the world. However, some have come to question their efficacy, especially following the collapse of Iraqi security forces to ISIS in 2014 after receiving billions of dollars in assistance and a decade of training by Western forces. Today's conversation examines cases when the U.S. successfully built a partner's military capacity to include with the Peshmerga in the counter-ISIS fight and with the Korean

military during the Korean War. The argument is presented that security force assistants often fall short of expectations, largely due to politics and interest divergence between actors. We conclude with some options to help policymakers and practitioners execute security force assistance more effectively given these constraints. Our guests for this conversation are Dr. Stephen Biddle and Matt Kansian. Dr. Biddle is a professor of international and public affairs at Columbia University and served as an advisor to Generals David Petraeus and Stanley McChrystal during the wars in Iraq and Afghanistan. His research focuses on the political challenges that influence the success of security force assistance. Matt Kantzian is a doctoral candidate in political science at the Massachusetts

Institute of Technology and a former Marine officer. His research focuses on whether Western efforts to build partner capacity among the Kurdish Peshmerga during the counter-ISIS fight improved the combat performance of Kurdish units. I am Shauna Sinnott. And I'm Kyle Atwell. And this is the Irregular Warfare Podcast, a joint production of the Princeton Empirical Studies of Conflict Project and the Modern War Institute at West Point,

dedicated to bridging the gap between scholars and practitioners to support the community of irregular warfare professionals. Here is our conversation with Matt Kantzian and Dr. Steve Biddle. Steve, Matt, really appreciate you being here with us today. Thank you. Thanks for having me. And I'd like to just jump right into this. So, Steve, why is the topic of building partner military capacity relevant right now?

Sure. Building partner capacity, security assistance, however we want to describe it, is a big policy issue and becoming a bigger one all the time. And lots of Americans are sick and tired of sending 100,000 Americans to Afghanistan, 160,000 Americans to Iraq, but they still believe that the world is a dangerous place and that there are situations around the world where there are threats to American interests that require some kind of response, but they don't want to send 100,000 or more soldiers. The solution for many Americans is building partner capacity. Let the locals do most of the work, reduce our scale of investment to something more sustainable. The Department of Defense is also engaged in a systematic shift

in the focus of U.S. defense strategy away from developing world counter-terrorist instability problems like Iraq and Afghanistan and towards great power conflict with especially Russia and China. But there are still all these other problems out there. And the way DOD wants to respond to them is building partner capacity. Well, I know it's also very personal for both of you as well. You both have experience working with this, which I imagine has informed not just your understanding of why it's important and how it can work, but also a lot of the challenges inherent to it. Matt, can you speak to your experience at the tactical level? I was initially ordered to get involved with security force assistance when I was a

young lieutenant. And so I got very interested in the question of what is making these Afghans who we're partnering with more or less effective. So the question was, why are they motivated to start in the first place? What's making them stop? What's making them more or less effective? And how can we shape that? Were there particular barriers you observed to being effective with that training? So there's stuff in our tactical DNA as American soldiers in terms of using cover and concealment, suppression, combined arms in order to defeat your enemy. And this is instilled in American soldiers and Marines at a very early point in your career. And you might think that it's then obvious and that everyone should get it. But this is actually

not an obvious process. And this is something that it took many obvious and that everyone should get it. But this is actually not an obvious process. And this is something that it took many years and hundreds of thousands of lives to figure out in World War I and World War II is exactly how to integrate these. And many third world countries still don't really get it. So I had this story of a Kurdish soldier during the Iran-Iraq war where he was in charge of a Sager missile, which is sort of like our tow. It's a wire guided missile. And he would get in his BMP, fire it, and then run away because he knew that the Iranians were going to shoot back at him. And he didn't have friendly forces suppressing the enemy. So in America, we know, well, if you have a missile launch, you're going

to stay in a cold position. You're going to come out to the hot position. You're going to fire. Other people are going to suppress so that you can stay there. But this is not obvious to third world militaries. And this is actually very hard. You need very good human capital in order to do this. Steve, can you describe your experiences at the theater level and barriers to effectiveness you observed? Yeah, I mean, I've been lucky enough or unlucky enough, depending on how you frame it, I suppose, to be involved in a variety of advisory efforts, mostly at the theater level in Iraq and Afghanistan. And when kind of parachuting into these combat zones as a tweed wearing civilian, you typically find yourself talking to the various parts of the headquarters that the command thinks are important. So they would typically throw me into the security assistance operation in these

