This review should begin with one important caveat. Although the title of the volume under scrutiny may suggest otherwise, this is not a history book. Nobel Memorial Prize laureate Robert Shiller, who wrote it, mentions historians and historical references on several occasions and repeatedly attempts to reconstruct past economic events historically, but overall this is an essay in economic theory, the main purpose of which is to propose a new research agenda in behavioral macroeconomics. More precisely, this is a follow-up on one idea Shiller introduced several years ago in the book he co-wrote with another Nobel Prize recipient, George Akerlof, Animal Spirits (Reference Akerlof and Shiller2009). In that book, Shiller and his co-author had suggested that among several other factors, narratives may help explain economic downturns. More precisely, they argued that narratives dissemination was part of the herding behaviors leading to economic bubbles during phases of economic expansion, eventually leading to the crash that follows. Having used the Akerlof and Shiller book for pedagogical purposes, I had found that part quite the weakest aspect of an overall convincing argument. One of my problems was that the notion of “narrative” seemed poorly defined or delineated, an issue that was not helped by the fact that I had read it in French and that in that edition narrative had been translated by histoire, a word that alternately means “story” or “history.” Therefore, what I expected from the present book was some sort of conceptual clarification. Unfortunately, my reading leads me to conclude that my reservations were not solely due to mere translation issues. But before I address these reservations, let me say more about the book’s stated objectives, its structure, and main assertions.

“Narrative economics,” Shiller tells us, is an old expression that, in the late nineteenth century, was mentioned in Palgrave’s Dictionary of Political Economy as a method of presenting one own’s narrative of historical events. What Shiller calls “narrative economics,” however, is a completely different matter: it is the use of narratives as a predictor of potentially damaging economic events. In his preface, the Nobel Prize theorist makes it clear that his purpose is not just to explain economic behavior using narratives but to provide “better forecast of major economic events” (p. xiii), the anticipation and prevention of which should be considered as a “moral imperative” (p. xv). Shiller situates this endeavor in the lineage of John Maynard Keynes but also of Tjalling Koopmans and Kenneth Boulding—he quotes the latter for arguing that “economics creates the world it is investigating” (p. xvi). In contrast, he invokes as one of the main proponents of the opposite thesis Irving Kristol, who once wrote as a criticism of Keynesianism that we should ignore what economic agents have to say and consider only what they do. The reference to Kristol, the American godfather of conservatism, is probably not incidental. Most of the stories Shiller holds responsible for creating speculative bubbles or causing major economic damages are conservative narratives, from those advocating savings and frugality during the first years of the Great Depression to those, like the Laffer curve narrative, that led to the tax cuts of the 1980s. Shiller’s central claim, which is not stated explicitly before we reach page 71, is that there is a dual causality between narratives and events. While new narratives drive major economic events, the latter favor the dissemination, reinforcement, and at times mutation of those narratives. Shiller makes that claim in Part II of his book, as he attempts to further elaborate on the notion of narrative, which he has introduced through several examples during Part I. Having done so, Shiller comes up with seven propositions or principles that rule the emergence and dissemination of narratives. Part III, the longest section, reconstructs a number of economic narratives, which Shiller says recur regularly throughout economic history. Part IV develops a program for future narrative economics research.

