David Autor, the academic voice of the American worker
IT IS hard not to notice it, when you first meet David Autor: the earring, there in the lobe of his left ear, in the shape of a gecko. It has become something of a theme among his students and colleagues, who have given him all sorts of gecko-related paraphernalia. The earring is one half of a pair bought with his wife, before they were married and long before he became the Ford Professor of Economics at the Massachussetts Institute of Technology. It is an eye-catching accessory, fitting for a man whose career has followed unconventional paths.
Mr Autor is an ever-so-slightly manic figure, who speaks quickly and engagingly. He is not a household name but he is enormously influential, in large part because of his groundbreaking work on the effects on American workers of China’s extraordinary rise. Economic models have long allowed for the possibility that trade between rich countries and poorer ones with an abundance of cheap labour could harm workers in the richer country. But economists generally assumed that any such negative effects would not be of much consequence. In a series of papers Mr Autor and his two primary collaborators—David Dorn, of the University of Zurich, and Gordon Hanson, of the University of California, San Diego—painted a very different picture.
Get our daily newsletterUpgrade your inbox and get our Daily Dispatch and Editor’s Picks.
In regions where the products manufactured by local firms were most vulnerable to an influx of cheap, Chinese imports, wages and labour-force participation remained depressed a decade after the onset of the “China shock”. Subsequent work by Mr Autor and his co-authors found that the shock contributed to deep social disruption as well: in marriage patterns, for instance, and in the appetite for fringe political ideas. The results surprised the authors and sent a jolt of disbelief through the profession. They also inspired a wave of new research into the effects of globalisation,and enjoined economists and policymakers alike to think more seriously about how workers survive and thrive amid major economic shifts.
Thirty years ago Mr Autor had no great ambition to launch new areas of economic investigation. Three semesters into an undergraduate degree at Columbia University he dropped out, in order to work out what exactly he wanted to do with his life. He had always been a tinkerer and interested in computers, and so he spent some time working in tech jobs: as a programmer for a hospital and then for a friend’s tech startup. The rigorously analytical way in which programming required him to think appealed to him, but the problems he was tackling sometimes felt trivial. Before the money he was earning grew too large to walk away from, he returned to university to study psychology.
That experience, too, frustrated him. His psychology work was the mirror image of his time spent coding: he was working on important problems which were maddeningly resistant to systematic analysis. Unsure what to do next, Mr Autor moved to the San Francisco Bay Area and began working with a Methodist church to teach computer skills to disadvantaged kids. That contributed to a budding interest in what exactly the computer age was likely to mean for workers. In the Bay Area, Mr Autor also met the woman he would marry, acquired the gecko earrings and decided, before long, to give academia another go. In 1992 he entered a public-policy graduate programme at Harvard University, and there, at the relatively ripe age of 28, he took his first economics course.
Economics hit the intellectual sweet spot. Here was the opportunity to apply systematic ways of thinking to important, real-world problems. And the problem in which he found himself most interested at the time was the way that new technologies affected labour markets. His thesis explored the impact of computerisation on temporary help workers. Upon finishing his work at Harvard he made the extremely unlikely leap from a public-policy graduate programme to the economics department at MIT, one of the most prestigious in the country.
From his new perch Mr Autor’s early work focused on when, how and with what effects firms decided to use new, and potentially labour-displacing, technologies. For a fascinating paper published in 2002, with Frank Levy, of MIT, and Richard Murnane, of Harvard, Mr Autor and his co-authors spent two years exploring every aspect of the operation of a large bank. Their investigation dug into the limits of automation: as in the case of the stubborn 3% of cheques which persistently elude efforts to computerise the entirety of the cheque-processing system, and which must be parsed by human workers.
As a result both of this research and his experience as a programmer, he realised that the way many economists were thinking about automation was flawed. Researchers often used the education level of the workers employed in a particular job as a good-enough proxy for the complexity of the role and, correspondingly, for how susceptible it might be to automation. On this reckoning workers with some post-secondary education ought to be less vulnerable to march of the robots than those who lacked even a college degree.
Mr Autor suggested that it was not the skill set of the worker which mattered, but the task to which the worker applied her skills. Some tasks that are trivially easy for machines (like multiplying 15-digit numbers) are very difficult for humans. Others that humans handle effortlessly, like manoeuvering around furniture while tidying an office, are incredibly difficult for machines. The “task approach” to labour markets that emerged from this work has become a critical tool in analysing the ways in which all sorts of disruptive events affect workers.
At the time, in the early 2000s, the importance of this contribution was not obvious, including to Mr Autor himself. America had just experienced one of the best decades for workers in a very long time. The disruptions wrought by the introduction of powerful new kinds of computing software seemed to be petering out. But as the decade wore on, the continuing relevance of Mr Autor’s work became clear.
The connection he drew between the extent to which tasks were routine or not, and the ease with which they could be automated away, proved important as policymakers began to grow worried about a “hollowing out” of the workforce. The production and office work done by respectably paid, “middle-skill” employees was disappearing; data entry and repetitive production tasks could be given to computers or cheap foreign labourers in a way that neither low-skill janitorial tasks, say, nor well-compensated professional work could be.
What’s more, by the early 2010s, it became clear that the tools Mr Autor had developed could be used to assess the effects of another big source of disruption: expanded trade with China. Along with Mr Hanson and Mr Dorn, he pointed out in a paper published in 2013 that routine sorts of work might also be susceptible to competition from abroad, either through offshoring or trade with low-wage countries. Western firms might not wish to send abroad complex jobs like design, but production work looks much the same whether it is done in Chicago or Shenzhen.
Where the economic affects of automation tend to be spread broadly across the whole of an economy, however, displacement of routine work by trade tends to have a more focused geographic impact. In subsequent work, the authors showed how it was these concentrated costs of trade with China which made the China shock so powerful. Communities hit hard by the shock could not recover easily. Trade left long-run economic, social and political scars.
The researchers knew their work would be explosive. They were themselves surprised by the results, and checked and rechecked their findings in preparation for the storm of criticism they expected to face. The profession was indeed sceptical at first. Mr Autor chuckles that economists’ initial reaction was to insist the work was wrong, then to shift to allowing that it was right but meaningless, before eventually grappling with the work’s implications.
Those implications were significant. Mr Autor’s China-shock work has heavily informed debates in Washington about how to manage globalisation. It has made it easier to discuss the real-world challenges associated with China’s rise—from the encouragement states give to cutting-edge industries to the national-security dimensions of trade in advanced technology—without being labelled a protectionist or a crank.
China’s integration into the world economy was a one-off. But the disruptions faced by American workers are not over. Even as Mr Autor and his co-authors were beginning their landmark China papers, fresh concern was growing about mass automation, brought about by striking improvements in artificial intelligence.
By now, however, the economics profession has become much more preoccupied by the shocks that ordinary workers suffer. At the annual meeting of the American Economic Association in Atlanta earlier this year, Mr Autor drew a crowd of thousands for a keynote address that focused on the bleak prospects of workers without a college degree. He has been asking himself questions such as these for two decades. But today the fast-talking economist with the gecko stud in his ear has everyone’s attention.