I’m going to start this with a confession.

I don’t know much about economics. It’s a blind spot for me.

And perhaps because of that ignorance, I’ve always viewed economics with suspicion. My main brushes with the so-called “dismal science” involve seeing reports in the mass media that refer to nameless economists making dire pronouncements about markets and investor confidence. I’ve always had a lot of discomfort with how the perspectives of those economists get elevated above all other discourse. “It’s the economy, stupid,” after all.

But I think one of the things that perturbs me the most about economics is how it’s often mistaken for science. It adorns itself in charts, graphs and numbers. But, it’s talking to me about things that, to me, are profoundly unscientific, or at least difficult to quantify and measure in a meaningful way. Economics—or at least the version that trickles down to my field of vision—makes a lot of forecasts and predictions about things like human behaviour that rarely follow a rational path. I don’t understand how to square that in my mind.

I feel a bit more comfortable with behavioural economics, which at least acknowledges the fundamental illogic that drives human decision-making. People, to borrow Richard Thaler’s phrase, are “predictably irrational.” And so, behavioural economics (rightly, I think) introduces human psychology into the conversation. It acknowledges that, at the end of the day, we’re dealing with human beings who are by and large messy constructs of innumerable independent variables and that don’t lend themselves to a mechanistic model of understanding.

The humanities: more real than you think

But I still think it’s long overdue that people who study human constructs like economies start to consider the contributions that fields in the humanities can make. I suspect the issue starts with a fundamental misunderstanding of what people studying the humanities are up to these days.

I have a PhD in English literature. To some people, that probably means I sat around a lot talking about poetry and grammar. But that’s a pretty naive mischaracterization. Yes, English majors and graduate students study metaphors and language. Yes, they think a lot about stories and how they’re told. But that doesn’t necessarily mean that their work is detached from anything real.

Literary studies today is deeply concerned with how stories, metaphors, and language itself reflect and affect our relationships with the world around us. As a field, it’s understood for a long time that the way people talk about things—the stories they choose to tell, and how they tell them—structure how we understand the in which we live. To pick a personal example, when I was a grad student I studied how Americans in the middle of the nineteenth century used the language of banditry to organize how they understood their invasion and occupation of Mexico. I was looking at the circulation of popular stories narratives, but those ideas had—and continue to have—a profound impact on the lives of real people

Summarizing Narrative Economics





All this preamble is to say that I was deeply interested when Robert Shiller’s book Narrative Economics: How Stories go Viral and Drive Major Economic Events popped up in my recommendations on Amazon. The blurb promised that Shiller would explore how “studying popular stories that affect individual and collective economic behaviour … has the potential to vastly improve our ability to predict, prepare for, and lessen the damage of financial crises, recessions, depressions, and other major economic events.” In other words, Shiller, a Nobel Prize-winning economist!—was taking seriously the role of story and narrative from the perspective of a discipline that in the past had, from my vantage point, been somewhat wilfully ignorant.

Shiller’s concept of “narrative economics” describes the study of contagion-like stories that influence economic behaviour, such as how, why, and where to invest; how much to spend and save; and whether to buy a home or take a certain job. He challenges the supposition that causality runs from the economic event to narrative rather than the other way around. In Shiller’s own words, “new contagious narratives cause economic events, and economic events cause changed narratives.”

Narratives, Shiller points out, mix emotion, human interest, and seemingly extraneous details but nevertheless form a strong impression in the human mind. When these stories are strong enough—when they are virulent enough—they can have a significant effect on the economy. Shiller provides many examples; for instance, he talks about how narratives in the latter half of the twentieth century that associated free markets with “efficiency” helped turn public sentiment against regulation and government intervention in the economy.

Stories as contagions

Shiller reads these trends through one operative metaphor: infection. He observes that the spread of economic narratives follows a similar pattern to that of contagious diseases. There’s an initial spike in the number of “infected” people who spread the story around, creating an all-too-familiar “hump”-shaped pattern of rise and fall over the course of months, years, or even decades.

These narratives are difficult to trace or predict. At any given time there are many conflicting narratives that inform economic decision-making. That some narratives spread quickly while others do not has nothing to do with quality, importance, or accuracy so much as random details and coincidence. The spread of some may be accelerated by super-spreaders, including the media or algorithm-driven marketing or social media channels.