At this stage, the reader of this review might be surprised that I have made no effort to properly define what a narrative is. There is a reason for this. In fact, in spite of all the references to narratives in the book, it is quite hard to understand what Shiller really means when using this term. The problem is not that he does not provide us with a definition but that in fact he gives several, seemingly self-contradictory, of them and creates even more confusion by occasionally bringing up facts and situations that occasionally deviate from the notion he has previously introduced. First, the subtitle of the book, How Stories Go Viral and Drive Major Economic Events, seems to imply that “narratives” and “stories” are used quite interchangeably. The blurb included inside the dust jacket teases that the book is about “popular stories,” yet it also mentions “internet troll farms” that “attempt to influence foreign elections.” To me, those troll farms do not look much like “popular stories” spreading out quite spontaneously like diseases but rather like carefully planned attempts to distort reality in order to undermine democracy. Now, I am aware that dust jacket texts are most often produced by the publisher’s marketing department, not by the author. The problem is that, in this specific case, the blurb is perfectly indicative of the confusion reigning throughout the book. In the preface, Shiller indeed tells us that “the word narrative is often synonymous with story” but immediately qualifies this by mentioning the Oxford English Dictionary definition of “narratives”: “a story or representation used to give an explanatory or justificatory account of a society, period, etc.” (p. xi). From this, we should infer that Shiller is mostly going to talk about “stories with meaning” and that, as the Oxford definition seems to imply, this meaning is most often intentional. But then Shiller adds that “a story may also be a song, joke, theory, explanation or plan that has emotional resonance” and that history itself can be understood as “a succession of rare big events in which a story goes viral” (p. ix). With a definition this large, there is fear that the author may encompass many things and quickly lose the thread of his argument. This fear is often substantiated in the rest of the text. In Chapter 4, which is supposed to offer further conceptual clarification and evidence of the virality of narratives, Shiller writes of narratives as “stories that seem to have entertainment value only” (p. 31). Those narratives, he argues, are part of the elements that differentiate us from other animals. He then asserts that our species could well be called homo narrator rather than homo sapiens, but then he follows this remark with a short digression on the role of music on the human mind, suggesting through the work of musicologist Anthony Newcomb that “the classical symphony” is a sort of “composed novel” (p. 35). Another sixty pages later, while dealing with how narratives can mutate and become contagious, the author is providing us with the example of the song “Happy Birthday,” which derived from the song “Good Morning to All” and began to become viral in many languages from the 1890s onwards and throughout the twentieth century. Shiller admits that “this song is probably not an important economic narrative” and that “[s]ome might say it is not even a real narrative because the words of the song do not tell a story.” Yet, he adds, “there is a story attached to the song in everyone’s consciousness” (p. 97). This made me lose the plot: Are we talking about narratives, popular stories, the virality of songs as analogies for the spread of those, or the “emotional resonance” that popular songs provoke in people? There are many passages like this in the book that feel a bit digressive and conceptually confusing.

Shiller’s definitional imprecision brings another issue: that of intentionality. While Shiller tells us about many examples of spreading narratives throughout the book, he seems to conflate those who develop out of sheer coincidence or unplanned opportunity with those that have been consciously designed for maximum virality and, at times, nuisance. When he evokes the spread of “fake news,” he justifies their existence by the tendency of storytellers to “provid[e] amplified tales” and to “spread titillating stories without making any clear moral decision whether they are spreading falsehoods or not.” Shiller even makes an analogy with the development of “fake wrestling matches” in the late nineteenth century (p. 84). Fake news, like the spread of narratives in general, is in this depiction nothing but the product of our frivolous, quite irrational minds, not that of a concerted attempt to influence events by a group of people. Where does that bring us? In the short chapter Shiller devotes to the Laffer curve, the author rightly attributes the origin of the curve’s narrative to conservative columnist Jude Wanniski’s 1978 book, The Way the World Works. Wanniski was indeed responsible for that colorful story of the curve allegedly drawn by Arthur Laffer for two White House officials, possibly Dick Cheney and Donald Rumsfeld. Yet, while Shiller seems to imply that the curve’s virality is mostly due to the vivid character of Wanniski’s story, he undermines the political background behind it. Wanniski was not a simple Wall Street Journal columnist, coming up with a good story to tell his readers, but an active member of conservative and neoliberal think tanks. The publication of his 1978 book was supported by Irving Kristol—yes, that same Irving Kristol whom Shiller is quoting in the preface—whose journal, the Public Interest, had published a couple of articles out of which Wanniski’s extended essay grew. Wanniski and Kristol subsequently used their influence networks to convert Reagan’s economic staff to supply-side economics. Supply-side economics spread and its origin in the Laffer curve spread as well in the press, most of the time in articles that derided both the idea and the curve that went with it. Then, further events explain why the curve began to be investigated more academically in serious economics outlets (Giraud Reference Giraud, Coopmans, Vertesi, Lynch and Woolgar2014). Therefore, to attribute the Laffer curve epidemic solely to the intrinsic appeal of Wanniski’s original story and to the general public’s tendency to appreciate this sort of narrative is quite doubtable.

