“We are about to run an experiment,” the international-relations scholar Robert Jervis recently observed of the Trump presidency. Scholars of international politics, he wrote, “bemoan the fact that our sub-field cannot draw on the experimental method.” But with an American president whose stated views on international relations differ so dramatically from those of his recent predecessors, even while many features of the international environment Jervis has studied for decades remain constant, “now we can.”

An important piece of the experiment has been ongoing for weeks as the Trump administration confronts the gathering threat of North Korea’s nuclear and missile programs. Kim Jong Un’s regime is now estimated to have as many as 20 nuclear warheads and could soon be able, if it isn’t already, to make some to fit on the missiles necessary to deliver them. The North has been accelerating missile tests—recently lobbing four in the direction of U.S. bases in Japan—in what the arms-control expert Jeffrey Lewis has called a practice run for a nuclear attack. In a New Year’s speech, Kim declared his country was in the final stages of developing an intercontinental ballistic missile (ICBM) capable of reaching the United States, a feat that Trump declared, via tweet, “won’t happen,” shortly before the U.S. accelerated the deployment of a missile-defense system to South Korea and kicked off annual joint military exercises with the South. On a visit to Seoul on Friday, Secretary of State Rex Tillerson left open the option of a military strike to prevent the weapons-development program from advancing too far, vowed to defend allies in the region, and ruled out negotiations with the North.

If these activities all test the theory that threats of violence can deter an adversary from engaging in violence of its own, they also test a concept Jervis articulated several decades ago, in the context of the Cold War nuclear standoff. His central insight in the 1978 paper “Cooperation Under the Security Dilemma,” was that “many of the means by which a state tries to increase its security decrease the security of others.” It’s not hard to find examples of this in East Asia. U.S.-South Korea joint military exercises are ostensibly designed to increase those countries’ security, but North Korea sees them as decreasing its own. American missile defenses on the Korean peninsula, meant to blunt the threat of the North’s missiles, have radars that look to China like they could help America spy on China’s own weapons, threatening Beijing’s ability to defend itself. North Korea itself maintains its missile development is to defend against the kinds of attacks the U.S. is now hinting at.

I recently spoke to Jervis about the security dilemma in East Asia, and the grand international-relations experiment of the Trump presidency. A condensed and edited transcript of our conversation follows.


Kathy Gilsinan: Walk me through how the security dilemma is working in the context of the U.S. and North Korea.

Robert Jervis: Let’s start with North Korea’s explanation for its missile program. It argues really that it is defensive and it doesn’t want to intimidate anyone else, except as that’s necessary for its defense. And after all, the U.S. had threatened to drop nuclear bombs on [North Korea] during the Korean War, and has implicitly or explicitly made nuclear threats in periods of tension. Secretary Tillerson has just threatened a pre-emptive strike, and while we see this only as a last-ditch defensive measure, I’m sure North Korea sees it quite differently.

North Korea argues and may believe that the reason it needs nuclear weapons and the missiles is to make sure that the U.S. doesn’t attack or, more likely, overthrow the regime. [The North Koreans] did explicitly point out after the Iraq invasion that the U.S. does overthrow regimes it doesn’t like, as long as they don’t have nuclear weapons. So there is a legitimate defensive rationale. Most people who know North Korea better than I do say that this isn’t the entire story, that North Korea [also] wants to exert greater influence in the region, to intimidate South Korea. Crazy as it seems, all the experts say that the whole Kim dynasty believes Korea will be unified—under North Korean auspices, a unified North Korean regime. As usual there are a mixture of motives and drives, offensive as well as defensive. But I think it’s easy for us to forget that there is a defensive aspect to this.

The next layer is that sometimes U.S. policy has sought to ameliorate any legitimate fears North Korea would have, and even the Bush administration finally decided, after a lot of internal debate, that we would reassure North Korea and make a promise that if negotiations were successful we would not seek to attack the country. Victor Cha [adviser on North Korea to President George W. Bush] tells the story in an article that he was tasked to tell the North Koreans this, and we really thought it would have some impact on them. These were verbal promises, but they represented a real policy change. The Russians went up to the North Koreans to say, essentially, we wanted this kind of guarantee during the Cold War and the Americans didn’t do it. Why this didn’t work not entirely clear; maybe the North wasn’t really worried, or it is the credibility problem. [Cha writes: “It seemed to me at the time that the DPRK finally received the security guarantee and the end to ‘hostile’ U.S. policy that they had long sought. Yet, after holding this out as a precondition for progress, in subsequent rounds of negotiations they proceeded to brush this off as a meaningless commitment, a piece of paper that guaranteed nothing for North Korean security.”]

