Today I’ll speak at an event run by the Stanford Economics Association. They asked me to talk about (micro)economic theory and doing research on it. You can read my slides here and my notes below. They draw from my posts on economic models and how we judge them.
What do theorists do?
If you read an economic theory paper, you will probably see a lot of math. Modern theorists are mathematicians. They write papers full of theorems and proofs. They use math as a common language for expressing ideas, making precise claims, and explaining their arguments clearly and concisely.
Economic theorists are also scientists. They use models to run experiments that are too hard or expensive to run in reality. This helps us design real-world systems, as Jackson (2019) argues:
One would never design a large airliner without carefully modeling its aeronautic properties, and testing it thoroughly via simulations and test flights of prototypes, before loading it with passengers. Why should designing a market for health insurance be any different? Models have the virtue of offering us insight in to what should we expect in scenarios that have never been tried before.
But the best theorists are more than mathematicians and scientists. They’re storytellers. They start their papers by setting the scene: a world in which people make choices. Next, they introduce characters: who they are, what they want, what they know, and what they can do. Then the theorist lets their story unfold. They take us on a journey from premise to conclusion. Along the way they share details about the characters’ world. Good theorists, like good storytellers, know which details matter to the story and which do not. They share the right details at the right time so that we don’t get lost or overwhelmed. They help us discover the moral of their story: the lessons we should take away; the meaning of all the math.
Two seminal stories
Akerlof (1970) tells a classic story. It’s set in the market for used cars. There are two characters: a seller (he) and a buyer (she). They want to transact at fair prices. The seller knows more about cars’ qualities than the buyer.
Akerlof’s story unfolds as follows: The buyer knows she knows less than the seller. She expects him to rip her off. So she rejects all his offers to sell at high prices. But if he can only sell at low prices, then the seller will only sell low-quality cars. Thus, in equilibrium, only “lemons” get sold. Whereas if the buyer knew cars’ qualities, then she could buy “peaches” at a high prices and lemons at low prices. And that’s the moral of Akerlof’s story: information asymmetries lead to market failures.
Kamenica and Gentzkow (2011)—hereafter “KG”—tell another story about information. Theirs plays out in a courtroom. Again, there are two characters: a prosecutor (he) and a judge (she). The prosecutor wants the judge to convict a defendant. The judge wants to convict if and only if the defendant is guilty. Neither knows if the defendant is guilty. The prosecutor can share evidence selectively, but the judge knows how evidence is selected. KG show that although the prosecutor has transparent motives that differ from the judge’s, he gets what he wants more often than not. Indeed, in KG’s leading example, the judge convicts 60% of defendants despite knowing 70% are innocent!
One moral of KG’s story is that people can “design” information to get what they want. Another is that people gain credibility by committing. In KG’s model, the prosecutor publicly commits to a method for selecting evidence. This contrasts with “cheap talk” models that lack such commitment. In those models, communication breaks down when the sender and receiver have different goals. This is because the sender has persuasion temptations that make him untrustworthy. That was the moral of Crawford and Sobel’s (1982) story. Kamenica and Gentzkow retell that story, but they start from a different premise and so lead to a different conclusion.
What makes a good story?
Good theory papers are like good stories. They describe and explore worlds similar to our own, and give us insights we can apply to our everyday lives. Akerlof and KG’s insight is that information matters. They share that insight by telling stories that are relatable: we can imagine ourselves in a car yard or courtroom, or in other situations where different people know different things.
Good theory papers, like good stories, also have an element of surprise. They tell us something we didn’t expect or know before. Akerlof surprised a top journal’s referees: they rejected his story because they couldn’t believe it. Kamenica and Gentzkow surprised me: I didn’t think people could be persuaded when knew they were being persuaded.
Finally, good theory papers are simple yet general. They are “simple” in that they contain caricatures of reality, highlighting a few features and ignoring the rest. They are “general” in that the highlighted features capture something fundamental and transcendent. Of course, such features are hard to find. That’s what makes doing theory so challenging. But it’s also what makes doing theory so rewarding. I love drawing connections between ideas and distilling them to their essence. If you love that too, then I recommend becoming a theorist!
Thanks to Bing Liu for reading a draft version of this post.