The Security Challenges of North Korean Collapse: A Conversation with Bruce Bennett and Jennifer Lind
This fall, International Security (October 2011) published Bruce Bennett and Jennifer Lind’s article about the collapse of North Korea. The article examines the problems that could result from a government collapse, the military missions that countries might take to stabilize the country, and the requirements of those missions. Bennett and Lind argue that stabilizing a collapsed North Korea could require 200,000 ~ 400,000 military personnel. Their analysis has implications for South Korean defense policy as well as regional diplomatic coordination.
38 North sat down with Bennett and Lind recently to find out more about their study.
38 North: How did you come to do this study?
Bennett: Jennifer and I worked together at RAND on a similar project more than a decade ago. We saw that many analysts had written on the questions of whether and how North Korea might collapse, but no one had really talked about what would happen afterwards. In our view, there could be so many difficult and dangerous problems associated with North Korea’s collapse that we thought it important to start a conversation about that. So we wanted to do an open-source analysis that examines the various problems that could be associated with collapse, the military missions associated with mitigating those problems, and the rough requirements for those missions.
Haven’t people been saying that North Korea will collapse for years? Do you two argue that collapse is imminent?
Lind: My previous work with Daniel Byman emphasized the many powerful tools that the regime uses to stay in power. Byman and I argued that because of those tools, especially the terror that the regime imposes on the North Korean people, Kim Jong Il’s regime seems very secure against popular revolution. The regime has also tried its best to “coup-proof” itself against challenges from elites. So Kim Jong Il’s position seems secure. However, Byman and I noted that the regime is preparing for succession, which is always the biggest challenge for an authoritarian government. Succession may go well, it may not.
So with succession on the horizon, it becomes important to talk about what might happen if it fails—if the North Korean government collapses (which is only one potential outcome from a failed succession; there are of course others). And the other reason to talk about collapse is that the problems that could occur are really frightening—in terms of the potential humanitarian cost, as well as the risk of crisis escalation in East Asia. Even if you think the risk of collapse is low, it is such a dangerous situation, it’s important to think about and plan for it.
In the article, what is your argument about how North Korea collapses?
Lind: Well, that isn’t the focus of the article—other analysts have already spent time thinking through different scenarios of what actually triggers collapse, or how various scenarios might develop. That said, to do the analysis that we wanted to do, we did need to start with a scenario. Everything pivots on the assumptions you make going in: depending on the assumptions, you can either generate findings that say managing this would be no problem, with minimal stabilization requirements, or that it would be Armageddon and will require a million soldiers to stabilize.
We chose to model a situation that was on the benign side of the spectrum. (This isn’t the true “soft landing”—a soft landing would be a managed transition in the style of East Germany. We’re talking about collapse, after all, which is always going to be somewhat unpredictable and dangerous.) The scenario we had in mind was an implosion scenario—succession fails as elites contest the successor, and the government sort of unravels. Elites are fleeing (instead of gathering up military units around them) and the military dissolves, as we saw in Iraq.
A lot of people would find that scenario really unlikely….
Bennett: For good reason. I myself have done a lot of work in which I describe and study collapse situations that would be far more dangerous. Most Korea experts can come up with a long list of reasons why collapse would be much more dangerous than what we modeled in our article. Government and military elites might compete for power, gathering up military units behind them. Regional warlords could emerge, who compete for territory and resources. So we could see a failed state like Somalia, but in a heavily militarized country in which most men have had military training, over a million men are serving in the active military and security forces, and over seven million people are in the reserves. As the elites hoard food, a humanitarian disaster could develop and North Korea could well descend into a civil war.
Furthermore, one might also be skeptical about how cooperative and docile the North Korean people would be in the event of U.S. and ROK intervention. There is deeply ingrained anti-Americanism and prejudice against South Koreans, coupled with very strong North Korean nationalism that may make the people fear and resist such intervention. So it’s not hard to imagine that North Korean government collapse could be very dangerous.
