Michael KreponKenneth Waltz

The debate between Ken Waltz and Scott Sagan, The Spread of Nuclear Weapons (2nd edition, 2003) is a great teaching tool. Waltz, the academic Grinch of proliferation studies, argues that more proliferation is better than any of the alternatives — as long as the Bomb spreads slowly. Sagan argues that more proliferation increases the likelihood that something will go badly wrong, whether by accident, miscalculation, or from deadly organizational biases. This is a completely unfair fight because Sagan has to be right just once, while Waltz has to be right every time.

Waltz’s views are a reprise of his classic 1981 Adelphi Paper, The Spread of Nuclear Weapons: More May Be Better, which is must reading for Wonks. Here is the unvarnished Waltz:

The gradual spread of nuclear weapons is better than no spread and better than rapid spread.

Possession of nuclear weapons may slow arms races down, rather than speed them up.

In countries where political control is most difficult to maintain, governments are least likely to initiate nuclear-weapon programmes.

Nuclear weapons induce caution, especially in weak states.

Minor nuclear states have even better reasons than major ones to accommodate one another peacefully and to avoid any fighting.

The biggest international dangers come from the strongest states.

The presence of nuclear weapons makes wars less likely.

Even if deterrence should fail, the prospects for rapid de-escalation are good.

In desperate situations, what all parties become most desperate to avoid is the use of strategic nuclear weapons.

It is not likely that nuclear weapons will spread with a speed that exceeds the ability of their new owners to adjust to them. The spread of nuclear weapons is something that we have worried too much about and tried too hard to stop.

Lest arms controllers feel deeply offended, Waltz also throws round-house punches at missile defenses in his debate with Sagan:

National missile defenses pose greater dangers to us and to others than the slow spread of nuclear weapons. The best thing about such defenses is that they won’t work. The worst thing about them is that merely setting development and deployment in motion has damaging effects on us and on them.”

According to [missile defense advocates, would be attackers] are smart enough to calculate the problematic effects of our future defenses, but too dumb to understand the risks of launching attacks that risk their own destruction.

The mere prospect of American missile defense promotes the vertical proliferation of nuclear weapons. It also encourages the horizontal spread of nuclear weapons from one country to another.

Waltz the supreme realist, assumes ever-present rationality… and the absence of Murphy’s Law, except when it comes to missile defenses.


  1. Chuck Thornton (History)


    I was an undergraduate at UC Berkeley* in the late 1980s. I took a course from Prof Waltz in my sophomore year that was my introduction to the world of nuclear weapons policies, deterrence theory, and all of the associated issues derived therefrom. He certainly argued the controlled-proliferation-is-good theory, but the course material he utilized was fairly balanced. Indeed, I still have all of the assigned texts on my bookshelf. I didn’t know any better at the time, but it turned out to be quite an introduction to the subject. And an influential one: my views on the subject, while continually evolving, will forever be influenced by Waltz’s methodological approaches and ways of thinking.

    All the best,

    *Despite UC Berkeley’s reputation as a hotbed of radical liberaleralism, the Political Science department at the time was one of the more (neo)conservative. I suspect that this was due to the creation of many other Poli Sci-like departments (Women’s Studies; Peace & Conflict Studies; etc.) that drew the more liberal faculty away from Political Science and leaving only the conservatives. It always struck me as odd that Ken Waltz would be at Cal, but who knows – maybe his theories were just radical enough to fit right in. But, I was only a naïve undergrad …

  2. MarkoB

    Waltz has another line in his debate with Sagan, perhaps his most important one. He states that the US opposes proliferation because the spread of nuclear weapons would “cramp our style”. In the post below the wonk cites Mark Hibbs as stating on Iran that the IAEA has discerned that, “some activities that would be appropriate for a nuclear weapons research program appear to have coincided with acute concern about external military threats faced by Iran.”

    I think the two citations might be related.

  3. MK (History)

    It’s great to hear from you.
    As Krishnamurti said, “True education is to learn how to think, not what to think.”