On our side with the missile defense, it is in a sense obviously defensive—the THAAD [terminal high-altitude area defense] missiles [which are designed to shoot down other missiles] can’t be rejiggered to attack ground targets. But of course to the extent that North Korea and even China sees [North Korean missiles] as protecting the North, anything that takes away their protection is seen as offensive, and makes them less secure. China says that the radars [in the THAAD system] reach deep into China and could be a threat to China’s second-strike capabilities against the U.S. Is this plausible? [Experts say] this is not entirely made up—it isn’t the [THAAD] radar alone, but this radar when hooked into or coordinated with all the other American radars could have, not an enormous, but non-trivial increase in our ability to degrade if not entirely thwart Chinese retaliation against the U.S., which they see as their ultimate deterrent.

The basic thing about the security dilemma [is that] when it comes to their security, even normal states are a little paranoid. They don’t do a full worst-case analysis, but they do bad-case analysis. [If you’re China] maybe the radars are more [powerful] than we let on, or if they let the first generation in, the second generation may be much more powerful.

The U.S. invasion of Iraq is an example of this kind of thinking. Even had the intelligence estimate [on Iraqi weapons of mass destruction programs] been right, nuclear weapons were at least eight years off, and then what we were worried about was [Saddam Hussein] giving them to terrorists. I think looking from the outside a lot of countries said, how do they believe that? That really is clinically paranoid. To Americans, even those who opposed the war, it didn’t seem paranoid, although it may have seemed wrong. I don’t think we were capable of understanding how others saw us as really being paranoid, or as using security as a pretense to cover other motives like the desire for Iraqi oil. And I don’t think many outside the U.S. in retrospect can believe the Bush people really were frightened; I think they were.

Gilsinan: You cite the example of the British Navy in the 19th and early 20th centuries, the ostensible purpose of its strength being to keep shipping lanes open, but it couldn’t, as you write “avoid being a menace to any other state with a coast that could be raided.” I wonder if you see that kind of dynamic playing out in East Asia—the U.S.-South Korea military exercises near North Korean coasts that could be raided; in the South China Sea the American priority of “keeping shipping lanes open” necessitating the presence of the U.S. Navy near territory claimed by China.

Jervis: I certainly do. I think the North Koreans do have to worry that one of these exercises isn’t going to stop wherever the stop line is, but will continue and roll over the [38th] parallel [dividing North and South Korea]. Probably for [the North Koreans] the idea that they would launch a conventional attack or land invasion of South Korea must look crazy, so what is the legitimate defensive purpose of these [exercises]? Whereas obviously we see [them] just as keeping our powder dry in the classic sense [and reassuring] South Korea. That said, there is a glimmer of hope. The Chinese proposed that we halt the exercises and the North halt nuclear tests and missile tests. And there may be deals like that possible, these exercises could be scaled back in a way that would seem to take North Korean [fears into account]. ... But that’s one ray of hope in [a not otherwise hopeful] situation.  Tillerson has just rejected this, but when he looks more carefully at the other options may find this is something he wants to come back to.

Of course that would require this administration to think there are legitimate North Korean interests in security and [concerns] we need to take seriously. Most people can’t get there, especially with a regime so unpleasant and provocative as North Korea’s.

Gilsinan: Your analysis generally deals with the setup of the international system, and how that affects states’ perceptions of their security. But what’s the role of individual leaders in these dynamics? How, for example, does turmoil in the South Korean leadership, and unpredictability in the U.S. leadership, change the analysis?

Jervis: I think there are three levels of problems. The first is intellectual, and also emotional—to really empathize with the other side and also understand how people can see you in a different way than you see yourself. That’s hard intellectually and it’s hard emotionally. You think people see you as cowardly; or you think people see you as too threatening. I was at a discussion at the Council on Foreign Relations recently where Mary Beth Long, who was in the Obama [Defense Department], was saying, well, our basic problem [with North Korea] is the credibility of our threats is zero—

Gilsinan: She was talking about the Trump administration specifically?

Jervis: No, she was saying just generally. We’ve never done anything toward North Korea, so the credibility of our threats is zero. I think that’s very exaggerated. After all, North Korea has staged few provocations recently, and it may well be that our promises not to attack if the North behaves itself are less credible than many of our threats. But she is right that it would be hard to make credible the threat to destroy missiles that are being tested, because this policy would be very dangerous.

The second hurdle is the personalities. Now it may be better with [H.R.] McMaster having replaced [Michael] Flynn [as national security adviser]; some people find it easier to empathize than others, and I don’t think Trump is [one who] grasps this. Reagan eventually did see how the Soviet Union could see the United States as threatening, and this helped end the Cold War.

The third level is domestic politics. Usually being tough is better domestic politics than not, and to explain to Congress and the American people that the U.S. has to take into account legitimate North Korean interests, when this is the most odious regime in the world—that’s not great domestic politics.

And of course it’s not clear that a deal is possible, but the overcoming of those hurdles are a necessary condition even though they’re not a sufficient one.

