by Michael Shellenberger
Recent decades have seen the growth of a large body of scholarly literature about why nations seek nuclear weapons, but far less on how they do it.
In 2016, Vipin Narang, an Associate Professor of Political Science at the Massachusetts Institute of Technology, wrote a landmark article, "Strategies of Nuclear Proliferation," for the journal International Security. In it Narang evaluated four strategies that nations pursue for nuclear proliferation:
- hedging (“develop the option for a weapon”);
- sprinting (“weaponize as quickly as possible”);
- hiding (weaponize without being discovered);
- sheltered pursuit (weaponize before patron abandons client”).
And Narang lists three kinds of hedging: “technical hedging, insurance hedging, and hard hedging.” They are listed in order of furthest to closest to pursuing a weapon
- Technical hedging involves having peaceful nuclear power but no weaponization work and only fringe elements discussing a weapon.
- Insurance hedging includes technical potential to create weapons-grade materials, limited weaponization work, and periodic discussion of a weapon.
- And hard hedging is having the capability of creating weapons grade materials, work on weaponization, and mainstream discussion of a weapon.
I reached out to Narang to ask how he would categorize nuclear energy-seeking nations in his typology. He said that he would categorize U.A.E. and Saudi Arabia as engaged in insurance hedging, and Egypt and Bangladesh in technical hedging.
“My reading of Turkey,” Narang told me, “is that it may already be anticipating a rupture with the US or NATO and is therefore laying the groundwork to replace the nuclear umbrella the US provides.”
- North Korea provides "the blueprint" for nations seeking to acquire a weapon;
- Sweden and Switzerland sought nuclear power in large measure to have the option of acquiring a weapon;
- Saudi Arabia may choose Russia over the U.S. or South Korea to build its nuclear plant to avoid obstacles to creating a weapon;
- Japan is already so weapons capable that it would likely need just three to six months to create a weapon;
- Nuclear weapons prevented full-scale war between India and Pakistan but also allowed lower-level conflicts;
- It is in U.S. interests to help nuclear weapons nation, including adversaries, to improve command and control of their weapons;
- In peacetime, there is very little risk of accidental use;
- The use of a nuclear weapon by any nation would more likely result in an immediate cessation of hostilities than an escalation;
- Proposals for the U.S. and other nations to put in place "no first use" policies and reduce readiness are unlikely to be adopted for strategic military reasons.
As an up-and-coming star in the field of nuclear peace and security studies, Narang's views are worth considering.
Michael Shellenberger: Your article changed our view of how nuclear energy diffuses into international system. It now appears to us that national security, having a weapons option, is often the most important factor in a state pursuing peaceful nuclear energy.
Vipin Narang: There are countries genuinely interested in power, and then it is possible that sometimes the temptation arises to think about military applications of nuclear energy — or, alternatively, they always had the temptation and pursued energy in order to have a future weapons option. [Former Indian Prime Minister] Nehru, for example, had no illusions and knew what the reactor was good for. The weapon option was always in the back of their minds. They took advantage of [nuclear technology sharing program] “Atoms for Peace” and others took advantage of the NPT bargain later on.
Western Europe is a classic example. Take Switzerland and Sweden. Both are neutral nations outside of NATO that had a very deep interest in weapons and a program through the 1960s. Today they are championed as nonproliferation nations, but both militaries were very interested in having the basis for a nuclear weapons program if necessary. Both used nuclear energy to explore those options.
[Texas A&M Professor Matthew] Fuhrmann’s argument is that energy tempted them into weapons. But your view that weapons drove nations to energy, not the other way around, may be more accurate given what we now know about many of these countries. The correlation may in fact go the other way. These countries used energy to generate weapons options, and nations like Switzerland and Sweden gave up the option but kept nuclear energy.
And we forget that Brazil and Argentina had enrichment programs. In Brazil it was run by the Navy. They said, “We need it for our submarines,” but that’s just hop skip and jump away from having weapons grade uranium enrichment.
MS: Has the discipline of nuclear weapons and security studies changed significantly recently?
VN: Up until 2016, Obama’s Prague speech calling for the elimination of nuclear weapons was the defining moment of 2010s. China was modernizing its arsenal but not increasing it. The rest of world thought the U.S. commitments were strong, and both superpowers were reducing their deployed arsenal. But we never actually substantially reduced our upload capability. It would shock people who think we reduced our arsenal size to learn how much we still have. We put a lot in storage not decommissioning.
