There is an oft-repeated story about the run-up to the Wall Street crash of 1929 in which Joseph Kennedy, JFK’s father, stops to get his shoes cleaned in lower Manhattan, and is shocked to find that the shoeshine boy is offering him tips on which stocks to buy. “If a shoeshine boy can predict where this market is going to go, then it’s no place for a man with a lot of money to lose,” he is reported as saying, before selling his holdings and avoiding the losses others suffered.
Unfortunately, like many appealing stories, this one turns out to be apocryphal. Kennedy does seem to have avoided the worst of the slump by well-timed share sales, but the Nobel laureate Robert Shiller, who has spent much of the last few years researching the stories, true and false, that drive economic events, can find no contemporary reference to his interaction with a shoeshine boy. The first mention of the story does not appear in print until 1957, and then it was attributed to the American financier Bernard Baruch, a friend of Winston Churchill.
We have all read, too, that the Wall Street crash led to a spate of suicides as ruined investors threw themselves off high buildings in despair. Except that it seems they did not. The suicide rate in New York was higher during the summer of 1929 when the market was rising strongly, than it was in the autumn after the crash.
In this country we are now sadly familiar with half-remembered facts and politically-motivated myth-making, whether about Brussels directives on straight bananas or the circumstances in which kippers may or may not be shipped by the Royal Mail. In spite of the worthy work of a number of fact-checking websites, we find it hard to keep up with what is fact and what is fable, where is Ruth and where is Mabel, as Alan Bennett put it. By force of repetition, £350m a week (or a month, or even a day, whatever) sent to Brussels has inflated our thinking about the cost of EU membership and the sums that could otherwise go to fund the NHS. Trenchant letters putting the record straight from the chair of the official statistics watchdog cut no ice at all.
Shiller is not so concerned by such short-term irritants. Brexit does not merit a walk-on part; indeed his research is almost entirely confined to the United States. His aim is higher—to identify the enduring narratives that influence the way we think about the economy, and may influence our patterns of spending and saving, and therefore become self-fulfilling prophecies. To do so, he makes extensive use of two sources: ProQuest, which allows one to word-search newspapers back to the 1700s; and Google’s Ngram Viewer, which enables researchers to look for words and phrases in books back to the 1500s. The results are fascinating, and sometimes startling. The stories widely told about economic events are often fantastical; yet they do not die and recur in uncannily similar ways from decade to decade. Once bitten, twice- or even thrice-bitten is the rule.
Perhaps the most obvious example of a recurring story is what Shiller describes as the narrative of “technological unemployment,” which currently centres on artificial intelligence, machine learning and driverless cars. Ever since the Luddite campaign in Nottingham in 1811, when textile machinery was destroyed by angry workers in fear of their jobs, there have been regular outbreaks of concern, even panic, that automation will remove the need for human labour, leaving millions idle and destitute.
Albert Einstein believed this narrative, asserting in 1933 that the Great Depression was attributable to “the improvement in the apparatus of production through technical invention and organisation [which] has decreased the need for human labour.” For a time, the key word was “robot,” which first appeared in the early 1920s in a play by the Czech playwright Karel Capek. The notion that robots in the home would de-skill householders has never quite become reality, but the sense of that threat has remained.
Since 1945 there have been multiple narratives about the destructive potential of artificial intelligence, as measured by spikes in the incidence of the words “automation” or “AI” in newspaper articles—in the 1960s, the 1980s, the 1990s and 2010s. The narrative of this decade looks likely to peak at the highest rate of mentions, and not only because Ian McEwan’s recent novel Machines Like Me was widely reviewed.
As Shiller notes, “each time, the narrative suggested that the world was only just now reaching a turning point when the machines take over.” Even though there are more people in work in the US (and the UK) than ever before, the narrative has shown no sign of relaxing its grip on the popular imagination, or even on the imagination of today’s Einsteins. (I suppose Michael Gove would put himself in that category.) It may have lain behind many of the fears about immigration from the EU that played an important role in the Brexit referendum. There is only so much work to go around, automation is cutting it further, and foreigners will take the lion’s share by being prepared to work for cut-throat rates of pay. Or so the story goes.
