Many of the specific questions raised about our article’s limitations by the commentators are, indeed, true, but they reflect the stated approach of the paper. North Korea is a country where the uncertainties are great, and this is no truer than in trying to anticipate a future North Korean government collapse and potential transition to Korean unification. Moreover, information on North Korea is scarce and difficult to interpret in large part because of North Korean information denial and falsification efforts. As a result, it is important to note our statement of the purposes of our article: “First, we seek to bring into the public debate a discussion of the scale of the problems that the collapse of North Korea’s government could create, and the potential for dire consequences, both humanitarian and strategic, if stability efforts were delayed or failed altogether. We describe the military missions that might be necessary to stabilize North Korea and estimate the force requirements for those missions. … Second and more broadly, this analysis sheds light on international intervention in collapsing states.” (86) With their comments, the reviewers have certainly contributed to our first objective, and their comments add to what we have contributed on the second. Moreover, we developed estimates of the military force requirements because we felt they would help motivate a public debate.
Author’s Responses to H-Diplo | ISSF Article Review 13
Edited for ISSF/H-Diplo by Diane Labrosse
George Fujii, H-Diplo/ISSF Web and Production Editor
Responses to H-Diplo/ISSF review by Brendan M. Howe of Bruce W. Bennett and Jennifer Lind. “The Collapse of North Korea: Military Missions and Requirements.” International Security 36: 2 (Fall 2011): 84–119. DOI: 10.1162/ISEC_a_00057. http://dx.doi.org/10.1162/ISEC_a_00057 , and to comments by Thomas Nichols, http://h-net.msu.edu/cgi-bin/logbrowse.pl?trx=vx&list=h-diplo&month=1204&week=c&msg=BNAPSGCjkq9%2bqv0yn2bq5w&user=&pw
Response by Bruce Bennett, The RAND Corporation
Response by Jennifer Lind, Department of Government, Dartmouth College
Published by H-Diplo | ISSF on 25 May 2012
PDF- https://issforum.org/ISSF/PDF/ISSF-AR13-Response.pdf
Response by Bruce Bennett, The RAND Corporation
Many of the specific questions raised about our article’s limitations by the commentators are, indeed, true, but they reflect the stated approach of the paper. North Korea is a country where the uncertainties are great, and this is no truer than in trying to anticipate a future North Korean government collapse and potential transition to Korean unification. Moreover, information on North Korea is scarce and difficult to interpret in large part because of North Korean information denial and falsification efforts. As a result, it is important to note our statement of the purposes of our article: “First, we seek to bring into the public debate a discussion of the scale of the problems that the collapse of North Korea’s government could create, and the potential for dire consequences, both humanitarian and strategic, if stability efforts were delayed or failed altogether. We describe the military missions that might be necessary to stabilize North Korea and estimate the force requirements for those missions. … Second and more broadly, this analysis sheds light on international intervention in collapsing states.” (86) With their comments, the reviewers have certainly contributed to our first objective, and their comments add to what we have contributed on the second. Moreover, we developed estimates of the military force requirements because we felt they would help motivate a public debate.
Turning to broad issues, we do not assume that the North Korean government “will collapse”, but rather that it could, and therefore that the implications of such a collapse need to be investigated. We also do not assume that “this collapse is imminent”, but rather that it could be. There is a serious debate in the North Korean expert community about the stability of the North Korean regime, a debate for which much of the needed information is not readily available perhaps even in North Korea itself, let alone outside of the country. We address a case of outside military intervention because we doubt the feasibility of non-military intervention into a state as heavily militarized as North Korea and because, as we note, some military planning is already going on between the ROK and the United States and by China. But we certainly encourage other scholars to make a case for non-military intervention if they feel that they can elaborate a feasible course of action. In addition, we do not make a case for “less benign collapse scenarios” than the scenario we describe because we think this scenario is far more benign than likely would be the case. This is a standard approach in the field of security studies for dealing with highly uncertain situations: Use the most benign case to establish the minimum requirements for action; “… this ‘best-case’ collapse scenario demonstrates that even benign assumptions produce extremely demanding force requirements for stabilizing a collapsed North Korea. More pessimistic scenario assumptions would increase force requirements and would lengthen the duration of stabilization missions.” (89) Note that the objective here is not operational planning for a collapse (which Professor Nichols rightly argues would not be done in this way), but simply trying to establish a minimum force requirement to inform discussions on the feasibility of such intervention and policy choices on whether to prepare for such a course of action or to pursue alternative options. This said, we welcome discussion of whether we have captured the minimum force requirements or whether lower numbers might be possible.
