I belong to the tribe that favors treaties that reduce the risk of war, add transparency and predictability to relations between well-armed competitors, and that depress impulses toward arms races. Not all treaties and treaty provisions succeed. Some have not only failed to accomplish these objectives, but have undercut them. What are the worst offenders?
In contemporary Republican circles, the 1972 Antiballistic Missile Treaty tops this list, but I strongly disagree. For three decades, the ABM Treaty braked one source of nuclear weapon requirements. Unfortunately, other accelerators were unchecked. If you are looking to apportion blame for MIRVs and other strategic modernization programs, the fault lies with the SALT I Interim Agreement, not with the ABM Treaty.
If we were to compile a list of the top mistakes in terms of provisions missing from arms control agreements, failure to stop MIRVs in the SALT I would certainly qualify. My list would also include the absence of a ban on underground testing in the 1963 Limited Test Ban Treaty. But this post focuses on what was agreed to, not what was out of reach.
What, then, are the three worst treaties or treaty provisions? Here’s my list:
1. The 1919 Pact of Paris that followed World War I. The mercurial William C. Bullitt, who staffed President Wilson and Colonel House during the Paris Peace Conference, quit the US delegation and subsequently testified against the Treaty, famously declaring that, “This isn’t a treaty of peace… I can see at least eleven wars in it.” It didn’t take long for Bullitt to be proven right.
2. The provision in the 1922 Washington Naval Treaty prohibiting new or enlarged fortifications in the area east of Singapore and west of Hawaii. This constraint was a corollary to establishing a 5-5-3 ratio of naval power projection among the United States, Great Britain and Japan. Tokyo accepted the short straw if western powers did not build up their infrastructure in the Pacific. This system of constraints did not turn our very well. I’ll return to the inter-war naval treaties in a subsequent post.
3. The 1928 Kellogg-Briand Pact not to use war as an instrument of national policy. Calvin Coolidge’s Secretary of State, Frank Kellogg, won the Nobel Peace prize for this agreement in 1929. Perhaps we should compile another list of regrettable Nobel Peace Prize recipients.
What I find notable about this short list is that all of these picks were negotiated during the inter-war period when high diplomacy sought to paper over hard reality. These agreements sought to constrain conflicts and eminently usable weapons without addressing the underlying reasons for war. They failed for understandable reasons.
Critics of nuclear arms control point to these failures all the time when lambasting treaties like New START or the CTBT. They cannot acknowledge two crucial differences between nuclear treaties and the inter-war accords. First, nuclear arms reduction and testing treaties haven’t papered over political differences; instead, they contributed to positive shifts in relations between well-armed, major powers that sought to avoid war. Second, nuclear weapons, unlike naval surface combatants, are not useful instruments of war.
Hmm… There wasn’t one ‘Pact of Paris’, there were several. And some of them worked out OK. If you’re going to be getting technical, better to single out the specific treaty (Versailles, I imagine, though it lasted far longer than Sevres, and not nearly as long as Trianon and St Germain) and perhaps even point out the objectionable parts.
Might I add Munich, 1938 and Molotov-Ribbentrop, 1939, as examples of confusing buying time vs stability or building a foundation for peace. I think it also shows what can happen with diplomacy when it confuses the tactical here and now vs the strategic long term, and fails to realize that time is not on someone’s side who should see the obvious. However given WWI and the Russian experience in WWI, and the Soviet civil war, their short sightedness should be understood with some context. None the less those diplomatic efforts were disasters.
CTBT Entry into Force provision is the worst, hands down….
Okay? So what nuclear munitions do you think need verification?
The NPT has the unfortunate effect of allowing a nation to maintain a “civil nuclear power industry” that is actually a bomb program holding at T-6 months or so, with sanctions awaiting the next overt step forward. With regard to a single potential breakout case, and presuming immediate and effective sanctions, that’s not so bad.
If we wind up with two potential adversaries each holding at at the six-month line, each understanding that effective sanctions will probably take six months or so to implement, and each hearing rumors that the other crossed the line into breakout last month, this seems more like a recipe for instability.
We managed to avoid an East Asian arms race over the DPRK’s nukes, and that’s a good thing. But then, that is simply to say we consider it a good thing that when North Korea violated and then abandoned the NPT, nobody cared enough to do anything, which is hardly a ringing endorsement of the NPT. Next up, the Arab world’s reaction to Iranian nukes, or Iranian almost-nukes…
It would be a recipe for instability if, in fact, it were a simple matter of a 6-month timeline, or any defined and predictable interval between initiation of breakout activity and, Presto… Bomb… Nuclear blackmail. But that is a caricature or at best a first-order approximation of reality, and in this case the higher-order terms make all the difference.
