Speaker: Torsten Bell, director, Resolution Foundation
Chair: Lucy Powell, UK Labour member of parliament and shadow secretary of state for education
Widespread ‘robot angst’ – how much should we worry?
Future fear: Frey & Osborne: Likelihood of jobs being automated (not just low-skilled jobs)
What has happened in the past? Jobs have gone in the middle and been created on top and bottom (increasing inequality). More lousy jobs and more lovely jobs but less middle jobs. This changes over time. What kind of jobs machines can do changes over time.
Strong relationship between prediction and actual trends.
Don’t panic (yet). Here’s why:
- OECD estimates are much lower than Osborne & Frey
- BCG thinks that only 10 per cent of cars will be driverless in 10 years. New tech has long lead time
- Mostly machines don’t replace jobs they change jobs
- Employment is not falling – indicating new jobs are still being created in higher numbers than jobs go away (net growth)
Is the labour market being polarised? No, not true, at least in the UK. Wages are not being polarised. Most of the earnings inequality increase happened in the 1980s. Hasn’t really come back. Also wage growth has stopped expect for at the very top (but this is not driven by technology).
Mid-skill jobs are mainly being replaced by high-skilled jobs. Meaning people just have more education and better jobs
Occupational structure change is complex, eg less secretaries and manufacturers, more customer service and health workers/ personal care. Also more business and service professionals but at lower wages. The old top is the new middle. See presentation.
But don’t be complacent either:
- Automation is real and is driving job transformation over time (technological unemployment). Strong argument for lifelong learning and other public policies
- Pain is not equally distributed – calls for distribution policies
In fact the problem is probably not enough robots, not to many
- Technology is not just changing jobs but whole value chains and industries and work processes.
- Will return to capital increase more than return to labor and as a consequence increase inequalities? So far earnings inequality explain a larger share of inequality but this will change. Then we probably need to tax robots or nationalise robots
- More precarious work? Gig economy is quite small but not clear that labor laws now are designed to deal with this
- Can we use robots more actively? Can they solve some of our problems? Digital health and care? Shared technology platforms to foster ne business growth? Can technology in schools help with literacy training and other social challenges?
- Our historical role was to deal with technological change. Have we forgotten this and how can we increase resilience?
- Progressives should offer policies to say what are the opportunities? How can we use technology for the common good? And what are the risks and subsequent policy needs?
- If wage growth and productivity is stagnant firms have less incentives to invest in technology
- Automatization requires skills investments
- The digital economy is difficult to measure – maybe productivity growth is actually slowing because improvements do not show up in the numbers. How big is this chunk? Hard to tell but seems unlikely that slowing productivity growth is all about measurement problems.
- Are we using the technology right? Or just making improvements in social life but not putting technology to good use in the economy?
Rapporteur: Sigrun Aasland