Why Sanctions Against Russia Are Falling Short

In new research, Sandeep Baliga, a professor of managerial economics and decision sciences at Kellogg, and his colleague Tomas Sjöström of Rutger’s University, apply game theory to this very question.

Kellogg Insight recently spoke with Baliga about his research, how it can be applied to sanctions against Russia, and what game theory suggests about what should happen in the next phase of the conflict.

The interview has been edited for length and clarity.

*

Kellogg Insight: How did you possibly have this paper on the optimal way to think through Russian sanctions ready in March, just days after Russia’s invasion of Ukraine?

Sandeep Baliga: We didn’t! It’s too complicated for it to have been done at the start of this conflict. I started it in 2020 and stuck it on my webpage in the summer of 2021. And then this conflict happened. The contents were quite pertinent even though we had written it earlier, because the same fundamental issues we study have been there forever.

Insight: Okay, so let’s set the stage for your research here. The premise is that you are a nation, and you are trying to get another nation to cooperate with your demands using sanctions.

Baliga: Right. And basically, there are two instruments you can use. You can either target the leadership to try and get them to change their behavior, or you can get citizens to put pressure on the leadership to get them to change their behavior.

Historically, we’ve mainly used the latter: comprehensive sanctions, where you are hitting the country as a whole, saying, “Look, can you rise up and overthrow this leader?” Or because the leader thinks the people are going to rise up and overthrow him, the leader changes his behavior automatically.

But there’s a big question of whether those work. And the question is impossible to answer in a scientific way because you have to know what would have happened without the sanctions. Take South Africa: Was it the sanctions that made them end apartheid? Comprehensive sanctions certainly don’t seem to work in a lot of cases, as with Kim Jong Un in North Korea or Saddam Hussein in Iraq. People looked at Iraq in particular and said, “Look we got nothing out of it: all we did was hurt a bunch of people who are innocent, so let’s switch to targeted sanctions.” So then the U.S. started just targeting sanctions to leadership.

But you have the same problem with targeted sanctions, where you can’t tell whether they work or not. So Tomas Sjöström and I wondered: Can we study this in some other way that does not just rely on a few historical anecdotes?

Insight: So you turned to game theory, and you’re trying to map out these different scenarios: how is this political leader or regime going to respond in the face of comprehensive versus targeted sanctions, as well as how is the actual population going to respond, in terms of whether they pressure the leader.

What struck me as unexpected was that, according to your model, you do not want to go all in on both targeted and comprehensive sanctions at the same time.

Baliga: Right. For comprehensive sanctions to work, you want to reduce targeted sanctions. You do not want to maximize both if you’re trying to make the citizens rise up.

There is a “rally around the flag” effect, where people become more sympathetic to a leader who has been attacked by an outside force. And the best example for this actually comes from World War II, when the Germans bombed Buckingham Palace. Because of this, support for the monarchy shot up among citizens in Britain. A newer example is the Brighton hotel bombing, where the IRA attempted to kill Prime Minister Margaret Thatcher and her cabinet. Support for Margaret Thatcher went up.

Now obviously bombs aren’t sanctions, but both fit into this framework where someone is trying to decide whether to attack the leadership or the population in order to persuade a nation to cooperate with your demands. And you have to realize that when you attack the leadership, the population may become more supportive. So you probably want to reduce the sanctions targeted toward leadership if you want comprehensive sanctions to work.

Now, on the other hand, if you just want to use targeted sanctions, then whether you add comprehensive sanctions comes down to considerations about the welfare of the country’s citizens. Adding comprehensive sanctions will hurt citizens, but it won’t necessarily matter in terms of getting the leader’s cooperation.

Insight: Right. So how well does this apply to Russia’s invasion of Ukraine?

Baliga: Well, the subtleties in our model may be less important because it is unlikely that the Russian people could rise up and replace Putin. In a democracy, comprehensive sanctions could be effective, but in what’s essentially an autocracy, they really can’t be. Years of comprehensive sanctions didn’t persuade people to overthrow Kim Jong Un.

So, in my opinion, they’re not going to work here. So all you have left is targeted sanctions.

Insight: So a question that I had—and this is the maybe the challenge of taking game theory and trying to apply it to a specific situation in the real world—but is there really a big distinction between a targeted and a comprehensive sanction? I’m thinking about, for instance, kicking Russia out of the global banking system. That is affecting everyday Russians, but it is also affecting the elite.

Baliga: It’s certainly true that in messy reality, sanctions can be a mixture of things. But the example you gave actually is more comprehensive than targeted. Because all the elite have money stored away, so the banking constraint won’t really affect them.

Now take something like sanctioning oil and gas exports. Yes, that’s going to hit regular people, because there are some people who work in oil and gas, and they’re not going to have jobs. But it definitely hits the elite, because the elite owns these companies. So it’s targeted: it’s just not being done effectively yet, because the EU is still importing oil and gas.

Insight: Given that Russia’s invasion happened after you’d developed your model, did the conflict cause you to revise your conclusions in any way?

Baliga: I’m still thinking about that. But what I will say is it has made me consider what happens next. And here I’m moving beyond sanctions, because, by this point, the sanctions are already in place.

But Georgy Egorov and I have been discussing the next step, a security guarantee that would help Ukraine if it is ever invaded again. And I’m starting to read things in the press about what this guarantee might look like—and my impression is that some of the things that policymakers are thinking about aren’t going to work. Because they replicate the errors of the past.

We offered Ukraine security guarantees in ’94 when it gave up its nuclear weapons. The UK, the U.S., and actually Russia said, “Oh we’ll guarantee your security,” and the net effect is that Russia has invaded, and the UK and the U.S. didn’t fully implement the earlier agreement in 2014 when Crimea was invaded.

What Putin learned from that is that we’re not going to help. And then he went even further, and even now we’ve had a hands-off approach because we do not want the conflict to turn nuclear.

So the upshot is this: you probably can’t have a nuclear country saying we’re going to provide security guarantees, because it hasn’t worked in the past. And for the same reasons that it didn’t work in the past, it’s not going to work in the future.

Insight: Ah, so you think that a non-nuclear power should serve this kind of guarantee instead.

Baliga: Right. Because if a nuclear power is scared of the conflict escalating into a nuclear war, then they can provide the guarantee but never act on it.