The IMF and OECD have drawn attention to the fact that, for the first time since the industrial revolution, the rate of technological change is increasing faster than new job opportunities are being created.
Unfortunately, when these warnings are distilled into media commentary they too easily become predictions of catastrophic job losses, especially in manufacturing. We should heed instead the example set by democracies where policymakers are intent on unleashing the opportunities in technological change while containing the risks.
Automation does not have to mean that the future will be a sci-fi nightmare in which robots perform most of the work while humans are reduced to precarious casual employment in sweatshops. Avoiding that will require action by governments, and it will require workers being properly engaged in deciding the future of work.
This is increasingly widely recognised. Even Elon Musk has made the extraordinary assertion – for him – that to avoid disaster governments need to regulate effectively.
In deciding how to shape the future, we must rely on fact rather than the predictions of doomsayers. We must challenge the notion that most jobs will disappear at the rate than some people have been predicting. The evidence to sustain such claims just is not there.
We frequently hear that more than 40% of existing jobs will be gone by 2030. Sometimes it is nearly 50%, depending on who is doing the doomsaying.
The source of all this gloom is a much-cited paper published in 2013 by two Oxford University researchers, Carl Frey and Michael Osborne. They argued that “about 47%” of jobs were at risk from computerisation. Their estimate went global, and was picked up in a CEDA report that has fuelled the media commentary here in Australia.
The disruption predicted by Frey and Osborne is real enough. But citing a bald figure, whether 40, 47 or 50%, can be seriously misleading.
David Autor of the Massachusetts Institute of Technology has argued that although automation may reduce average employment (the number of jobs per unit of output) it does not necessarily reduce aggregate employment (the total number of jobs).
According to Autor, robots and computers substitute for human workers in repetitive tasks while increasing the value of tasks that workers uniquely perform – tasks involving problem-solving, adaptability and creativity.
That is also the view of the International Federation of Robotics, which argues that in future manufacturing robots will complement and augment labour. Robots will perform routine or dangerous tasks while raising the demand for higher-skilled forms of human labour. New lower-skilled jobs may also be created, through spillover effects in other sectors.
This is not wishful thinking. It is already happening in some advanced economies. The Centre of European Economic Research has reported that between 1999 and 2010 the replacement of routine work by new technology actually raised labour demand across the 27 EU economies by 1.9 million jobs.
It is essential that this bigger picture be kept in mind each time the Frey and Osborne research, or some garbled media version of it, is mentioned.
The European Union argues that with the correct policy settings opportunities can actually be expanded in the new age of automation. The EU has set a target: by 2020, it wants 20% of total economic value to come from manufacturing. That is a 25% increase on present levels.
This is no rosy-eyed optimism. EU planners accept that there are difficult problems to be overcome, such as deciding the amount of investment needed to attain the target.
Questions about legal liabilities, intellectual property and the harmonisation of standards will have to be resolved. Above all, there are questions about the transfer of skills that will be required in the transition to the smart factories of the future. That includes questions about how people at the bottom end of the labour market can be reskilled.
But, even with these acknowledged difficulties, EU policymakers are still talking to the public about opportunities. The Germans, in particular, estimate that there will be a 6% increase in manufacturing jobs in the next decade.
The Germans are already demonstrating that smart factories can be compatible with lower unemployment. The unemployment rate in Germany is 3.9%, and the number of robots per 10,000 workers is 300.
Compare that with Germany’s neighbour, France, where the number of robots per 10,000 workers, 126, is much lower but the unemployment rate, 9.5%, is much higher.
Australia should heed the German example. We cannot be luddites, rejecting new technology. But nor should we assume that technological change means an insecure future for manufacturing and those who work in it.
Senator Kim Carr is the Shadow Minister for Innovation, Industry, Science and Research. This article is adapted from his speech to the Australia Institute’s Future of Manufacturing Summit, held in Canberra in June.