Over time, the curve flattens and spread diminishes as people forget and lose interest in the narrative. However, a narrative is almost impossible to eradicate completely; resurgences are always possible. Details may mutate, but the underlying fundamentals persist. For example, Shiller argues that narratives about bitcoin recall nineteenth-century narratives about bimetallism. Similarly, contemporary narratives about machine learning and automation share DNA with stories circulated by Luddite movements and in stories about “cybernation” in the 1960s.

It’s interesting stuff. I found the parallels that Shiller was able to draw between different narratives quite compelling. In particularly, I appreciated his analysis of shifts in language used to describe the stock market after the crash of 1929.

But after reading the book, I found myself underwhelmed. None of this struck me as terribly groundbreaking. Much of it seemed obvious.

Shiller’s blind spot

And I felt that the argument itself was flawed. Shiller seems to be stuck between two modes of thought. On the one hand, he wants to bring an understanding of narrative and story to economics. That seems to me to be a very humanistic endeavour, and a necessary intervention into a field that sometimes takes too seriously its commitment to its pseudoscientific trappings.

But Shiller’s efforts toward that are compromised, I think, by his use of virality as his operative metaphor. It’s a bit of a bifurcated thesis: Shiller seems to want to embrace narratology, but he’s uncomfortable with its messiness and ambiguity. So he pairs it with the language of epidemiology. I’m not sure he’s able to cross the bridge between the two, though.

It’s too bad. Shiller encouragingly talks about the idea of “consilience,” a term coined by William Wheelwell in the nineteenth century and adopted by Shiller to talk about the unity of knowledge across academic disciplines. (As an aside, I find it much more pleasing to my ear than more recent neologisms like “multi-” or “interdisciplinarity.”)

But just as I have a blind spot when it comes to economics, Shiller has a pretty huge blind spot when it comes to the discipline that could best inform his study of narrative: literary studies. He’s talking about story and metaphor, but his alarmingly brief survey of the lit department is confined to a very particular moment that is represented, to Shiller, by people like Vladimir Propp and (amazingly) John G. Cawelti. These two represent a profoundly narrow perspective on literature that’s been out of vogue in English departments for probably forty or fifty years. This isn’t frustrating because it’s an affront to academic fashion; it’s frustrating because so much of the work done since Propp and Cawelti’s time is directly germane to the kinds of phenomena that Shiller’s interested in.

The meaning of metaphor

That brings us to the other fault with Shiller’s book. Stories aren’t diseases; but, that’s the metaphor Shiller uses to understand “narrative economics.” The implications of that choice aren’t thoroughly examined. What does it mean to introduce this concept through that particular metaphor? What does aligning narratives with illness communicate to his readers? It’s a question that’s all the more pertinent today, when an epidemic has become a structuring narrative in all of our lives.

These aren’t questions that Shiller answers. He provides plenty of nGram graphs showing the recurrence of phrases over time, and compares them to similar graphs charting the progression of different infections throughout history. He picks up a number of stories that circulated in popular culture and offers a kind of loose, surface-level literary analysis to explain their lineage and speculate on their “virulence.” But unfortunately, it’s the sort of analysis that I think could have been done much more thoroughly had Shiller thought to venture a bit further down the hall of the humanities wing at Yale to talk to the folks working on literature and cultural studies.

I think that even while fields like economics have begun to accept that human beings are not rational creatures, that we’re not “meat machines” that process inputs and produce outputs, there remains a profound discomfort with things that can’t be easily quantified or measured in a lab. Ambiguity is muddy. Numbers are (on the surface, at least) clear.

So while I appreciate Narrative Economics for its willingness to acknowledge the power of story, Shiller’s retreat to scientific metaphors is telling, and a huge missed opportunity.

There’s already mounting evidence that data and numbers aren’t the hard, objective facts we sometimes like to pretend they are. When we select which data to collect, we’re making subjective decisions and, yes, constructing new narratives about what’s important to us and what’s not. We’re seeing that supposedly neutral, unfeeling mechanisms like machine learning and algorithms can actually compound human bias. Garbage in, garbage out.

If it takes work like Shiller’s to help us grow acceptance of alternative, more humanistic frames of knowledge, so be it. But hopefully it’s just the first of many steps in that direction. There’s much that a humanistic perspective has to offer.