  4. Daryl Press (History)


    People often criticize the Waltz argument with two responses, which I think are both incorrect. You allude to them in your post.

    1)I don’t think it’s correct to say that Waltz has to be right every time for his view to trump Sagan. If, for example, the existence (or even spread) of nuclear weapons prevents 1 great power war per century (on average), or several regional wars per century, and the cost of that deterring effect is 1-2 Sagan-esque accidents per 100 years, this would presumably be a very good tradeoff from the stand point of humanity and international order.

    Major conventional wars have killed so many people that vastly reducing their likelihood is potentially worthwhile even if there are accidents.

    The devil is in the details — how likely are those wars absent nuclear deterrence, how reliable will the deterrence be, and how ghastly will the accidents be? Are the accidents of the “accidental detonation” type, which might kill a 100-200 k people (or far fewer if it occurred in a remote location)? Are the accidents of the type that would release contamination, with perhaps smaller lethal effects? Or are the mishaps the triggering of accidental nuclear exchanges, which could be mind-bogglingly bad. In my mind, this mixed-Waltz-Sagan analysis (wars deter powerfully and some accidents will likely happen) lead to a pragmatic view toward proliferation in which one judges how good/bad a new proliferant would be depending on how good a job they will likely do as stewards (as well as how responsible their government is).

    But it is not right to say that nuclear weapons must be accident-free for Waltz to be right.

    2) You refer to Waltz believing in “ever present rationality.” Yes and no. You can certainly find quotes of his that make it sound that way. In my opinion, however, the best version of his argument is that as long as leaders are somewhat sensitive to costs they will be powerfully inhibited from taking any actions (eg, launching a nuclear strike) which are likely to trigger massive nuclear retaliation. I think that’s the stronger version of his argument — not that people are emotionless, perfectly-rational calculators.


  5. William (History)

    The idea of nuclear deterrence re MAD only works when the parties involved are relatively sane, i.e., not suicidal.

    This unfortunately is not the case in Iran, with the mullahs and Ahmadinejad being, to a greater or lesser degree, devoted to bringing about the conditions whereby the Hidden Imam [the Mahdi] might return, only one of the reasons why the country must be prevented from ever obtaining nuclear weapons.

  6. MK (History)


    Well said. Much more nuanced than my post.

    Waltz does pretty well in his debate with Sagan. He hasn’t been proven wrong — yet.

  7. yousaf

    As usual, the risk of deterrence failure are underplayed.

    For a more accessible version, see here.

    Just like nuclear weapons have kept us safe, the space shuttle programme also kept astronauts safe — until 1986.

    Nuclear Optimism (“nuclear weapons keep us safe”) is not based on any statistical analysis and therefore its conclusions are ill-founded, and downright dangerous.

    Martin Hellman has an email distribution list where he recently sent the following:
    September 7, 2009

    How Confident Should a Nuclear Optimist Be?
    Martin E. Hellman*

    Nuclear optimism is a school of thought which argues that more nuclear weapons make the world safer. Given that our nation and Russia each have around 10,000 such weapons in its arsenal, such thinking is more widespread than might be thought. The following assessment is therefore much more than an academic exercise, and has vital implications for humanity’s future.

    In a five-page essay in the September 7 issue of Newsweek, Jonathan Tepperman explains Why Obama Should Learn to Love the Bomb by quoting the dean of nuclear optimism, Prof. Kenneth Waltz: “We now have 64 years of experience since Hiroshima. It’s striking and against all historical precedent that for that substantial period, there has not been any war among nuclear states.” Tepperman calls for “coldblooded calculations about just how dangerous possessing them [nuclear weapons] actually is.” This response rises to that challenge and shows that the data used to justify nuclear optimism is highly misleading.

    In the same way that life-insurance companies utilize statistical analysis to produce cold blooded projections of fatality rates for individuals, statistics tells us that, to be 95% confident of our statements, we cannot project the last 64 years of nuclear non-use more than 31 years into the future. Even if one drops the required confidence level to 50%, that only increases the time horizon from 31 to 44 years. And, with the fate of the earth at stake, a higher confidence level would seem appropriate. If we want to be 99% confident about our statements, the 64 years of non-use that we have experienced cannot be used to justify a time horizon of even 14 years. Statistics does not rule out that we might survive significantly longer than these time horizons, but it does say that the data thus far cannot be used to justify such hopes with any degree of confidence.

