Automation and Artificial Intelligence (AI) are clearly the major game changer in the world economy today, and will be for decades to come. Society, and the labour market in particular, have to take precautions to manage the challenges that stem from these developments, rather than to try to bring back manufacturing jobs, which themselves are facing a gloomy future.
The Trump administration’s job market policy
The Trump administration is convinced that it is because of international trade deals putting the interests of the insiders of the Washington elite first, that “blue-collar towns and cities have watched their factories close and well-paying jobs move overseas, while Americans face a mounting trade deficit and a devastated manufacturing base.”
Looking to the manufacturing sector, this observation is actually true. Between 1990 and 2015, US manufacturing has lost 30% of its jobs. Similar development patterns can be observed in other advanced economies like the UK.
On another point, the Trump administration is right in that job losses partly occurred due to import competition from China. According to research findings between 1999 and 2011, a 2-2.4 million job loss resulted from the pressure Chinese imports put on the US manufacturing sector.
Therefore, Trump concludes that renegotiating better trade terms for the US will bring manufacturing jobs back to American shores and thereby increase wages. Accordingly, Trump has set ambitious mid-term goals: 25 million new American jobs in the next decade, and a return to 4% annual economic growth.
How does Trump plan on getting the jobs back?
One case in point is the example of Carrier. The CEO of Carrier, Greg Hayes, after being pressured by Trump – by way of ‘Twitter-diplomacy’ – not to move manufacturing jobs to Mexico, decided to invest $16 million in a plant in Indiana, allegedly saving some 1000 jobs. This is definitely good news for the workers at the factory.
The second part of Hayes’ statement is telling, however, for the future of the manufacturing sector. In the same breath, Hayes announced that in order to be competitive the $16 million will be used for automation. As a result, fewer jobs will be available at the factory in the future.
Trade agreements or automation: Which is the biggest job killer?
A recent study by Ball State University indicates that roughly 13% of US job losses can be traced back to increased trade competition. The biggest chunk of jobs lost, however, is associated with productivity growth, which is often associated with automation.
Automation and labour market: Are more jobs being created than destroyed?
Researchers disagree on the pace and impact of automation on the labour market. One main current of debate argues that productivity gains and secondary effects ultimately benefit society as a whole, whereas another current sees robots replacing human labour in the long term, simultaneously anticipating a rise of unemployment that will accelerate impoverishment. However, there is consensus about the fact that a substantial number of jobs are going to be automated, while a number of new –mostly highly skilled – jobs will be created.
Addressing Automation: Is good education enough?
No. It goes without saying that it is crucial to increase investment in high-quality education, beginning with early-education and continuing with the retraining of workers and lifelong learning, so as to prepare people for the requirements of the future labour market. But taking into account what a challenge it is even for advanced economies to improve educational upward mobility (upward mobility has been slowing down in OECD-countries), this isn’t an easy task.
An OECD study suggest that 9% of all US jobs face a 70% risk of being automated in the near future. What is more, people with primary or lower secondary education are at a much higher risk by virtue of working in occupations that are automatable. So possessing a high level of educational attainment is still currently the best safeguard against being automated out of work.
The fact that due to steadily increasing computing power, technological change is proceeding at an exceptional pace, leads to the question of whether retraining or education in general will be able to keep pace with ever-accelerating technological progress.
A third area which needs further attention is the possibility that future AI will increasingly advance into non-repetitive tasks and high-skill occupations, replacing parts of the human labour force as robots become more sophisticated, possibly even rendering high educational attainment increasingly obsolete with respect to successful labour market participation in the long term.
Can American manufacturing become great again?
In terms of the numbers of humans employed in the manufacturing sector the answer will most probably be ‘No’.
Occupations that are instead expected to witness a sharp rise in demand are associated with activities in design, construction, programming, maintenance of intelligent machines, which could result in further job opportunities in professional fields such as engineering, IT, and teaching. These are occupations which typically require a high level of skill.
What should human societies do?
In various media, a black and white picture is often painted where readers are urged to either fully embrace technological change, with all its allegedly positive consequences, or where modern technology is seen as a job killer, and automation-induced changes that put even the most repetitive occupations and tasks at risk are rejected altogether.
While neither of the aforementioned options – fully embracing technological change led by pure market forces without any regulation, nor becoming 21st century luddites – will work, human societies certainly need to reconsider which role robots and AI should play in their societies.
If we assume that the future human labour market will contract and that humans will not only be complemented, but will also be competing with intelligent machines, the question is whether the average human being can still earn a decent salary from participating in the labour market. Looking at rising inequality and wealth concentration in advanced economies, as well as at the increasing influence and monopoly-like power of global companies like Google, Facebook, and Amazon, the answer could be ‘no’.
The core issue that needs to be addressed is the distribution of wealth within societies. As participation in the labour market may no longer guarantee a decent living, an unconditional income could be part of the solution. This would give humans the possibility of leaving a number of repetitive or uninspiring occupations, and devoting their time to something more meaningful. Moreover, a tax on labour carried out by intelligent machines could be imposed, and companies could be legally obliged to share profits with their employees. All of these solutions deserve more thorough analysis, but have the potential to form part of a human society where automation-induced productivity gains benefit all members of society and not only higher income groups.
An Expert Comment by Klemens Witte (LL.M. oec. int.) on the implications of automation for the labour market will be published in the Dialogue of Civilizations Research Institute’s newly launched E-library. It will critically assess policy recommendations developed by decision-makers, and elaborate on the concept of ‘fair automation’.