Lind: It is important, though, to remember that back in the 1980s, you wouldn’t have found many analysts who would have thought that in a few years, East Germany would go down the way it did….
Bennett: That’s certainly true. But, most importantly—we didn’t adopt this assumption about collapse based on our predictions of how it would occur. We assumed a benign scenario because it was the most analytically useful way to structure the project. We could have adopted, as Jennifer said, “Armageddon” assumptions for collapse (the elites are fighting a civil war; war breaks out with South Korea; nuclear weapons are used; etc.), and missions for a scenario that dangerous would have generated massive requirements. But then the obvious rejoinder is, “well, it might not be that bad.” So our idea was basically to do something like a “best-case” scenario. We say, ok, here’s the way it will look if we are very lucky and North Korea collapses in a relatively benign way.
If collapse occurs in a much more dangerous manner, our analysis is still a useful guide. When we use force ratios to calculate requirements, for example, we note how you would adjust those ratios if the environment is more dangerous. So our analysis serves as a useful starting point for a conversation about collapse.
Lind: What I tell people is, if you strongly believe North Korea’s collapse will occur in a more dangerous way, you would need to add new missions, plus tens of thousands of troops, to our estimates. For example, we estimated the requirements of a stability operation using a number for a “garden variety” operation—13 troops per 1,000 people in the population. If planners believe that a North Korean stability operation would be much more dangerous, they might go higher than this, to 16 or to the 20 that some analysts recommend.
Why are your estimates for force requirements so high—given that, as you said, your scenario assumes a relatively benign collapse?
Bennett: We model several different missions—the missions that we believe would be the most urgent for stabilizing the peninsula. But one mission in particular, the humanitarian stability operation, is the big force driver. The goal of this operation would be to secure the lines of communication (so food and medical supplies can be sent in via roads, ports and airports); to distribute humanitarian relief supplies; and to take over normal public security functions (substituting peacekeepers for the North Korean police). As Jennifer just said, we estimated the force size at 13 troops per 1,000 people; given North Korea’s population, that generates a requirement of over 300,000 troops.
Given how daunting (both diplomatically and logistically) it would be to organize and insert such a force, we look at other ways to stabilize the country that require fewer forces. For example, we modeled an operation in which we divide North Korea into tiers and stabilize each tier sequentially. But there are also, as we point out, drawbacks to doing it that way. What you save in requirements may translate to greater instability or a longer operation.
Do you envision a big role for the United States in stability operations in post-collapse North Korea? Would the war-weary American taxpayer go for this?
Lind: We do the basic analysis in a generic way—we count how many troops are needed to perform these missions. We aren’t saying what color uniforms they should be wearing. So it’s a military analysis, not a political one. We say, here are some serious problems that might arise after North Korean collapse—if some country wanted to conduct a military mission to alleviate those problems, what would those missions be, and how many people would be required to perform them?
We then get into the politics to some extent: we note that there are some countries that are likely to get involved—based on geography, perceived interests or capabilities that they bring to the table. Those countries of course include China, South Korea, and the United States. Russia and Japan are also proximate and have interests engaged. But the point of the article is not to advocate anyone’s involvement; the point is to calculate the requirements for these missions in a generic sense, in order to help countries in their individual planning, and to inform the public debate.
How do you model going after the nukes? Who’s going after them?
Lind: While the paper doesn’t make recommendations about involvement by any particular country, realistically, we can imagine that China and the ROK/US Combined Forces Command (CFC) would be very interested in making sure that the nukes don’t slip across borders after a collapse. The United States has specialized capabilities for “WMD elimination” that other countries just haven’t developed. We hear media reports that CFC is already training for this kind of operation. So it’s not a stretch to imagine that those countries would get involved.
