The secret of effective market game-playing is to recognize that the market game hinges on the Narrative, on the strength of the public statements that create Common Knowledge.
Epsilon Theory Manifesto
Nobel-winning economist Robert Shiller recently published Narrative Economics, a book on “How Stories Go Viral and Drive Major Economic Events“. Shiller gave a talk at LSE on the big ideas (video, audio, related 2017 paper).
Context: This article is part of the Big Ideas series, where I synthesize takeaways from the world’s best experts in multiple disciplines. This article is a special in the series, because unlike other articles that are synthesized from Discover magazine expert interviews, this piece is largely inspired by a public lecture.
What is a Narrative?
Let’s start with definitions. According to Shiller:
- Narrative = a telling of a story that attaches significance, meaning or emotions to it;
- Story = a chronology of events.
What is Narrative Economics?
Shiller makes a key distinction between narrative economics as defined in the dictionary vs. defined by himself. The textbook definition of narrative economics is “economics research that takes the form of telling a narrative about economic events”.
For Shiller, narrative economics should have a narrower focus, i.e., only investigating popular economics narratives that “went viral”, “changed things” and “became contagious”.
Shiller thinks economics narratives are powerful in affecting (& shaping) economic decisions. He identifies 9 perennial economics narratives:
- Panic vs. confidence narratives – e.g., the Big Depression is a panic narrative;
- Frugality vs. conspicuous consumption – e.g., Trump’s book “Think Like a Billionaire”;
- Monetary standards – e.g., the Gold Standard vs. Bimetallism debate;
- Technical unemployment, i.e., labor-saving machines replace many jobs;
- Automation & AI replace most jobs;
- Real estate booms & busts;
- Stock market bubbles;
- Boycotts, profiteers & evil business;
- The wage-price spiral & evil labor unions.
Broadly speaking, the 9 narratives above focus on the macro economics momentum / “culture” (1-3), employment (4-5), investment (6-7) or actors in power (8-9).
Shiller argues that data sources are at the root of economics evolutions. He believes the recent “digitization of search” is and will bring shifts to narratives. Moreover, Shiller claims that big events occur often not because of a single narrative, but because of a “confluence of narratives“, i.e., as a result of the chemical reaction of multiple narratives.
With an interesting twist, the word “narrative” appears less frequently academic articles in economics & finance compared with other subjects – see this analysis of JSTOR articles below:
Studying Narrative Economics via the Virality Model of Epidemics
If we think of a narrative as a disease, then we could study its spread by borrowing patterns from research on epidemics. In other words, we could leverage research on how viruses “go viral”, and try to figure out how narratives get popular.
The Kermack-McKendrick (1927) mathematical theory of disease epidemics is a breakthrough in medicine, because it “gave a realistic framework for understanding the all-important dynamics of infectious diseases” in the words of Shiller.
The Kermack-McKendrick model divides the population into three groups: susceptibles, infectives, and recovered. Importantly, the model suggests the curve of the number of infectives to take a “humpback” shape, i.e., rising sharply before declining at a similarly fast speed:
We could see similar “humpback” shaped curves in data that could serve as proxy measurements for how popular an economics narrative is.
Here’s an example on how frequent the phrase “stock market crash” appears in news & newspapers:
Here’s another example on how frequent the phrase “Great Depression” appears in news & newspapers:
The Future of Narrative Economics
Shiller is hopeful that ” the advent of big data and of better algorithms of semantic search might bring more credibility to the field”.
Meanwhile, narrative economics faces challenges, including:
- On data collection, we need to move beyond “passive collection of others’ words, towards experiments that reveal meaning and psychological significance”, e.g., via focus groups or social media – though the proper design & implementation of such experiments is not easy;
- Dealing with the overlap & “chemical reactions” of multiple overlapping narratives is difficult;
- Causality is tricky. As Shiller says, one challenge is in “distinguishing between narratives that are associated with economic behavior just because they are reporting on the behavior, and narratives that create changes in economic behavior.”
Nevertheless, the challenges make the field more interesting. I am particularly interested in predicting which narratives will gain momentum. Perhaps the narrative machine will serve, to some extend, as a crystal ball that offers a narrow glimpse into the future.
Fully Booked Club 