The Work of the Future: Building Better Jobs in an Age of Intelligent Machines

The Work of the Future

The Work of the Future

Building Better Jobs in an Age of Intelligent Machines

David Autor, David A. Mindell, and Elisabeth B. Reynolds

Foreword by Robert M. Solow

The MIT Press Cambridge, Massachusetts London, England

© 2021 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher.

Library of Congress Cataloging-in-Publication Data

Names: Autor, David H., author. | Mindell, David A., author. | Reynolds, Elisabeth B., author. | Solow, Robert M., writer of foreword. | Massachusetts Institute of Technology. Task Force on the Work of theFuture, issuing body.

Title: The work of the future : building better jobs in an age of intelligent machines / David Autor, David A. Mindell, Elisabeth B.Reynolds ; foreword by Robert M. Solow.

Description: Cambridge, Massachusetts : The MIT Press, [2021] | Includes bibliographical references and index.

Identifiers: LCCN 2021010583 | ISBN 9780262046367 (hardcover)

Subjects: LCSH: Employees—Effect of technological innovations on—United States. | Technological innovations—Social aspects—United States. | Automation—Social aspects—United States. | Labor market—UnitedStates. | Income distribution—United States. | Technological innovations—Government policy—United States. | Labor policy—United States.

Classification: LCC HD6331.2.U5 A98 2021 | DDC 331.250973—dc23

LC record available at https://lccn.loc.gov/2021010583

d_r0

Contents

Foreword by Robert M. Solow

PART I

1 Introduction

2 Labor Markets and Growth

3 Technologies and Innovation

PART II

4 Education and Training: Pathways to Better Jobs

5 Job Quality

6 Institutions for Innovation

7 Conclusions and Policy Directions

Acknowledgments

Notes

MIT Task Force on the Work of the Future Research Briefs

MIT Task Force on the Work of the Future

Index

Foreword

Robert M. Solow¹

I am writing this in the last week of January 2021. Sixty years ago, almost to the day, my family arrived in Washington, DC, and I started a year’s work on the staff of President Kennedy’s Council of Economic Advisors. The US economy had not yet emerged from the typical postwar recession of 1960. The unemployment rate, as I remember it, was a hair under 7 percent.

But another, more difficult problem had arisen. Each of the last three typical postwar recessions had taken place at higher unemployment rates than the ones before. Some economists and many in Congress and the financial press were suggesting that this higher unemployment rate was not the usual sort. It reflected not a lack of demand for goods and services but the fact that the unemployed workers were unqualified for employment: they were in the wrong place or had the wrong skills or no skills at all, or inadequate education. The usual fiscal and monetary policy maneuvers would do no good at all.

There is a tendency, whenever there is unexpectedly high or persistent unemployment, for simple monocausal explanations to circulate. Blaming the unemployment rate on the characteristics of the unemployed is one such. There is a certain immediate plausibility to such an explanation. The unemployed do tend to be less qualified than the employed. But whatever the true source of the unemployment, the normal process of turnover and selection will eventually focus the unemployment on the least qualified. This certainly does not mean that training the untrained will increase employment.

Here is a simple analogy: think of a high school basketball game played in a gym with a fixed number of seats bolted to the floor. Tickets are free, and more spectators arrive than there are seats. The seats will go on the average to the quick and aggressive. For the slow and passive there is standing room. Now suppose you train the standees to be faster and more aggressive. At next week’s game more of them will get seats. But the total number of seats does not change at all. Getting employed in a modern industrial economy is a lot more complicated than getting a seat at a basketball game, but you see the point.

This was an important matter when the Council of Economic Advisors tried to calculate appropriate fiscal and monetary policy. My first assignment from Walter Heller, the council chairman, was to evaluate this theory of rising structural unemployment. This being Washington and not MIT, I think I had three weeks. My conclusion was that there was certainly an element of structural unemployment but there was no evidence that it was increasing.

Of course, blaming the characteristics of the unemployed is not the only simplistic way to account for unexpectedly high and persistent unemployment. Dramatic changes in technology are another and equally common theory. I first heard the word automation during that 1961 debate. We are already beginning to hear that the robots are coming, the robots are coming (and some day they will actually come).

The current situation is different. Leaving the COVID-19 pandemic aside, there has been no long-run rise in the unemployment rate, at least not yet. What we have instead is something more complex.

