The calculation of probabilities of automation for individuals on the Annual Population Survey (APS) allows us to produce detailed demographic breakdowns, by taking the average probability of automation within each demographic group. To produce probabilities of automation at four digit occupation level on the APS, we created a pooled dataset of seven years of APS data (2011 to 2017), to ensure the sample sizes were large enough to calculate a probability for each four digit occupation code.
We then group the probabilities of automation in three categories: low risk of automation for probabilities lower than 30%, medium when the probability falls between 30 and 70%, and high risk if the probability is greater than 70%.
In 2017, out of the 19.9 million jobs analysed in England, 7.4% (1.5 million) people were employed in jobs at high risk of automation; a fall in absolute terms of 46,000 employees, 0.7% fewer when compared with 2011. Conversely, the number of employees that were in jobs at low risk of automation in 2017 was 5.5 million, equating to 27.7% of all employees, a rise of 2.4% since 2011.
Chosen excerpts by Job Market Monitor. Read the whole story at The probability of automation in England – Office for National Statistics