    To understand why we can only be confident of surviving time horizons significantly shorter than the 64 years of non-use already experienced, it helps to consider related “space shuttle optimism” arguments that led to the loss of Challenger and her crew. The engineers who had designed the shuttle’s booster engine tried to delay Challenger’s final launch because the weather that morning was unusually cold, and previous cold weather launches had a higher incidence of partial “burn through” on O-rings designed to seal the booster. But those at NASA responsible for the launch decision suffered from the common misperception that the shuttle’s prior 23 successful launches provided ample evidence that it was safe to proceed with launch number 24. Instead, as we now know, that launch suffered catastrophic burn through of the O-rings, with resultant loss of the shuttle and her entire crew.

    NASA’s optimistic reasoning was literally dead wrong. Even 23 perfect launches would not have provided sufficient evidence to confidently predict success for launch number 24, and previous near misses, in the form of partial O-ring burn through, made optimism even more outrageous and unsupportable. The unassailable, cold blooded conclusion provided by statistics and Challenger’s deadly lesson is that 64 years of nuclear non-use, particularly with near misses such as the Cuban missile crisis, is no cause for nuclear optimism.

    * Martin E. Hellman is a member of the National Academy of Engineering and Professor Emeritus at Stanford University. His current project applies risk analysis to nuclear deterrence, and is described in detail at NuclearRisk.org. A summary statement has been endorsed by:

    • Prof. Kenneth Arrow, Stanford University, 1972 Nobel Laureate in Economics
    • Mr. D. James Bidzos, Chairman of the Board, VeriSign Inc.
    • Dr. Richard Garwin, IBM Fellow Emeritus, former member President’s Science Advisory Committee and Defense Science Board
    • Adm. Bobby R. Inman, USN (Ret.), University of Texas at Austin, former Director National Security Agency and Deputy Director CIA
    • Prof. William Kays, former Dean of Engineering, Stanford University
    • Prof. Donald Kennedy, President Emeritus of Stanford University, former head of Food and Drug Administration
    • Prof. Martin Perl, Stanford University, 1995 Nobel Laureate in Physics

    To create greater public awareness of this critical issue, I hope you will forward this email to friends who might be interested and encourage them to sign up for future updates via the JOIN US box at http://nuclearrisk.org/.

    To better understand the problem and solution check out:
    “Soaring, Cryptography and Nuclear Weapons” at http://www.nuclearrisk.org/soaring_article.php
    Frequently Asked Questions at http://www.nuclearrisk.org/faq.php.

    If you missed earlier emails to the group, they are at http://nuclearrisk.org/resources.php along with other sources of information. Emails #4 and #5 describe recent events which had the potential to produce a crisis comparable to the 1962 Cuban crisis.

  8. FSB

    maybe Iran is not suicidal.

    Maybe Iran sees that neighbors on either side of it have been attacked by the US.

    Maybe Iran sees that Israel has attacked 3 of its neighbors in 3 years: Lebanon, Gaza, and Syria. Before that Iraq too.

    Maybe, just maybe, Iran also feels threatened.

    Maybe Iran has greater reason to fear the actions of the US and Israel, than Israel has to fear the words of Mr. Ahmedinejad.

    Maybe the implications of a nuclear armed Iran will not be so dire.

    Maybe, just maybe, the Hidden Imam will return at the End of Days. That will be one party I’d like to be at.

