Droids won't steal your job; they could make you rich

Popular economics moves in cycles, and each cycle gives rise to a new worry about the future. Back in 2008/9 the fashionable concern was about debt, a fear driven partially by Carmen Reinhart and Kenneth Rogoff’s This Time is Different. By 2014, cooler kids were clutching copies of Thomas Piketty’s Capital in the Twenty-First Century, and fretting about inequality. Nowadays the zeitgeist anxiety is that robots are about to take all of our jobs. Propelled by a never-ending stream of books and op-eds, it is a fear that is surprisingly well spread, encompassing a spectrum from billionaire technologist Bill Gates to French Socialist presidential contender Benoît Hamon. Thankfully though, while a robot takeover is capturing a lot of column inches, it isn’t showing up in the data.

Management consultants McKinsey are the latest to, er, drone on about the drones. Their Global Institute has published a report on automation that looks at 46 countries covering around 80 per cent of the worldwide jobs market. They argue that with a combination of robotics, artificial intelligence and neural networks, almost half of the activities that workers are currently paid to do could be automated by 2055. That would have consequences for 1.2bn workers, who between them earn around $14.5 trillion in wages. These are big and scary numbers, even in global terms.

The report follows in the footsteps of thought-provoking, but under-questioned, work by two Oxford academics, Carl Frey and Michael Osborne. In 2013 they claimed that almost half of all US occupations were at a high risk of automation in the next 20 years. They later extended their analysis to the global economy, and found that an outright majority of jobs in advanced economies, more than two-thirds in India and fully three-quarters in China were susceptible to replacement. The annual Alpine jamboree for the rich and powerful in Davos was last year dedicated to discussing the so-called “fourth industrial revolution.” An associated report warned that 5m jobs would go by 2020 across 15 major economies, all driven by technological change.

But there has also been one major recent study of technology, automation and the future of work that hasn’t grabbed the headlines at all. A 2016 OECD report came to a starkly different conclusion: across the developed economies “only” around nine per cent of jobs are at a high risk of being replaced by machines in the coming 20 years. That may still sound uncomfortably high, but it is well within the normal range of industrial history. One in 10 jobs going would pose big adjustment difficulties for some, and challenges for government, but it is fundamentally a manageable problem: the kind of thing we’ve dealt with before.

The crucial difference comes from a more thorough methodology. Rather than looking at broad occupations and whole classes of activities, the OECD stripped jobs down into individual tasks. Machines are less likely to be able to replicate creativity, social interaction and the need for human-to-human contact anytime soon, and a surprising number of jobs involve these attributes.

Take the US category of “retail salesperson.” Frey and Osborne gave that occupation a 92 per cent risk of “computerisation.” But the OECD points out only four per cent of such workers perform their job without face-to-face interaction or group work. It turns out that three-quarters of bookkeepers, who Frey and Osborne give a 98 per cent “robotisation risk,” could not carry out their role without human contact and collaboration.

The demon robot is starting to look less terrifying. Indeed, he has been maligned before. The BBC archive contains a 1978 Horizon which explained how a new brainwave called the microchip was “the reason why our children will grow up without jobs to go to.” Back in the 1810s, the Luddites destroyed new cotton producing machinery which they saw as throwing thousands of workers on to the scrap heap. In reality, the story of human progress since the birth of agriculture around 10,000BC has been the story of humans replacing labour with technology and freeing up time to do new, and different, jobs. That process speeded up after the 1750s, and—even if sometimes falteringly and unevenly—it eventually drove living standards to be transformatively higher.

Labour-saving technology has two economic effects. The first and more obvious is displacement: new inventions displace workers and throw some people out of employment. But secondly and less obviously, those who keep working are now more productive—churning out as much stuff as many more workers did previously. Those newly productive workers should soon earn more, and so spend more as well. That fuels demand and creates new jobs, compensating for the posts that have gone. Before too long, history suggests that this compensation effect will be the dominant one. Of course, to use the most dangerous words in economics, this time may be different. But betting against 12,000 years of history is an almighty call. If a trend has existed since roughly the Neolithic era, it is surely sustainable.

But the debate is still shot through with anxiety—witness the calls from Gates, Hamon and others for a “tax on robots” to slow technological progress and protect jobs. This is certainly a more sophisticated Luddism than the original—“tax the looms” rather than “smash the looms.” But ultimately it comes from the same assumption, that productivity growth is a threat to jobs rather than a kicker for wages and living standards. That runs against not just economic theory but all historical experience.

We are debating a problem we don’t have, rather than facing a real crisis that is the polar opposite. Productivity growth has slowed to a crawl over the last 15 or so years, business investment has fallen and wage growth has been weak. If the robot revolution truly was under way, we would see surging capital expenditure and soaring productivity. Right now, that would be a nice “problem” to have. Instead we have the reality of weak growth and stagnant pay. The real and pressing concern when it comes to the jobs market and automation is that the robots aren’t taking our jobs fast enough.