A Game-theoretic Solution to the North Korea Problem

Marcus Noland (PIIE)



A couple of colleagues forwarded to my attention an interesting piece by University of Chicago professor Harald Uhlig. In the Chicago School tradition of making simple assumptions and rigorously following them to bizarre conclusions, he uses game theory to analyze the North Korean nuclear issue and concludes that the optimal policy consists of immediate South Korean surrender, withdrawal of US troops from the peninsula, and the conclusion of a Unified Korea-US Free Trade Agreement. Sufficed to say, with such prescriptions I doubt that Professor Uhlig will be joining the Trump NSC staff, though perhaps he could make the People's United Party candidate’s list in the next South Korean national assembly election. I think that he is serious, though I am not absolutely sure: this could be some kind of Teutonic humor that doesn’t translate well. But I’ll assume he’s serious: if he’s kidding, well, the joke’s on me.

So how does Uhlig reach such an eye-opening conclusion? First, he posits an objective function or goal of the North Korean elite: “A nice life. Not being bullied around. More power.” Then he asks how the North Korean elite can obtain these things. Well, fast-forward the development of nuclear weapons and associated missile delivery systems to shake down your neighbor and deter its ally. “Where is the money near North Korea?” Uhlig asks, “Tough question? Really?!  C’mon.” Answering his own question, “It is in South Korea, of course,” (In point of fact, Japan is a lot richer than South Korea, and has pacifist tendencies to boot, but I think that the same logic applies to both, so let’s just ignore this nitpick.)

So how to get on with the rape of South Korea? Well, one way would be to threaten them with nukes, perhaps taking out a South Korean city (along with the local US military base). But this is really unnecessary (and perhaps risky—one really cannot be certain how the US will respond to a nuclear attack on one of its forward bases). If the threat is credible, it should induce the South Koreans to surrender without a shot being fired, though a few shots may have to be fired to demonstrate that this isn’t a bluff.

But couldn’t the US pre-empt, before the North Korean develops the capacity to hit the US?  (The logic of pre-emption is laid out in Joseph DeThomas’ recent essay at 38 North.) Hmmm…on this one Uhlig obfuscates, “The generals have already told Trump that they cannot guarantee to wipe out the entire weapon arsenal of North Korea! The South Koreans already beg for Trump not to do anything! The U.S. cannot possibly throw an atomic bomb first! A pre-emptive strike will be really, really messy, millions will die! Military options must be off the table! The U.S. must seek peaceful solutions!” Really? I’ll come back to this claim in a moment.

But for the sake of argument, let’s suppose the US doesn’t pre-emptively destroy the North Korean WMD capability. So what’s the deal? Uhlig argues that “Option 1 is for South Korea to surrender to North Korea, for the US to withdraw its military bases and to make the unified Korea a preferred trading partner, in exchange for stopping their development of intercontinental missiles, and to arrange this all quickly: 2017 would be good. If the upper 10,000 in North Korea can subsequently have a peaceful life in luxury and with power, and if the US and the unified Korea both benefit from the trade agreement, there will then be little reason for Unified Korea and its rulers to fear the U.S. and vice versa, guaranteeing the not-being-bullied part.” This outcome, he argues, would be superior to a US pre-emptive strike, “Most South Koreans wouldn’t see much changes in their daily lives: after all, North Korea would want a rich South Korea, not destroy it. They can no longer vote for their leader for the next 30 years or so: oh well. It wouldn’t be so bad.”

If the weapons are basically defensive, meant to secure deterrence—then neither pre-emptive bombing nor surrender are the correct policy.

So, why do I find this piece unpersuasive? It can be criticized from “inside” the model, as well as from without. The basic issue from inside the model is that if one assumes the goal of US policymakers mirror those of their North Korean counterparts, it is not at all clear that withdrawal and surrender dominate a pre-emptive strike: the millions dead would be Koreans, and I don’t recall them entering into our utility function. In fact, if one pursues Uhlig’s backward logic, it is not clear that any South Koreans would die. The North Korean goal is survival; when the US missiles begin hitting the choice is to retaliate against South Korea and face obliteration or stand down and maintain a positive possibility of survival. North Korean surrender wins. Pre-emptive bombing doesn’t seem like such a bad play.

But the deeper criticism is the simple supposition of elite goals; the assumption of complete certainty, and the counterpart dismissal of risk; and lastly that the US is playing a simple Korean peninsula, and not a more extended, global game. All are relevant.

The first issue goes to what the North Koreans want. Over the decades there have been three broad sets of motivations attributed to them. The first, typically put forward by left-leaning commentators, is that the nukes are a “bargaining chip”—the North Koreans simply built the weapons to be bought out. As North Korean weapons development has progressed and the regime has displayed little interest in bargaining, support for this perspective appears to have waned. I suppose that from Uhlig’s perspective, getting bought out would address the obtaining a good life goal, but not address the issues of bullying and power.

The opposite perspective, typically associated with right-leaning observers is closer to Uhlig’s—the weapons are meant for conquest—but the conclusion that these analysts normally reach pushes toward sanctions and pre-emption, not surrender.

Reflecting on the experience of 2002, when the North Koreans floated the idea of a significant conventional troops demobilization, while at the same time pushing forward with missile and nuclear weapons development, reaching out to Japan from which they could expect post-colonial claims, and initiating economic reforms, I proposed a third interpretation: that the development of WMDs and their delivery systems possibly signaled the beginnings of a strategic reorientation. Recognizing that the correlation of forces was such that they were not going to unify the peninsula on their own terms for a long time if ever, the North Koreans were battening down for an extended period of peaceful coexistence. The WMDs were meant to sustain double-sided deterrence (i.e. escape from being bullied in Uhlig-speak), which achieved, would permit them to pursue better lives: releasing men from the military, engaging in reform to provide them employment, and getting money from the Japanese to keep goods on the shelves during the transition. From this perspective, the nuclear weapons were fundamentally defensive in nature, protecting the North Korean elite from bullying, helping them achieve materially more satisfying lives, while maintaining power.

If I was correct in my assessment 15 years ago, that the weapons are basically defensive, meant to secure deterrence—then neither pre-emptive bombing nor surrender are the correct policy.  The notion of asymmetric escalation, complicates but does not change this logic. The North Koreans fear invasion. The thought is that if invaded and losing a conventional war, the North Koreans could use tactical nukes defensively if they thought that the possession of nuclear armed ICBMs would deter the US from responding in kind. Such capability could facilitate the demobilization some of their gigantic military—more than 1 million troops under arms, the fourth largest army in the world, supported by a population of 25 million—while at the same time enhancing their confidence that they could deter an invasion: a more efficient, less costly route to deterrence.

The less benign interpretation is that the North Koreans would be the ones to launch the invasion; if they began to lose they would use tactical nukes, sue for peace, and threaten to strike the US, if the US and its South Korean allies did not stand down. Again, immediate capitulation is not the correct policy.

Finally, there is the issue that while North Korea may be a peninsular power, the US is a global power. Uhlig contemplates a scenario where “just to make sure they know you mean business, [North Korea drops] an atomic bomb on, hmmh, say Daegu, killing nearly 3 million people and wiping out the U.S. military base there, in one go.” But this logic could just as well be applied to the US: from the standpoint of a global superpower, using nuclear weapons against North Korea might underscore credibility. Indeed, bombing North Korea might be a relatively low-cost way to achieve this end, since no one really likes the North Koreans. And in President Donald “totally destroy” Trump, we may have just found the leader who finds this logic compelling.

More From

Related Topics