Driverless cars. Robotic doctors. Algorithms which write journalistic copy. It seems like every day brings a new story about how robots could be set to take our jobs. But before we panic about the end of employment, we should take the opportunity to look back at the history of previous technological change.
In focusing on the future potential of robots and artificial intelligence, we sometimes forget just how much has changed during our own lifetime. Tesco first introduced the automated checkout (and the phrase “unexpected item in the bagging area”) in 2003. It’s a mere ten years since the invention of the iPhone, and smartphone use becoming almost ubiquitous in our daily lives. And in the past five years, the number of people who have never used the internet has almost halved.
These shifts in how we work, following on from the information revolution that made computer use widespread from the 1970s onwards, have not seen a fall in the number of jobs. The UK is currently experiencing a record level of employment – though too many of the jobs are still poor quality and insecure. That doesn’t mean that there hasn’t been huge shifts and disruption in where people work. Looked at over a longer period, people have increasingly moved away from working in heavy industry, and towards more professional jobs. In 1950, almost one in three workers worked in manufacturing, while one in 12 worked in professional and technical services. By 2016 these shares had reversed.
But the biggest problem working people face today is not whether they have a job, it’s with how much they’re paid. Real wages still haven’t recovered to the levels seen around the time the iPhone came in, with the average worker earning £20 less a week in real terms than before the financial crash. The rapid technological change we’ve experienced over the last decade has not delivered improvements in business efficiency: productivity has flat-lined since 2008. That’s led to many businesses saying they can’t afford to pay their workers more, even in the face of rising inflation.
Could robots and wider technology help us to turn this round? Bold claims are made for the impact of artificial intelligence on productivity. For example, the consultancy PwC say that AI could deliver a 10 per cent uplift in GDP by 2030 – an extra £2,300 a year per person.
This would follow historical precedent. The technological developments of the first and second industrial revolutions – the widespread use of steam and then electricity – enabled living standards to rise. Perhaps this boost to GDP could be used to help with the costs of an ageing population. Take the campaign to reverse the planned increases in the state pension age, for example. A 10 per cent boost to GDP would make the cost look negligible.
But the experience of the third industrial revolution, which can be traced from the 70s onwards, should give us pause for thought. How will the benefits of rising productivity be shared, if they do indeed show up? It’s not just that a greater share of national income has gone to those at the top, although that’s certainly the case. Across the developed world, the share of national income going to workers as a whole – as opposed to capital – has fallen. As technology has done more, it looks like those who own it have increasingly monopolised the rewards.
That doesn’t mean we should fight the robots. New technologies could deliver huge benefits in areas as diverse as making us healthier, to helping tackle climate. But it does mean we should plan to make sure that the financial benefits are fairly shared.
First we need to talk about what we want from technological change. Germany convened a year-long inquiry into “industrie 4.0” to talk to unions, business and workers about how to manage the impact of industrial change. The British government should set up a similar commission. We also need to make sure that those who may have their jobs disrupted are fully prepared. Two-thirds of those who will be in work in 2030 are already working. Equipping them for the jobs of the future means rapidly upping our investment in workplace skills – which is currently just half the European average.
Some have argued that fair shares will require us to tax the robots, perhaps to fund a universal basic income. We don’t think that’s the right approach – we want to embrace the potential of new technology to deliver better quality, and better-rewarded work. And there’s simpler ways to use the tax system to redress inequality. At present, big cuts in corporation tax will come at the same time as big cuts in the benefits which support working families with children – and inequality is set to rise. Reversing that trend would be much simpler than identifying which forms of technology to tax.
Finally, we need to put in place the institutions we know are necessary to ensuring a fair distribution from work. Countries with higher levels of collective bargaining have seen wage inequality rise to a lesser extent than those where fewer workers have the benefits of union representation. Unions are working hard to find new ways of organising workers – including using new technologies to up our membership. But government could make it easier too – for example by enabling unions to access workplaces where we’re not yet represented. As a paper commissioned by Barack Obama put it, “technology is not destiny”. It’s politics that will ultimately determine whether we can make the robots work for us.