Jobs in construction are among those at “high risk” of being replaced in the next 20 years as a result of the onward march of technology, automation and robotics, the latest research has revealed.
The study, carried out by Oxford University researchers Carl Benedikt Frey and Michael Osborne for professional services firm Deloitte, used detailed analysis and insights from businesses to assess the impact of increasing digitisation and automation on employment.
It found that 35% of existing jobs in the UK are at high risk of disappearing over the next 20 years due to technology, covering roles in construction and extraction, manufacturing, office and admin support, sales and services, and transportation. The figure was 30% for jobs in London.
In contrast, 40% of jobs were considered at low or no risk of being replaced, covering roles in skilled management, financial services, computing, engineering and science, education, legal services, community services, the arts and media and healthcare. The figure was 51% for jobs in London, mainly because a greater proportion of the workforce is already in high-skilled positions.
"While the scope of potential job automation is rapidly expanding, London’s workforce is relatively resilient to these developments."
Carl Benedikt Frey, Oxford Martin School at the University of Oxford
Lower paid roles in the UK are most in jeopardy, with jobs paying less than £30,000 a year more than five times more likely to be lost to automation than jobs paying over £100,000, said the report. In London, the figure is eight times more likely.
In future, computerisation will no longer be limited mainly to repetitive tasks easily specified in software and performed by machine, the report states: “The capabilities of computers are expanding beyond routine work: as a result, tasks that were once considered too complex for coding will be converted into well-defined problems capable of digital solutions… Improvements in the ability of sensors to capture data are an example of this process of change: better sensors make possible further developments in robotics.
“The ability to collect and analyse Big Data allows Machine Learning algorithms to perform some cognitive tasks more efficiently and effectively than labour, making possible, and desirable, the automation of some knowledge work.”
However, positions likely to withstand the drive to automation include those that require a high level of perception and manipulation, in which people can see and respond to circumstances in ways that computers and robots cannot, as well as jobs requiring creativity and social skills.
The report found some evidence that London businesses are already planning to increase headcounts, attract new skills and redesign their workplaces to adapt to the predicted changes.
Of 100 businesses, based in London, asked how technological and labour market changes will impact their employment, skills and property needs in the next five years.
- 73% said they plan to increase their headcount in the next five years and half said they will add at least 10% to current staff numbers.
- 84% said the skills of their employees will need to change over the next ten years, with skills in “digital know-how”, “management” and “creativity” increasingly in need, and “processing”, “support and clerical work” and “foreign languages” less in need.
Where businesses said jobs will be reduced, automation was the leading causal factor, followed by nearshoring and then offshoring.
The dominant factors cited by businesses for London being attractive for their recent recruits were the financial package, the workplace environment and the ability to learn new skills.
Carl Benedikt Frey, James Martin Fellow, Oxford Martin School at the University of Oxford commented: “To remain a world-leading city, London needs to manage the transition of its workforce into new occupations and industries, as it has done so successfully in the past. While the scope of potential job automation is rapidly expanding, London’s workforce is relatively resilient to these developments.”
To compile the report, Frey and Osbourne adapted similar research done in the US, which identified three “bottlenecks” to automation that reduce the risks to jobs: the skills of perception and manipulation; creativity; and social intelligence. For each of 369 classifications of jobs, they systematically identified features corresponding to the degree of each of these skills and from this they estimated a probability of automation.
