Jobs disappear and robots take over. Or everything remains as usual. These are two completely different scenarios for the future of the labour market. They cannot both be correct.
There is fairly broad consensus among economists that technology has so far been positive for work, but the question is what happens in the future. During the industrial revolution machines destroyed jobs. Among artisans, several crafts were replaced by machines and less skilled workers. But throughout the 20th century, technology and jobs have mostly reinforced each other and have led to need for better skills, higher wages and productivity.
Despite comprehensive automation and restructuring of the whole economy, work has not disappeared. However, there has been significant job polarization in many OECD countries in recent decades, resulting in a shrinking middle class.
In Sweden, this trend has so far taken place mostly through an increase of highly-paid jobs and strong real wage growth over the past 20 years. This is in contrast to the United States, where large groups have not received any increase in real wages over long periods of time. It is likely that the trend of job polarization will continue.
Is there a risk that large groups in Sweden will experience weak wage growth and that robots will take over? This will likely depend on how the challenges are managed – by politicians and by the labour market partners. In the long run, technological developments will lead to greater prosperity, but the path could be tumultuous if unwise decisions, or no decisions at all, are made.
READ ALSO: Why Sweden is a long way from six-hour days
People understandably treat the claims that robots are taking over with some scepticism, since many may not see any direct evidence of this in their own workplace. But the idea of a robot sitting at a desk (with a coffee cup) is misleading, when much automation is in fact about software in the cloud, physically located in some anonymous server hall. It is worth bearing in mind that today's smart phones have replaced a variety of devices and services you previously had to pay for: GPS navigation, music players, scanners, voice recorders and so on. The applications available via cloud services for managing huge amounts of data are well developed and could easily be used to replace human labour. Most people will probably be surprised at how much actually already has and easily could be automated.
Automation of simple services continues, but what is new is a wave of automation also for skilled work. It covers everything from simple services, legal services to administrative services. Research has identified the potential to automate jobs where tasks are repeated in a predicable manner and can therefore be programmed. Think about what tasks you encounter every day that are more or less predictable. They are increasingly often going to be automated.
Some real-life examples:
- Restaurants and hotels almost without staff; a machine cooking several hundred hamburgers per hour.
- Software can replace great numbers of junior lawyers, can grade essays in school and automatically generate text so well that it is difficult to see that it was not written by a human.
How should Sweden stop jobs being lost to robots? Photo: AP Photo/Peter Dejong
The limit of what can be automated can these days not even be drawn at the tasks that require creative or cognitive ability, such as composing music or researching empirical relationships.
The main driving force behind the ongoing trend for automation is the aspiration to create something better and more efficient. In Sweden and several other OECD countries this aspiration is reinforced by the fact that we are entering a period when many people are exiting the labour market for reasons of old age, while fewer young people are entering. Combined with skills mismatch in the labour market, this provides a strong incentive for automation. It is not a coincidence that Japan, with the oldest population in the world, may have taken the most steps in automation in the sector devoted to care of the elderly.
But even if technology allows an increasing degree of automation, other factors can sometimes be more significant. A considerable part of school education, for example, could be automated, but social norms and values can be more important. In some countries in Asia there are driverless trains, but so far no autonomous passenger flights. In healthcare, the demographic trend is probably a more important a factor than automation. A small country such as Sweden, with its own language and own institutional framework, also does not enjoy the same kind of benefits of economies of scale when it comes to automation.
What then is the sum of the various driving forces? Will 50 percent of today's professions disappear? This figure was presented in a report by Oxford researchers Frey and Osborne. The short answer is that this picture is incomplete and does not take into account that new jobs are being created all the time. Some existing jobs increase and new jobs that we cannot even imagine today are created. The question should instead be if the new jobs are being created at the pace needed.
There is a real risk that we are facing a period of increased tensions in the labour market, particularly as wage bargaining for large parts of the economy is under way at a time with low confidence in the Riksbank's (Sweden's Central Bank) inflation target. The biggest income inequality is created by long-term unemployment, a risk that increases in times of rapid change. To ensure positive development while taking advantage of new technology and reducing its downsides, the following should be done:
1. Reduce the cost of labour by cutting taxes on (human) work. Sweden's high tax on labour unnecessarily increases the already strong incentive for automation. Tax deductions for household services should increase in scope rather than being slimmed down.
2. Facilitate the so called sharing economy by reducing uncertainty in how regulation is applied. Allow competition in the first instance by reducing unnecessary barriers rather than forcing digital companies into old regulation unsuited to a digital world.
3. Improve the opportunities for lifelong learning and make it easier for people to accumulate new knowledge. Without new knowledge the risk of being overtaken by technology increases, resulting in a risk for poorer prospects in the labour market.
4. Reduce the difference in social security systems between employees and self-employed.
These measures are good for the economy irrespective of how strong digitalization turns out to be. There will be new jobs when the robots take over our old ones – provided that politicians do not construct unnecessary obstacles.
This is a translated version of a debate article which originally appeared in the Dagens Nyheter newspaper.
Mårten Blix, PhD in Economics, is a guest researcher at the Swedish Research Institute of Industrial Economics. He is a former secretary in the Commission on the Future of Sweden. He is working on a report on the effects of digitalization on the economy.