Monday, September 28, 2009

The Realist goes M.A.D.

Over the past two days, I've made some fairly bold claims about nuclear power on my twitter account. Here's my claims:
  • I trust the sanity of people who have the most to gain by not engaging in nuclear war. Being king > being dead. And yes, even Kim Jong Il has more to gain by not nuking, and he's as close to insane as we've got. It's in Kim Jong Il's interest for people to think he is crazy - he (and those in power around him) gain nothing, however, by actually engaging in Nuclear War.
  • The main argument I've heard for Mutually Assured Destruction (MAD) not being stabilizing: Bernard Lewis, who claimed that the Iran govt was apocalyptic, and saw the certain destruction of nuclear war as an inducement, rather than a deterrent. I spent a good part of last year writing a response for my War on Terror class. My core counter argument was, essentially, that while Ahmadejinehad (and Iran proxy) have a lot to gain through nuclear posturing, and even through the possession of nuclear weapons, they are first and foremost a government. And as a government, things are better for them if they both a) stay in power, and b) stay alive. Religious belief may be strong enough to motivate a terrorist to kill himself on behalf of his community, but very few people are genuinely willing to risk initiating the actual death of their community. Altruistic motives fuel suicide terrorism; if you think your death will benefit the community, you may well do it. But it does not extend far enough to risk the entire community, because nothing is gained by that.
  • M.A.D. works precisely because even if the leader has more to gain in a war, and leaders usually have the most to gain, gains in war become impossible w/nuclear second strikes. Were a nuclear nation to initiate war against another nuclear nation, the damages that resulted would be, well, apocalyptic. You'd get two devastated nations, and the cost + time involved in rehabilitating them is certain to be expensive. Not going to war is, in this day and age, always cheaper and the better economic prospect for a nuclear armed nation. Plus, any government that initiates such a war is sure to either die, be deposed, or be greatly reduced in power within minutes.
  • Finally: a single person may be irrational (Ahmadejinehad, Kim Jong Il). But a small group of people (say, the rest of the governing bodies in Iran with a special emphasis towards the Supreme Leader, or the bureaucratic elite of the DPRK) errs towards rationality.
  • Governments of States, as collections of people with a vested interest in preserving the status quo, are going to be more rational and more restrained that the sum of their parts. They might posture, and they may well pursue nuclear weapons as a deterrent, but actually engaging in nuclear war is not in their interest.
That is, more or less, the whole of my argument as regards states. Nonstate actors are generally perceived as less rational. Fortunately, however, nonstate actors can't really produce outright nuclear weapons on their own (dirty bombs being another matter). Nuclear forensics, while not as developed as it should be, mitigates the risk of a nonstate actor obtaining (and therefore using) a nuke. Nuclear forensics, ideally, makes the nation that gives a nuke to a nonstate actor (read: terrorist) responsible for any deaths caused, and therefore vulnerable as though the state itself had launched the weapon. So rather than MAD deterring the use of weapons by nonstate actors, it deters nations unloading nuclear weapons on terrorist groups. And so that makes the governments of all nuclear nations put a premium on tight control, small stockpiles, and encourages nuclear actions to be controlled as a state, rather than enacted by nonstate proxies. Following the thesis of states as rational actors, MAD is an effective deterrent, so long as nuclear war is seen as fundamentally unwinnable.

Qualifiers and Postscripts:

Nuclear forensics is valuable, but it isn't as effective a deterrent as MAD. MAD is purely a "between states" thing, and works on the premise of states as rational actors. Since states have a lot to gain by posturing, posturing through nonstate loose cannons is a fancy little risk in this day and age. That said, nuclear forensics still holds promise of making deterrence continue to be relevant, two decades after the cold war.

Security through proliferation? The topic itself came up because a friend made an aside about how giving everyone nuclear weapons was not the path to peace. Another friend interjected that I "might advocate giving all STATES nuclear weapons. Some silly thing about rational actors." So, I then went out and kind of babbled my way through a rough version of the argument you see above. I genuinely trust states rational actors, and I stand by the value of deterrence in a world with nuclear weapons. What I omitted in the above argument but included in my conversation are two fairly important asides:
  1. The only nation that was nuclear and actively, unilaterally disarmed itself was the apartheid government of South Africa. The reasons for this were multiple - the cold war was ending, South Africa really didn't see a need for itself to be nuclear, and the outgoing government really did not trust the rationality of the people they were handing the reins of power over to. It's an example of disarmament, which is a net win for everyone, through an explicit distrust in the rationality of states, which is probably fair but makes me a sad panda. I'm not sure what relevance this has, beyond being basically a silver bullet counterargument to my stated claims. Seemed worth mentioning, any way.
  2. Having nuclear weapons protects a single nation while increasing the risk to all other nations, resulting in a net lose of security. This is a macro-scale effect of the SUV phenomenon: if you drive an SUV, you yourself are safer, but every SUV on the roads makes the roads less safe. More nuclear weapons among more states doesn't actually provide much in the way of stability to anyone outside the most recently nuclear state, and greater proliferation comes with a greater risk of loose nukes and nonstate actors using them. For this reason, while I don't begrudge a nation like Iran seeking to protect itself with a nuclear deterrent, I'm really not all that fond of greater proliferation. I understand it, and don't see it as leading to the end of the world, but in absolute terms it's not a good thing.


Conor said...

