The Technology Trap Read online

  THE TECHNOLOGY TRAP

  THE TECHNOLOGY TRAP

  CAPITAL, LABOR, AND POWER IN THE AGE OF AUTOMATION

  CARL BENEDIKT FREY

  PRINCETON UNIVERSITY PRESS

  PRINCETON & OXFORD

  Copyright © 2019 by Princeton University Press

  Published by Princeton University Press

  41 William Street, Princeton, New Jersey 08540

  6 Oxford Street, Woodstock, Oxfordshire OX20 1TR

  press.princeton.edu

  All Rights Reserved

  Library of Congress Control Number 2018966069

  ISBN 978-0-691-17279-8

  eISBN 9780691191959

  Version 1.0

  British Library Cataloging-in-Publication Data is available

  Jacket art: Diego Rivera, Detroit Industry Murals (1932–33), detail from north wall, fresco © 2018 Banco de México Diego Rivera Frida Kahlo Museums Trust, Mexico, D.F. / Artists Rights Society (ARS), New York / Alamy

  For Sophie, with love

  CONTENTS

  Preface ix

  Introduction 1

  PART I. THE GREAT STAGNATION 29

    1  A Brief History of Preindustrial Progress 33

    2  Preindustrial Prosperity 60

    3  Why Mechanization Failed 72

  PART II. THE GREAT DIVERGENCE 93

    4  The Factory Arrives 97

    5  The Industrial Revolution and Its Discontents 112

  PART III. THE GREAT LEVELING 141

    6  From Mass Production to Mass Flourishing 147

    7  The Return of the Machinery Question 174

    8  The Triumph of the Middle Class 189

  PART IV. THE GREAT REVERSAL 223

    9  The Descent of the Middle Class 227

  10  Forging Ahead, Drifting Apart 249

  11  The Politics of Polarization 264

  PART V. THE FUTURE 297

  12  Artificial Intelligence 301

  13  The Road to Riches 342

  Acknowledgments 367

  Appendix 369

  Notes 373

  Bibliography 425

  Index 453

  PREFACE

  Future historians may wonder why we failed to learn from the past. Historically, when large swaths of the population have found their livelihoods threatened by machines, technological progress has brought fierce opposition. We are now living through another episode of labor-replacing progress, and resistance is seemingly looming. According to a 2017 Pew Research Center survey, 85 percent of Americans now favor policies to restrict the rise of robots.1 And Andrew Yang has just announced his bid for the White House in 2020 on a campaign to protect jobs from automation.2 The underlying concern is not hard to understand. Aided by advances in artificial intelligence (AI), robotics, machine vision, sensor technology, and so on, computers have become capable of performing a wide range of tasks that could be done only by humans a few years ago. Top-down programming is no longer required for automation to happen. In the age of AI, computers can learn themselves. What used to be distant moon shoots in computing are now reality.

  In September 2013, my Oxford friend and colleague Michael Osborne and I published a research paper estimating the potential impacts of advances in AI on jobs. We found 47 percent of American jobs to be at high risk of automation as a consequence.3 A few months later, I was invited to speak at a conference in Geneva. I was in good company, with one former prime minister, one chancellor, and a couple of labor ministers. After my talk, a well-known economist in the audience—let’s call him Bill—approached me and dismissively remarked, “Is this not just like the Industrial Revolution in England? … Didn’t machines displace jobs then as well?” Bill was right, of course, but it was only on my way back to the airport that I realized how right he actually was in suggesting that things are no different this time around. Some jobs will disappear, but people will find new things to do, as they always have, and therefore there is nothing to worry about. Unfortunately, that is only half the story.

