Introducing Story into Economic History and Policy

Robert Shiller, 2013 Nobel laureate in Economic Sciences, recently published a book with the surprising title Narrative Economics: How Stories Go Viral and Drive Major Economic Events. The intellectual journey that took him there is a fascinating one. Always interested in forecasting, Shiller began his career as an econometrician but soon became skeptical about traditional approaches such as large-scale multiequation models and rational expectations models (Shiller 2013). In his early research, Shiller showed that actual fluctuations in stock prices were much more volatile than would be expected if investors were motivated solely by rational expectations. Subsequently, Shiller went beyond traditional economics to psychology for insight into how economic agents form their expectations and has even used investor surveys as a methodology. More willing than traditional economists to call out economic irrationality, he was one of the few who predicted the bursting of the Internet stock bubble in 2000 and the housing bubble a few years later. Shiller’s purpose in writing Narrative Economics is “to improve people’s ability to anticipate and deal with major economic events, such as depressions, recessions, or secular (that is, long-term) stagnation, by encouraging them to identify and incorporate into their thinking the economic narratives that help to define these events” (Shiller 2019, 71). He feels this is essential because “economic research has not emphasized the stories that people tell to one another and to themselves about their economic lives. … By missing the popular narratives, it also misses possibly valid explanations of major economic changes” (Shiller 2019, 276). Shiller is addressing policy-makers as well as private sector actors and therefore his work should be of interest to public administration scholars. Though Shiller cites some of the voluminous literature on narrative, his focus is on economic forecasting. The definition of narrative he starts with is distinctly casual: According to the Merriam-Webster online dictionary, a narrative is ‘a way of presenting or understanding a situation or set of events that reflects and promotes a particular point of view or set of values.’…. Narratives generally take the form of some recounting of events, whether actual or fictional, though often the specific events described are little bits of color brightening a concept and making it more contagious. (Shiller 2019, 36) Most of the book consists of Shiller’s exploration of the following nine “perennial economic narratives” in the context of both recent and distant American economic history: panic versus confidence; frugality versus conspicuous consumption; gold standard versus bimetallism (with a nod to bitcoin); labor-saving machines that replace many jobs; automation and artificial intelligence that could replace almost all jobs; real estate booms and busts; stock market bubbles; boycotts, profiteers, and evil business; and the wage-price spiral and evil labor unions. Shiller’s version of macroeconomic history is different from the traditional, which focuses on changes in key economic variables, explorations of causal mechanisms, and the decisions of policy-makers. In each chapter, Shiller presents discussions of a given narrative when it went viral, drawn from both media and other sources (books, letters, even sermons). Shiller also incorporates an analytical engine inspired by the mathematical modeling of epidemics. These models show how an epidemic at first gradually but then rapidly spreads through a population, peaks, and then recedes. His hypothesis is that when narratives go viral, awareness spreads in a similar way. Shiller takes words or phrases that express his narratives such as “automation,” “artificial intelligence,” “Great Depression,” and “American dream” and measures the frequency they appear in popular media over time. He employs Google Ngrams to show the percentage of words in all books and Proquest News and Newspaper archives to determine the percentage of all newspaper articles published in a particular year that include his chosen word or phrase. Shiller concludes his book with an ambitious research program that involves interviewing large population samples to hear their stories about their economic decisions, convening regular focus groups to discuss economic narratives, and building historical databases of focus groups’ deliberations, sermons, and personal letters and diaries that deal with economic issues. Shiller is near the end of his career, so this is a research program that his students and subsequent researchers would have to carry out. Though this research program represents a new direction for economists, it is standard practice for other disciplines such as marketing and political science. Marketers and political strategists use in-depth interviews and focus groups to enhance the effectiveness of their marketing and election campaigns. These, of course, are deeply partisan purposes. Economics, however, endeavors to serve the public good and avoid political partisanship. If, as Shiller claims, these research techniques could produce better economic policy decisions, for example, by identifying asset bubbles before they lead to painful crashes, then the question that follows is which institution should be empowered to use them. In the context of the Trump presidency, the only major economic institution that can still claim with some credibility that it is nonpartisan is the Federal Reserve. The Federal Research also has deep pockets to support research such as Shiller’s. The Federal Reserve is a research-oriented bureaucracy that is highly transparent, for example publishing detailed minutes of the Federal Open Market Committee