The last few years have seen a number of reports attempting to explore how the future of work will unfold, and one of the most commonly cited was that produced by McKinsey in 2017.  The report predicted that while less than 5% of occupations were fully automatable, some 60% of jobs had tasks that could be automated around 30%.

Of course any such report inevitably includes a large dose of generalization, as automation is likely to apply differently in different industries, and indeed other reports have examined the potential impact of automation in different regions, skill levels and even classes.

It’s perhaps no surprise, therefore, that McKinsey should return to the future of work with a new report that attempts to fill the gap and explore whether the risk of automation is shared equally between men and women.

Gender differences

The report is largely a rehash of the data used in 2017, but with a particular focus applied to the female portion of the workforce.  It suggests that women might actually fare better in our age of automation, with the data suggesting that they have a slightly lower risk of their work being automated than men, with the potential for moving into new jobs at a similar level also higher than men.

As with any form of disruption however, there will be an inevitable period of transition, and the authors suggest that women may struggle to make this change because they are less mobile than men, and may lack some of the technical skills men have.

These challenges perhaps underpin findings revealed in a new report from IE Business School looking at AI in the workplace, which showed that women in the UK were considerably more concerned about the prospects of AI-based technologies taking their work than their male peers. This was perhaps explained by stark differences in knowledge about AI, with approximately 50% of men saying they were confident in their AI expertise, versus just 32% of women feeling similarly confident.

“In addition to internal communication, the need for upgraded investment in people and training is a major outcome from this study,” the authors explain. “Not only do employees need to feel safe and understand more about AI so that they can maintain higher motivation and morale, they also need to feel confident in engaging with this new technology.”

Skills gap

McKinsey estimate that between 40 million and 160 million women around the world are likely to need to make a career transition of some kind as a result of AI-based technologies.  A hallmark of the 4th industrial revolution is the hollowing out of the workforce, with an ever greater premium placed on high skill levels.

Indeed, recent data from the Stanford Center on Poverty and Inequality highlights how unskilled millennials earn around $2,600 less per year than Gen Xers, and nearly $10,000 less per year than Baby Boomers at the same age, with technology playing a major role in this discrepancy.

The perilous nature of low skilled work was also emphasized in research from the JRC and the University of Salamanca, which looked at automation and migration.

The report explores the kind of jobs that are currently being done by migrants, and then examines the make up of those jobs in terms of the tasks they perform.  It finds that as many as 50% of migrants, especially those from outside the EU, are performing tasks that are not only relatively easy to automate, but also commercially worthwhile automating.

“Both EU mobile citizens and third country nationals have a higher likelihood of being employed in jobs with high automation potential than nationals, even when socio-demographic characteristics are taken into account,” the authors say. “However, the likelihood decreases as educational attainment increases, for all, but more so for migrants.”

Indeed, research led by Oxford’s Carl Benedikt Frey went as far as to suggest that Trump and Brexit supporters are more susceptible to automation, with those whose jobs have been disrupted by technology significantly more likely to support Donald Trump in the 2016 presidential election.

Risk of automation

Whether work is held by migrants, women or Trump voters, the characteristics are likely to be similar, as research from Stockholm School of Economics pointed out last year.

The report outlines five factors that are believed to be crucial in the pace and extent of automation:

1. Commercial availability – many of the technologies hitting the press are confined to lab conditions, with few commercially available. Indeed, many are at a very early stage of their development, and it remains to be seen how successful they will be at scaling up. The differences between technical feasibility and commercial adoption mean that many will fall by the wayside.
2. Cost of implementation – even when technologies have made it to market, the cost of implementation will remain a factor, as the new technology will need to present a suitably robust use case to warrant upgrading from the existing tech.
3. Economic Benefits – this use case analysis will also examine the economic benefits of implementation. The rhetoric makes it sound very easy, as humans will be displaced, therefore the technology must be more cost effective. As full substitution is unlikely, it renders the economic equation slightly more complex, especially as higher profits can be re-invested in new areas and thus create new work.
4. Labor market dynamics – the adoption of technology is influenced heavily by the labor market dynamics in a region. For instance, in Japan demographic pressures are resulting in a shrinking workforce that is forcing many employers to turn to technology. Equally however, despite the risk of automation being high in industries such as food, the low wages of workers have thus far held back substitution of humans for machines.
5. Social, legal and ethical acceptance – the final factor concerns the social, legal and ethical acceptance of autonomous technology. This is likely to be one of the most important factors, certainly in determining the pace of change. The legal pace of change is typically a lot slower than technological change, and social attitudes can be equally slow to adapt. This can add years to the roll-out of any new technology, even when it’s technologically and economically ready.

“All five of these factors have a significant influence on the speed and scope of technology adoption. In particular, a lack of applied research, low wages, high costs, and legal and ethical boundaries hamper the adoption of technology,” the authors say.

These factors are less beholden to certain characteristics of the job holder as they are of the job itself, and as Frey points out in his latest book The Technology Trap, these are not disruptions that are unique to our current time.  Whereas previous periods of disruption have failed to provide people with the support to help them adapt to technological disruption, we are beholden to try and break that cycle and do far more to help people adapt to the changes facing them.

The political disruptions around the world suggest we have not gotten off to a great start, but it’s not too late, providing we start to act now rather than continue sitting on our hands.