Ukraine and the economics of world policing

In the past decade or so, the United States has started to pull back from its role as “the world’s policeman”. Should we regard it as a positive or a negative development? Do we need world policing at all?

Consider the following problem: Russia is, according to certain intelligence reports, planning to invade Ukraine. Suppose “the Russians” expect a positive net benefit b from the invasion. On the other hand, if they invade, they may face a sanction (f), either a military or “economic” one, multiplied with the probability of “apprehension” (p). In the case of an overt invasion, we could make the assumption that the probability of detection and hence apprehension is 1, so that Russia invades if and only if b>f. Notice that I assume risk-neutrality in this very simple version of the model, which, however, does not influence of what I am about to say.

Why would we want to punish an aggressor? The easiest way to answer this question is to ask what would happen if we didn’t. We could reason from the fact that political leaders, like everyone else, respond to incentives, be they monetary or nonmonetary. Then with f=0, aggression occurs whenever b is greater than zero. But suppose we don’t “intrinsically” care about the well-being of Ukrainians. Why then should the West care if Ukraine is invaded? I think the reason we should care even if we aren’t otherwise concerned about Ukraine is that absent a security guarantee, the Ukrainians themselves would try to deter aggression coming from Russia or other countries. This would potentially involve developing nuclear weapons or other weapons of mass destruction. Indeed, the 1994 Budapest Memorandum established a security guarantee for Ukraine in exchange of Ukraine forgoing the option to develop nuclear weapons. Also, it likely wouldn’t stop with Ukraine. If it is known that there is no “world policeman” to punish aggressors, many countries around the Globe would find it in their interests to heavily arm themselves, leading to nuclear proliferation and likely instability. Furthermore, planning invasions and spending resources on avoiding being invaded is a wasteful activity as it does not increase total well-being. Most countries benefit if they could spend fewer resources on planning and defending against attacks.

Thus some of the “free riding” by countries on the world policeman may be “optimal”; that is, it may be better to have one superpower, alliance or organization protecting countries around the world than having many countries engaging in an arms-race with each other. However, world policing is still plagued by moral hazard issues. Countries that are threatened by invasions will rationally be less precautious in their behaviour if they expect the world policeman to protect them from would-be aggressors. Possible solutions to this problem could be the world policeman policing not only aggressors but also potential “victims”. For instance, the United States often intervenes into Israel’s affairs, partly in order to deal with this moral hazard problem. The world policeman can also expect countries to share some of the burden of protecting them from aggression, as suggested by Presidents Obama and Trump.

What determines the optimal f? First, suppose that the aggressor is a literal “madman”. This may imply that a given f will have a much lower deterrent effect in his or her case. Against such actors preventive measures, including the now somewhat discredited notion of preventive war, may be more appropriate than ex post retaliation. Second, consider the cost of the intervention. Third, we need to examine what is at stake in the given situation. For instance, if Ukraine is invaded, how much would Ukrainians suffer? How much would the stability of the region deteriorate? Thus the appropriate f might vary from case to case, which however, creates a problem: unlike in the case of criminal law, where f is partly determined by law, in world policing f is selected in an almost fully discretionary manner. This creates a risk of unpredictability and misunderstandings. If, for instance NATO or the United States select a low f in the case of Ukraine, does this reflect a low value placed on the Ukraine crisis, beliefs about Vladimir Putin, the high social costs of the intervention, or simply a lack of political support for such interventions? In the case the latter is a plausible cause, a relatively unprotected Ukraine could make Taiwan and other countries rather nervous, which could lead to all the problems we previously discussed. Thus in world policing, some kind of rule or commitment (see also the work on Thomas Schelling on commitment in international relations) is needed to uphold a peaceful World Order. On the other hand, there is also an obvious case for prioritizing threats as resources that can be used for world policing are scarce. Prioritizing should preferably be transparent enough to achieve the desired goals. At this point we could smuggle p back into our framework, as uncertainty about priorities might occur even in the case of relatively straightforward communication. As p goes below 1, it may be counterbalanced by a greater f. Of course, one may go so far as to suggest a very low chosen p combined with a high chosen f which would be attractive under risk-neutrality as it would save policing resources (see the related discussion in Gary Becker’s classic 1968 paper on crime and punishment). On the other hand, the importance of marginal deterrence would point toward the “punishment fitting the crime”. For instance, threatening to punish every aggression with say a nuclear strike would obviously be a bad idea: first, such a threat would be rightly seen as “empty”, and second, if deterrence were not achieved, the aggressor would have an incentive to commit an even bigger aggression as he or she could already expect to receive the highest punishment even for lesser “crimes”.

To sum up, world policing is a complicated business. Nevertheless, it would be bad if the US pulled back completely from its role as a world policeman. At the same time, the US is right to expect protected countries to share in the financial burden of world policing. Preferably this, and not a retreat to isolationism, will be the future of world policing. I myself believe that the US won’t choose a completely isolationist path: a stable and peaceful world is obviously beneficial to one of the world’s richest and largest countires, which is one reason why it has taken on a world policing role in the first place.