Tian Xuan, Associate Dean and Chair Professor of Finance, PBCSF, Tsinghua University
On the afternoon of October 11 (Beijing time), the Royal Swedish Academy of Sciences announced the 2021 Nobel Prize in Economics. The prize was jointly awarded to Joshua Angrist and Guido Imbens, "for their methodological contributions to the analysis of causal relationships”, and David Card, "for his empirical contributions to labor economics."
Although the contributions of the 2021 Nobel Laureates seem to be different (Card for his empirical contributions to labor economics; Angrist and Imbens for their methodological contributions to the analysis of causal relationships), they are essentially the same: to test causal relationships by natural experiments.
Why causal relationships?
There are always people questioning Nobel Prizes in Economics. Some argue that, "the Nobel Prize in Economics is obsessed with the empirical and experimental study that is far from reality. It shows none emphasis on theoretical innovation, with little remarkable contribution achieved."
However, what the researches have set off is a "credibility theory revolution" that will lead to a change in social practice paradigms. Practice is to test the truth. The core societal issues can only be properly solved on the premise that causal relationships are made clear. Hence, the significant contribution of the 2021 Nobel Prize in Economics.
What we need is not conjectural conclusions, but constructive methodology. With the development of modern statistical techniques, especially statistical software, it has been increasingly convenient to find the relationship between two variables. However, the "endogenic" variable interference has been a stubborn issue in economic studies, especially in econometrics, where the research subjects are mostly conducted by focusing on how to address and mitigate the endogeneity of economic data.
To eliminate the endogenous interferences, economists need to use techniques and tools to set a plausible sample of "what if this event had not happened" and then compare it with real data.
This is exactly what this year's Nobel Prize emphasizes on. Generally speaking, the logic of the achievements is: if we want to prove that x causes y, and not vice versa, we should identify the related and "unexpected" exogenous shock(s) in x, and see how the y changes as the result of the random perturbation, from which the causal relationship between x and y could be studied and tested.
Now, let's forget the terminologies, and try to understand the concept with an example:
Wine drinking = Successful Career?
We may find such phenomenon in our daily life: some people love to drink exquisite wines, and they also have successful careers. If you simply put "wine drinking" and "successful career" together, you will find that the two variables seem to be positively correlated.
But can we conclude that "it is because they love drinking exquisite wines, so they possess successful careers"? Obviously, we cannot draw such a simple and rough conclusion.
The reason is simple. Let's assume that the person who loves drinking wines and succeeds in his career is A. The first possible assumption is: as A has a successful career, his wealth allows him to buy exquisite wine to drink. In other words, it is not A's love for wine that leads to his successful career, but the other way around. A's love for wine is triggered by his successful career. In this case, the previous conclusion of "the successful career is because of the love for wine" has reversed the relationship between cause and effect.
There is another assumption: the wine drinking A is probably an outgoing and enthusiastic person. Hence, A is more likely to succeed because of his outgoing, enthusiastic and more aggressive personality. The second possible assumption is: because of his outgoing and enthusiastic personality, A has a more successful career, which has nothing to do with A's love for wine. In other words, A's love for wine is just a representation of his outgoing and enthusiastic personality. In this case, the previous conclusion of "the successful career is because of the love for wine" has a missing variable.
This example is to give you a general understanding of the difficulties in establishing causal relationships. So, is there a way to accurately identify the relationship between "wine drinking" and "successful career" , while avoiding problems, such as "reversed causal relationships" and "missing variables"?
If we use the research methodology to solve the problems in the above case, we can try the following:
We find some exogenous or highly random impacts on wine drinkers that keep them continue to drink while others stop drinking wines. In this "quasi-natural experiment", the experimental group and the control group are compared in terms of their future career success (e.g., taking income as a simple measure of career success). Then, an accurate causal relationship between "wine drinking" and "successful career" could be identified.
Taking the “unexpected factor" as a new approach, the Nobel laureates apply their perfectly resourceful experimental and calculative methods to set up a distinctive benchmark to explore causal relationships and the underlying profound economic principles.
Every time the Nobel Prize in Economics is announced, people are keen to associate it with the current economic problems, which inevitably leads to the conclusion that the prize seems to be increasingly far away from pragmatism, and fewer disruptive original contributions. The basic logical flaw therein is that the Nobel Prize is never awarded to solve the economic problems of the day, not even during the financial crises.
The significance of the Nobel Prize in Economics is that it advances theoretical research to help address the core issues that persist in any society. It gives in-depth enlightenment on the boundaries of human knowledge.
There are fewer talent insights and original contributions in economic theories than in the past few decades. Geniuses like Samuelson, who has made significant contributions, is long to be seen. However, this does not mean the decline of economic research in the social significance, but rather the maturity of economic theories.
When economic research increasingly focuses on the micro level and the present issues, the Nobel Prize in Economics is changing from theoretical to empirical research. The research is generally divided into two categories: theoretical research that aims to portray economic and social phenomena through mathematical modeling and derivations; and empirical research that describes economic phenomena and reveals the laws of economic operation by collecting economic and social data. For example, the empirical research of causal relationships of the Nobel Prize in Economics this year allows us to infer from numerous data that x and y are causally related and establish the relevant causal relationship through scientific and rigorous natural experiments, supplemented by some instrumental variables. In this way, we are able to better verify and understand the theories and logic of the economic and business society. Another example is the research of the 2019 Nobel Prize in Economics, which uses field study to develop solutions to poverty alleviation and is essentially an experimental approach to development economics (alleviating global poverty).
The change from theoretical to empirical research, does not mean the Nobel Prize in Economics is getting far from pragmatism, but serving the society in a more robust and profound way.
(This article is based on The Beijing News' interview with Tian Xuan)