Misbehaving: The Making of Behavioral Economics by Richard H. Thaler is a book about the research process. It discusses how he and other behavioral economists came to their place in life, what mistakes they made along the way, and what lessons they learned from those mistakes.
This article discusses nine of those lessons for readers who are interested in learning more about the experience of being an academic or scientist.
1. Originality is overrated.
Thaler makes this argument in his discussion of the development of prospect theory, one of the cornerstones of behavioral economics. He and collaborator Amos Tversky developed their version of expected utility theory, which was initially rejected by mainstream economists who dominated economic journals at the time. However, subsequent research by others showed that these economists had their own biases and that the version of the expected utility theory they preferred ended up being wrong.
Thaler notes that there are many examples of landmark ideas in economics emerging simultaneously or nearly so, including general equilibrium theory (developed by Arrow, Debreu, and others), game theory (Osborne and Solow), rational expectations (Muth), and the efficient markets hypothesis (Fama).
2. Borrow other people’s tools
As Thaler points out, academic economics is a highly formalized field, where graduate students are expected to master complex mathematical modeling very early in their careers. Consequently, many of those who choose an academic career path struggle with applying these techniques and formulating their ideas in a formal, rigorous way.
Thaler suggests that one can skip this step and still be a productive academic by borrowing the techniques developed by other researchers and applying them to different problems. He notes that economics journals have long been filled with examples of recycling another researcher’s idea into a new journal article, including papers he had written in the past.
3. Take advantage of your ignorance (for formal models)
Thaler notes that the mainstream initially rejected many academic ideas in economics because they did not fit into formal mathematical models that dominated journals and university departments at the time. These included behavioural phenomena such as prospect theory, strategies for implementing efficient markets based on irrational investor beliefs, and macroeconomic phenomena such as time-inconsistent preferences.
As Thaler points out, he benefited from this state of affairs because he developed his way of thinking about these topics that did not fit into the existing journals. He also benefited because graduate students interested in these ideas could find employment at universities that hired non-mainstream economists.
4. Start with the real world; don’t make assumptions if you don’t have to
This lesson follows directly after “Take advantage of your ignorance (for formal models).” As Thaler points out, many economic topics do not require complicated formal modelling and can be studied by looking at what occurs in the real world. For example, he notes that many of the papers from his dissertation research on how people make choices do not use any mathematics to explain their ideas.
5. If you have a good idea, try to replicate it before moving on
This lesson follows directly after “Start with the real world; don’t make assumptions if you don’t have to.” Unfortunately, it is hard for any economic scholar to replicate their research, as they have a natural bias in favour of their work. As a result, many replication efforts are not reported until years later and only after the original researcher has moved on to other topics.
This can be problematic because it can take a long time before the original researcher eventually admits that their work does not replicate, and even more time before they acknowledge their mistakes.
6. Avoid collecting data just because you can
As Thaler points out, one of the attractions of running experiments is the ability to generate many observations of people or systems without investing much effort into such data collection. However, he notes that this also means that academics are predisposed to focusing on whatever data is available, regardless of how good it is.
As Thaler points out, economists should try to collect the best possible data before studying any phenomena. But, unfortunately, academic incentives may push researchers towards using whatever data is available first, rather than spending time and effort collecting high-quality data.
7. Let other academics know what you are doing
This lesson follows directly after “Avoid collecting data just because you can.” As Thaler notes, scholars need to communicate their findings so that others can test them against their experiments or observations, identify any mistakes in the original work, and build on them to make discoveries.
8. Decide what you want to test before looking at the data
This lesson follows directly after “Let other academics know what you are doing.” As Thaler notes, this is a good rule-of-thumb because it forces researchers to be specific about their predictions and hypotheses upfront, making it easier to avoid falling into ad hoc explanations when experimental results do not match expectations.
9. Try harder to disprove your theories
This lesson follows directly after “Decide what you want to test before looking at the data.” As Thaler notes, most scientists try to produce results supporting their hypotheses, but economists should do the opposite. He notes that it is always easier to explain events after they have occurred than predict them before they occur.
Every scientific finding should be viewed with suspicion and tested against alternative explanations deriving from other hypotheses or theories. As Thaler notes, a theory can be considered particularly strong if it requires unusual conditions to succeed.
Key Takeaways: The Making of Behavioral Economics by Richard H. Thaler
The following points summarize the lessons in this article:
- Do not rely too much on formal models.
- Start with the real world.
- Avoid making assumptions if possible.
- Try to replicate research before moving on.
- Do not collect data just because you can.
- Communicate your findings with other academics early and often.
- Decide what you want to test before looking at the data.
- Try harder to disprove your theories.