Navigating automation and what it means for unemployment is, along with climate change and an ageing population, one of the great challenges of the twenty-first century.
A 2015 study by the Bank of England showed how technological change, in particular the rise of machine learning, may mean the loss of up to fifteen million UK jobs. A year later the Bank’s Governor, Mark Carney, returned to those forecasts saying that without decisive action livelihoods could be “mercilessly destroyed.”
Such a conclusion is increasingly common. In 2013 a study conducted by Oxford University modelled that 47 per cent of U.S jobs were at “high risk” of being automated, with a further 19 per cent at medium risk. Peter Sondergaard, research director for the consultancy Gartner, claimed that by 2025 one in three jobs will disappear as the result of an emerging “super class” of technologies.
For those sceptical of the ‘rise of the robots’, such reasoning is merely the latest expression of the “Luddite Fallacy”—the mistaken belief that capitalist expansion, driven by new technology, will eliminate more jobs than it creates. For two centuries, they reply, market exchange has confounded such predictions delivering rising living standards, enhanced productivity and work for nearly everyone. And all while the Earth’s human population grew from one to 7.5 billion.
Yet this time really could be different. As many as 80 per cent of the jobs we have today existed at the turn of the last century. While the world economy may be much bigger now than it was a hundred years ago, employing more people and enjoying far higher productivity, the kinds of work nearly everyone performs—from drivers to nurses; teachers to cashiers—aren’t new.
There are some exceptions to this, such as the percentage of people working in social care increasing over the last twenty years as people live longer. While that will continue, reflecting an important demographic shift, such growth will be rare. What is more, as professions are eliminated elsewhere, labour moving into the few areas of net job creation will serve to push wages down.
Whether its autonomous vehicles or Amazon rolling out almost entirely-automated supermarkets—cashier is the second most popular job in the US—the next several decades will likely lead to the elimination of whole professions.
Today five million Americans drive for a living, and while I suspect autonomous vehicles will take longer to diffuse than some realise, it’s only a matter of time. For those that point to safety concerns, 30,000 die in motor accidents in the United States every year. Surely that argument will be inverted once technology is sufficiently advanced, with anything other than autonomous vehicles viewed as as needlessly risky. New technology is coming, and the world of work will change radically as a result.
***
To their credit Britain’s trade unions are taking the spectre of automation, and what it means for employment, seriously. Cases brought by the GMB and IWGB against, respectively, taxi firm Uber and courier services Excel, City Sprint, eCourier and Addison Lee, have laid the groundwork for intervention in the so-called “gig economy.”
Some union leaders are planning longer term, too. In a speech given last year Unite’s General Secretary, Len McCluskey, spoke of how “automation could be a good thing for industry and society, if handled in the right way.”
He added that, rather than enhancing profit, “it should instead be an opportunity for a shorter working week with no loss in pay, or the gateway to a nationwide programme of re-skilling and up-skilling existing workers.”
Presaging that concern was the expectation that automation seemed set to ravage industries previously enjoying high levels of union density with technology displacing labour: “an issue facing all parts of manufacturing and many service sectors but … presently focussed on the motor industry above all.”
Such a view is seductive—if for no other reason than we typically view automation as being synonymous with industrial robotics and car assembly lines with remarkably few people.
But while the car industry was the first sector to see significant automation, the challenges of the next several decades will extend far beyond traditionally blue-collar jobs.
In fact, its leading edge will be most keenly felt in precarious professions where unions are either weak or non-existent—think logistics, minicabs and retail—as well as historically privileged white collar work, like healthcare and law.
Opportunities to “skill up” will be limited. Ultimately, increasingly complex machines and software will prove capable of performing any task which is repetitive—from driving to searching legal documents or even performing certain surgeries.
As a result the sphere, of uniquely “human” skills will contract so significantly that re-training for new jobs won’t be an option. Some think this means we will all become interior designers, carers and “creatives”, but while some expansion in these spheres is likely one doubts that will provide work for a world of ten billion people.
Such uncertainty about the scale and impact of automation led to Unite calling for the government to set up a “Future of Automation Commission” involving unions, employers, researchers and academics.
***
One source of inspiration for that would inevitably be Germany’s IG Metall union. The largest industrial union in Europe, IG Metall’s workers recently won a two-year guarantee for a 28-hour week alongside a 4.3 per cent pay rise.
An increasingly confident labour movement here—and a prospective Labour party in government—would undoubtedly want to imitate that. (Full time workers in the UK average around 42 hours a week.) There is also scope to draw on “digital trade unions” that are cropping up in the United States.
But that should be just the beginning—because something profound is unfolding.
McCluskey is certainly right for much of the economy, and attempting to replicate the recent gains of IG Metall are crucial in building a hopeful future, but we also need to think about distributing the dividend of such technology beyond work too.
In the case of autonomous vehicles, that might include the demand for free public transport—starting with local buses—as the intersection of permanently cheaper (non-human) labour meets a similar trajectory with renewables (whose costs are falling so quickly they are already hitting parity with fossil fuels in a number of countries). Both will get progressive cheaper with each passing year for the rest of your life. A decade from now paying for a local bus might feel as odd as paying for an email account does today.
There is also a pressing need to respond to how automation could aid post-industrial cities that have suffered from a lack of investment since at least the 198ps. Although research shows that automation could hit these places hardest, with a changed political project it would instead mean new opportunities and greater common wealth.
How would left behind cities and towns transition to capital-intensive industries like this? The answer is rolling out extensive universal basic services—in healthcare, housing, transport, information and education - built on the ‘extreme supply’ made possible by new technology.
Alongside that, the remaining labour intensive work—in areas like catering, construction, care and cleaning—would be subject to forms of municipal protectionism, with only worker-owned firms within a particular area able to win public contracts (these co-ops would themselves be capitalised by a network of regional investment banks).
Building on the ‘Preston Model’ that would mean that alongside higher levels of automation and improving public services, money would stay within communities rather than flow to company shareholders elsewhere. This is the left’s answer to not only automation, but the collapse of Carillion too.
Changed technology should be met by unions fighting for fewer hours alongside higher pay. Ultimately, however, we need to start thinking about a world beyond work altogether. Whether that happens is not a question of technological inevitably but political will.