India and Pakistan were having arms race, but few people outside of us who study South Asia were paying attention. But then Trump mentions tippy top weapons, North Korea is off to the races, and we‘re building missile defense systems that scare Russia and China. Missile defenses are portrayed as largely defensive systems, but they can support highly offensive strategies against states with small nuclear arsenals, since it allows us to potentially neutralize a handful of incoming warheads. We may say they are for North Korea, but if we can make the Chinese nuclear force look like North Korea’s by attacking their nuclear systems on the first, missile defenses—if they work, which they don’t yet—can theoretically intercept the “residual” force.
So missile defenses, more than our nuclear systems themselves, terrify the Russians and Chinese if we could ever get them to work. So now they think they need to build up to defeat missile defenses. Putin is describing nuclear-powered cruise missiles to evade them.
Add to that the fear that the U.S. won’t live up to its extended deterrence commitments, and some of our allies start considering acquiring their own nuclear weapons for deterrence. So everyone is fending for themselves.
MS: Is the system of extended deterrence falling apart?
American foreign policy goes in cycles and we are in retreat mode. When America goes into retreat, the question our allies ask is “who will defend us.” The answer sometimes is “ourselves.” They go into self-help mode, and weapons are ultimate insurance against invasion.
The debate in Germany about whether to get a weapon is reigniting. I would be surprised if they did it. But after World War II, Chancellor Adenauer was serious about a French, Italian, German shared nuclear weapons program. This idea is back and it’s not totally fringe any more.
Sometimes our allies will publicly flirt with it in the hopes that the US will provide stronger reassurances to them. But the flirtation also prepares for a world in which they actually face US abandonment and may have to develop their own nuclear deterrents. Japan has been playing that game for decades, and now it seems some elements in Germany are considering it as well.
MS: What’s happening in Turkey?
VN: Because Turkey is a NATO ally, it enjoys a formal extended deterrence guarantee from the United States. In addition, we store about 50 B-61s [nuclear bombs] deployed at Incirlik in secure vaults, albeit under US control. Any flirtation with a nuclear weapons program should—according to my theory—be aimed at eliciting greater assurance from the US or preparing for abandonment.
But Turkey may be sui generis because our relationship is so fraught and now they are playing footsie with Russia. When the coup attempt happened last year people worried about the 50 B-61s stored at Incirlik because there were reports that the base commander was a Gulenist. It would be difficult for anyone to access the US nuclear weapons stored there, but it is important to remember that the safety features on these weapons are only designed to “delay” unauthorized use, not “prevent” it. I don’t see any good reason for us to continue to maintain tactical nuclear weapons on Turkish soil anymore. They are increasingly a liability, not an asset.
And my guess is that Turkey may be preparing for the day that withdrawal occurs. My reading of Turkey is that it may already be anticipating a rupture with the US or NATO and is therefore laying the groundwork to replace the nuclear umbrella the US provides.
It would be a classic case of [weapons] insurance hedging. But it’s not clear what threat there is to Turkey. It’s not from Russia, but maybe they are pricing in a future nuclear Iran, if my theory is right. But wouldn’t surprise me if Turkey abandons the US or vice versa, and its nuclear efforts put them closer to the Saudi and UAE model.
MS: How would Turkey do it?
VN: North Korean President Kim Jong-un has shown the blueprint. If you want a meeting with the president of the U.S., and insurance against an invasion, then get a nuclear weapon. Do it secretly. Make it ambiguous.
North Korea has provided the blueprint for a lot of these countries. Build a reactor, pull out of NPT, kick out the inspectors. Last year was the critical moment for North Korea’s weaponization as they tested long range missiles and higher yield nuclear weapons designs. Once they did so, they declared that they achieved completion of their deterrent force. And countries know that the US is very cautious militarily against states with nuclear weapons. An Indian general in fact remarked that the most important lesson of the 1991 Gulf War was that no one should ever fight the United States without a nuclear weapons capability.
The most fruitful pathway for Turkey is uranium enrichment, not a reactor, which would have to be under international safeguards. Uranium enrichment is probably easier to hide, though perhaps a less efficient pathway. You can try with civilian reactors. Taiwan had a hole in its spent fuel pool. Syria tried. But it’s much harder.
