Sense and nonsense in the public discussion of the future of work

Richard Baldwin 01 November 2019

a

A

Whether or not the robots are coming for our jobs, they are coming for our politics. In the US Democratic Party debate held on 15 October, the political candidates debated the issue of automation. Erin Burnett, one of the moderators, started it off:

"According to a recent study, about a quarter of American jobs could be lost to automation in just the next 10 years," she said, and asked the candidates what they would do about it. 

The candidates offered an assortment of policies. Several suggested a basic income guarantee. The frontrunners used it as an opportunity to talk about their existing commitments. So, Sen. Elizabeth Warren suggested that reforming tax policy and strengthening social security was the priority, while Sen. Bernie Sanders proposed government-funded programme of infrastructure work.

Whether or not these politicians have adequate answers, at least the tone of the debate has shifted since previous election campaigns, during which the debate was customarily centred around lost jobs in manufacturing. And this time too, the moderators soon pivoted back to familiar ground, asking whether the politicians would enact policies to repatriate manufacturing jobs from Mexico. 

But this time it is different, because automation is no longer a story about lost jobs in unskilled or semi-skilled manufacturing. Most of us in the US or Europe have been shielded from the sharp end of globalisation and automation because our jobs involve doing things, not making things. 

The robots are thinking

This is changing. 

Here's an example: Walls Fargo has just released a study that concludes that 200,000 US bank jobs will be lost in the next decade. Cuts will come in back offices, bank branches, call centres and corporate employees, with between a fifth to a third of staff (who are half of US banks' costs) losing their jobs.

Globalisation and automation are about to get personal for hundreds of millions of North Americans and Europeans who don’t work in factories. Our ability to gather, transmit, store, and process digital information is doubling every year or so, and this, in turn, is creating a new type of globalisation that is virtual, and a new type of robot that can think. 

‘Virtual presence’ technology and instant machine translation will enable talented foreigners sitting abroad to provide services in our offices and workspaces. It will be almost as if these remote workers were actually there with us. A form of artificial intelligence (AI) known as ‘machine learning’ is teaching computers to automate tasks involving experience-based pattern recognition – tasks that are common in professional, white-collar and service jobs. The fastest-growing part of the business software market, you may be surprised to learn, is robotic process automation (RPA) – the applications that use machine learning to observe and then copy these tasks.

Exponential change

These changes will disorder the professional and service sector jobs radically faster than globalisation disrupted the manufacturing sector. Traditional globalisation was driven by a straight-line rise in our ability to ship goods. AI and RI can grow exponentially. 

Our brains – which evolved to understand a walking-distance world – find exponential growth to be implausible. Exponential growth means that each step is twice the size of the previous one. If you could double the length of your stride each time you took a step, your 23rd stride could take you from New York to Los Angeles. 

Globotics is advancing exponentially, due to three digital laws: Moore’s Law, Gilbert’s Law, and Metcalf’s Law (see my sequence of blogs). It is also why so many people are either unaware of how revolutionary the changes will be, or are living in denial about how fast they will come. 

Because the new forms of virtual globalisation and thinking robotics are driven by such a wildly different growth process, they deserve a new name. I have called it 'globotics' in my book, The Globotics Upheaval (first chapter free on Google Books). 

The great upheaval and a neo-Luddite backlash

The globotics revolution promises a wonderful future. We might all prosper in a fairer, more humane society. The danger lies in a neo-Luddite backlash.

With its pace and focus on service and professional jobs, globotics could destroy the foundations of middle-class prosperity in North America and Europe. In today’s job-centric capitalism, prosperity is based on good, secure jobs – and the stable communities that are built on them. 

Many of these jobs are in the sectors that ‘globots’ will disrupt. If millions of lives are thrown into disarray, and communities are disrupted, we probably won’t see a ‘keep calm and carry on’ attitude. There is a substantial chance that there will be a backlash. 

This is the story that one of the candidates for office in 2020 is telling. Andrew Yang has written about it in his book, The War on Normal People. And it is no coincidence that he was the only candidate during the debate who focused on automation's effect on the service sector:

"I've been talking to Americans around the country about automation. And they're smart. They see what's happening around them. Their Main Street stores are closing. They see a self-serve kiosk in every McDonalds, every grocery store, every CVS. Driving a truck is the most common job in 29 states, including this one; 3.5 million truck drivers in this country. And my friends in California are piloting self-driving trucks."

We need the political debate to face the different reality of the globotics revolution, not pay lip-service to ideas that are more suited to the last great upheaval caused by industrialisation. That upheaval created a world where job loss meant poverty and perhaps starvation for landless workers. Now professional, white-collar and service-sector workers may seek to slow or reverse the threat to their own job security – as taxi drivers and truck drivers have already started to do. If mishandled, the protests could spin out of control. We have seen this before.