I have a few points of disagreement. First, MAD only works if the other side can actual generate the will to destroy you. There are several cases where this is not true. An ethnic cleansing is a good example. If Saddam (or any other leader who engaged in genocide and actual had a nuke) used nuclear weapons internally, the consequences would be only slightly worse than using chemical weapons. It would work as a stabilizing force: no country would want to dedicate the resources to destroying a leader when there was the guarantee of nuclear attack on their invading troops. Citizens in the country would fear the government in new ways. With the exception of it being terrible, there is little deterrent here.

Also, MAD only works when both sides can dish out significant causalities. If Iran were to use nuclear weapons against say Israel, Israel lacks the number and automation to destroy all of Iran. America would certainly attempt to intervene and crush the Iranian government, but the area as a whole would be strengthened against the west. The leaders would be safe in any and every country they went to in the area. This is doubly true if Israel used a nuclear response. We would not 'wipe Iran off the map'. The worldwide environmental wreckage of such a move on our part would make it ridiculous unless immediate destruction was threatened.

I don't disagree with your fundamental point that states are rational and assuming a state is irrational is a bad move. I also think stopping other countries from joining the nuclear club while we hold an arsenal is ridiculous. I also don't think Iran actually wants to wipe out Israel or that North Korea wants to wipe out South Korea. I was just using country names that are topical as examples. (Saddam was chosen for the use of chemical weapons against the Kurds) I just don't think the use of a nuclear bomb is always an irrational move, in the same way a genocide sometimes makes sense to a government.

Kelsey Atherton said...


To your first point: internal conflicts are states vs. nonstate actors (as factions in civil wars count as nonstate). So that makes puts it outside the scope of MAD, which is purely limited to relations between nuclear armed states. And it falls into a wholly other category, the messy arena of sovereignty vs human rights. As for war between nuclear and non-nuclear nations, it is just not imitated by non-nuclear nations. Which makes attacking a nuclear nation impossible, and is perhaps the most asymmetric part of modern interstate war. It also helps explain why things like Afghanistan and Iraq are about a nuclear nation versus non-state actors (who aren't as vulnerable to nuclear weapons), rather than a nuclear nation versus non-nuclear nation.

Second point: Israel is a nuclear nation, so nukes from Iran would be met by Israeli nukes within minutes of launch, probably. For that conflicting dyad, it would be worth it to launch everything.The environmental destruction takes a back seat to the immediacy of making good on threats in the event of annihilistic war. The United States itself doesn't need to act for these threats to work, or for that scenario to play out.

Also: I meant to include more of this in my piece, but the main stabilizing part of MAD is second strike capability, while having only first strike capability is the really destabilizing part. Second strike guarantees MAD, as an attacked nation is then still able to retaliate. Only being able to launch before being hit is a much harder thing to manage.

Also: there are circumstances where nuclear weapons could be used, but for my purposes MAD is a between-states function, and I had forgotten about internal uses.

Lastly: I've earlier touched upon the gains the US can make by reducing it's nuclear forces, making its conventional superiority all the more striking. That's the disarmament I like.

Conor said...

I guess my issue is with second strike in the case of two nations without the ability to destroy each other. For example let's say Iran detonates a nuclear weapon non-conventionally. Then Israel is left in the tactical gray zone of launching a revenge strike with all 70-400 of it's nukes. If it does, it turns the Middle East and much of the world against it for it's asymmetry. If it doesn't, then Iran goes unpunished. Sure America would come in and wipe out Iran as a nation but the Iranian leaders who escaped would be popular enough to hide. Regardless, the outcome isn't mutually assured it's one side gets away with a small strike or another side gets away with a major revenge strike.
Basically, Iranian government is destroyed but if they are actually demagogues they gain celebrity and the promise of future power (don't believe Ahmadinejad is but that doesn't mean a leader can't be). Second strike is too delayed to work as anything but revenge.
One of the reasons for MAD was the separation of US and Russia. It would take 24 hours to get here from Russia, in that time, bombs could be flying the opposite direction. We also both had the ability to cause death on the scale neither side could accept. It is not clear the same is true in the case of Israel v. countries that dislike it.
The other major issue I have with proliferation that I did not mention last time is the stability of states. Generally if a state can make a nuclear weapon it is stable. But the question then becomes, how long term? Do we believe the Iranian government which has been overthrown at least four times in the past 30 years is not at risk to collapse? The same could be said about Pakistan, Russia, Kazakhstan(unclear how disarmed it is), N. Korea and even possibly China. If these countries collapse, tracking nuclear weapons becomes a headache. Non-state groups that believe they can break troops with attacks like this can get their hands on weapons or on the material in the case of a dirty bomb.

Conor said...

I guess I basically agree with you. My only issue is that MAD is based off a system where each side weighs the numbers the same. Most governments will do that, some will not. Governments have made silly moves that have doomed themselves based off ideology in the past (still have never seen a good reason for Afghanistan not extraditing bin Laden, or Japan launching an attack on Pearl Harbor). The moves probably seemed reasonable at the time, they jut weighed the numbers differently and incorrectly. MAD assumes no government will do that, and expects a rational government to punish the people of an irrational one.

Kelsey Atherton said...


MAD assumes the kind of hyperratioanlity that essentially leads to hesitancy. Conventional military moves allow you to correct course after war's been launched. I think we knew the immediate objective in Afghanistan (and Japan knew their's at Pearl Harbor), and figured that we'd have enough time to figure out and endgame. With nukes, you debate the endgame at the exact same moment you think about attacking. It focuses the mind a bit, and the consequences for making a wrong move are absolute.

That said, MAD works better the fewer actors their are, as every new nuclear nation increases the risk of human error.