And I would be surprised if the Israelis allow Turkey to have an operational reactor. Israel wouldn’t attack NATO country. But if Turkey left NATO all bets would be off. Israeli policy has essentially been no reactors in the region—it bombed Iraq’s and Syria’s. So if history is any guide, Israel won’t let anyone in the region get an operational reactor.
MS: Most nuclear energy industry experts say it would be a no-brainer for Saudi Arabia to hire South Korea to build its nuclear plant. But Saudi Arabia is taking a long time to pick between South Korea, the US, Russia, and China. What’s going on?
VN: It may be that they’re worried about proliferation-resistance in South Korean nuclear plant design. Russian reactors may not have that same resistance. They just don’t care as much. The U.S. Westinghouse design is very difficult to use to create a weapon if under safeguards. If they go with the Russians, that might indicate their intent. If they breakout of nonproliferation treaty (NPT), nothing stops them. But my guess is that Saudi Arabia would prefer a US extended deterrence guarantee to an indigenous nuclear weapons program. At this point, however, this is all speculative.
MS: How would Saudi Arabia do it?
VN: You have to get through hard phase of building a reactor, getting it hot, having a reprocessing facility, and then getting the safeguards out of the reprocessing side. North Korea pulled out of the NPT early — 1993. Israel, India, and Pakistan never signed.
Saudi Arabia has long been rumored to have a deal with Pakistan to — in an emergency — have Pakistani nuclear weapons be deployed on Saudi soil. The Saudis invested billions in Pakistan’s nuclear program. The rumors have persisted, though the Saudis don’t talk about it, and the Pakistanis deny it. But ultimately, the path of least resistance is if the Saudis could say to Trump, “we’ll stop [pursuing a weapon] if we get explicit guarantee from you.”
MS: So then is the civilian nuclear reactor a waste of time for nations pursuing a weapons option?
VN: I’d push on both doors. The North Koreans and the Pakistanis did this also. The advantage of reactors is an infrastructure if you do decide to pursue the option. In many ways, the plutonium pathway is more attractive and potentially more efficient from a weapons standpoint. And you don’t have to scale up enrichment and need massive uranium deposits. The centrifuge approach to enrichment isn’t particularly easy given how strict the export control lists are these days for components. The Iraqis tried so many different ways with enrichment in the 1980s and had little success.
Saudi Arabia might have been [Pakistani nuclear scientist] A.Q. Khan’s fourth customer beyond North Korea, Iran and Libya. It wasn’t Iraq. Nobody knows who it is in the open source. I’ve long suspected Saudi Arabia or Syria because they would be obvious candidates. If it’s the Saudis, they may have some uranium enrichment designs and capability already. But, in general, a determined state would probably try to push on both doors. Maybe there are reasons you’d prefer plutonium. I’d try both, because if you lose one you may still be able to hide the other.
MS: How else might those nations pursue a weapon?
VN: You can do what Pakistan did in the 1980s and develop the capability but stop just short of weaponization—though Pakistan in the 1980s had the advantage of a US administration that was willing to look the other way. They went right up to component separation and said, “we don’t have weapons” because they aren’t assembled. We have the pit [core of the weapon] and a non-fissile [nuclear] component. But it’s not assembled. And the US in fact turned a blind eye when Pakistan did that in the late 1980s. We said they didn’t have weapons because they aren’t assembled, and Reagan certified as much every year. But for all intents and purposes they could rapidly assemble them. They were much closer to a weapon than simply having the basis for fissile material production. And closer than Iran was in the early 2000s when they were just doing the theoretical and low-level experimental work on warhead, wiring, triggers, and other theoretical and engineering work to put yourself in a position to compress time to becoming weapons power. Pakistan was the last screw away. Iran was much farther, and there is a lot in between, such as Japan.
MS: You point to Japan as a model of a nation that is very close. Are they just six months away from a bomb?
The lore is three to six months, but no one really knows. An X-box has more sophistication than a crude nuclear weapon. Japan is totally different from any other hedger because they have control of the full fuel cycle.. No other NPT state has full control of their fuel cycle. At the same time, Japan even more than South Korea is terrified of becoming the next France, and have to provide its own security. Recall that French Prime Minister Charles De Gaulle said, once the USSR developed ICBMs that could hit the US homeland, “You won’t trade Pittsburgh for Paris. We’re not interested in sharing. We want our own nuclear weapons force that isn’t subjugated like the British nuclear weapons are to NATO.”
