First, some theory.
Many countries promise to undertake some future action if some condition comes to pass. Sometimes, they are simple. If Iran ceases its nuclear program, then the West and others will lift sanctions. At times, these promises come as threats: If India attacks Pakistan, then Pakistan may launch nuclear reprisals. At other times, they look like more complicated and diffuse relationships: If the United States maintains a cooperative posture toward the rest of the world, then the rest of the world will be more likely to follow U.S. leadership. A major problem for countries trying to use expectations about the future to shape the present is that it can be difficult to make these promises or threats credible. Would the United States really have engaged in nuclear war if the Soviet Union had invaded Western Europe? Estimates about whether this promise is credible shape how countries behave in the present.
Now, some history.
For academics, these questions are, well, academic. For policymakers and others, they are immediate and their resolutions can’t wait for peer review. De Gaulle was incentivized to get an estimate of what the United States would do in a conflict right, and he thought it unlikely that Washington would trade Peoria for Paris, leading to a French nuclear deterrent independent of the United States; other Western European countries either trusted the United States more or had more constraints on their actions, and so they acted as if they thought yes. The United States, for its part, tried to make the Soviet Union believe that this was credible (and note that projecting credibility can be different from actually being credible), and so it employs a complex strategy of troop deployments and signals about what it would do should some hypothetical situation be realized. Indeed, a great deal of the Cold War in the Global North was about these forms of hypothetical conflicts and arranging signals so that Moscow and Washington would recognize, understand, and respond to those signals in the right way. (In the Global South, meanwhile, tens of millions of people died in the hotter, bloodier conflicts that the Cold War spawned or encompassed.)
And now, how theory and history collide.
In this situation, there was a large market for theories that purported to explain how to make signals more credible and, especially, how to make credible threats more efficiently (that is, more cheaply). After all, maintaining large troop deployments abroad is expensive, and if the same or substantially the same effect can be generated for less cost, then it’s worth investigating that idea. (Of course, if one seriously believes the premises of the Cold War as seen from Washington and elsewhere, the costs of being wrong would be pretty large, maybe even larger, and so there were also incentives to be conservative about this.)
One outgrowth of this demand, as a group of historians explore in How Reason Almost Lost its Mind, was the arrival of economists and others who supplied a variety of theories addressing simplified versions of these scenarios. This is ecology where the stock, canonical games of undergrad lectures—Prisoner’s Dilemma, Stag Hunt, and so on—came from. And it is also where many of the explanations for actual U.S. policy decisions arrived from.
Actual U.S. policy regarding strategy, nuclear policy, and so on was shaped by titanic forces: politics, bureaucratic interests, interservice rivalries, actual competition with the Soviet Union, allied demands, and so on. To make sense of these collisions, academics offered either thick descriptions of what was happening or simple models that kind-of explained what happened. The simple models sold pretty well, and now this era looks like a golden age of relevance to some who wish that political scientists had more policy prowess. It’s not clear how much academic scribblers really mattered at the moment, but it is clear that over time those theories became accepted as justifications and even prescriptions for what happened—in the same way that, say, the nuclear triad, originally an accident and a compromise, eventually became enshrined as doctrine.
One such example of this, I think, was the doctrine of tripwire forces. The basic idea, as the Nobel memorial prize winner Tom Schelling wrote in a few places, was that having a small number of troops in places like West Berlin made threats to go nuclear more credible. Should those troops be attacked and die, then there would be a wave of support calling for retaliation and escalation—and knowing this, the adversary would be less likely to attack in the first place. Deterrence would work, and relatively on the cheap.
Now, it’s a little hard to see what Schelling thought would happen. In his central exposition of how this attack would generate a response, his writing, never terribly specific, became unusually vague. Perhaps that’s because it’s in a section of his arguments about “threats that leave something to chance”, and black-boxing the mechanisms served his purpose. Perhaps it’s because—for other reasons—he actually knew this claim was on weak ground, and maybe he was even just offering cover for his political allies. We’ll find out some day. But the most straightforward account of this argument holds that the reaction will be sparked by public outcry that will demand that leaders respond under penalty of removal from office for being weak or failing to retaliate. That anticipated situation should thus make the whole contraption work and thus make credible these deterrent pledges.
Now, let’s move forward to the present. Does this mechanism work?
Researchers would love to know. In cognate situations, particle physicists can simulate high-energy conditions, for instance at the beginning of the Big Bang, by smashing particles together or otherwise artificially creating such situations with similar properties. Political scientists, fortunately, can’t quite simulate crises with nuclear stakes. But we can use experiments and surveys creatively. And given the stakes of these questions, it’s important to do so, because what if the cheap and easy version of deterrence doesn’t actually work?
In a new article in Foreign Policy Analysis, which you should be able to read from that link, Steve Ward (who bears no responsibility for this post!) and I look at the question of whether tripwires work. We find that tripwires are much less important to our survey respondents than the strongest versions of theory would imply. Indeed, tripwires are less important not only than the costs of a subsequent conflict or the likelihood that the United States would win, they are also less important in swaying public opinion than whether there’s a formal alliance with the attacked country the United States would be acting to defend. In other words, it looks like the pressures that such an attack would generate automatically are not particularly intense. Credibility may not be quite as cheap and easy as it appears!
We are very happy the article is out. I think it supplies valuable evidence to a now decades-old discussion. It’s not the last word (no, seriously, it’s not even the last word from our research project) but it is a challenge to raise the bar over a key element of what countries do and are doing to make their most important promises credible.
If you’re really interested, you can hear more about what we think on the Power Problems podcast.
Surveys at least begin to take seriously that nations aren't abstract entities who somehow immaculately act on self-interest or some other principle, but instead act through a series of institutional decisions, deliberations, policies, etc at different levels of the state, sometimes in contradictory ways. (As per How Reason Almost Lost Its Mind, etc.)
But to me the real test is what we can't see (and can't iterate, because official and civic actors aren't experimental test subjects), which is the in situ citation of various doctrines that are articulated in part to try and influence or direct state and interstate action--both the formal mention of something like "tripwire doctrine" or "MAD" in an actual conversation that is actually informing a contemplated or ongoing action or practice--and more folkloric discussion that is effectively synonymous with the formal idea even if the person speaking doesn't know about it. (e.g., sometimes we formalize an idea in scholarship that has been around in informal and folksonomic ways for a long time). It's not even enough to have transcripts or after-meeting documents or other kinds of writing that serve as evidence about what was said, because we've all been in meetings where the after-meeting archive somewhat fictionalizes what was said, or where the after-meeting archive excludes all the sidebar conversations and lunchtime whispers that were far more determinative, or where key participants spoke informally before the formal meeting and came to an understanding, or spoke informally after and essentially agreed to ignore what was said formally, etc. E.g, an ethnography of how decisions happen and how decisions inform (or don't inform) official action is what we need; it also what we will almost certainly never have.
I'd have to check the source, though I recall seeing this in a prior discussion and in figures presented at the Acquisition Research Symposium that for *U.S.* *operational* expenses, are actually lower for troops abroad than troops at home. So as a practical matter, bringing troops home doesn't save *DoD* money.
That said, I suspect this is true because of offsets by host countries plus sunk cost capital expenses, so more expensive in the big picture probably holds. It just is a meaningful caveat for U.S. budget battles.