For generations, the real wage rate in the US had grown at more or less the same rate as output per hour of work. This meant that their ratio, the share of output paid out in wages and salaries, had no trend. There were short-run variations but not much else. That seems to have changed in late 1960s or early 1970s. The real wage trend began to fall short of the productivity trend. It was not that the productivity trend accelerated; that might suggest something technological going on. The difference was that real wages fell behind. That involves a lot of economics, especially when it is set beside David Autor’s famous finding that the economy was providing lots of low-wage and high-wage jobs but was losing the middle-skilled employment that had been part of the American Dream. The dramatic increase in inequality of income and wealth fits in here, too.

Now there were many candidate causes, and they were not mutually exclusive. Those middle-skilled jobs may have been lost to workers in poor, low-wage countries. Workers were clearly losing bargaining power, as evidenced by the virtual disappearance of labor unions from the private sector. Employers hardened their attitudes. The general market power of large firms in concentrated industries was probably increasing, maybe by a lot. The problem was not to pick a cause but to reckon how much weight to attach to each of a list of causes, and that is a very hard thing to do. It follows, of course, that fitting remedies to the disease, if it is a disease, is equally complicated.

By the way, I do not want to leave the impression that education and training are minor factors in all this. Certainly maintaining a skilled and adaptable labor force makes a necessary contribution to productivity. Second, easy access to education can function as an equalizing factor, although it is pretty clear that it does not perform this function very well in the US. Finally, the system of education and training maintains a common culture and a common understanding of citizenship. The point I was making just now was only that more training or even better training does not necessarily lead to higher employment.

During the first seventy years or so of the twentieth century, American capitalism pretty regularly delivered about three-fourths of the national income in the form of wages and salaries. That is the trendless number I mentioned earlier. During the past forty years or so that number began to diminish, and it had reached something like two-thirds when the pandemic arrived. That is still much the larger part of national income. At that scale, any major change in the labor market is bound to have consequences for the rest of the economy, consequences that will then reflect back on the labor market. Disturbances arriving from outside the labor market will directly influence outcomes in the labor market. That is why, instead of being a dreary recycling of the work skills needed for oncoming technologies, this book turned into a wide-ranging survey of the economy as it now appears. No doubt we will need another such report someday, perhaps when the robots arrive. In the meanwhile, read on and learn what the grown-ups are thinking.

I

1

Introduction

A decade ago, powerful mobile phones were still a novelty, driverless cars were never seen on public roadways, computers did not listen to conversations or respond to spoken questions. The possibility of robots taking jobs seemed far off, save for an assembly line or two. But as the emerging capabilities of robotics and artificial intelligence (AI) began capturing headlines and the popular imagination, researchers and commentators began warning that jobs long thought to be immune to automation—those demanding expertise, judgment, creativity, and seasoned experience—might soon be better accomplished by machines. Citizens of industrialized countries took notice, reacting with mounting trepidation.

In this context, in the spring of 2018 MIT president L. Rafael Reif commissioned the MIT Task Force on the Work of the Future. He tasked the group with understanding the relationships between emerging technologies and work, to help shape public discourse around realistic expectations of technology, and to explore strategies that might enable a future of shared prosperity. The Task Force was co-chaired by this book’s authors, Professors David Autor and David Mindell and Executive Director Dr. Elisabeth Reynolds. Its members included more than twenty faculty members drawn from twelve departments at MIT and more than twenty graduate students. The Task Force commissioned and conducted numerous research studies, many of them published as working papers and research briefs, on which we draw heavily for this book (a complete list of Task Force publications is presented at the end of the book).

In the three years that the Task Force delved into the future of work, autonomous vehicles, robotics, and AI advanced remarkably. But the world was not turned on its head by automation, nor was the labor market. Despite massive private investment, technology deadlines have been pushed back, part of the normal evolution of breathless promises as concepts are tested in pilot trials, integrated into business plans, and actualized in early deployments. These are the diligent, if prosaic, steps toward making real technologies work in real settings to meet the demands of hard-nosed customers and managers.

Our research did not confirm the dystopian vision of robots ushering workers off factory floors or AI rendering superfluous human expertise and judgment. But it did uncover something equally pernicious: amid a technological ecosystem delivering rising productivity and an economy generating plenty of jobs (at least until the COVID-19 crisis), we found a labor market in which the fruits are so unequally distributed, so skewed toward the top, that the majority of workers have tasted only a tiny morsel of a vast harvest.

For most US workers, the trajectory of productivity growth diverged from the trajectory of wage growth four decades ago. This decoupling had baleful economic and social consequences: low-paid, insecure jobs held by non-college-educated workers; low participation rates in the labor force; weak upward mobility across generations; and festering racial earnings and employment disparities that have not substantially improved in decades. While new technologies have contributed to these poor results, these outcomes were not an inevitable consequence of technological change, or of globalization, or of market forces. Similar pressures from digitalization and globalization affected most industrialized countries, yet their labor markets fared better.