*** A second example is the bubble narrative. Again, this has deep roots, dating back at least to the celebrated tulip mania in 1630s Amsterdam. The dot-com boom in 1990s followed the tulip playbook very closely with a vertiginous rise in prices, unrelated to solid financial data, followed by a sharp correction. The latest version of the story surrounds Bitcoin, Tether and other cryptocurrencies, which incorporate a heady mixture of “get rich quick” attractions, combined with blockchain technology, which is so complicated and costly that it sounds as though it must have some intrinsic value. Add a dash of libertarian anarchism, and you have in Shiller’s words “a narrative that is well crafted for contagion… part bubble story, part mystery story [which] allows non-experts and everyday -people to participate in the narrative… Having a Bitcoin wallet makes the owner a citizen of the world and in some sense psychologically independent of traditional affiliations.” On this analysis, the remarkable cryptocurrency boom seems neither so alien nor so hard to explain. But the related problems are likewise familiar—there have often been dramatic busts, and some not-so rich dreamers have lost their shirts as a result. The fact that they have provided useful material for economic historians may not be adequate compensation.
There are several other recurring narratives mapped by Shiller. Away from “the robots” and the financial markets, he also looks at recurrent stories in the property market, and in repeated waves of popular irritation with greedy workers (when wage push inflation is identified as a culprit) or rapacious capitalists (when income inequality increases). To use a phrase which he sources to a small, Indian marketing company in Nagpur, formed in 1991, whenever these themes and ideas go “viral,” he finds the same pattern as in his Big Data textual crunching—the references rise sharply, then retreat.
But where does this analysis lead, aside from generating a number of very similar frequency distribution graphs which dot the text? Does it matter that mankind is susceptible to these repeated narratives, which promise to explain economic phenomena, yet in practice obscure them?
*** Shiller is careful not to exaggerate the analytical power of the new field of narrative economics, which he has been ploughing for some time. Like all academics worth their tenure, he ends with a plea for more research funding. He believes more analysis of the feedback loops between the stories we tell ourselves and real economic variables would improve our understanding of “the deliberate manipulations and deceptions we have experienced, and will help us formulate economic policies that take narratives into account.” If we could forecast booms and busts more reliably, it could be a great leap forward, not only analytically but also more practically in terms of prevention and cure. But we cannot be sure that such enlightenment would flow from further deep mining of the ProQuest and Ngram databases.
Shiller does, however, give a hint of what he sees as the potential explanatory power of narrative economics. He argues, for example, that the wage-price spiral narrative, which incorporates elements of the greedy unions and rapacious capitalists stories, coming together in an insider conspiracy to extract rents from the broader public, peaked in 1980 and has mostly been in decline ever since. That decline likely has something to do with falling rates of unionisation.
While it ran strongly, the wage-price spiral engendered high expectations of inflation to come, which in turn held up interest rates. Workers tried to bid up their own pay in anticipation of expected price rises. Shiller argues that “the dynamics of this worldwide narrative epidemic likely provide the best explanation for these epochal changes in trend of the two major economic variables, inflation and interest rates.” Faced with the power of this change of trend, central banks may be struggling in vain to push inflation expectations higher by reducing interest rates or engaging in round after round of quantitative easing. They might do better to fund Facebook campaigns featuring imaginary angry shop stewards shouting outrageous demands for unjustified pay rises, which would engender fears of price rises to come.
Another fascinating conclusion is that the American Dream narrative, which recurred several times in the 20th century, especially in the 1950s and 1980s, has re-emerged under President Trump. This time it is in the form of “Make America Great Again,” his campaign rallying cry, printed on millions of baseball caps. But it is more than just a device to elect a populist president. The message justifies people’s desire to purchase expensive cars, extravagant homes and other lavish consumer products and services. If the good times are rolling, why not? The rhetoric itself creates higher consumption and consumer borrowing, which in turn boosts the economy. It is the classic self-fulfilling prophecy.
It does, however, have a less happy offshoot when it comes to generosity: if everything is going well, why waste one’s money helping the poor, when they should be able to help themselves? Shiller points to drops in certain charity donations and deduces that the “declines in the modesty and compassion narratives... extend to a lower willingness to help the world’s emerging countries.” That does seem to be a plausible concomitant of the Trumpian narrative, even if the decline in charitable donations he cites actually occurred before the ostentatious New Yorker arrived in Washington.
There may be a danger, here, of over-egging the narrative pudding. There may be other forces at work. The stories do not emerge fully formed from a manipulative journalist’s pen. They surely require some fertile soil in which to germinate. But Shiller has found a rich seam to mine, one which Brexit campaigners, wittingly or not, worked to some effect. A British edition of Narrative Economics would no doubt be similar to the US version in many ways, but there could be intriguing differences. Perhaps somewhere in the bowels of the LSE there is a team of narrative economic historians struggling to emerge into the sunlight. I hope so.
Narrative Economics: How Stories Go Viral and Drive Major Economic Events by Robert J Shiller (Princeton University Press, £20)