In reality, the state initiating breakout cannot be confident it will not be caught early, either due to observable activity or an internal leak, nor can it be confident that what may take it 4-8 months to accomplish will not take its adversary less time. The particular details of each party’s technology, and the level of effort it makes or is able to make, matter, and are difficult to assess with any precision. Then there is the question of what preemptive moves the adversary may make once it knows or suspects that you are breaking out. In short, breakout implies taking the full risk of war, and the combined probability of losing the race or descending unprepared into uncontrollable chaos is almost certainly greater than that of winning such a foolish gamble. The Germans and Japanese can tell you something about this (wrt both preemptive rearmament and preemptive attack).
This is a very general principle which applies not only to nuclear weapons but other kinds such as other WMD, space weapons, conventional rearmament, etc. It is likely to become even more salient in the future as we consider preventive arms control regimes in an era of automated production, robotic weapons, etc.
Of course the details are unpredictable. What astounds me is, you say that like it is a good thing. Unpredictability is generally considered a Very Bad Thing when it comes to deterrence, or diplomacy, or avoiding wars in general.
W/re nuclear proliferation, when some local leader asks, “Will our enemies have nuclear weapons when this crisis comes to a head later this year? Will we?”, the
rest of us want the answer to be unambiguous. Ideally an unambiguous “no”, but “yes to both, for sure” leads to stable deterrence, and even “they will, we won’t” will have the diplomats looking for the right accomodation to the new reality. If the answer is “we’re not sure”, the next question starts to look like, “would it help if we kicked the bomb program into high gear right now?”, possibly asked on both sides of the border at the same time.
Wars, and arms races, rarely happen because an aggressor is certain that they are the path to victory. They more often happen because someone is no longer certain that they can avoid defeat by not playing the game. Treaties that create persistent uncertainty regarding existential threats, are not a recipe for peace.
“Wars, and arms races, rarely happen because an aggressor is certain that they are the path to victory. They more often happen because someone is no longer certain that they can avoid defeat by not playing the game.”
That is an interesting thesis, but a very general one, and entirely consistent with my comment.
Your specific, hypothetical example was an arms control regime which, as long as it is observed, enforces a predictable 6-month delay, for either of two adversaries, between initiation of breakout and acquisition of nuclear (or some other “winning”) weapons.
In this scenario, instability results from (unrealistic) certainty as well as (realistic) uncertainty. Either side has an incentive to initiate breakout covertly and bet that the other side won’t catch on until it is too late, given the certainty that a small lead is enough to ensure victory. Uncertainty about whether the other side has already preempted or will preempt is another important ingredient here.
The fact that the certainty is unrealistic is consistent with your observation that “Wars, and arms races, rarely happen because an aggressor is certain that they are the path to victory.” So much for your hypothetical, and anything it might imply.
What about the realistic case in which the exact length of time between initiation of breakout and possession a trump card, whether and how soon the adversary will detect or suspect the breakout, and how the adversary will react, are all uncertain? The safest bet is usually to stay close enough within the terms of the regime so as not to destabilize it and provoke a destabilizing response from the adversary, while remaining constantly watchful and diplomatically engaged with the adversary.
“Unpredictability is generally considered a Very Bad Thing when it comes to deterrence, or diplomacy, or avoiding wars in general.
I do question this; my argument is that the irreducible uncertainty of the outcome of aggressive first moves is an important ingredient of stability in arms control and other regimes of peace. Whether unpredictability is a “good” or “bad” thing in general, it is nevertheless a condition of our existence, within which we must make any strategic calculations; and I do think it can have positive as well as negative implications and uses.
Let me second Kevin (CTBT EIF provision), but add two more: PNEs in the NPT and the verification provisions of the BWC.
The situation following the horror of WWI was one of nations who had spent too much to kill too many for too long.
The Treaty of Paris was a disaster because of it’s use as a weapon of revenge.
The Washington treaty was an attempt to limit military expenses and worked until the Great Depression thrust reactionary governments into power in many nations. By 1935 the withdrawal of Japan and Italy had rendered it useless.
To the extent it led to reduced military spending in major and minor warships by all signatories in the decade after WW I it was a success.
The Kellogg-Briand Pact, which btw is still legally in force, was quite quixotic and ineffective aside from providing the legal basis for prosecutions of “crime against peace.’
It is easy to view the ‘failure’ from hindsight, but ignore the context.
I agree about the Treaty of Paris, but disagree about the other two efforts.
Your criticism of the Washington Naval Treaty seems odd. As you quite rightly point out in your subsequent blog, the objective of this Treaty was to provide each signatory with security in their home waters. Equality of security would remove threats of coercion, and allow peaceful development. The WNT reduced the Japanese fleet to a size that it could never threaten US home waters – and it was quite reasonable for our ally Japan to ask for and recieve provisions to prevent the US fleet from being a threat to Japan in their home waters.
Furthermore US criticism of the non-fortification clause is hard to take seriously when there was never going to be the money to build the fortifications that were sacrificed in the Treaty. As Roosevelt said – the US traded fortifications she would never have built for fortifications that the Japanese would almost certainly have built.
Regards
David