different theaters. And the perspective I tended to take was so absurdly different than the perspective that they tended to take, that it also tended to kind of pique my interest in those. Can you, like, how is your perspective different then? Well, I mean, as a social scientist, you tend to come at a problem like this, oddly enough, thinking about social science. And the social science on this tends to suggest that the performance of allied militaries is not as much an engineering problem, oddly enough, as it is a social science problem. The politics of this matter. And yet most of the organization organizations I was dealing with,

when I would sort of sail in and say, hi, I'm here from theater headquarters, and I'm a civilian, and I'm here to help. Always a good beginning to any conversation. The people I was interacting with in these kinds of settings tended to see it as more or less an engineering problem. The problem is the ally doesn't have enough stuff. They don't have enough capacity. Our job is to build the partner's capacity. That means we take stuff that we've got and we give it to them. We take knowledge we've got and we train them in it.

We organize them. We set them up properly to run a defense ministry. Basically, we just transfer resources to an ally who's not cutting the mustard because they don't have enough resources. And as a social scientist, rather than an engineer, you look at this and you say, well, maybe part of the problem here is the politics of this relationship rather than just an absence of stuff. I mean, is it really the case that the ally wants exactly what we want and the whole problem is that they just don't have enough stuff to realize our common joint interest? Or is there something else going on? Short answer, there's something else going on

most of the time. So my exposure to this downrange in these various headquarters kind of reinforced the suspicion that there was a broader need for a different understanding of how this process worked that might not just improve the way the theater commands were doing it in the field, but might also have some influence on the way defense planning was done. Does that transactional approach at the theater level translate to the policy realm as well, where it seems like small number of forces committed equals a greater return on investment, so it's more cost effective? Well, people think it's cost effective. How cost effective it is is an issue for research, right? I mean, generally speaking, my take is that it's not useless, but that it's less cost effective than its proponents say. And therefore, the circumstances

in which we use it will be non-zero, but need to be targeted carefully. And the way we do it needs to be oriented towards trying to reach its ceiling. But part of the issue, especially for defense planners in Washington, is understanding what the ceiling is. Steve's exactly right that people don't really think about the sort of social science aspects, but I would push back that the cost is lower. The question is whether it's more effective or just as effective or less effective. And part of the question is, what are you trying to get out of your specific partner relationships where some of the ones we're trying to get increased combat effectiveness in order to destroy an adversary. Other partner relationships, we're

trying to just keep lines of communication open with a regional partner or ally to keep them on our side and to keep access to these areas that are of maybe not primary importance like China or Russia, but that are of secondary or tertiary importance. So that even if you're having an advise and assist mission with the Chadians and they don't get all that much better at fighting, that's fine because then the Chadian Ministry of Defense is still open to you and that you're still going to be able to get something out of this relationship? Well, I mean, at some level, I don't disagree. I mean, the less you ask of a tool, the greater the likelihood that tool will be able to do what you're asking it to do, right? So if all we're asking of the ally is talk to us, then this is a tool that can help them be willing

to talk to us. We're transferring stuff to them, right? That creates some willingness to talk. If what we're asking them to do is to defeat the Taliban for us, which is what we've been asking the Afghan National Defense and Security Forces to do for almost 20 years now, that's not so easily accomplished. And it's not clear that this is as cost-effective as it looks. After all, we've spent, what, $61 billion plus transferring capacity to the Afghans so far, $25 billion plus transferred to the Iraqis so far. If what we're asking them to do is the central

problem that the defense strategy is assigning to this job, frankly, deal with threats to American security that we don't want to send 100,000 soldiers to deal with, right? That means roll up the terrorist havens or defeat the insurgents. And for that kind of mission, which is demanding, yes, but it's a mission that we routinely ask partners to do for us. The tool is not as cost effective as it looks. We've spent a non-trivial amount of money trying to do this with remarkably little payoff in those kinds of theaters. If what we want from the partner is something closer to the ability to deliver humanitarian aid to the population, for instance.