MS: Shouldn’t we be glad India and Pakistan have weapons since it has reduced wars?
VN: That’s probably fair at one level. Since both states acquired even covert nuclear weapons capabilities, they have had the longest period of peace across the international border. Nuclear weapons have, at one level, stabilized the region. They cured the previous disease, which was massive conventional war, at least for now. It didn’t solve all the problems. But just because medicine has a side effect, you don’t not give the medicine.
Pakistani provocations persist but that they haven’t gone to full-blown conventional war. Because of Pakistan’s nuclear weapons, India hasn’t retaliated for those provocations with military action across the international boundary. The downside is that Pakistan may now believe it has a shield behind which it can launch terrorist attacks against India, and that is a real problem. But for now, India has been restrained and stayed it retaliatory hand.
The question is how long can that persist? India said maybe we could retaliate below a certain threshold, but hasn’t tried it yet. The worry is that India’s leaders, or the public, say at some point “enough is enough.” How many terrorist attacks can you accept in your major cities without doing something? India has, and continues to, experiment with options to enable limited or total retaliation. At the conventional level, however, Pakistan’s threat of using battlefield nuclear weapons on Indian forces should they cross the international border has thwarted those efforts. But India is searching for answers to that paralysis. There aren’t many good options for it.
MS: Is it the case that more nuclear weapons states there are, the better?
Nuclear weapons do deter. I understand why weak nations want them. They do provide deterrence against invasion. They do provide existential protection. The question is are there some states, with certain regime types or civil-military relations, where the risks outweigh the perceived deterrence benefits?
But states like North Korea, Pakistan, and Egypt have potentially more volatile domestic political situations than, for example Japan or Germany or India. And even India is very opaque about its management and security procedures and the US has been concerned about lax oversight even there—and even the US itself is not immune to the risks of accidents, having had quite a few snafus of its own recently. So even in the most stable of states, the risk of accidents is real. Add to that mix the potential for violent domestic upheaval and one has to question whether having nuclear weapons possessed by a state at risk of coup or revolution is a good thing. You start getting into a world where more countries have them, there’s simply more systemic risk.
MS: But wouldn’t getting the bomb demand greater command or control from nation-states?
Pakistan may be one or two senior radicalized officers from having a threat to, or breakdown of, command and control. We assume there will be continuity in government, and regular transitions. The trouble is chaos or irregular leadership transitions, and uncertainty about the control of nuclear weapons in the state. Kim Jong Un has signaled that he has sole authority over nuclear weapons. But when he flew Air China to Singapore to meet with Trump, what if there had been rumor the plane had been shot down en route? What is his command and control? What if he feared being shot down and put in place a “dead hand” procedure which means, “If I’m shot down, you fire a nuclear ICBM at New York?” Rumors can go viral and there have been no way for those in Pyongyang to reach Kim, and they may have assumed the worst. These are the kinds of things that scare me.
MS: So on the one hand you don’t think Pakistan and North Korea will give up their nuclear weapons, on the other hand they make you nervous. What do you do about that?
VN: It’s heresy to say this in Washington, but we have strong incentive once a nation gets the bomb to help them put in place robust command and control systems so that they are confident in it. We want nations with nuclear weapons to only become more stable, and not fear for the survivability of their force, which can give them very itchy trigger fingers by putting them in so-called “use them or lose them” situations.
The dirty secret is we don’t want a nation with just tens of weapons because then they will fear that we’ll disarm them in an attack [by finding and destroying all their nuclear systems, completely disarming them and freeing us to topple the regime]. Pakistan is more confident today because it has 150 - 200 weapons. It isn’t worried about being in a “lose them or use them” situation against India. And so we worked with them on command control. The US government did a lot with Pakistan on safety, security and command and control, though there were limits as to how much we could do and how much Pakistan would accept. But everyone took it seriously and there’s an understanding that a nuclear state is unlikely to disarm, so the challenge is ensuring robust and safe command and control of the weapons and removing fears that it might lose its nuclear arsenal in an attack.
