Michael KreponConflicting Blueprints

Sometimes it helps to frame choices about familiar subjects in different ways. By way of prologue – and to dust off the arguments in my book, Better Safe than Sorry (2009) – here’s another way to think about future arms control choices:

The global system created over many decades to prevent nuclear proliferation can be likened to a construction project. The construction is only as sturdy as the common resolve of the five nations with nuclear weapons that also enjoy permanent seats on the United Nations Security Council. As the world’s strongest power, the United States has the most responsibility for building maintenance. If Washington walks away from this job, the construction site will become unsafe. But even if the United States does its job properly, Russia, China, France, and Great Britain still have to support the structure. When the P-5 works in concert against the perils of proliferation, the construction provides reliable shelter. When they place other national security and commercial interests ahead of proliferation concerns, the construction becomes wobbly.

The building’s load-bearing walls consist of agreements, rules, and norms designed to prevent proliferation. Treaties that set legally binding obligations constitute the steel beams that keep this structure erect…

Some construction on the first floor was only partly completed. One room for a treaty that would end nuclear tests for all time was built but never occupied… Other planned construction was never undertaken, especially a treaty banning the production of fissile material for weapons.

Constructing the first floor of this edifice required consensus, not only between the superpowers, but also between weapon possessors and abstainers. During the first nuclear age, the foundation remained strong. Even though the nuclear arsenals of both superpowers rose to absurd levels, new additions to the nuclear club were kept reasonably in check…

The second nuclear age began in 1991 with the demise of the Soviet Union and the surprise discovery in Iraq of an advanced bomb program… The structure built to prevent proliferation during the first nuclear age was not designed to deal with new members of the nuclear club or the threat of nuclear terrorism… At the same time, none of the P-5 acted like strong stakeholders during the second nuclear age…

The first nuclear age was an exercise in establishing norms against proliferation. The norms helped to apply leverage on states that were fence-sitters. The norms did not prevent rule breaking, but they did make it easier to isolate or sanction rule breakers. During the second nuclear age, these norms were weakened, and there was less discipline to reinforce them…

The inner circle of the George W. Bush administration held jaundiced views about the effectiveness of global nonproliferation norms and the utility of treaties. They wanted to build a second floor to address new proliferation challenges, using different tools. Rooms on the second floor would not require consensual building permits, because the second floor required the coercive tools that the first floor lacked…

Some of this construction worked reasonably well…Constructing a second floor of the global nonproliferation system was needed, and the Bush administration deserves much credit for these initiatives. The Bush administration’s design, however, had a fundamental flaw: It focused almost entirely on the second floor while leaving the first floor to its structural weaknesses. The administration focused on the second floor precisely because it fundamentally disagreed with previous builders over what constituted the first floor’s load-bearing walls….

Moreover, construction during the first nuclear age was built on the core principle that rules needed to apply to all… The entire first floor could never have been built had its builders chosen to construct one set of rules for responsible states and another for bad actors because the construction crew couldn’t agree on who belonged in which category…

The master builders of the global nonproliferation system considered these load-bearing walls to be the norms against using or threatening to use nuclear weapons, testing these devices, and producing the fissile material that made these weapons so lethal. Some – but not all – of these norms are embedded in treaties. The architects of the first floor believed that the structural integrity of their building required the rebar provided by treaties – especially the treaty banning nuclear tests…

The Bush administration used fundamentally different building principles for the second story. It postulated a new guiding principle for proliferation: that the fundamental problem was not related to the Bomb, but to the bomb holder. The character of the state mattered most… The Bush administration also significantly qualified a norms-based approach to proliferation: Rules and norms were good, except where norms constrained U.S. freedom of action or the actions of U.S. friends and allies…

No reputable construction company would build atop a shaky foundation without strengthening its load-bearing walls.

This analysis still holds, at least in my view. Fundamental disagreements continue to exist on Capitol Hill over the value of treaties and strategic modernization programs. Democrats and Republicans have two sets of blueprints for reducing nuclear dangers, with insufficient overlap. Democrats now feel comfortable in some rooms on the second floor, such as the one housing the proliferation security initiative, but the structure’s foundation will remain shaky as long as Republicans in the Senate oppose new strategic arms reduction treaties and the CTBT.

The old instruments of suasion to bring Republicans on board – hugely expensive commitments to new projects at the weapon labs and strategic modernization programs – can no longer be replicated, given their track record of uncontrolled cost growth, their Cold War legacy, and competing demands for defense dollars. Republican hawks still think they can hold treaty ratification hostage to multiple big ticket items – as was the case for New START — but they are dwelling in the past. The Obama administration wasn’t able to deliver on its promises and its successors will also be hard-pressed to find the sums needed to recapitalize the Triad, the labs, and the stockpile. Nuclear hawks now have to deal with deficit hawks as well as arms controllers.

The era of linkage between multiple, expensive strategic modernization projects and ambitious treaties is over. U.S. nuclear forces will have to do less with less, and besides, treaties are no longer ambitious. Their modest offspring aren’t worth exorbitant price tags.

