Donald Trump, accidental master of game theory

Unlock the White House Watch newsletter for free

Your guide to what the 2024 US election means for Washington and the world

It’s Donald Trump’s world now and the rest of us are just living in it. Or perhaps that’s not putting it strongly enough: we’re living not just in his world but in his head. Still nearly two months away from actually taking office, the US president-elect has been catching the rest of the world (including seasoned US politicos) off guard with one gobsmackingly unconventional cabinet nomination after another.

Trump’s nominations, as well as his not-always-coherent policy statements in the past, have caused a spate of questioning by well-meaning analysts. Does Trump really think (now-withdrawn) Matt Gaetz and Tulsi Gabbard could do a good job in the positions he picked them for? Or are these provocations designed to separate those among his followers who retain an independent mind from those who will kiss the boss’s ring no matter the outrage? Or is the intention to provoke a stand-off in Congress that would let Trump invoke the never-used power to suspend the legislature?

Similarly for policy questions. What is Trump’s plan on Ukraine? Is he really going to sell out Kyiv to get some unknown benefit in trade from Russian President Vladimir Putin? And what about tariffs: will they be put in place to strengthen US bargaining power with other economies, or does Trump just want to have them regardless of any concessions others may grant?

I have little idea how to answer these questions with confidence. As my colleague Ed Luce puts it, on a whole lot of big issues, “Trump could go either way”. But I believe it’s unwise to assume that there is a secret intention to discern. Instead, the uncertainty is the intention.

There is a plausible view that Trump’s behaviour is no more than what it looks like on the surface — sheer policymaking incompetence by an impetuous, mediocre business executive. But this misses out the ways in which unpredictability serves Trump’s interests, and that Trump knows this.

At a basic level, autocrats (actual and wannabes) and gangster bosses thrive on fear and reject the rule of law (in the broadest possible sense of a system of rules that applies consistently to everyone). As seasoned Brexit commentator Chris Grey explains in a thoughtful post, the unpredictability based on a leader’s personal caprice is central to their way of running things — a feature, not a bug.

It’s in this light, I think, that we should see the new “government efficiency” programme headed by Elon Musk and Vivek Ramaswamy. Their made-to-sound-sensible Wall Street Journal op-ed says their imminent culling of the civil service “won’t target specific employees” and that those fired “deserve to be treated with respect”. But Musk has already unleashed the ex-Twitter mobs on named civil servants who can now suddenly find themselves exposed to online abuse, and Ramaswamy has proposed entirely arbitrary firing by odd or even social security number digits, as the Journal reports. Again, unpredictability is the intention here.

There is a branch of economics that makes perfect sense of this. Game theory studies strategic behaviour: situations where the best thing you can do depends on other people’s actions which, in turn, they choose depending on what they think you may do. A basic concept in game theory is that of a “mixed strategy”: choosing randomly between possible actions (not all actions need to be chosen with the same probability).

A mixed, ie randomised, strategy can be better for you than any non-random one (ie predictable behaviour) because it prevents your opponent from moving the “game” in a direction that is better for them and worse for you. It can even be the only approach that stabilises a situation where the interests of different “players” are badly aligned.

Think of sports: the most secure way to prevent your opponent from predicting where you will aim your penalty kick or serve from having observed the patterns of your past behaviour is to literally randomise — choose the direction based on whether the stadium clock’s second counter is odd or even when you glance at it, say. Similarly, creating uncertainty — the “fog of war” — even where there isn’t any to begin with, is often advantageous.

Putin and the Russian spy service understand this well. Putin is an “emotional man”, some analysts say — and while I have no problem believing that he truly and passionately believes in the historic destiny of Russia that he put out before assaulting Ukraine, the point here is that even if he doesn’t, he has an interest in being seen as non-rational. This is true at a systemic level as well: the point of misinformation campaigns is not so much to make people believe a particular falsehood as to make them unwilling to believe that anything can be true at all, in a kind of ultimate ontological uncertainty.

True randomness is hard to manufacture. So it helps to be seen as a little bit mad, or a little bit out of control. Indeed, the beautifully named “trembling hand equilibrium” is a game-theoretic concept showing how optimal strategies must treat others’ behaviour as constantly subject to random perturbation: the finger that pushes the button could always tremble. (In my opinion, the name illustrates how game theory is by far the most creative branch of economics.)

But back to Trump. If you agree to see him as an accidental but instinctive expert game theorist, what lessons should you draw? I can think of two: one about him, one for the rest of the world.

One is that even he is not without self-limitation. He seems to carve out financial markets (unlike banks, as his business career shows) from the who-knows-what-he’ll-do madness. Both in his previous term and this time around, he has appointed Treasury secretaries who can be relied on to run the Treasury in fairly conventional ways: Steven Mnuchin then, Scott Bessent now. Strikingly, Mnuchin even drove (together with France’s finance minister Bruno Le Maire) the multilateral process to fundamentally reform global corporate taxation. As for Bessent, do read Shahin Vallée’s argument that he may just be the man to pull off a repeat of the 1980s Plaza Accord on realigning global currencies.

The lesson for others is that it’s unproductive to spend much effort trying to divine Trump’s true intention behind the chaos. What should you do instead? In the game-theoretic models, the answer will often be to randomise over your own options. But countries that hold on to the rule of law value the trust and efficiency they get from predictability, so beyond some narrow negotiations (in trade, perhaps) that hardly works for Europe.

A better answer must be to inure yourself to Trump’s game playing. In terms of attitude, that means what Grey calls “playing it long and cool” and avoiding jumpiness. In practice, it means the harder task of identifying actions to which it doesn’t matter so much what the US does. For example: Europe should work to redirect its export surplus to domestic investments in its own goals, as I argued last week.

The important thing is to lucidly choose the things that you need to achieve and can achieve without counting on US support. From there, just plough on.

Other readables