Figuring out how one would conduct a counter-WMD operation in a collapsed North Korea is really tricky given how little we know about its WMD programs and infrastructure. We don’t have a good idea how many sites there are, or where they are located. People have also talked about the existence of thousands of tunnels and underground facilities that, if there, will need to be searched. So it’s impossible to know right now to know with much certainty how demanding this kind of operation would be.
In our analysis, we argue that a counter-WMD operation should have four parts (performed simultaneously): First, containment—monitoring and intercepting vessels leaving North Korean waters and aircraft leaving North Korean airspace, and monitoring overland traffic at borders. Second, surveillance of known sites to make sure that no one is moving or looting materials from them. Third, securing the highest priority, known WMD facilities—or places where any movement has been detected. This could be done using assault teams inserted from South Korea, from off North Korean coasts, and perhaps from China. We model this operation akin to a “raid” like the Bin Ladin raid in Abbottabad, and estimate you’d need 3,000-10,000 troops for this component—on the lower end if missions can be prioritized and sequenced, rather than performed simultaneously. Finally, one would want to do a methodical sweep through North Korea’s WMD facilities in tandem with the stability operation and a disarmament operation (some of the other missions we model).
Lind: Yeah. That’s the reaction we usually get. It’s always a bit surreal to give a talk on this and bring the audience from where we are today, to a world in which China or CFC are inserting assault teams into a collapsed North Korea! But that’s the point of the analysis—to give leaders and the public the tools for how to think about this, and to plan for it—and to hope the world never comes to this.
In the article we explain that the consequences of a poorly planned response to North Korean collapse would be calamitous. Many of these missions are time-sensitive—for example, the longer it takes to organize humanitarian efforts, the more North Koreans might perish or decide to take flight. The longer North Korean WMD sit unsecured, the greater the risk that they will disappear. So plans should be in place well in advance, in case collapse occurs, and in case leaders want to act.
Joint planning is also essential because uncoordinated efforts to stabilize North Korea have severe escalatory risks—for example, if China and CFC send military forces into North Korea to locate WMD or to secure borders. Imagine the Chinese People’s Liberation Army racing south, while American and South Korean soldiers are racing north. China is a nuclear-armed country with which we have challenging relations—accidental conflict and subsequent escalation could be catastrophic.
What sort of planning is currently going on?
Bennett: When Jennifer and I were working on our project at RAND a decade ago, there wasn’t much planning. This started to change when people were so shocked by Kim Jong Il’s stroke in 2008. Since then, media reports indicate that CFC has begun combined planning, and the Chinese appear to be planning on their own as well. But there hasn’t been combined planning with China—Beijing has been very unreceptive to the idea. China, after all, is North Korea’s key ally, and Pyongyang would not look kindly upon Beijing making plans for the North’s demise! China’s principal objective with regard to North Korea is to maintain stability; Beijing believes that openly planning for the collapse of the North Korean government would undermine that objective and could increase the probability that North Korea collapses, which Beijing wants to avoid at all cost. The Chinese also seem to think that this kind of planning smacks of planning for imposing regime change. But given the potential instability that our paper examines, it’s essential that the Americans and South Koreans create a dialogue with China on this.
Diplomats have been saying for a long time that the Americans and South Koreans should be talking to China about this. Do you really think there is any way to get China to agree?
Lind: We hear that there have been discussions with the Chinese at least at the Track II level, such as among scholars or retired military personnel. Such discussions are important, but the most productive dialogue would be confidential discussions at the Track I or I.5 level. (I hope that such discussions are already going on, and we just don’t know that they are happening.) Regardless, the US government should keep asking Beijing for inter-governmental discussions—given the risks, it’s imperative that the governments communicate directly.
Do you envision a role for China in a stabilization effort?
Bennett: Well, again, our paper doesn’t make recommendations of this kind. However the truly daunting manpower requirements should give the South Koreans pause (they are the ones on whom the burden will fall the most). There may be ways in which Chinese intervention could help, if stabilization is done multilaterally, perhaps through UN auspices. China is, after all, far closer geographically to likely North Korean nuclear weapon sites and other key threats that need to be resolved in a time urgent manner. In the paper we note that Chinese forces could, for example, guard China’s own border, and could participate in a multinational humanitarian operation.