We did eventually learn how to make industrialisation work for the majority, but the process was spread over two world wars and the Great Depression. During that period, many ordinary people embraced populist politicians who promised authority, justice, and economic security – just as a new generation of populists do today. To avoid a repeat of the worst social problems that came with industrialisation, our governments need to ensure that globotics seem more like a decent development that is fair, equitable and inclusive, and not divisive. 

These is nothing wrong with globotics’ direction of travel – it’s the speed and the unfairness that pose the problems. Job displacement will be driven at the pace of digital technology, because job displacement is the business model. Job replacement, by contrast, is driven at the pace of human ingenuity and entrepreneurship, which will be much slower.

How to thrive in the upheaval

Every great transformation creates triumphs for those who can seize the opportunities, and tragedies for those who can’t. Preparing society for the upheaval requires a thorough understanding of the technical, economic and political forces shaping globotics, since there will be no universal recipe for success. The changes are coming just too fast and in too many ways for a one-size-fits all formula to work. Individuals will have to think through what globotics means for their particular situation. I can think of three principles:

  1. The old rules won’t work. The standard rule for thriving in the age of globalisation was to get more skills. This won’t work in the age of globotics, because the globots have lots of skills. 
  2. Don’t compete with globots. Move away from skills based on experience-based pattern recognition and towards tasks that require frequent, in-person contact, which globots can't copy. 
  3. View your humanity as an edge, not a handicap. Humans have unique advantages over globots in things like local cultural knowledge, judgement, empathy, intuition, creativity and comprehension of complex interactions among teams of humans. With the right preparation, we should be able to use globotics’ fast-learning and time-saving features to make ourselves and our companies more productive, and more responsive to competitive opportunities and challenges.

Robotism in the long run

If we do make it to the long run – a future in which we all work fewer hours and globots do all the nasty work – we will need a new economic system, with policy innovations beyond the scope of any political debate we are hearing today.

In today’s economic system, traditional jobs are how people get the purchasing power they need to buy the things that the economy produces. If the ‘means of production’ are distributed widely, so value added is recycled, there will be no issue. But if robot ownership is consolidated in the hands of a few, we have a problem. Without traditional jobs, we’ll need a new mechanism. As the candidates speculated, universal basic income may be one way to do it, but a better way would be to pay people for the value they create on or off the job. This would mean valuing unpaid work like raising children, caring for the sick and the elderly, creating Wikipedia pages, or even making great YouTube videos.

The redistribution critical to today’s socioeconomic harmony works through the 'over-pricing' of services. For example, waiters and bartenders in Germany get paid much more than those in Kenya, but they are not much more productive. This is possible because German bars and restaurants charge high prices by international standards. 

The high bartender wages, in other words, are based on prices, not productivity. The high prices are an implicit channel for ‘taxing’ the German citizens who are globally competitive in a way that subsidises those who aren’t. But what happens when digital technology allows the price differences to be arbitraged away (as traditional globalisation arbitraged away the differences in goods prices)? When globots take over this sort of service sector job – and, as Wang observed, this is already happening in fast food restaurants – we’ll need a new way of sharing the gains and pains between workers who are globally competitive, and those who are not. 

Finally, globot ownership in the 21st century will become what land ownership was in the Middle Ages. It will be the key to wealth, social prestige, and political power. This means that we will have to find a system of globot ownership that supports a politically and socially sustainable society. 

It’s our choice

The radical transformations that came with the industrial revolution, and the shift from feudalism to capitalism, destroyed a centuries-old social fabric based on reciprocity and hierarchical relationships. As Karl Polanyi wrote in The Great Transformation, the commoditisation of labour and mass migration to urban and industrial areas disturbed traditional values to such an extent that the people pushed back, embracing communism or fascism. That process of push and pushback took many decades. The Industrial Revolution started accelerating around 1820, but communism and fascism became political forces a century later.

Things are moving much faster this time. The globotics revolution will avoid these societal problems only it advances at a human pace, and the disruption will be viewed by citizens as a price worth paying. 

The pace of progress is not set by some abstract law of nature. We can control the speed of disruption. We have the tools, and now we need policymakers to use them. It’s our choice. 

References

Baldwin, R (2018), The Globotics Upheaval, Oxford University Press. 

Polanyi, K (1944), The Great Transformation, Farrar & Rinehart.

Yang, A (2018), The War on Normal People: The Truth About America's Disappearing Jobs and Why Universal Basic Income Is Our Future, Hachette.

a

A

Topics:  Labour markets Productivity and Innovation

Tags:  automation, globotics, robots

Professor of International Economics at The Graduate Institute, Geneva; Founder & Editor-in-Chief of VoxEU.org; exPresident of CEPR

CEPR Policy Research