Number of Workers in Jobs That Can Be Automated Falls (ft.com)
Employment has fallen in jobs that can be easily automated and risen in those which are trickier for robots, damping hopes that higher minimum wages could unleash a wave of investment in automation. From a report: Statistics from the Office for National Statistics published on Monday showed that in 2011 about 8.1 per cent of workers were in jobs with a "high" risk of automation, but by 2017 that had fallen to 7.4 per cent. [Editor's note: the link may be paywalled; alternative source and original study.] The ONS report highlighted that fewer workers remain in areas that can be easily automated, such as dry cleaning and laundry jobs, which fell by 28 per cent between 2011 and 2017, and retail cashier work, which fell by 25.3 per cent over the same period.
Since the financial crisis the UK has enjoyed rapid growth in employment combined with one of the lowest rates of investment spending of any large rich country. Many economists have suggested that hiring cheap labour instead of investment in new techniques is behind the country's weak rate of productivity growth. Policymakers had hoped that increasing the minimum wage would spur companies to replace low-paid jobs with machines, in turn boosting growth in productivity. [...] But the ONS analysis suggests the increase in employment over the past decade has not come from jobs that could easily be done by machines. Instead, since 2011 the UK lost jobs with a high or medium risk of automation, like shelf fillers, but gained them in areas with a low risk, such as physiotherapy.
Since the financial crisis the UK has enjoyed rapid growth in employment combined with one of the lowest rates of investment spending of any large rich country. Many economists have suggested that hiring cheap labour instead of investment in new techniques is behind the country's weak rate of productivity growth. Policymakers had hoped that increasing the minimum wage would spur companies to replace low-paid jobs with machines, in turn boosting growth in productivity. [...] But the ONS analysis suggests the increase in employment over the past decade has not come from jobs that could easily be done by machines. Instead, since 2011 the UK lost jobs with a high or medium risk of automation, like shelf fillers, but gained them in areas with a low risk, such as physiotherapy.
Employment has fallen in jobs that can be easily automated and risen in those which are trickier for robots, damping hopes that higher minimum wages could unleash a wave of investment in automation.
It's been this way since the start of the industrial revolution and even before. Some jobs get automated by new technology and those people have to find or create new jobs. As a species we're quite good at that. Generally the automation enables jobs that didn't exist previously. If you need evidence look at the very device you are using to read these very words. Smartphones are a multi-billion dollar business that didn't even exist 20 years ago. The web didn't even exist 30 years ago yet you'd be hard pressed to argue it hasn't been a net creator of jobs and prosperity. (note I didn't say a uniform creator of prosperity) That's not to say there aren't some bumps and bruises along the way for some people but in the end our society ends up better off pretty much every time.
There is this fear that somehow this time it will be different. That somehow devices have finally gotten clever enough to replace people permanently with nothing left for people to do. Problem with that idea is that it presumes there is a finite amount of economically useful work which is an idea that has never been true in the history of man. It also presumes that we have no control over automation politically, economically, or physically which also isn't true. One of our defining traits is that we are tool makers. Our tools enable us to do more than we could do without them. Instead of just making a product we use machines to help us make it so we can spend more time selling it or improving it or funding it.