How would South Korea feel about Chinese intervention? What about the United States?
Lind: Many people believe that, in the event of instability in North Korea, the Chinese will move to stabilize their border—not just move to their border, but drop down into North Korea. Either way—if CFC wants China in, or if it doesn’t want China in—this needs to be coordinated in advance. But ultimately Seoul and Washington may have no say whatsoever about Chinese involvement.
Bennett: Many of my South Korean colleagues firmly oppose the prospect of Chinese intervention. They worry that the Chinese will occupy North Korean territory and either refuse to withdraw or will make very serious demands to withdraw. The Chinese “Northeast Project” which identified much of Korea as being historically Chinese territory has caused a great deal of worry. So many South Koreans are very nervous at the thought of Chinese forces on the peninsula. But if South Korea wants to avoid Chinese intervention, it needs to have the military capabilities to prevent instability following a collapse—in order to shield China from that instability and prevent China from taking matters into its own hands.
Will collapse be easier to manage if it happens sooner rather than later?
Bennett: Yes and no, but mainly no. In favor of collapse happening sooner, we’re seeing a trend of very low birthrates in South Korea plus big reductions in the size of their military. Seoul has already reduced the size of the military from 700,000 in 2000 to about 650,000 now, and may reduce it further to 400,000 or so by 2025. Most of these reductions are concentrated in the Army, which would be tasked with these missions. So over time, the Army will be less and less able to provide the number of personnel that would be needed to handle a North Korean collapse. And while South Korea has a large reserve force, most reserve personnel get no more than three days of training per year—not the kind of training one would need for the sensitive missions that we describe.
Lind: On the up side, if collapse doesn’t happen for years and years, this allows more time for preparations. South Korea can make sure it fields the forces necessary to perform the missions we describe. CFC would have more time to train together. And—back to our big point—with more time there can hopefully be more joint planning. More broadly, the logic of the “Sunshine Policy” was for South Korea to take action today that could reduce the burdens of unification. The peninsula hasn’t looked very “sunny” for a while, but perhaps that will change.
What else can South Korea and others do right now that might make collapse easier to manage?
Bennett: One strategy, which the United States did a bit of prior to the invasion of Iraq, was to attempt to establish contacts with North Korean elites. We assume in our article that the North Korean elite generally accept a ROK-led unification (as in East Germany, where elites accepted West German-led unification). As I mentioned before, many people would think that would be pretty unlikely. But perhaps North Korean elites could be influenced through contact. Furthermore, in general, more and more information from the outside is penetrating into the North, and the North Korean elites are learning just how much things are better in the South.
Another big part of collapse preparation should be intelligence collection. If South Korea and the United States want to go after North Korean nuclear weapons after a collapse, they need to know where WMD are stored.
So what happens to the Kim family after collapse?
Bennett: When I talk with South Korean audiences about unification, they talk more about punishing North Korean offenders (admittedly some of which needs to happen) and seem reluctant to think about selective amnesty or providing pensions to senior North Korean leaders to get them to agree to change.
Lind: And you can imagine why they feel that way—the Kim regime has brutalized Koreans and people will want to hold them accountable. As we’ve previously seen in transitioning countries, Koreans will have to wrestle with a tough dilemma between the desire to bring these dictators and other culpable elites to justice for their terrible human rights violations, versus the need to get these guys out of power, and to stabilize the transition—which usually means, as Bruce notes, amnesties. The issue of justice in a post-North Korea world is just one of the many daunting challenges Koreans may have in front of them.
Bruce W. Bennett is Senior Defense Analyst at RAND in Santa Monica, California. Jennifer Lind is Assistant Professor of Government, Dartmouth College.