The issue I just raised is, I think, related to a deeper problem in Shiller’s framework: the fact that in spite of an alleged purpose to unite knowledge and to borrow from history and the social sciences, his worldview is firmly grounded in the natural sciences’ reductionist methodology. This is well illustrated in Chapter 2, which I must say irritated me so much that I briefly considered to stop my reading there and not write the present review. Shiller, in this chapter, explains to us that he believes in the unity of science and that through this work, he wants to initiate a multidisciplinary approach that borrows as much from the humanities as from the natural sciences, which looks promising. However, the word he uses for this, consilience, which he borrows from William Whewell and Edward O. Wilson, does not bring much doubt about the sort of knowledge unity Shiller is favoring: that which can be attained only when the humanities conform to the natural sciences rather than the other way around. The rest of the chapter confirms my foreboding. Shiller attempts to compare the occurrence of the word “narratives” in various social sciences and in the humanities, using a graph to do so. The word “narrative” appears to be used quite sparsely in economics and finance but used more frequently—and increasingly so—in anthropology, history, political science, psychology, and sociology. For the period 2010 to 2019, it appears in 30% of history articles on JSTOR. Yet what Shiller fails to realize is that the meaning of the word "narrative" differs across all of these disciplines and that it is also subject to variations within them. Shiller, however, decides to ignore this issue and instead takes a quick tour of these disciplines and of how they make use of the concept. I will focus only on history here to show how far off he is in understanding the field. Shiller writes that “[h]istorians have always displayed an appreciation for narratives” and argues that “a deep understanding of history requires inferring what was on the minds of the very people who made history” (p. 14), mentioning historians’ efforts to include popular stories into their accounts. While he is right to argue that in recent years historians have made a concerted effort to give more consideration to their protagonists’ own historical reconstructions, I think Shiller misses the elephant in the corridor. Historians do not just “display an appreciation for narratives,” they also produce narratives, which probably explains in part why the word occurs so much in their academic output. Narratives are both the raw material from which historians write and their end products. During the second half of the twentieth century, historians have become increasingly aware that narratives are not “things” but constructs and that this is true both for the stories they collect from various sources and for those they produce using them. Historical constructivism has led historians like Michel Foucault and Paul Veyne to take a firm stand against the “philosophers’ histories,” these historical reconstructions that attempt to bring together historical facts as a means of theorizing, and instead of offering stories that explain, they advocated stories that merely narrate (Veyne Reference Veyne1984). Taking this stand implies that the narratives produced by historians, though often detailed and grounded in archival research, are not by essence “better” than the histories people tell about themselves. One historian of economics with an actual “appreciation for narratives” and a remarkable awareness of their constructed, communities-dependent character is Evelyn Forget, who has written at length on the subject (e.g., Forget Reference Forget2005).

