New technologies, including advanced computing systems and the automation of various job activities, will change about 30% of activities in about 60% of occupations, and will require new skills across most industries, management consulting multinational McKinsey Global Institute global economic and technology trends research head Dr James Manyika said on Friday.
Speaking at the University of the Witwatersrand School of Business, he noted that the activities – which represent only portions of various occupations – that are susceptible to automation are both high and low wage activities that are predictable and structured.
For example, activities that require some level of data collection and basic processing, such as office work, and physical work that is structured and predictable, such as certain manufacturing and manual work, are open to being automated.
“Predictable and structured physical work includes, for example, manufacturing lines. Manufacturing activities in the US that are susceptible to automation constitute about 50% of the pro rata wages of about $2.2-trillion, whereas in South Africa, these activities constitute about R600-billion or about 45% of wage-hours worked.”
He added that manufacturing employed only about 11% of the workforce in the US.
While about 10% of jobs, which include high wage jobs such as office workers and payroll specialists and low wage jobs such as cleaners and food preparation workers, may be made redundant because of automation, more new jobs will be created, in aggregate, he said.
However, the creation of jobs differs very broadly and is dependent on the availability of labour in the country. For example, Japan, with an ageing population, will see a reduction in occupations, such as teachers, while countries such as India will see a dramatic increase in these occupations.
“The rise of computing power, the availability of data to train algorithms and system-level innovation, including system integration advances, have produced machine-based systems that outperform a person at a specific task. An example is machine vision, which has recently achieved an error rate of 1% down from 26% in 2011, compared with the average human error rate of 5%.
“The productivity increases made possible by automation are critical for ageing populations where the labour availability is falling,” he said.
“While the normal model for estimating economic growth, as well as employment and wage growth facilitating distribution of resources in an economy, is the growth of labour and labour productivity, there has been a decoupling over the past few years where productivity improvements have not directly led to wage growth. This effect might require a new social contract for resource distribution.”
An example of this is the stagnation of wages over the ten years between 1995 and 2005, when the US only experienced about 2% wage stagnation, compared with 2005 to 2015 when about 70% of wages stagnated.
Wage polarisation could be exacerbated in advanced economies, but developing countries will see a growing middle class.
However, some economists worry about premature deindustrialisation in developing countries, owing to automation.
Manyika touched on some key factors that contributed to countries being able to sustain good economic growth over the past 50, 20 and 5 years, which can also help underperforming economies, such as South Africa, tackle its high unemployment rate and how it adapts its economy to the impact of technological advances.
“The most important question is the question of inequality and the Fourth Industrial Revolution. Growth depends on skills, and growth remains fundamental to economies. Countries that maintained consistent growth typically had a pro-growth agenda, maintained low inflation and increased government efficiency and thereby resource allocation.
“Further, these countries typically increased their participation in the global economy. An example is the difference between India and South Africa. India maintains high levels of connectedness to the global economy, whereas South Africa has much lower levels of participation in the global economy.”
Strong competition within the private sector, and generally a robust small and medium-sized enterprise sector, were also features of the countries that maintained good growth. However, even if sectors were concentrated, as long as there was strong competition between the large private companies the growth trajectory held firm.
In developing countries, the largest number of new jobs will be those requiring secondary education or technical skills.
While many people may have to switch occupations, additional labour demand trends will offset more job displacement owing to automation within occupational categories.
“There are no specific barriers to retraining people. However, one key feature we have identified in the success of training centres to produce skills required by industry is their geographic proximity to the industries that they provide training for.”
Meanwhile, Manyika also highlighted a non-technological improvement that could support economic growth, which was gender parity.
Gender parity across the world was generally low, with no country that was studied exceeding 74% of gender parity. Full gender parity would add $24-trillion to the global gross domestic product.
However, a more achievable estimate is that if countries worldwide can achieve gender parity equal to their best performing neighbouring countries, $12-trillion can be added to global gross domestic product.