So what are the options? Option I is reduced ambitions for both strategic modernization programs and treaties, using a more modest form of linkage to maintain a functional domestic consensus in both domains. Option II is no new treaties anytime soon and continued shrinkage of the labs and nuclear force structure. Option II will remain more likely, at least until the Republican Party regains its balance. Arms controllers won’t like the absence of treaties, but they will be able to accommodate themselves to Option II more easily than staunch defenders of nuclear deterrence.


  1. j_kies (History)

    Following your construction based simile, attendees at a recent APS short course might have drawn the following conclusion:

    The foundation of the building is in a swamp that requires continuous stabilization. The NNSA is in a domain where they appear to expend most of their resources to maintain the safety / security / reliability of the aging stockpile in the presence of an aging workforce. A lot of effort is expended to move forward merely to prevent uncontrolled regression. Its reasonable to expect others to have similar problems in their nuclear enterprises.

  2. Rob Goldston (History)

    I like your turn of phrase, “less for less”. This is pretty much baked into the cake for now. Those of us who want CTBT and FMCT to move forward need to find a new path. The old method didn’t work for CTBT anyway.

    It is insane that a country with the majority of the world’s nuclear tests in one pocket, and more than more than enough fissile material in the other, can’t agree that everyone should stop producing these.

  3. Hass (History)

    Interesting how you seem to think that world nonproliferation laws and norms are something that the US has some sort of inherent right to manipulate, and you just assume it does so for the greater good of nonproliferation, when in fact we’ve witnessed that the US is very very good at being selectively outraged over that issue. It seems that the WMD wonks have their own nationalistic biases.

    • MK (History)

      Do you read what I write?

  4. Bradley Laing (History)

    —Mr. Krepon, could you clear this claim up:

    1. A Nuclear Reactor 1.8 Billion Years Old
    In 1972, a French factory imported uranium ore from Oklo, in Africa’s Gabon Republic. To its surprise, it found the uranium had already been extracted.

    They found the site of origin to be a large-scale highly advanced nuclear reactor that came into being 1.8 billion years ago and was in operation for some 500,000 years.

    Scientists gathered to investigate, with many explaining it away as a wondrous, yet natural, phenomenon.

    Dr. Glenn T. Seaborg, former head of the United States Atomic Energy Commission and Nobel Prize winner for his work in the synthesis of heavy elements, explained why it definitely wasn’t a natural phenomenon, and thus must be a man-made nuclear reactor.

    For uranium to “burn” in a reaction, very precise conditions are needed.

    The water must be extremely pure, for one. Much purer than exists naturally anywhere in the world.

    The material U-235 is necessary for nuclear fission to occur. It is one of the isotopes found naturally in uranium.

    Several specialists in reactor engineering have said the uranium in Oklo could not have been rich enough in U-235 for a reaction to take place naturally.

    Furthermore, it seems the reactor was more advanced than anything we could build today. It was several miles in length and the thermal impact to its environment was limited to 40 meters (about 131 feet) on all sides. The radioactive waste is still contained by surrounding geological elements and has not migrated beyond the mine site


    • Cthippo (History)

      I remember hearing about that in school.

      The key difference is that when the Oklo natural reactor was operating, about 1.7 billion years ago, natural uranium consisted of about 3% U-235 instead of the 0.7% today. U-235 has a shorter half life than U-238 and to the percentage of 235 in uranium has been slowly decreasing over the millennia. 3% U-235 is perfectly good light water reactor fuel, and all it took was for water to seep in and act as a moderator, slowing down the neutrons and making them more effective at causing fission. The reaction taking place in the rocks would boil the water, forcing it out, and the reaction would stop until the water condensed and flowed back in. Lather, rinse, repeat.

      It has been calculated that the Oklo natural reactor operated on a 3 hour period, “running” for about half an hour and then cooling off for about 2.5 hours.

      The natural reactor was actually predicted by Paul Kuroda in 1956, but wasn’t discovered until 1972.

  5. Bradley Laing (History)


    Mark Fitzpatrick, who has followed this story through the years as a non-proliferation expert at the state department and now at the International Institute for Strategic Studies had this to say:

    ‘It is not a new story, of course, but Urban came up with some new data points: a Saudi belief that it could obtain nuclear weapons from Pakistan at any time, and reported intelligence that Pakistan has prepared nuclear weapons for delivery to Saudi Arabia. The first part is probably true: The Saudis helped to finance Pakistan’s nuclear weapons programme and believe that they were given a promise that the weapons would be used to defend the Saudi kingdom if need be.

    The second part is probably false:…”

    David Albright, the head of the Institute for Science and International Security, broadly agrees. This was his comment:

    Would Pakistan give them [nuclear weapons]? There would be real punishments for that and they would want to avoid those. For Saudi Arabia to take possession it would mean withdrawing from the NPT [nuclear Non-Proliferation Treaty]. Any US military sales would have to stop. That could not be ignored. Only in a very dire situation in which Iran has a nuclear weapon and is being confrontational, could you imagine something like this.

    This sounds about right to me. If the Saudis are constantly calibrating the costs and benefits of shipping in Pakistani nukes, I imagine they are still a long way from calling in their chit, and the Pakistanis likewise. But it serves both countries, as well as Israel, to have the story front and centre while US and Iran sit down in Geneva.

    • Mark Pontin (History)

      ‘But it serves both countries, as well as Israel, to have the story front and centre while US and Iran sit down in Geneva.’

      That there is the size of it.