We recognize that the ROK political authorities and public may not support such an intervention based upon the potential economic cost and other reasons. While public opinion polls help inform the view of whether the ROK would intervene in day-to-day circumstances, the ROK populace can be very responsive to major events. Thus if China were to intervene first, trying to halt refugee flows, the ROK leadership and populace could significantly change their preferences (as discussed below). We believe that the ROK could decide to intervene, but if it does, it should have the benefit of also understanding the military requirements and challenges of such an intervention. With our article, we are attempting to stimulate a discussion on this topic.
With regard to specific concerns:
First, the community should indeed assume that there would be “boat people” in addition to refugees across land borders. We do not estimate the requirements for dealing with them (or for dealing with people leaving North Korea by aircraft) because the major force requirement for dealing with boat people would be for naval personnel; our estimates of force requirements are focused on ground force personnel.
Second, China may well intervene in a North Korean collapse; indeed, the ROK may be reluctant to intervene (as noted by the commentators) until it sees a Chinese intervention. A Chinese intervention into the North would be viewed by many in Korea as an attack on Korean territory (since the ROK constitution does not treat North Korea as a separate country), and would likely provide the spur for ROK action. And yes, there should be coordination with the Chinese to reduce the chances of accidental war should both sides intervene, and successfully reaching a mutually acceptable approach to such multilateral intervention would be very difficult. As Jennifer Lind and I have noted that in the past South Koreans have not wanted to discuss a Chinese intervention, often assuming that China would understand how distressing such an action would be to the ROK. But the ROK has not built a military force to prevent the need for a Chinese intervention (e.g., a force able to promptly take control of the North Korean side of the border with China and prevent the flow of North Korean refugees into China). ROK leaders appear increasingly willing to talk about such a difficult situation—a needed first step for the kind of accord that Professor Howe seeks.
Third, our article does not directly discuss how the potential for accidental conflict between China and the ROK/U.S. could be reduced; other work being done by one of us (Bennett) is addressing that issue. The purpose of the article’s force composition section was to identify other countries that could contribute to the required intervention into North Korea, not to describe the force sizes or concepts of operation required.
Fourth, our paper mentions the ROK defense planning [reducing its ground forces from “550,000 now (Marines and Army combined) to about 420,000 by 2020”], but we did not explain why these changes are planned. The size of the ROK military is primarily a function of available manpower and the conscription period; note that the ROK Army is still composed of about 75 percent conscripts. The ROK drafts almost all young men who do not volunteer for military service. The ROK faces a demographic problem here: The low ROK birth rate has already reduced the size of the age cohort that is drafted from over 400,000 young men per year in 2000 to just over 300,000 in 2008. While that number will be increasing in a typical birth rate cycle for a few more years, by 2025 the number will be not much more than 200,000 per year, roughly half the available pool of young men that existed from the mid-1970s through around 2003. In addition, President Roh Moo-hyun reduced the Army conscription period from 26 to 24 months, and then set in on a track to progressively decline from 24 months to 18 months in 2014. But President Lee Myung-bak halted this decline at about 21.5 months in the aftermath of North Korea’s shelling of Yeonpyeong Island in late-2010. As Professor Howe suggests, increasing this conscription period would be difficult (within the powers of the ROK President but politically highly unpopular), making the steady decline in the size of the ROK military fairly certain absent unusual circumstances. The ROK could offset this decline by strengthening its reserve forces that now mainly serve three days per year at most, which means that they are not very viable.
Professor Nichols criticizes what he calls our “dispensing with politics in order to get to modeling”, as well as the poor fit between the cases we cite and the likely case in North Korea. Here, he runs into a third key problem with North Korea (beyond the uncertainties and lack of data cited above): the lack of really analogous cases to a North Korean collapse. Interestingly, he seems to argue that all of our cases would tend to underestimate requirements, consistent with our choice of a most benign case to identify minimal requirements. More importantly, he does not suggest any cases which are more analogous, reflecting the conclusion we reached while preparing the paper that there were no truly analogous cases and that the existing cases underestimate requirements. We therefore believe that the difficulty here is not so much with our paper as it is with the availability of appropriate cases: We were forced to use what is available.
Professor Nichols also properly argues that in an actual scenario, the assignment of stabilization forces to an area would be based on many issues beyond the population. He is, of course, speaking of operational/tactical planning, whereas we are attempting to use strategic planning to roughly estimate countrywide requirements. And of course his points will be true to the extent that further data is available and viewed as reliable. We are not aware of such data (our second major point above), and have to recognize that even if it does exist, it would be changing. Think simply of the issue of local political and military leadership around North Korea: How will it react to an intervention? While many might be prepared to fight an intervention today, how would they feel if they were already under attack from a neighboring North Korean military faction trying to take away their food supplies? How would they feel if, over a period of several years, appropriate information operations are taken to help them understand how much better their life could be as part of the ROK?