  The long-term economic benefits of the Industrial Revolution, to which Bill alluded, are uncontested. Before 1750, per capita income in the world doubled every 6,000 years; since then, it has doubled every 50 years.4 But the industrialization process itself was a different matter. While economic historians are still debating whether the pains inflicted on the workforce by the Industrial Revolution were worth it, for later generations they surely were. Yet many contemporary laborers, who saw their livelihoods vanish as their skills became obsolete, would just as surely have been better off had the industrial world never arrived. As the mechanized factory displaced the domestic system, traditional middle-income jobs dried up, the labor share of income fell, profits surged, and income disparities skyrocketed. Sound familiar? Indeed, so far our age of automation largely mirrors the early days of industrialization in economic terms. It took over half a century until average people saw the benefits of the Industrial Revolution trickle down. And unsurprisingly, as many citizens experienced a reversal of fortunes, the consequence was cascading opposition to machines. The Luddites, as they were called, raged against mechanization, and they did everything they could to resist it. If this is “just” another Industrial Revolution, alarm bells should be ringing.

  The idea underpinning this book is straightforward: attitudes toward technological progress are shaped by how people’s incomes are affected by it. Economists think about progress in terms of enabling and replacing technologies.5 The telescope, whose invention allowed astronomers to gaze at the moons of Jupiter, did not displace laborers in large numbers—instead, it enabled us to perform new and previously unimaginable tasks. This contrasts with the arrival of the power loom, which replaced hand-loom weavers performing existing tasks and therefore prompted opposition as weavers found their incomes threatened. Thus, it stands to reason that when technologies take the form of capital that replaces workers, they are more likely to be resisted. The spread of every technology is a decision, and if some people stand to lose their jobs as a consequence, adoption will not be frictionless. Progress is not inevitable and for some it is not even desirable. Though it is often taken as a given, there is no fundamental reason why technological ingenuity should always be allowed to thrive. The historical record, as we shall see, shows that technology’s acceptance depends on whether those affected by it stand to gain from it. Episodes of job-replacing technological change have regularly brought social unrest and, at times, a backlash against technology itself. In this regard, the age of automation, which took off with the computer revolution in the 1980s, resembles the Industrial Revolution, when the mechanized factory replaced middle-income artisans in large numbers. Now, as then, middle-income jobs have been taken over by machines, forcing many people into lower-paying jobs or causing them to drop out of the workforce altogether.

  To capture attitudes toward technology over the centuries, this book brings together much of the technical economics literature with historical accounts of technological change and popular commentary. Though it concerns the future, it is not a prediction of it. Prophets may be able to foretell the future; economists cannot. The objective here is to provide perspective, and perspective we get from history. As Winston Churchill once quipped, “The longer you can look back, the farther you can look forward.”6 Thus, before looking forward, we shall begin by looking back. The Industrial Revolution was a defining moment, but few people grasped its enormous consequences at the time. We are now in the midst of another technological revolution, but fortunately this time around, we can learn from previous episodes. Bill dismissed our study as Luddite. And indeed, parallels are often drawn between the Industrial R
evolution and now to suggest that the Luddites were wrong in trying to halt the spread of the mechanized factory. Artisan craftsmen, whose feelings were stronger than their judgment, rebelled against the very machinery that came to deliver unprecedented wealth for the commoner, or so the story goes. This story is an accurate description of the long run, but in the long run we are all dead. Three generations of working Englishmen were made worse off as technological creativity was allowed to thrive. And those who lost out did not live to see the day of the great enrichment. The Luddites were right, but later generations can still be grateful that they did not have it their way. History is made in the short run because the decisions we make today shape the long run. Had the Luddites been successful in bringing progress to a halt, the Industrial Revolution would probably have happened somewhere else. And if not, economic life would most likely still look similar to the way it did in 1700.

