Automation brings changes in the techniques of production that are considered a central ingredient in improving economic efficiency and increasing productivity. However, it also works with a massive force effecting structural transformation in the economy. These structural transformations affect the availability of jobs across genders, age groups, and skill sets. It also affects the share of labour in incomes, and is giving rise to new inequalities owing to divergent levels of human development across and within countries.

However, research results show a mixed picture of the impact of automation on jobs. A recent study by McKinsey claimed that “about half” of the activities in the world’s workforce “could potentially be automated by adapting currently demonstrated technologies.” However, they also claimed that only around five percent of jobs can be “fully automated.”

Although the recent surge in robotics has rekindled the fear of job losses, the historical data does not warrant worries about any massive and chronic joblessness. Having looked at industries like textiles, steel, and auto manufacturing in North America, it has been concluded in research studies that if demand for the products remains high then job losses are not daunting.

The channel of productivity gains ultimately helps lower the prices and increase demand for the products. But it is not fair to conclusively conclude that automation does not negatively impact jobs. It creates winners and losers. Those who can acquire new skills are less affected by machines. As Alvin Toffler once predicted, the main task in the future shall be to learn, unlearn, and relearn. This shall be dynamics of new literacy.

James Bessen, Boston University economist and researcher, claims that the alarmist predictions of job losses are false. He argues that advances in technology generally lead to more and diversified jobs, although the nature of work changes overtime. He brings in data about proliferation of ATMs but shows that it has not reduced the employment of tellers. He also argues that “automation of textile mills in the 19th century led to an increase in weaving jobs because it created more demand”. “Robots can replace humans in certain tasks but don’t entirely replace humans,” he says. But there is a caution that that automation “is destroying a lot of low-skill, low-wage jobs, and the new jobs being created need higher skills.”

Let us have a look at the jobs from various skill levels. For example, in those sectors which do not use information communication technologies (ICTs) as basic means of wealth and value creation, low skilled and females get employment, however ICT-based firms create employment for almost all kinds of workers. Interestingly, the software industry is a little discriminatory. It favours the highly skilled and does not accommodate low and medium skilled workers including the very young ones.

Just like the software sector, import of industrial robots decreases employment in the manufacturing sector for the low skilled workers. However, robot-intensity increases the income share of high and medium skilled workers, the old workers, and men.

Those who lose in the process are the low-skilled workers, but they are compensated in those sectors which emerge as incomes increases takes place, and new occupations crop up in the shape of new jobs. Such vacancies are often advertised in the services sector where automation has not really gone deep and low-skilled workers are needed to maintain services.

[Automation] has the potential to create new inequalities, lower the income share of labour, and possibly cause social and political disruptions; asking humans to think innovatively and create stable societies. 

Is there a China factor impacting automation in different economies? Such a question is always asked with curiosity. Yes, there is an impact of the rise of China, and the impact is through the channel of international competitive pressure. For example, Danish manufacturing has traditionally been more manual than automated. However, the firms exposed to increasing competition from China in their product destination markets have shown a tendency to become more automated. This trend for investment in automation has increased since 2010.

There is a growing concern about the relative income share of labour and capital owners. Here the issue is not jobs being replaced by robots and automation, but becoming less rewarding for workers as compared to the capitalists. Many researches by international organisations such as the UNCTAD, and independent scholars show that during the last 30 years, the income share of labour has gone down both in the advanced and emerging economies. It is interesting to note that according to the neoclassical economics the factors (labour in the case of emerging economies) which provide comparative advantage must be rewarded more. There are some research studies which show that in advanced economies routinisation and technological progress have given advantage to capital over labour.

However, in less developed economies the cause behind low share of labour in income is global value chains which reward greater capital intensity.

What can be the potential implications if automation starts creating “superfluous people”? In the words of Yuval Harari, “wealth and power might become concentrated in the hands of the tiny elite that owns the all-powerful algorithms, creating unprecedented social and political inequality.”This can potentially create social disruptions, and this is why the argument for “universal basic income” is getting stronger.

It is being argued that universal basic incomes may allow people to be more creative and follow their passions. However, these are estimates and jury is still out on the potential impact of robust automation and robotisation of work. There is clearly potential to create new inequalities, lower the income share of labour, and possibly cause social and political disruptions asking human to think more innovatively and create stable societies.

The writer is Executive Director IMPACT Research International