Gilsinan: You also write about how vulnerability, or perceived vulnerability, informed the different foreign policies of Britain and Austria following the Napoleanic wars—basically that Britain, not being surrounded, could afford not to be too concerned with other European states’ conflicts in a way Austria, being close to them, could not. I wonder if you could look at America almost shifting from a British-like position of semi-invulnerability, to an Austria-like one of more vulnerability, if and when North Korean missiles can reach the U.S. mainland.

Jervis: Yes, and that does partly recapitulate the Cold War—until the mid-1960s the United States was largely invulnerable, but once the Soviets developed [intercontinental ballistic missiles]—nuclear weapons against your homeland remove any geographic buffer you used to have. That said, would a North Korean ICBM really be a “game changer”? We’ve lived with Soviet nuclear weapons, we’ve lived with Chinese nuclear weapons; no one wants the North Koreans to get ICBMs, but it isn’t the end of the world. And there’s a big split in the U.S. government about how big a change would it be, because those who think “oh my god if they get ICBMs it is a game-changer” are more inclined to shoot the missile down on the launch pad or in the air, both of which are much more difficult than you would gather by reading press accounts.

Gilsinan: So the policy the U.S. pursues would come down to which view is better-represented within the government.

Jervis: It would be hard even for a thoughtful decisionmaker to say, look, they have a missile that can reach the West coast; this is not pleasant but we shouldn’t exaggerate the danger of the situation. I think intellectually if they get an ICBM that’s bad. There are various things we can do to slow it down. Arguing that this is not trivial but not crucial—it’s a hard sell even to reasonable decisionmakers who are responsible to the fate of their country.

Gilsinan: So in this context, is there any advantage to Donald Trump’s much-vaunted unpredictability? If the North Koreans think he’s crazy and just might do anything, is that any kind of deterrent?

Jervis: Well this goes back to Tom Schelling, and the basic “rationality of irrationality,” which means that you can strategically appear irrational in order to force concessions from a rational adversary who fears your willingness to take risks. We don’t have really good research on that. [Richard] Nixon seems to have believed it, but he did not really get advantage from it, and as far as we can tell he was not seen as wildly erratic. Of course with Trump, who knows? I’d love to read the analyses being written by foreign governments.

Gilsinan: So Nixon’s madman theory didn’t work because it didn’t look like he was crazy enough?

Jervis: He would talk about doing crazy things, and he would instruct [his national-security adviser and secretary of state Henry] Kissinger to tell [the Soviet ambassador Anatoly] Dobrynin about how crazy he was. His bark was generally worse than his bite. He backed away from the big escalation [in Vietnam] in the fall of 69. I’m glad he did. He bombed Cambodia but that wasn’t a big deal [in terms of escalation with the Soviets]. We went on nuclear alert in the fall of ‘69 but the Soviets never noticed [until] the last few days, and then they attributed it to our reaction to the Sino-Soviet border conflict. But Nixon acted crazier at home than abroad, and it didn’t seem to have the desired impact. I’d be surprised if North Korea restrains itself from further testing on the belief that Trump is just too unpredictable. Not impossible, maybe they empathize with strangeness.

Gilsinan: You think Trump’s bark might be worse than his bite? He’s already backed off some of the more provocative moves he’s made on Asia, reaffirming the One China policy after the Taiwan call, for instance. He’s been all over the place on North Korea though—there was the tweet about how North Korean missiles “won’t happen,” he’s said nuclear weapons are the biggest danger but vowed to expand the nuclear arsenal.

Jervis: So far, trite statement, work in progress. North Korea may have gone the other direction; [during the campaign he said he’d] sit down over hamburgers with Kim Jong Un. And we haven’t heard much of that, the principals committee has been meeting; I’ve gathered from people with contacts in Washington that they’re spinning their wheels, which is not surprising. And of course the people at the top have no expertise, and I doubt that they know the full history. Policymaking in this area is particularly hard, and the Obama administration hasn’t left them anything particularly solid to build on. It’s hard, and compounded by at most lack of experience and worst bad instincts.

In his recent remarks in South Korea Tillerson rejected the Obama policy of “strategic patience,” which makes some sense because this wasn’t much of a policy, but it isn’t clear that he has more than bluster to put in its place. We might not succeed if we tried to shoot down a North Korean missile, and trying would require convincing our South Korean and Japanese allies that this is appropriate. Military escalation would be a real possibility. Are we ready for war? He’d better talk to Secretary of Defense [James] Mattis. Tillerson says he doesn’t believe “conditions are right for talks,” but it’s hard to see any resolution without talks, and having flatly rejected the Chinese proposal I think you’d have to be very optimistic to believe that China is going to put much more pressure on the North unless we show a willingness to meet them, if not halfway, then at least partway.

Gilsinan: Any final words of restraint or comfort from you as a Cold War scholar?

Jervis: There were a number of times the Cold War looked desperate and the worst never occurred. Problems looked insoluble and by god they worked themselves out. There were big wars in Korea and Vietnam and many people killed, but the worst never happened. Deterrence, as a basic feeling, is a powerful inhibitor.