That's a less demanding mission. We may very well be able to accomplish that. If what we want the partner to do is something the partner also very much wants to do, then the tool can be really quite effective. So part of my message with security assistance in general is because we need to understand where the ceiling is lower than people think, we in turn need to use that to do a better job of knowing when to apply it. But I think that the ceiling can be pretty high if our objectives are very much aligned. In certain cases, when our objectives are very much aligned with the partner force, the ceiling actually can be very high. But there aren't all that many situations in which what we want and what the ally wants

align closely enough for the ceiling to be as high as policymakers in Washington often think. The Kurds were actually an example here, right? Our interests and theirs aligned pretty well in Mosul. Our interests and theirs were headed to a significant misalignment in Raqqa. But when it came down to the issue of, are our Kurdish allies going to be nice to non-Kurdish Sunni Arabs in Syria, it was at best an open question as to where that was going to lead. But even with the, partly what's interesting about the Kurds is that it's in a situation of unusual interest alignment. And I think that's really important for understanding the Kurds as a case study. But Matt, how do you bridge these broader concepts of interest alignment with the more

tactical competencies that you were evaluating? Yeah, well, I mean, my research is much more focused on this micro level engineering problem. So the first question you have to ask is, regardless of this, the greater level of objective alignment, what strategic objectives are you trying to accomplish? The first question you have to answer is, does this actually do what you want at the small unit level? And I think it's significant that I get started at this,

you know, at the company level, while Steve is parachuting into the theater level headquarters, and then I'm studying these very micro level issues at the small unit level. And it actually seems really good too, because you studied the Kurds, which we're saying is kind of an ideal case

in interest alignment. So maybe this shows us the best version of partner force training that we might expect. Is that accurate? Well, I think that they have a lot of things going for them. One, we have the shared enemy who is an existential threat to them, both in Iraq and Syria, right, where they have been very much pushed back by the Islamic State, and that they

are, they've seen what the Islamic State does to people who it considers kafirs, unbelievers who they conquer, and it's not good. And so they are very, very motivated to avoid this. You know, you have a lot of people joining who are talking about their Kurdish identity and their Kurdish nationalism and not talking about, well, I just needed a job. and then you're trying to transfer this engineering problem to them of, can I transfer this knowledge and supplies to you, then you would say that the raw material is probably better in this case than in other cases. So what factors in this case were you able to actually assess? So during the summer of 2017, I did a survey of 2,300 Peshmerga, who are the Iraqi Kurds. of 2,300 Peshmerga, who are the Iraqi Kurds. And we found that sort of a third of them had no training whatsoever, that they had just showed up at the front lines and started fighting. So they're completely untrained. It's just all OJT on the job training. Then you have a third

who get this internal training where they are sent to formal training sites run by other Peshmerga. internal training where they are sent to formal training sites run by other Peshmerga. Problematically, these Peshmerga are like the veteran who I spoke with from the Iran-Iraq war. They come from the Iraqi military, which as we know from its performance in 1991-2003, is not very capable, does not have this modern system down. And so they're not very good at training their new recruits in modern system tactics. And then you have a third who get training by the anti-ISIS Western coalition countries. And these third of Peshmerga are significantly more confident in their battlefield performance. And they're more likely to actually fire back and engage with the enemy on the battlefield, where sort of of the non-coalition trained Peshmerga, 17% of them

are reporting that they're hiding in combat, that when they take fire, it's like what Marshall reported in World War II, this very infamous statistic that 75% of American soldiers didn't fire back. Almost certainly an exaggeration, but everyone who's been in the military on deployment sort of gets the sense that not everyone is responding to danger equally. And that a big part of that that comes out in my research is that training can significantly improve this and that getting trained in the Western system, in the modern system, is making these Peshmerga more confident, more willing to fight back and less likely to hide. And so at the small unit level, the engineering question is that, yes, under some circumstances, this can work in actually increasing small level effectiveness. But then the larger question that Steve does a lot more research on is, well, does this aggregate up? And to what extent is this a unique situation with

these Kurds? And to what extent does this pertain even in Burkina Faso in 2020? So what kind of training did the one third who reported that they were more willing to participate in combat? What kind of training were they receiving? You know, it's really a good story of the NATO allies come in there and each of them takes a different block of training where sort of the Germans built a mount of military operations in urban terrain town to run the Peshmerga through. I think the British were doing marksmanship, if I recall correctly. The Italians had sort of classrooms about land navigation. And then they also would send out special forces teams to individual units to conduct overall training in not just marksmanship, but basic fire and maneuver. Okay, you shoot,

and then I'm going to advance with your suppression. And the Kurds, this was sort of revolutionary to them in a lot of ways. At Mosul Dam, I talked with some Kurds who were just graduating from a Dutch course, and they went out to fight against an ISIS attack the week before. And they said, well, we were really surprised, but this actually works, you know, where you have one Hunvi suppress them, and then we advance, and then the other Hunvi advances, and then we go back and forth. This actually helps. And so I do think that there is a certain element in a lot of these contexts where at the small unit level, what we do is effective.