MS: Isn’t that a paradox, though? Aren’t we saying we’ll try to stop you from getting a weapon but once you get one we’ll help you?
VN: We have spent so much time trying to stop nations from getting weapons so that we don’t have to accept other nuclear powers outside of the NPT. Our entire grand strategy is premised on stopping most nations from getting weapons, though we turned a blind eye to Israel and Pakistan. But we have forcibly stopped, or induced, adversaries from acquiring them, and one should not forget that one of the most powerful nonproliferation tools in our kit is extended deterrence, which dissuades our allies from pursuing nuclear weapons. But if nonproliferation fails, then challenge then becomes ensuring that any new nuclear state manages the weapons safely and securely. So the paradox is that we try hard to stop them, but if it fails, then we have to tacitly accept that they have them and not necessarily help them, but at least encourage robust command and control institutions and survivability so they won’t have fears of us, or others, disarming them in an attack.
MS: Is it the case nuclear weapons only end, prevent, and reduce the size of wars?
VN: There is a debate over whether they have coercive value as well. That is, do nuclear weapons help you change another state’s behavior and get them to do something they otherwise wouldn’t? It’s hard, because it is often not credible to threaten them with nuclear weapons if they don’t stop, for example, harassing your ships. But China did justify its acquisition of nuclear weapons in the 1960s on the basis that the US was coercing it in East Asia. But it was also the case that we were a superpower, so it’s hard to attribute that perceived coercion to nuclear weapons. But there is a persistent belief amongst some policymakers that nuclear weapons may provide some coercive value., even though in Arms and Influence, Thomas Schelling said deterrence was easier than compellence.
MS: What about this idea of a “virtual nuclear arsenal” where nuclear-armed nations make it harder to use them by removing warheads from missiles, or disarming submarines.
VN: I call it recessed deterrent. The Chinese and Indians did it because they were conventionally superior militarily, and did not need nuclear weapons to deter conventional attacks on their homelands. For the U.S. to invade China, it would have to launch a prohibitive amphibious assault. So the Chinese concluded that they could have a more relaxed posture. It would have assured retaliation if it were ever attacked with nuclear weapons, but it would not necessarily have to be immediate. This allowed the party to keep the nuclear weapons demated from the missiles and assert highly centralized control over the force. They didn’t have submarines at first, so they did not have to put fully ready systems out to sea. So it could declare a no first use policy, and rely on a handful of (plausibly survivable) long range nuclear missiles to deter nuclear use on China itself. The Chinese did that and then the Indians followed.
But once you have submarines you can’t do that. For a variety of reasons, one will probably have the submarine launched ballistic missiles fully mated when the submarine deploys at sea, so the system is premated and ready from the outset. The Indians have likely decided that it is too difficult and chaotic to persist with disassembled and demated forces at sea, and as a result, they are revisiting the arrangement on land as well and moving to a more ready, canisterized force. The Chinese may be revisiting this as well. There is technological pressure, too, since a more sophisticated command and control structure also allows for a more ready force. So early on in the system you may get recessed deterrence, where the force is easier to manage and it is easier to prevent unintentional use. But when you have more robust command and control, and more advanced systems in which you don’t have to sacrifice safety, why not go to a more ready disposition? And that is what we are probably seeing in China and India today.
MS: How much realism is there in worries over accidental use?
VN: In peacetime, the risk is probably very low. The risks are there, but they are lower because one can design the system to fail safe in times of peace, at low alert levels. The risk in peacetime is lax oversight, as the US Air Force experienced when it unknowingly transported nuclear missiles between two bases on the homeland. That was not good. In other countries they are so scared of inadvertent use which is why procedures tend to favor high centralization and systems that may be designed to fail safe. But in a crisis, as alert levels sharply rise in some countries, the risk of accidental or inadvertent use can really spike as safety procedures and features are moved and any officer that has a nuclear weapon could potentially use it if he thinks the world is about to come crashing down on his head. In North Korea, for example, an inadvertent tweet could set off Kim Jong-un into thinking that the Americans are coming after him, and he or his deputies could launch missiles to stave off the attack or just out of revenge. One can get use out of sheer misperception or miscalculation. As long as the weapons exist, it’s possible.
MS: Everybody imagines that if a nuclear bomb of any kind is used, even against soldiers in a tactical way, then it will automatically escalate. But that’s not the case, is it?
