Realism vs Pareto efficient outcomes in economics

Realism vs Pareto Optimality

Since the events in Ukraine and Palestine, I have been following John Mearsheimer more closely – a professor of international politics. He is a “realist.” In international politics there are several schools of thought, which I admittedly don’t understand deeply. But the gist, according to Prof. Mearsheimer, is that “realism” holds that what matters in inter-state relations is relative power. That is, what counts is how strong our country is compared to our neighbors. This arises from the belief that above the state, there is no higher institution that will protect it if attacked by another state. Therefore, our country must be stronger than the rest.

The emphasis here is on the word “relative.” In other words, it is fine if we are not all that strong, as long as others are much weaker than us.

This is interesting for economists because mainstream economics typically teaches Pareto Optimality early on. Pareto optimality essentially says that an outcome from exchange/trade/economic activity in general will occur as long as at least one party gains without reducing any other party’s income. This arises from a model where each entity (person or country) acts solely based on its own self-interest.

This yields the argument that if we let everyone optimize their decisions based on personal welfare, the final outcome will be better because it is guaranteed to be Pareto optimal – nobody loses, though some may gain extra income. If there are 100 people acting in self-interest, and 50 see their income rise while the other 50 see no decline, then total income (i.e., GDP) increases.

Note that “considering only oneself” clearly conflicts with the “relative” thinking of Mearsheimer’s realism. Considering only oneself means asking only whether I am getting richer, not whether I am getting richer MORE THAN my neighbor.

Example

Say you and your neighbor each have 10 million rupiah. Your neighbor offers you a project worth 10 million, where they would hire your labor for 1 million, keeping 9 million for themselves as the one who sourced the project. If you agree, you are 1 million richer, while your neighbor is 9 million richer. Would you take it?

The “Pareto optimality” school says this transaction is perfectly fine because you are 1 million richer than you would be without the deal. The “realism” school would say this might not be good, because it leaves you with 11 million while your neighbor has 19 million – nearly twice as rich as you. Before the deal, your wealth was equal; now it is not. You should be envious!

This can be taken further: under Pareto optimality, a person would not make a decision that hurts themselves. But under realism, a person might willingly take a loss, as long as their neighbor loses even more. This is why the US restricts semiconductor capital exports to China. Not only does this policy cost the US money – the US itself loses because it sells resources used to make semiconductors and buys products containing them. But the US accepts its own losses as long as China’s losses are even bigger!

Implications

Economists model optimal decisions based on utility. Pareto optimality says a person’s utility depends only on their own consumption, not their neighbor’s. But realism says we should care about our neighbor’s consumption. Why? Because of externalities. Under realism, if your neighbor gets 10 million richer and you only 1 million, they can buy a rifle while you can only afford a machete. Therefore, we must always stay vigilant and envious of our neighbors.

Returning to the jealous-neighbor scenario, at the individual level, the probability of being invaded by your neighbor is extremely low. This is because there exists an entity above individuals that will protect us if a neighbor does harm. In theory, that entity is the state. That is why we have police.

But at the state level, there is no entity higher than the state that will protect us if another country attacks. That is why, according to Professor Mearsheimer, liberalism and democracy were able to spread everywhere back in the late 80s, 90s, and 2000s. The reason is that the no. 1 entity was the United States. In those days – what Prof. Mearsheimer calls the “unipolar moment” – no entity was more powerful than the US. Everyone feared the US. The US then created NATO. This entity served as protector of its member states. The police of the world was the US.

Now? The US is weakening RELATIVELY to other countries. US GDP as a share of global GDP keeps declining, naturally because other countries (Russia and especially China) keep growing. US manufacturing capability (a proxy for military production capacity) continues to decline. Unipolarity is gradually eroding, and we are entering a “multipolar” moment. The world’s police is no longer as strong as before. In this environment, the realism school will return to the mainstream.

Of course, in economics itself, Pareto optimality has been further analyzed for its weaknesses, and we continue to strengthen models in this space. Pareto optimality is a very simple model, well-suited as a thinking framework for students (similar to simplified load assumptions in civil engineering semester 1). But sometimes I still encounter people using it for serious discussions. Nowadays, incorporating externalities (and other assumption violations) into models is nothing new. If you are an economist, always keep this in mind!

Second, the world is not run (solely) by economists. There will naturally be many political decisions that consider factors beyond economic efficiency. So there is no need to get angry when the government adopts policies that are not economically efficient – there may be other, more important objectives. But it is admittedly strange when a leader claims to prioritize the economy while producing policies that make no economic sense.

Oh, there’s a meatball vendor outside. Let me go buy some meatball soup, bye!

By the way, the part 2 post on realism is here.