Will robots and Artificial Intelligence ‘steal’ our jobs?
Let’s start with unemployment. No two experts can agree on how many people are at risk of losing their job to automation in the future, but some say it will be millions. Keynes’ work on technological unemployment is often hurled towards robot and AI enthusiasts. But proponents are quick to point out that this technology can solve some serious economic problems without having detrimental effects on employment.
Firstly, some of the sectors that are particularly prone to automation suffer from chronic labour shortages, or at least, employers have big problems with retaining labour (rubbish collectors, hospital porters and workers serving and preparing food, to name a few). Employees working these jobs frequently quit as they suffer from stress or work injuries. As a result, if these jobs are automated, business efficiency improves with no huge change to employment opportunities. Secondly, countries with ageing populations have rapidly shrinking workforces. Robots have proved to be an inexpensive solution, to not only replacing the gaps in the labour force but also for reducing the monumental debt burden faced by government.
Advocates of this technology also remind us that, historically, even as existing jobs were destroyed, new types of work were created, with no evidence to suggest that this trend won’t continue, although retraining might be needed. For instance, driverless cars demonstrate the balance between job creation and destruction. Although we are not quite at the point of truly autonomous vehicles, employees at Uber and people who drive delivery vehicles are at risk of displacement. Driverless vehicles struggle to deal with the unexpected so if an unpredictable situation arises, the driverless car will be helpless – creating a need for people to oversee this.
Self-evidently, demand for robots and AI itself generates new jobs, not just for producing them but also for maintenance purposes – technology is notoriously poor at preserving itself.
More importantly, though, job-destroying technologies indirectly create new jobs via their impact on wealth. In competitive markets, the reductions in production costs induced by labour-saving technologies are quickly passed on to consumers in the form of lower prices. Consumers then have more cash to spare, and new markets will be created to satisfy the resulting extra demand. There are far fewer people working in manufacturing today than there were a generation ago. But the fact that we no longer have to reserve huge chunks of our household budgets on e.g. household goods like fridges and washing machines means that we have much more to spend on e.g. dining out, joining a gym or going on holiday. Consequently, the number of people employed in these leisure industries has surged.
Those citing Keynes’ theory fail to mention that he actually saw technological unemployment as an opportunity. If the labour market is sufficiently flexible, a machine that can cut necessary human labour might lead to shorter working hours for the existing staff rather than redundancies. The future might be a future of more leisure rather than more unemployment – would that really be so terrible?
But what about income inequality? Employers of the future will require staff to be more digitally literate. Highly skilled workers can expect to do well, but what about blue-collar workers and white-collar workers in routine occupations? As technological change is skill-biased there is the very real possibility that, as the use of Robots and AI becomes more widespread, income inequality will increase. Some have even predicted the emergence of an unemployable underclass – ‘people who do not know what to study because they have no idea what skills will be required, who can’t work because there’s always a cheaper and better robot and spend their time living off the state’.
Alternatively, robots and AI can help in closing the gender wage gap. Prof J.R. Shackleton argues that ‘Men are more likely to work outside in all weathers, work unsocial hours […] and suffer much higher rates of industrial injury’. With robots able to do dull, dangerous and dirty work, this effectively removes the problem of men having higher wages because of compensation for less attractive work.
When we talk about widening income inequalities, we are not just talking about within-country inequality. The gap between developed and developing countries is predicted to widen – reasons as to why will be investigated in a future blog.
Income inequalities present a challenge, but we should not worry too much about the employment effect. There might not be much of an impact on overall levels of labour demand at all, and to the extent that there will be, it will probably be of the beneficial kind, leading to more leisure rather than more unemployment.