Finally, the commentators do not suggest that our article overestimates the potential consequences of a collapse. These consequences would also be important inputs into any decision to intervene. If the consequences tend to be greater than what we postulate, there will be significant incentives to pursue intervention once these consequences are recognized. As our we stated when we outlined the purposes of our article,, this is an important part of the public discussion because policy debates often fail to adequately anticipate the implications of restraint that prevents intervention.
Response by Jennifer Lind, Department of Government, Dartmouth College
Dr. Bennett and I are pleased that our International Security article, “The Collapse of North Korea: Military Missions and Requirements” has generated debate on H-Diplo.
To review: our article notes that for a long time, analysts speculated about the collapse of North Korea, but there were few if any unclassified assessments of the challenges associated with stabilizing a collapsed, anarchic North Korea. Our article posits a relatively benign collapse scenario, outlines several missions that countries may need to perform if such a scenario arises, and calculates the force requirements for those missions using a variety of historical and theoretical metrics. We found that even in this optimistic scenario the force requirements are enormous – implying that multinational plans should be coordinated well before collapse.
Our critics raised three main objections: that we ignored the politics of the situation to focus only on military modeling; that our overarching assumption for the study was too optimistic about how North Korea might collapse; that the cases we relied upon to create our metrics were too different from that of North Korea (and thus do not provide useful analogs).
First, Dr. Howe argues that we ignored the politics of the situation, specifically arguing that the South Korean people strongly support reductions in their military and would not support maintaining a larger force structure for the purpose of a North Korean stabilization mission. The purpose of our analysis is to let the political leaders, military planners, and foreign policy analysts in all countries who may be affected by North Korean collapse know about the scale and breadth of the calamity that such a collapse might create; and to let them know about the large force size that may be required to prevent the worst disasters in such a scenario. We hope that by publishing this analysis and stimulating debate on the topic, political and military leaders in South Korea, United States, China (and others) will think hard about what military forces they need to maintain for such scenarios, and what multinational planning they should be engaging in to facilitate crisis management. If—even when confronted by analyses such as this—the people of South Korea choose neither to build the forces to deal with this problem nor to develop contingency plans with their neighbors, that is their right—but they may one day suffer serious consequences from these choices. In sum, these kinds of analyses are important inputs into the political process. We of course do not expect that an isolated article in an American academic journal will turn South Korean politics on a dime, but if this analysis rings true with the leaders of Combined Forces Command, they should commission their own studies of the problem, which can be discussed with the policymakers and people of their countries as well as those in neighboring countries.
Dr. Nichols in his critique says that our analysis is overly optimistic. This is a peculiar criticism: as we describe on pp. 88-89, we explicitly chose a relatively benign collapse scenario because even those optimistic assumptions led us to dire conclusions about force requirements and the need for peacetime multinational planning to mitigate the worst outcomes. Analysts frequently bound complex problems by using optimistic assumptions to scope out the lower bound of requirements, or by using pessimistic assumptions to estimate the upper bounds. Nichols’ critique is therefore puzzling. We say that we adopted a relatively rosy collapse scenario; Nichols agrees. We say that, therefore, the vast force and planning requirements we derive probably understate the real scope of the problem; again he agrees. Our bottom line from our analysis is that those countries that will be most affected by a North Korean collapse should be planning for contingencies at least as bad as the one we described. We do not think Nichols disagrees with this conclusion either, so it is not clear what his objection on this point actually means.
But there is a broader and more important point to make about the use of assumptions and the way we constructed this analysis. We intentionally set up the analysis to provide an architecture for a range of follow-on studies. In future work, some analysts could alter the assumptions about the process of collapse and re-run the analysis to explore how different collapse scenarios affect intervention timelines, force levels, and the needed scope of peacetime planning. Other analysts could highlight additional missions that countries in the region might try to perform during a stability operation – or highlight missions that we examined but perhaps underestimated. (For example, Howe correctly notes that our analysis of refugee control does not take into account that boat people may flee North Korea to the South.) And military planners could apply alternative metrics to assess the force requirements for the range of missions that might be required.