VN: If a nuclear bomb is used anywhere I think the world will come to a screeching halt. The first use of nuclear bomb since 1945? I think people will stop and ask “What the hell just happened?” and the international community will race to try to stop escalation.
MS: What if North Korea fires a nuclear bomb at us?
VN: I think the US military and leadership would both prefer to end the regime with a conventional military attack, not nuclear. That would be the first instinct I imagine. We’re so powerful militarily that we could destroy any state, except for Russia and China, that used a nuclear weapon against us with conventional weapons.
MS: How might a tactical nuclear weapon be used?
VN: Pakistan had settled on a NATO-style war winning strategy. “I want war to end, I don’t want to break India or destroy its cities, I just want to stop the Indian Army from slicing Pakistan in half. If we use our own nukes, on our own territory, in the desert, against an Indian strike corps, we haven’t given them justification to use nuclear against our cities.” But even then, it would be an event of such magnitude that the world would race to stop it from escalating.
MS: Why do people so often raise the threat of Saddam Hussein acquiring the bomb and using it against Iran?
VN: Saddam is the go-to example of risk-accepting dictator. The idea is that we can’t have somebody who would so brazenly invade Kuwait get the bomb. He’s the poster boy for that fear. People used it to argue against Kim Jong-un having the bomb, but he’s showed himself to be hyper-rational thus far
Iraq had a crash program in the 1980s. They got a lot further than anyone knew—though it still wasn’t particularly far. They had fissile material under safeguards at Osirak that they could have ripped the off of and used one-time for a nuclear bomb. They had separate enrichment programs in their infancy which, had Saddam not invaded Kuwait, may have clicked at some point. Their target was always imagined to be, I suppose, Iran and Israel. Would Saddam have used a nuclear bomb against Iran if he had it? Maybe. Iran would have retaliated with chemical weapons. Saddam was risk-accepting, but was he so risk accepting? I think he probably would have tested a weapon at a remote site to signal that this is it. If you test you hope the international community comes in and stops Iran.
MS: What about no first use policies?
VN: China and India say they have it but it is ambiguous. India’s is already officially caveated on being able to retaliate with nuclear weapons in the event of chem or bio weapons use. India also injects ambiguity around preemptive use, and the Chinese may have the same view.
In terms of the U.S., we are conventionally superior, but our allies are not. So if not for extended deterrence, we could conceivably consider a no first use pledge, though I doubt any of our adversaries would ever believe it. But because our allies depend on the nuclear umbrella, and at least the possibility of American first use on their behalf to stave off, for example, a Russian invasion, they tend to be allergic to any discussion of no first use declarations. When the Obama administration leaked that they were considering it, the Japanese reportedly freaked out. Relatedly, US tactical nukes are a critical part of our extended deterrence commitment, since the first use would likely be battlefield use on behalf of the allies. We would not want to use nuclear weapons against the Russian or Chinese homeland in any way, which would then expose the US homeland to retaliation, so the idea of limited nuclear options and battlefield nuclear weapons arose in order to credibly defend the allies without risking a massive nuclear war between the US and Russia/China.
So if you have extended deterrence, you need a lower order option. Some argue we should give up extended deterrence for our allies and have no first use. We would be able to potentially give up our tactical weapons in Europe. It’s all up for debate right now. But if you look at the Pentagon nuclear posture review, it assumes extended deterrence will persist indefinitely. The proposal for low yield Tridents and cruise missiles are about defending our allies in Europe and Asia against a Russian or Chinese land/island grab. We are modernizing our B-61 weapons and their delivery platforms on the assumption that we will be there for a long time. So the Pentagon imagines they will persist. The President seems like he just wants the Europeans to pay more.
MS: Should we allow the President to have sole authority over the use of nuclear weapons?
It’s dangerous, but there’s a reason why we have it, which is in the event of a Russian first strike. There would have been no time to get consent from the joint chiefs or the secretary of defense if the President needed to order a preemptive or retaliatory strike. They needed less people in the loop, not more. So sole authority was a feature, not a bug. You could revisit it, but there was a good reason why we had it. In the post-Cold War world you could argue that it is antiquated, and there should be a required two-person rule to release any nuclear weapons. But I doubt it is in the cards.. In a lot of ways, if it can withstand Trump, it can withstand anyone.