Flexibility, slag heaps and bombs

Foreign policy minefields in the era of Trump

I don’t have anything useful to say about what will happen in the Middle East now that the U.S. has dived, eyes closed and head first, into someone else’s war. Other people will have better analyses, though perhaps not on social media. I expect that online discussions are going to be particularly useless for the next while, thanks to the convergence of attention dynamics, lack of specific understanding, and desperation on everyone’s part to understand what is happening in a dire situation. I’ll try not to contribute to that confusion, by not talking in public about things I don’t understand, even if they are the most urgent topic of conversation. Quite possibly, I’ll fail.

What I can talk about, a little, is how Donald Trump’s style of decision making helped get us into this mess. My understanding, for what it is worth, of how Trump makes decisions is that it has one and only one true North Star: making sure that nothing stops Donald Trump from doing whatever he wants to do in the short term. One of the few good predictions I’ve made was was this one, shortly after he was first elected in 2016.

the successful practitioner dominates the public space and public argument as everyone tries to interpret what the hell you have done, paying attention to you and no-one else but you, so that you can continue to play center stage in the theater of politics while everyone else is reduced to Waldorf and Statler, carping from the critics’ box.

If this is right, the key qualities of presidential politics over the next four years will be instability, frequent policy change, palace intrigues, and Trump looking to reign triumphant above it all, not particularly caring (a la Padgett and Ansell’s Cosimo) about attaining specific goals, but instead looking to preserve his position at the center of an ever shifting spider web of political relations, no matter what consequences this has for the integrity of the web.

Trump is not completely unpredictable. For example, he likes to use tariffs as an instrument, and he has a crude set of beliefs about what winning and losing involve. But he is willing to raise tariffs by staggering amounts, or to lower them, as needed to keep attention focused on him. So too for immigration - he has broad desires, but in the last several days he has made two U-turns in rapid succession on imprisoning and deporting farm workers without legal status. For both tariffs and immigration there is a ratchet effect, in which things that get worse tend to stay worse. But that ratchet seems to me to largely be a side effect (perhaps a welcome one for Trump) rather than an intended purpose.

Many of the past constraints on Trump have disappeared. Much of the US foreign policy apparatus, for better or worse, is supposed to ensure consistency, across the different branches of the federal government, so that different agencies and departments don’t trip each other up and across time, so that the U.S. is able to stick to long term goals. But as all the above suggests, Trump sees little value to internal consistency. It constrains him from acting with the flexibility he wants.

It’s no surprise that the National Security Council, which is the part of the US policy apparatus that specializes in ensuring consistency, has gone through so much chaos and upheaval in the first months of Trump’s term. So too, for many other parts of the government apparatus. Every administration is trying to build the plane as it flies. This may be the first administration that is yanking random pieces out of the engine, and chucking them out of the cargo bay in mid air.

Some of the downsides of this are already widely understood. Internal inconsistency across the government means that there are going to be many blunders, and different bits of the government working at cross purposes. Inconsistency over time provides a lot of short term flexibility, but it forecloses strategies that are based on long term predictability.

For example: if you aren’t going to stick to deals over the long run, no-one will want to make deals with you if they have any choice in the matter. And if they do make deals, they will likely be empty of real content, intended to distract rather than to bind. More subtly, you are going to have a hard time deterring others from doing stuff that you don’t want them to do. Deterrence is all about making binding commitments ex ante to do things that might be costly for you ex post. If you set a red line, that someone might cross over, you deter them by making it clear that you will punish them for it, even if delivering the punishment is painful for you. It will be hard for you to convince others that your threat is credible, if your policy style is all about bluster and endless revision.

I think that there are two other, subtler ways in which policies based on boundless flexibility are self defeating. First, unless you are extremely canny, you are likely to lose your flexibility over time, by creating facts on the ground. This is what Dan Davies calls the “slate heap” problem. The side-effects of your actions tend to heap up into giant piles of slag. The Trump administration’s decision to bomb Iran’s nuclear sites is likely to trap it into further actions that it doesn’t want to take. Second is that others may create facts on the ground for you, dumping slag in places that will force you to go in one direction rather than another. That helps explain how Netanyahu succeeded in making Trump do something that he’d likely have preferred not to do.

The slate heap problem is an argument about how side effects accumulate. As Dan describes the operation of slate quarries:

At the very start of a quarry operation, it must seem like it hardly matters what you do with the waste material; the site is wide open and there isn’t enough spoil to interfere with anything. As time goes on, the heaps get more noticeable … a while after that, they have become bigger … they’re increasingly an important constraint on the business. Finally, you reach the point at which you have to accept that you have huge mountains of unstable rock, which can collapse without warning if blasting work is carried out nearby and which can’t be moved; … the whole quarry is now too dangerous to operate.

As Dan points out, solving this is an inherently complex problem - you simply cann’t know at the beginning of a complex project how things are going to look at the end. Equally, some approaches will be much worse than others. Specifically, if you build your entire policy process around a principal who always demands maximal flexibility in the moment, without any strategic vision for how to preserve flexibility over time, you are going to find it really hard to avoid massive unstable heapings of side effects eventually hemming you in from all sides. Indeed, you will have a hard time even seeing the slag heaps piled up around you, until they have started to collapse.

Bombing Iran is piling up great heaps of unexpected consequences that will merge with each other and possibly fall on America’s head. Again, I’m not an expert on the Middle East, but there are many intersecting piles of likely consequences - retaliation from Iran, other governments rushing into the power vacuum, breakdown of national authority, spreading instability, and uncertain energy markets for starters. Dealing with any or all of these will require further actions, which may in turn set off further cascades of unexpected consequences, requiring further action and so on. The road to sticky foreign policy quagmires is paved with temporizing immediate decisions.

Perhaps this will all shake out into a better world, but that is not what I’d bet on myself. Given the evisceration of the National Security Council, and the incredibly short time between Israel’s initial bombing, and the decision of the U.S. to dive in, I’m prepared to bet a lot of money that there hasn’t been much serious analysis or wargaming of what might happen after the bombing. That in turn means that the U.S. will be less able to anticipate when things start to go pear-shaped, or to react usefully if they do. Current actions will pile upon past ones and the responses of others, until the whole thing possibly topples over.

This is, very obviously, not the product of accident. Netanyahu clearly anticipated that the U.S. would be pulled in if he started a war. In a sense, he built his own slag pile, calculating (or gambling) that it was placed so as to force the U.S. to go in one direction rather than another.

That reflects another downside of flexibility. If a country has institutions that oblige it to be consistent, it will miss out on some short term opportunities. Equally, it will be better able to resist some forms of external manipulation, committing in advance that it will behave according to some predictable logic regardless of what others do. If the U.S. had had a clear Iran policy, it might have been better able to refuse Netanyahu. That, in turn, might have made Netanyahu less likely to take a risky gamble on unilaterally starting a war that Israel could not succeed in winning on its own.

All this may possibly have broader lessons. In a world where the U.S. prizes consistency and is at least somewhat willing to live up to its commitments, it will lose out on some short term benefits. Equally, it will be able to withstand some efforts by soi-disant allies to pull it into conflicts that are not in its interests. It will be able to wargame out some (far from all) of the likely consequences of big dangerous actions.

In a world where the U.S. does not value commitment, few others will truly bargain with it, although they may go through the motions of pretending to. What is the point of making deals with a country that may renege unexpectedly on the president’s momentary whim. Instead of negotiation, they may try other means of constraining the U.S, creating facts on the ground that it has to respond to, in the hope that the response will point in the right direction.

The latter world seems to me to be worse for America and worse for everyone else. It is the world we are now living in, and it continues to go down hill.