But there's also a problem where the people who are conducting the training on the Western military side also have strong incentives to report that our training is working and you don't have other people going in and checking their homework. And so you just get these reports from the trainer who says, I did a great job. You don't go and verify that with any sort of objective analysis and that the only thing you have to go off of is somebody whose incentives are strongly aligned with reporting that they had a good job. It's also hard to report that you actually improved the capability. It's a lot easier to report the number of people that you ran through a course. Right. I mean, there's a big problem with inputs, outputs, and outcomes, right? Where the input that you put in is however many dollars we spend, how many soldiers or Marines we assigned to train. The outcome, the output rather is much, is also pretty easy. How many people you're actually training, how much tons of supplies you

transfer, how many AK-47s you give them. And then you have the outcome of, did this actually help them on the battlefield? And that's much more difficult to measure. But Steve, are these tactical payoffs relevant to the strategic perspective you take in your approach? Well, ideally, you want it to go end to end. If this is going to work, if it's actually going to help in achieving those policy goals, at some point at the point of the sphere, down at the tactical level, some recipient unit has to do better. The point of the analysis, if one is doing the analysis right, is to make an argument about underlying characteristics of something. And the argument that I've been making in this work since I've been doing it is that, in fact, some fairly profound features of the political organizations of these

states ripple through the system down to the point where they affect the ability of a platoon in combat to function and the ability of a platoon in contact to motivate themselves to get off the checkpoint, go out and patrol, kill Taliban, and do all the rest. Part of the reason for looking at that whole chain, though, is some parts of it are susceptible improvement by folks on the ground doing the hard work with the ally, like Matt, who can train them this way as opposed to that way, who can motivate them this way as opposed to that way, who can motivate them this way as opposed to that way, who can monitor them more aggressively or less aggressively. Harder performance at the point of the sphere, though, is these broader issues of the political system of the country, which it's unreasonable to ask Matt to fix. Matt, would you please go get the Afghans a judiciary that works?

I mean, I think that's very, you know, very true. And I guess the question is, for people who are at lower levels than the theater, is this just an issue of reporting up for, you know, sort of field grade, company grade officers reporting back to the higher level decision makers, like these people are not making progress. There's not a lot I can do. It sort of runs, cuts antithetically to the military's can-do attitude. So I guess the question is, what's the policy recommendation there? Yeah, and I think they're important in different policy implications at different levels of command. There's some big issues that need to be worked at the theater level and that are four stars job description, interacting with the state and the other country to get them to fire corrupt corps commanders. That's a four-star job. If you go all the way down to the advisor level, the captains, the majors, the sergeants who are working with the allied military, part of what I think a more political understanding of this implies is the monitoring information that's

necessary to inform conditionality, which is the usual recommendation from a more political view of this. If you view this as politics, at least as much as engineering, the way you create political incentives for the ally to do what we want, and put it to what they want, is through conditionality. As an illustrative example, a wonderful doctoral dissertation just completed by now Dr. Bryce Lloyd-Ole, George Washington University, looks at this question of if you take this more political understanding of the enterprise, what does that imply for how advisors should do their jobs? And he looks at how aggressively they should monitor. And part of the reason he looks at this is there's a lot of variance out there in the field. Some advisors are

monitoring really aggressively. They check everything. They do inspections seven times an hour. They're in the knickers of the advised unit in a very aggressive way. Others think this interferes with the building of rapport. And it makes the advised unit think that we don't trust them, oddly enough. And therefore, some advisors take a much more hands-off view of their role in monitoring because they don't want to destroy the relationship. And without a lot of doctrinal guidance, therefore, advisors kind of do what seems best at the moment, and there's all sorts of variation there. So Bryce looks at that variation and tries to associate it with better and worse outcomes for the advised units and comes

up with an argument that the right answer, oddly enough, is in the middle. Part of the recommendations that I often draw from the work I'm doing is that a doctrine needs to reflect a more political understanding of this so that all of the advisors that end up in places like Afghanistan or Burkina Faso or Kurdistan or wherever, bring with them a consistent understanding of how the process works and therefore what their job is and what they need to emphasize and what they need to do to do that. There are lots of other implications too, but that's a start. I mean, the problem is that a lot of that is kind of idiosyncratic to individual personality. You can set rules and regulations for how to monitor, but at the end of the day, you can tell an officer to not micromanage his team and then half of them are going to micromanage anyways because it's within their nature.