  9. Daryl Press (History)


    1) The existence of 64 years of non-catastrophe might be a poor basis for extrapolating that nuclear deterrence will always hold in the future if the argument were merely based on that evidence — but this is not merely an inductive argument (ie, one based solely on an observed pattern in the data). Rather, the arguments about nuclear deterrence are also supported by pretty strong deductive arguments about human behavior: if people are remotely sensitive to costs, they will be very unlikely to take actions which vastly increase the odds of triggering massive nuclear retaliation. That’s a pretty reasonable statement — every human I’ve met (though I’ve never wandered through an insane asylum) is sensitive to costs — even my 3 year old son. So we don’t merely have 64 years of non-use to make us expect that deterrence will be pretty robust: that data goes hand-in-hand with a pretty reasonable theory of how the world works.

    2) “64 years” understates our evidence. We have 64 years of US behavior, nearly as many years of evidence about Soviet behavior, 20 years of Russian behavior, decades of French and British behavior, same with Indian and Pakistani behavior, and Chinese behavior, and depending on when one starts counting we are accumulating data on NK behavior too. I forgot Israel. There’s even pretty nice variation within that data for type of govt, even religion.

    3) Your argument looks only at the risks associated with nuclear deterrence failure. What about accounting for the risks that nuclear deterrence reduces? A deterrence failure that kills 10 million people would be terrible — but how does that tragedy compare with the benefit of avoiding the “missing” great power war of the 2nd half of the 20th century? How many lives will be saved if nuclear deterrence mitigates the mutual fears and dangers associated with China’s rise? Remember that great power war — which killed roughly a hundred million people in the first half of the 20th Century — stopped happening when people developed nukes.

    4) Why assume that a deterrence failure would put the survival of the world at stake? In most plausible circumstances — a Pak – India exchange — or a terror attack on Israel (or the US) followed by a dozen or so weapons in response, lots and lots of people would die. It would be a tragedy. But “world at stake” is at the far end of the continuum of possibilities — when we weight the national security benefits of these things, and their pacifying effects on world politics — against the risks, we need to maintain a clear view of the scale of tragedy if nuclear deterrence fails.

    — Daryl

  10. Daryl Press (History)


    What’s the evidence that the Iranian leadership values NOTHING in this world? For the nuclear deterrence argument to be undermined by claims of Iranian irrationality, you’d have to convince me that the people in Iran who control the weapons don’t value

    (a) their own lives;
    (b) their children’s lives;
    c their hold on power;
    (d) the continued existence of the Iranian state as a political unit;
    (e) the Iranian population;
    (f) the historic treasures of Persian civilization;
    (g) some of the most important religious sites in Shi’ite islam;
    (e) more.

    Those are the things that would likely be wrecked if Iran got itself into a nuclear war with the United States or Israel. IF Iran’s leaders value NONE of those things, then you’re right that deterrence won’t work. But what’s the evidence of that?

    We certainly see examples of people willingly giving their lives for their causes — but that’s not evidence of their irrationality, let alone their leaders’ craziness. When we think of people giving their lives, our minds immediately leap to the murderers who flew planes into buildings on 9/11. But you should also think about the beautiful lines of grave stones at the American Cemetery at Normandy. Many of those men fought in circumstances in which they essentially knew they would die. Almost all US Medal of Honor winners did some act of heroism that essentially guaranteed their own death. But that doesn’t mean that the US leaders don’t care about (a) through (g). Frankly, it’s hard to succeed in the rough-and-tumble politics of authoritarian states if one doesn’t have a pretty sensitive risk / threat meter.

    Perhaps FSB is right and Iran wants the bomb because it feels threatened – or perhaps he’s wrong and Iran wants the bomb to help them extend their power in the region. (Or perhaps they don’t want it!) I don’t know — neither do any of us. But as long as Iran’s leaders care dearly for SOMETHING on the list above, they’re not the “irrational” undeterable folks that people often assert.

    (When James Baker tried to deter Saddam from using CW or BW in the 1991 war, he seems to have casually turned the conversation to Saddam’s and Tariq Aziz’s children. We all care about something.)

    — Daryl

  11. lizz (History)

    Iran is not only not suicidal, there’s simply no evidence that the Iranians want nuclear weapons.