  This brings us to the second theme of this book: whether replacing technologies will be blocked depends on who stands to gain from them and the societal distribution of political power. During the Industrial Revolution, the Luddites and other groups did what they could to stop the spread of labor-replacing technologies, but they were unsuccessful, as they lacked political clout. In fact, as we shall see, one of the reasons why the Industrial Revolution first happened in Britain is that for the first time, political power was with those who stood to gain from mechanization. The hegemony of landed wealth was challenged by the mobile fortunes of merchants, who came to form a new industrial class with growing political influence.7 The mechanized factory was deemed critical to Britain’s competitive position in trade and thus to merchants’ fortunes, which its government would do nothing to jeopardize. But for most of history, the politics of progress were such that the ruling classes had little to gain and much to lose from the introduction of labor-replacing technology. They rightly feared that angry workers might rebel against the government. In the seventeenth century, for example, the craft guilds had become a force of growing political power in Europe, and they vehemently resisted technologies that threatened their livelihoods. And fearing social unrest, European governments typically sided with the guilds. Consequently, economic incentives to invest in labor-saving technology were few. And because mechanization would put the incomes of parts of the population at risk, prompting social unrest and possibly a challenge to the political status quo, the ruling classes did their best to restrict it.

  One reason economic growth was stagnant for millennia is that the world was caught in a technology trap, in which labor-replacing technology was consistently and vigorously resisted for fear of its destabilizing force. Could countries in the industrial West experience a return of the technology trap in the twenty-first century? While it may seem unlikely, it certainly looks more likely than it did when I began writing this book four years ago. Proposals to tax robots in order to slow down the pace of automation now feature in the public debate on both sides of the Atlantic. And unlike the situation in the days of the Industrial Revolution, workers in the developed world today have more political power than the Luddites did. In America, where Andrew Yang is already tapping into growing anxiety about automation, an overwhelming majority now favor policies to restrict it. The disruptive force of technology, Yang fears, could cause another wave of Luddite uprisings: “All you need is self-driving cars to destabilize society.… [W]e’re going to have a million truck drivers out of work who are 94 percent male, with an average level of education of high school or one year of college. That one innovation will be enough to create riots in the street. And we’re about to do the same thing to retail workers, call center workers, fast-food workers, insurance companies, accounting firms.”8

  The point is not fatalism or pessimism. And it is surely not that we would be better off slowing down the pace of progress or restricting automation. The Industrial Revolution was the beginning of an unprecedented transformation that benefited everyone over the long run. AI systems have the potential to do the same, but the future of AI depends on how we manage the short run. If we seek to understand the challenges ahead rather than glossing over them in the belief that in the long run everyone will come out ahead, we will be in a much better position to shape the outcome. Yang may be extremely unlikely to be elected president, but as the observant Rana Foorohar writes, automation is likely to become a major topic in the 2020 election.9 As there has been a populist backlash against globalization, we should be concerned that populists might easily and effectively tap into growing anxiety about automation as well, unless we address it. Fortunately, for all the parallels between the age of automation and the days of the Industrial Revolution, which I shall push shamelessly throughout this book, there are also many differences. But our fascination with our own day and age, and our preoccupation with both the promise and peril of new technology, often leads us to think that we are experiencing something entirely new. Seen through the lens of the long record of human history, however, this seems unlikely to be true.

  INTRODUCTION

  Progress would be wonderful—if only it would stop.

  —ROBERT MUSIL

  When looms weave by themselves, man’s slavery will end.

  —ARISTOTLE

  Had it not been for the deeds of six hundred lamplighters, the streets of New York City at night in 1900 would have been lit by nothing but the moon. Equipped with torches and ladders, they were the force ensuring that pedestrians could see more than a burning cigar a block off when they left their homes. But on the night of April 24, 1907, most of the twenty-five thousand gas lights in the streets of Manhattan were never lit. The lamplighters, who would normally start carrying the torch of civilization around 6:50 P.M., left the lights out and went on strike. No violence was reported. But as it grew darker, New Yorkers poured in complaints to the gas companies and the local police. Policemen were sent in to light up the neighborhoods, yet without ladders this proved a difficult task. Many officers were too obese to climb the lampposts. And they got little help from the public. In Harlem, crowds of boys invented a new sport: whenever an officer was successful in firing up a lamp, they would climb the post, turn out the light, and run. On Park Avenue one youngster was arrested after having put a light out after an officer got it burning. Few lamps burned for long. Even by 9:00 P.M., the only bright public spots were a few transverse roads in Central Park, which had been equipped with electric streetlights.1