So I think that's why advising is such kind of a complex process because you've got to vet personality along with conduct the training to get them to the right place. One of the reasons why it's an important thing that smart, thoughtful officers like Matt are out there doing this is this is a hard project. It's a different enterprise than a lot of things that relatively junior field grade officers do in the military. It requires judgment, and it requires a thorough grasp of the larger cause and effect relationship of the project they're engaged in. I guess one question is, so it comes up about conditionality, right? Where if the partner does not do what we want them to do, then we're going to cut you off. At what level should that conditionality be happening? Because

obviously in Afghanistan, we've exhibited that we have no conditionality and that no matter what the Afghans do, we're going to keep writing you a check. Yeah, I think the kinds of policy issues about conditionality that you're talking about are jobs for 06s and 4-stars. And it's got to start with the 4-star level, right? So part of the problem with this is conditionality means necessarily denying some resource that's useful for fighting a war to an ally who's fighting a war. So you've got to be careful and discriminating about where and when you use it, or else you end up undermining the whole war effort with it. So usually what you would want is when at the four-star level, you're building the campaign plan for the theater. One of the lines of operation in a typical campaign plan is the governance.

And that should be approached the way you would approach any other part of a campaign plan. It should be a main effort. There should be economy of four sectors. There should be places where we're going to accept risk. There should be places where we're going to devote resources. And similarly, take Afghanistan as a theater, for example, right? You're not going to be able to apply rigorous conditionality simultaneously across every regional command in the country, because A, we don't have the resources of attention to do that right. B, we would undermine the military campaign. What you need to do is make a decision at the theater level that we are going to start with Kandahar. And we are therefore going to engage in a program of conditionality in Kandahar. Elsewhere, we are going to tolerate

corruption. So you, Kansian, over here in Helmand, your job is keep monitoring, but hold the line. No conditionality, just provide the stuff, but fill the intel pipeline with information on how it's being used, right? But we're not asking you to imply conditionality. We need that district chief of police to stop underwriting land seizures by the Malign Actor Network associated with the government in Kandahar. And to do that, we need you to tell him if any more pomegranate orchards get seized by the government, We are shutting off your money. And we are on the lookout and we are monitoring land use in your AO. And A, I want you monitoring land use in your AO. And B, I want you delivering the message to that DCOP, that district chief of police, that either it's this or no more funds. But you don't want the local officer making up their own mind on this. Well, I think you bring up an interesting point because it's not just who's making up the mind, but even in situations where that type of conditionality may have been used in the past with kind of a strategic objective, I'm not sure that gets communicated

down to the officers at the tactical level. Like, hey, you are doing this. Instead, we just try and turn on and off spigots. And the person on the ground is still just trying to operate to the best of their ability and trying to maximize their performance because they think they're going to get measured based on the outputs, not necessarily on the outcomes. Well, and this gets to the point you made earlier about different personalities pursuing this rather differently. And this can happen in time series too. So, you know, Captain Schmedlap shows up and he's really tough on the local district chief of police. And he's hammering the guy with conditionality and he's monitoring really aggressively. And then he rotates out. And

the next person in is Captain Biddle. And Biddle's just a softie and a schmo. And it's important to have rapport, right? And you need to avoid being the ugly American and oppressing people. And so he's nice to the guy. And now the guy is just totally confused. And he can't figure out these Americans because one of them behaves one way, another behaves the other way. The guy in the AO next door is really tough, the guy in my AO is really soft. If this isn't coordinated, all you do is you confuse everybody. And if what you're trying