    FT: There is an argument that Iran should have nuclear weapons, like at least one of your immediate neighbours [Pakistan].
    Salehi: In matters of national security we are not timid. We will assert our intentions. If nuclear weapons would have brought security, we would have announced to the world that we would go after them… We do not think a nuclear Iran would be stronger… If we have weapons of mass destruction we are not going to use them – we cannot. We did not use chemical weapons against Iraq.
    Secondly, we do not feel any real threat from our neighbours. Pakistan and the Persian Gulf, we have no particular problems with them, nor with Afghanistan. The only powerful country is Russia in the north, and no matter how many nuclear weapons we had we could not match Russia.
    Israel, our next neighbour, we do not consider an entity by itself but as part of the US. Facing Israel means facing the US. We cannot match the US.
    We do not have strategic differences with our neighbours, including Turkey.

  12. yousaf

    your methodology does not account for mistakes, accidents and inadvertent and/or unauthorized use of the stockpile. Just because it did not happen yet does not mean it cannot happen. Indeed we have come close many times. Your arguments completely ignore this fact.

    Further, I do not see much evidence of their “pacifying effect on world politics” — this may be true from a narrow US/Soviet viewpoint, but not, say, from an Angolan or Vietnamese perspective.

    Rather than venting their rage on each other during the Cold War, the superpowers merely destroyed a series of dirt poor, sorry peasant nations in a long series of proxy wars (due to nuclear weapons). If you were a resident of Vietnam or Afghanistan (or even a US GI), nuclear weapons did not necessarily keep you safe.

    Lastly, it is not only my viewpoint, but rather also that of the people quoted: Martin Hellman, • Prof. Kenneth Arrow, Stanford University, 1972 Nobel Laureate in Economics • Mr. D. James Bidzos, Chairman of the Board, VeriSign Inc. • Dr. Richard Garwin, IBM Fellow Emeritus, former member President’s Science Advisory Committee and Defense Science Board • Adm. Bobby R. Inman, USN (Ret.), University of Texas at Austin, former Director National Security Agency and Deputy Director CIA • Prof. William Kays, former Dean of Engineering, Stanford University • Prof. Donald Kennedy, President Emeritus of Stanford University, former head of Food and Drug Administration • Prof. Martin Perl, Stanford University, 1995 Nobel Laureate in Physics.

    I suggest you read the linked articles in my earlier message for further details, in case you have not already — they are rather clear eyed as to the dangers of nuclear optimism.

  13. yousaf

    PS: I don’t agree with looking at the behaviors of various countries linearly as you have done. The highly interdependent deterrent calculus of the various powers (e.g. pakistan impacts india impacts china impacts USA) cannot be neatly unfolded to a linear string of time greater than actually transpired. It is a clever way to artificially increase the time-horizon, but I do not agree with it.

  14. Daryl Press (History)


    Nuclear “optimists” are often portrayed as being naive — e.g., they simply ignore the possibility of accidents, as you wrote. It’s just not true.

    Accidents will happen, and when they do lots of people may die.

    The argument of nuclear optimists is different: nuclear weapons may have ended once and for all one of the most destructive human activities: major, great power wars of conquest. If so, that achievement may be well worth the continuing risk of accidents or war. You may not think the tradeoff is worth is, which is fine, but don’t mischaracterize the arguments of the other side as ignoring the risk of accidents. (Though I think you and MK are right that Waltz himself understates the risk of accident, in my view.)

    I should add that the argument I’ve been making about reducing the likelihood of major wars is an argument that’s particularly appealing to American nuclear “optimists”. In other countries — particularly small ones who live near enemies — the argument is much more visceral and far less abstract: nukes are the final backstop which (nearly) ensures that their lands won’t be overrun, their houses won’t be seized, and their families won’t be killed. Whether you like the Israelis or not, that’s how they see their nukes. It’s why South Korea may want them soon, too. It’s why Iran may want them. For Americans to say that nuclear weapons are an abomination may not be adequately empathetic to the real security problems that others face — others who don’t have the blessings of two big oceans and docile neighbors.

    Many people may prefer a world without nukes because of fears of nuclear war, but we should understand what that will do to those who feel weak and vulnerable.


Pin It on Pinterest