  Citizens who took up work as lamplighters that year were unlucky. Oil and gas lamps had always required personal attention, but with the mysterious force of electricity, the touch of the lamplighter was no longer a skill that had any value. Electric streetlights brought light and nostalgia. Many citizens still felt that a young man must turn lights on at dusk and off at dawn. In New York City, lamplighters had become a neighborhood institution alongside the police and the postman. Their profession had existed since the first streetlights were inaugurated in London in 1414, but it was about to become a distant memory. As the New York Times noted in 1924, “The lamplighting business in the great metropolis has been victim of too much progress.”2 To be sure, the first electric streetlights in New York City had already been installed in the late nineteenth century, but they had hardly made lamplighters redundant. Each lamp was equipped with its own switch, which had to be turned on manually. Early electrification just made the job easier, as lamplighters no longer had to carry long torches to ignite the lamps. Still, the men who used to light the gas lamps were not the beneficiaries of progress. The mastery of light had once allowed a working man to support his family. Now, turning on the lights had become a task so simple that it could be done by young boys on their way home from school. And as so often in history, simplification was merely a step toward automation. As electric streetlights were increasingly regulated from substations, the jobs of lamplighters were cut in large numbers. By 1927, electricity had a monopoly on illumination in New York City, and the last two gas lamplighters left their craft, ending the story of thei
r profession and that of the Lamplighters Union.3

  Thomas Edison’s invention of the light bulb surely made the world better and brighter. In his laboratory in Menlo Park, oil lamps and candles still polluted the air on the day of his breakthrough. As William Nordhaus, winner of the Nobel Prize in Economics in 2018, has shown, the price of light fell dramatically thereafter, as electricity spread to Chicago’s Academy of Music, London’s House of Commons, Milan’s La Scala, and the trading floor of the New York Stock Exchange.4 For the purpose of streetlighting, even the New York lamplighters, some of whom were forced into early retirement, willingly admitted that the new system was more expeditious. One lamplighter could at best attend to some fifty lamps per night. Now, several thousand lamps could be switched on by one substation employee in seconds. Yet nothing could be more natural than resisting a threat to one’s livelihood. For most citizens, their skills are their capital, and it is from that human capital that they derive their subsistence. Thus, despite all the virtues of the new system, it is not surprising that electric light wasn’t welcomed by everyone everywhere. When the municipality of Verviers in Belgium announced the switch to electricity, for example, lamplighters took to the streets in fear of losing their jobs. To banish the tyranny of darkness, the local government enrolled another team of lamplighters, but they were soon attacked by the strikers—who threatened to keep breaking lamps till doomsday. Intervention by local police ended with angry lamplighters raiding police headquarters. The Belgian government had to call in the army to resolve the situation.5

  Some surely paid the price for progress. But over the course of the twentieth century, the vast majority of citizens in the West have accepted technology as the engine of their fortunes. They have recognized that it improved working conditions by eliminating the most hazardous and servile jobs. They realized that their wages depended on the use of mechanical power. And they benefited from the continuous flow of new goods and services that became available to them. Revolutionary technologies like automobiles, refrigerators, radios, and telephones—to name just a few—were all unavailable to European monarchs in the Renaissance, but by 1950 they were common features of Western life. In 1900, the average housewife could still only dream of living like the upper classes, who had servants to do the most tedious household tasks for them. In the following decades every home suddenly got equal access to the electric servant. Washing machines, electric irons, and a host of other electric appliances took over hours of drudgery in the home. In short, the capitalist achievement, as the great economist Joseph Schumpeter observed, did not consist of providing “more silk stockings for queens but in bringing them within the reach of factory girls in return for steadily decreasing amounts of effort.”6