to do is create political incentives to change behavior, confusion is your enemy. It needs to be clear what you want them to do and what you need them to do and what the consequences of doing it or not doing it are. If you leave this up to the judgment of the local officer, which will be influenced inevitably by that local officer's personality, it will be influenced by all sorts of other things. And what you get is randomness. them this. Now, preferably, you want these kinds of things to obey the principle of being divisible, revocable, and therefore potentially contingent. So wherever possible, you want the assistance program to be such that you can divide it into little bits and turn it off and on at the margin rather than all at once, right? Ideally, what you'd rather not do is go to a district chief of police and say, we're pulling all the assistance out tomorrow

unless you behave yourself. It would be much better to say, we're pulling out 10% of our assistance. Right. I mean, and that's a big challenge with if officers are being graded on the output of their system. And so that this is not just incentives that we're creating outside for the partners, but also inside where maybe then instead of being judged on the capital that you transfer to your partner or that you spend in civil affairs, that you get judged on the amount of monitoring you do and the quality of these reports that get sent up to hire. No, I mean, this is a job for doctrine, right? I mean, doctrine should imply what are the standards of evaluation going to be for officers doing these jobs. If the doctrine is based on an assumption of interest alignment, which is never the case, but that tends to be the way our doctrine reads, and you see this as an engineering process,

you will therefore evaluate captains and majors on the basis of how much resources they transfer. Most of the conversation we've talked about, you know, kind of the major cases that most people think about, which is Iraq, Afghanistan. The security force assistance that I'm more familiar with is the other 130 something countries that we have military advising missions go on. So sometimes we don't have a representative in every district. We have a small team of individuals. So I think going back to something Matt said earlier is that the question of are small payoffs okay in some circumstances? I think that when you only have maybe a team of 10 individuals, our objectives really are not at all training the partner force.

That's just a diplomatic tool to get us access. Our actual objectives are relationships, influence, maybe overflight, maybe intelligence collection. It's very easy for us to go to the big cases that we've all talked about, but there's a lot more that security force assistance and training equipment is used for out there. Well, I want to be clear that I'm not arguing that we should never do this. I'm arguing that we overdo it. I'm arguing that we rely on it too much, especially in these big Muslim theaters like Iraq, Afghanistan. But that doesn't mean that there's no role for it. The ceiling on what it can offer for us isn't zero. It's lower than people think it is. Therefore, we should use it more discriminatingly than we often do, but it's not zero. And I'm not sure I have a huge problem with sending 12 Green Berets to Burkina Faso. And it's not just special operations forces who are conducting these light footprint advising

missions. Advisors come from across the joint force to conduct security force assistance in countries around the world to include, of course, the new Army Security Force Assistance Brigades who have deployed to not only Afghanistan, but also Africa, and reportedly will be deploying to Colombia soon. Well, the Security Force Assistance Brigades were another matter, right? Now we're not talking about 12 Green Berets. Now we're talking about a serious chunk of resources. If I'm going to restructure the Army to produce a bunch of security force assistance brigades and send them places, now I want to know whether they're doing anything really worthwhile with it. Redirecting whole brigades of the army to this mission, if all they're doing is persuading people to talk to us and allow overflights, I'm not sure that's worth the money. But Steve, are there instances when security force assistance has been applied successfully on a larger scale?

I tend to find the case of the Korean War interesting. The Korean War involved a surprisingly large amount of security force assistance. The Republic of Korea's army at the beginning of the war was small and remarkably poorly equipped and remarkably ineffective by the end of the war, by the end of the active phase of the war. It became large, proficient, and very well equipped. And that process was largely a business of security force assistance, or if you prefer, building partner capacity. And it was, I think, about as effective an example of that as I can think of offhand, with the possible exception

of Matt's work with Kurds. What was it that made the Korean exercise so much more effective than, say, the Iraqi exercise or the Afghan exercise? And I would argue that there are at least two big distinguishing features about Korea that are important to keep in mind. One is there was an unusual degree of political interest alignment between ourselves and the Syngman Rhee government, thanks to the North Koreans. The North Koreans conveniently aligned Rhee's interest with ours by threatening Rhee and his government with extermination, which oddly enough Rhee didn't want either. How would you say that that's different than the case of the Taliban versus the Jiroa? Because the Jeroa doesn't think that, for most of the war anyway, the Jeroa has not thought that the Taliban were a proximate threat

to destroy them. Maybe manana, eventually, someday, but not soon. The interesting thing about the North Korean invasion of South Korea is it was an acute, chronic, existential, obvious, inescapable threat. And in a period of weeks or months, it could have destroyed this government and assassinated the leadership. Secondly, they stayed in alignment. Often in situations like this, there's an initial interest alignment if an insurgent or a cross-border invasion or whatever looks like it's going to topple the government. In the El Salvador case, early on, it looked like an insurgent or a cross-border invasion or whatever looks like it's going to topple the government. In the El Salvador case, early on, it looked like the insurgents were going to take San