within a month of its meetings. It has been the subject of considerable research by public administration scholars; if it began to use narratives in its decision-making, that development would elicit further public administration research. As a researcher on business and political narrative, I share with Shiller and others in our field the hypotheses that people habitually think in terms of narratives and that narratives evoke powerful emotions that impel action. I am, therefore, sympathetic with Shiller’s intent and approach. Based on my research, in particular the recent book Negotiating Business Narratives (Borins and Herst 2018), I have some suggestions. Shiller’s definition of narrative is decidedly minimalist. When Shiller discusses what he calls narratives, he is actually discussing concepts. He may include occasional events that illustrate the concept, but rarely does he outline sequences of events that would constitute a typical narrative. One of the few sequences of events he presents is that of a typical birthday party culminating in the singing of “happy birthday to you” (Shiller 2019, 97–98). Of course, that narrative is irrelevant to his theme. A second atypicality in Shiller’s economic narratives is that they rarely include individual human protagonists. Although Merriam-Webster and other definitions of narrative, such as that in Wikipedia (2019), include only events, most actual narrative texts (novels, biographies, histories, and documentary and docudrama films) include human protagonists. The reason for employing human protagonists is to provide a focus for audience identification and to facilitate their emotional experience. Consistent with the economics discipline’s focus on social forces, the protagonists in Shiller’s narratives are anonymous. The rare exceptions to Shiller’s anonymous narratives are those that involve celebrities, such as the apocryphal story he cites in which either John D. Rockefeller, Bernard Baruch, or Joseph Kennedy concluded the stock market had peaked in 1929 and decided to sell his portfolio when a shoeshine boy gave him a stock tip (Shiller 2019, 236–7). Shiller appears to attribute agency to celebrities, for example claiming, without any apparent evidence (at least to me), that “the Trump narrative epidemic contributes to the upward turn in home prices in the United States starting after 2012” (Shiller 2019, 225). If Shiller believes that this and other celebrity narratives are widely emulated by economic agents, he should focus on narratives of celebrities engaged in economic transactions (whether actual transactions or product endorsements) and attempt to measure their impact. A third issue concerns the coherence of Shiller’s narratives as a set. Shiller (2019, 15–16) recognizes efforts by literary scholars to enumerate a mutually exclusive and collectively exhaustive set of basic types of plots, genres, or fables. In Negotiating Business Narratives, my coauthor and I attempted something similar, hypothesizing three axes of meaning for business narratives (the individual protagonist, the firm, and the society) and two plot trajectories (success or failure) for each axis. This generates a typology of eight business fables. We found it easy to classify over 60 narrative texts dealing with three major industries (information technology, automobile manufacturing, and financial trading) in terms of that eight-fold typology. Shiller gives no indication that he used a set of principles or concepts to derive his list of nine economic narratives, nor did he indicate whether he thinks that the list of nine includes all possible economic narratives. It would be worth his while to develop a conceptual scheme to classify economic narratives. My fourth concern is about Shiller’s use of Google Ngrams and Proquest News and Newspapers to generate his measures of the virality of economic narratives. The mathematical epidemic model that he takes as inspiration shows epidemics peaking at a substantial percentage of the population. In contrast, his economic narratives constitute less than one percent of all newspaper articles in a given year as measured by Proquest and, at most, one tenth of a percent of all words in books in a given year as measured by Google Ngrams. The mathematical model is, at best, useful in predicting the shape of the diffusion paths for Shiller’s economic narratives. It would be instructive for Shiller to search for other viral narratives that could explain the other 99% of the topics of newspaper articles. This would provide some sense of how salient economic concerns are relative to others such as war and peace, disease and health, and politics. In addition, Shiller’s focus on the virality of narratives assumes that it is the best measure of a narratives influence. The range of possible responses to a viral narrative is likely larger than either adopting or ignoring it and could include partial adoption or strategic reaction to the widespread adoption of a narrative by others.1 Perhaps the interview-based component of Shiller’s research program could shed light on such responses. Robert Shiller, eclectic economist par excellence, has delivered a thought-provoking study making the case that better understanding by policy-makers of the thinking of economic agents could lead to better macroeconomic management. One compelling instance of this would be the identification of asset bubbles and constraint of the forces creating them before they burst and create massive economic and social dislocation. Shiller has relied on ideas taken from the study of narrative to facilitate his study of the thinking of economic agents. But a deeper exploration of those ideas would only strengthen his research and advocacy.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.