Our bottom line is this: there is no way for anyone to predict how a complicated disaster like a North Korean collapse might unfold, or forecast with precision the problems that collapse would create for the international community and the North Korean people. Our article provides methods for analysts and planners to conduct studies across the plausible ‘scenario space’ to explore not just the number and type of forces that might be required – but even more importantly, the topics that the United States may wish to be discussing with its regional allies (and with China) about military operations that might need to be conducted soon after collapse.
Regarding the second claim: Nichols’ criticisms of our historical comparisons are not compelling. He is skeptical about the use of metrics from previous wars to explore the requirements for future operations, but his criticisms of specific claims in our paper are vague, and it’s impossible to tell why precisely he objects, or what metrics he thinks military planners should be using instead. For example, he notes that a comparison between Iraq after the fall of Saddam, and North Korea after a collapse, are suspect because “North Korea is not comparable to Iraq as a case.” But our only use of the Iraqi case was to warn that the chaos in the days after Saddam’s toppling is “a vivid reminder of the dangers associated with collapse of a government and the need for detailed planning before it occurs.” (p. 89) Does Nichols disagree with that statement?
We did not use the Iraq stabilization case—as Nichols implies—as a single case study from which to derive force requirements for North Korea. As we discussed clearly in the paper (pp. 92-93), the metric we used to estimate the requirements of a stability operation came from two seminal analyses by James Quinlivan (RAND) and by John McGrath (Combat Studies Institute, U.S. Army Combined Arms Center).[1] These works rely on a large number of historical cases and capture the military analysis community’s best assessment of the lessons of previous stability operations.
Nichols also objects to our use of the East German case as a source of insight into the demands of disarming the North Korean military. He introduces the acronym “DDR” (for Deutsche Demokratische Republik) to critique our discussion of “DDR” (Disarmament, Disarmament, and Reintegration”)—this will confuse all but the most stalwart of readers, and suggests that he himself was confused. (We offer two full pages of discussion about the challenges of “DDR” on pp. 104-106.) In this section we explain the metric we relied upon to estimate the manpower requirements for a North Korean DDR operation—and that metric (contrary to Nichols’ criticism) was not from the East German case.
Our article does mention the case of East Germany as an example of the vast manpower requirements for disarming massive, antiquated arsenals. It’s certainly debatable how similar a North Korean DDR operation would be to the East German one (although East Germany provides a far more useful case than the vast majority of DDR operations that have been performed worldwide, which typically were much smaller in scale, and involved more lightly armed soldiers). In the end, though, if Nichols believes that the East German case is probably easier, and that the North Korean case would have greater requirements, that’s consistent with our analysis which was designed to produce lower-bound estimates.
A more compelling critique of our own DDR analysis is that it relied on a metric from the single case of Cambodia. We wish we had a broader set of cases from which to draw (we discuss our selection of this metric on page 107). However, we were unable to find detailed manpower data on other major DDR operations, even the one conducted in East Germany. Hopefully future scholars can find detailed data on the East German and other DDR cases, and use these to improve upon the metric we used here.
To conclude, we were puzzled by Nichols’s lament that this type of analysis is why policy makers ignore academic work; he implies that we are ivory-tower academics who know nothing about the policy process. What is particularly strange about this aside is that one of us (Bennett) has worked for 35 years at RAND, and in that capacity has conducted scores of studies for the Defense Department on U.S. military planning around the world. For the past two decades he has focused on contingencies on the Korean peninsula, and regularly interacts with Defense Department officials and the U.S. and ROK military officers who would carry out such operations.
The other author (Lind) worked as a consultant at RAND on topics related to this paper, and has presented versions of this analysis to a broad range of audiences in Washington and Seoul, which included officials at the U.S. Defense Department, U.S. State Department, ROK Ministry of National Defense, and Korea National Defense University. Lind gave a briefing on this topic to the U.S. Special Operations Command, Korea and discussed the analysis at length with U.S. military personnel responsible for planning for contingencies on the Korean Peninsula. The audiences in Washington and in Seoul offered suggestions to improve our analysis – many of which were incorporated into the International Security version of the article. None shared Nichols’ view that the analysis, or our analytic approach, is disconnected from the real-world problems that they work on each day. In sum, the problem of academic drift away from policy relevance is a very peculiar criticism to level against these authors and this analysis.
© Copyright 2012-2015 The Authors.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License.
Notes
[1] John McGrath, Boots on the Ground: Troop Density in Contingency Operations, Global War on Terrorism Occasional Paper, No. 16 (Fort Leavenworth, Kans.: Combat Studies Institute Press, 2006); James T. Quinlivan, “Force Requirements in Stability Operations,” Parameters, Vol. 25, No. 4 (Winter 1995-96), pp. 59-69.