Salvador and topple the government. Usually in situations like that, you get early interest alignment. U.S. aid is reasonably effective. The ally gets just good enough to prevent being conquered, and then they're not willing to reform any more than that. But the second interesting thing about Korea is an intensive degree of monitoring. Individual U.S. advisors that were working with South Korean Army small units controlled the unit's budget. They had all the paperwork. They were monitoring all the money flow. had all the paperwork. They were monitoring all the money flow. They were in these guys' knickers to an extraordinary degree, and the theater command was using that information to turn off

resources to units that were corruptly misdirecting our assistance. When you get that combination, the monitoring to create conditionality coupled with the interest alignment creates an ally that in just a couple of years becomes an extremely proficient conventional military force. We've been at this in Afghanistan for close to a generation. budget and very acute is very different than our policy now, which is always by, with, and through, where I hear that in some meetings repeated like a mantra that's going to ward off evil. By, with, and through. We're by, with, and through. We're empowering the local government. And in fact, that's not what you want because empowering the local government, they have these sort of internal clashes that they want to use their resources for more than the external threat that you want. And so that actually this whole idea of by, with, and through, we're going to work through the local government actually can be very counterproductive.

Absolutely. Right. And as we tie all this together, what do you think are some of the big takeaways, at least for that practitioner level? Right. I mean, the first question is, how do we know if we're doing the job that we say we're doing? And that it never worked to grade your own homework. You always give yourself, well, okay, I was close enough on that one. And that this is the same thing with security force assistance and building partner capacity, is that this is something that needs to be monitored, that the government accountability office, other people have called on to be monitored, but that it's really hard to do. You really need to build a monitoring infrastructure. The second thing is that people need to not be afraid to fail and to report that for this conditionality to work, report that, you know, my local partner is really not that great, you know,

and that don't be afraid to say to hire, I don't think you should keep funding my effort. And Steve, from your vantage point? Yeah, there are a couple of things I would suggest. The first is if you take a more political understanding of the phenomenon, that implies you need conditionality. Conditionality is all about understanding the interest divergences between yourself and the ally that you're working with. So the first thing I would do in light of that is I would rewrite written doctrine to reflect a more systematic thoroughgoing understanding of the political nature of this problem, the importance of interest alignment, and the need for conditionality to create incentives in the presence of interest

misalignment. And that then, if the doctrine is rewritten appropriately, will do things like change the mission statements and the evaluation criteria for the officers that are doing the advising as well as their chain of command. So the first point is to rewrite the assistance programs preferably from the beginning to be divisible, revocable, and therefore potentially contingent to conditionality. And we often don't do that because it means designing them for flexibility rather than efficiency. Well, that's all we have time for today. We really appreciate the conversation. This has been helpful in understanding how building partner capacity fits into irregular warfare and how we apply it and how we can do it better. So thank you for being here. My pleasure. Thanks for having me. Thank you. Thanks again for listening to episode three of the Irregular Warfare podcast.

We release a new episode every two weeks. Up next, Nick and I discuss irregular warfare oversight in D.C. with former Acting Assistant Secretary of Defense for Special Operations in Low-Intensity Conflict, Mark Mitchell, and former Congressional Staff Member, Pete Villano. Following that is a conversation on how armed rebel groups manage human resources with Dr. Vera Miranova, who interviewed over 600 Syrian fighters and civilians on the front line in the Syrian Civil War. conversation on how armed rebel groups manage human resources with Dr. Vera Miranova, who interviewed over 600 Syrian fighters and civilians on the front line in the Syrian civil war. Please be sure to subscribe to the Irregular Warfare podcast so you don't miss an episode.

You can also connect with us on Twitter, Facebook, or LinkedIn to start a conversation about past or future episodes we have lined up. One last note, what you hear in this episode are the views of the participants and don't represent those of West Point, the Army, or any other agency of the U.S. government. Thanks again, and we'll see you next time.