Yet we know that history and economics show no intrinsic conflict among technological change, full employment, and rising earnings. The dynamic interplay among task automation, innovation, and new work creation, while always disruptive, is a primary wellspring of rising productivity. Innovation improves the quantity, quality, and variety of work that a worker can accomplish in a given time. This rising productivity, in turn, enables improving living standards and the flourishing of human endeavors. Indeed, in what should be a virtuous cycle, rising productivity provides society with the resources to invest in those whose livelihoods are disrupted by the changing structure of work.

When innovation fails to drive opportunity, however, it generates a fear of the future: the suspicion that technological progress will make the country wealthier while threatening numerous livelihoods. This fear exacts a high price: political and regional divisions, distrust of institutions, and mistrust of innovation itself. This anxiety has been laid bare in US politics as a growing gulf between the haves and the have-nots has driven a deepening national schism over how society should respond to the needs of those at the bottom of the economic ladder.

The central challenge ahead—indeed, the work of the future—is to advance labor market opportunity to meet, complement, and shape technological innovation. This drive will require innovating in our labor market institutions by modernizing the laws, policies, norms, organizations, and enterprises that set the rules of the game.

The labor market impacts of technologies like AI and robotics are taking years to unfold. But we have no time to spare in preparing for them. If those technologies are deployed in the labor institutions of today, which were designed for the last century, we will see similar effects to recent decades: downward pressure on wages and benefits and an increasingly bifurcated labor market.

This book suggests a better alternative: building a future of work that harvests the dividends of rapidly advancing automation and ever more powerful computers to deliver opportunity and economic security for workers. To do that, we must foster institutional innovations that complement technological change.

We are living in a period of significant disruption, but not of the kind envisioned in 2018, when the Task Force was launched. The final phases of researching and writing this book occurred during the 2020 months of COVID-19, when citizens of many countries were in a state of pandemic lockdown. Our technologies have been instrumental in enabling us to adapt to these new circumstances via telepresence, online services, remote schooling, and telemedicine. These tools for performing work remotely don’t look anything like robots, but they too are forms of automation, displacing vulnerable workers from low-paying service jobs in such industries as food service, cleaning, and hospitality. We face a labor market crisis stemming from the COVID-19 pandemic. Millions are unemployed. But technological advances did not cause this crisis.

Long before this disruption, our research on the work of the future made it clear how many in our country are failing to thrive in a labor market that generates plenty of jobs but little economic security. The effects of the pandemic have made it even more viscerally and publicly clear: despite their official designation as essential, most low-paid workers cannot effectively do their jobs through computing platforms since they must be physically present to earn their livings.

Some forecast that robots will soon take over those roles, though few have to date. Others see the indispensable role of human flexibility since it is human, not machine, adaptability that has allowed us to reorganize work on the fly during the pandemic. Still others see COVID-19 as an automation-forcing event—a catalytic force that will pull technologies from the future into the present as we learn to deploy machines in jobs that humans cannot safely perform. However it plays out, the effects of COVID-19 on technology and work will last long beyond the pandemic, although those effects may look quite unlike what anyone envisioned in 2018.

Other forces have also roiled the 2018 visions of the future, including the rupture between the world’s two largest economies and a surge of political turmoil and economic populism that culminated in a violent attack on the US Capitol in the wake of the 2020 election of President Joe Biden. These pressures are reshaping alliances, breaking apart and reorganizing global business relationships, and spurring new forms of cyberwarfare, including disinformation, industrial-scale espionage, and electronic compromising of critical infrastructure. The US and China had friction before, but nothing like the fracture that is now occurring. What began as a trade war has morphed into a technology war. China’s whole-of-government approach to tackling major industrial and technological goals poses a competitive challenge for Western economies, which typically take a decentralized, often business-led approach. It remains to be seen whether China’s focus on government-driven domination of data accumulation yields technological advances beyond creating powerful tools for monitoring and controlling its own population.

The clash with China is rippling through the economy and threatens to hinder innovation, which increasingly emerges from countries around the world, often by researchers who are collaborating across borders and time zones. How can we make sure that technological advances, whenever they come, yield prosperity that is widely shared? How can the US and its workers continue to play a leading role in inventing and shaping the technologies and reaping the benefits?

To address these questions, this book is divided into two parts. In part I, we look at the evolution of work and the status of key technologies that are poised to shape its future. Part II suggests how to