It is pretty clear that Shiller is uninterested in this type of constructivist methodology. After his quick, reductive, tour of the humanities and the social sciences in this chapter, he turns to the discipline that really interests him: epidemiology. I have not addressed that point so far but it is pretty clear from the title of the book and from the wording I have used that the epidemiological analogies are everywhere in Narrative Economics. Narratives and popular stories are viral or contagious, they spread and mutate. Shiller spends several chapters—the bulk of Part 2, actually—to show how it works. Of course, reductionism is not a problem per se. I myself tend to like to have my humanities and social sciences quite constructivist and thick, and my economics reasonably reductionistic. Yet, this becomes a problem here when Shiller introduces references taken from the more constructivist social sciences and humanities and attempts to impose his reductionistic views unto them. Let me take one brief example. In Chapter 4, Shiller evokes the work of sociologist Francesca Polletta, who studied students’ protests against racial discrimination in the 1960s. He uses her work to illustrate that “sociologists have begun to think of the contagion of narratives as central to social change” (p. 32). Yet, in fact, Polletta does not use epidemiological analogies in her work as a way to explain the spread of students’ protest. On the contrary, she deals with the social construction of narratives and challenges the very idea that protests spread “like a fever.” She writes that “although the sit-ins and the groups they galvanized were not without prior organizational affiliation, their narrative construction as spontaneous was central to an emerging collective identity” (Polletta Reference Polletta1998, p. 153). In her conclusion, she emphasizes the need to grasp “not only the formal features of narrative that distinguish it from other discursive forms, but the social conditions in which stories are more and less familiar, easily communicated, and authoritative” (1998, p. 155). Shiller’s use of Polletta’s work not only displays his misinterpretation of her actual intent but also illustrates the gap that exists between the methodology of the recent social sciences and humanities and his own reductionist endeavor. Likewise, the conceptual frameworks used by historians of science and STS scholars to study knowledge dissemination—for instance, “translation,” “trading zones,” or “performativity”—are absent although some of them would have been quite relevant to Shiller’s endeavor. The lack of reference to performativity is particularly puzzling because, first, the author, a specialist of financial markets, is necessarily aware of the vast sociological literature on the subject, and, second, his allusion to Boulding’s claim that “economics creates the world it is investigating” (see above) should have led him to resort to that notion—or at least explain why he prefers not to do it.

Reflexivity, which is certainly not the most common characteristic among economists, is definitely in short supply here. Shiller does not seem to realize that, while his book is concerned with studying narratives, it is also filled with narratives of his own. The virus analogy is of course one of them. Comparing narratives to viruses brings a normative perspective. Viruses are bad and provoke diseases, just as narratives provoke economic downturns. Humanities scholars have long studied the narratives involved with epidemics, which include metaphors of social disorder, guilt, and punition (see, for instance, Sontag Reference Sontag1989, and more recently works in medical anthropology such as Leach and Tadros Reference Leach and Tadros2014). Therefore, epidemiology is anything but a neutral explanatory device when it comes to reflect on narratives dissemination. The other big narrative in Shiller’s book is that of the social scientist like himself who is able to stand above the crowd of storytellers, internet trolls, and passive economic agents who spread contagious and harmful narratives, most of the time quite unknowingly. This narrative is told throughout Part 3, which tells stories about depressions, technological changes, “profiteers,” and “evil” labor unions. These are the recurring, perennial narratives, which in Shiller’s perspective display the qualities of the most viral stories, as they “recur,” “mutate,” and form “constellations” of stories that reinforce each other. Most precisely, these narratives are those that entertain some sort of causal relationship with the “biggest economic events in the United States as defined by the NBER” (p. 111). Nine of the latter are mentioned—from the depression of 1857 to 1859 to the “Great Recession” of 2007 to 2009. Accordingly, Shiller identifies nine “perennial” narratives, chosen because of their mutation and recurrence features. Yet Shiller does not say much about the actual process of selection. Also, no justification is provided as to why the number of narratives should correspond to the number of previously identified “major events.” Therefore, we are tempted to believe that these categories were just made up. In addition, the substance of those chapters does not really match the author’s stated objectives of using narratives as a predictor of future economic events. Shiller’s stories are told mostly verbally, using the epidemiological concepts he has brought forth in Part 2 only sparsely, and the only quantitative elements are graphs plotted using Google Ngrams with sources from processed databases such as ProQuest News and Newspapers, Google Books, or Google Trends. These historical reconstructions are informative and entertaining from a general public’s perspective but seem to display little academic expertise. Shiller is quite aware about this as, in the last section of the book, he devises a program for future research, which should include better efforts at quantifying narratives using not only historical databases but also interviews and focus groups. While this is very promising, this also makes the book look more like a giant grant application and not a full-fledged treatment of the subject under scrutiny. And while the research agenda discussed in the concluding chapter is promising, Shiller’s unfortunate habit of using social science and humanities knowledge as the fatty sauce that helps one to swallow an otherwise reductionist pill makes me strongly doubt that his plea for more interdisciplinary work between the natural and the human sciences is genuine and that it will specifically help foster collaboration between economists and historians.