What are the main social, economic and technological drivers of change in the labour market in the long-term?
What are the main social, economic and technological drivers of change in the labour market in the long-term? What skills and competencies will be required for the types of jobs that the US and UK economies will need in 2030? Where will the main skills gaps be and what can educators and policymakers do to anticipate them?
These are the questions that Nesta will be addressing in a new research collaboration with global education company Pearson, in partnership with Associate Professor Michael Osborne at the Oxford Martin School and independent researcher, Philippe Schneider.
Predicting future jobs and skills demand over a fifteen-year horizon is a daunting task. Think of the controversies that have arisen just around the effects of widespread automation, with estimates of the share of the US workforce at high risk ranging from 47 per cent in Michael Osborne’s 2013 study with Carl Benedikt Frey to 9 per cent in the case of a recent OECD study. In our research, we are taking a more comprehensive look at the drivers of jobs, including technological progress beyond computerisation, socio-demographic change and the turn towards more flexible and remote working practices. We will tease out how these drivers interact with each other when assessing their labour market impacts.
Fortunately, all the evidence suggests that there is a high degree of persistence in the occupational make-up of the workforce. That is, the composition of the workforce – a social institution – evolves only gradually over time. This is important, because it suggests that looking back at past experience is a good starting point for making judgements about the future.
So, in the first stage of our research we will conduct a historical analysis of how the US and UK workforces have changed over time. In the second stage, we will convene expert foresight workshops to label a training set of occupations as high, medium or low future demand as well as the extent of uncertainty around these labels. And in the third stage of the analysis, we will train a machine learning classifier to predict the likelihood of all occupations being in high, medium or low demand in the future, based on the modelled relationship between the occupation labels in the training set and occupational skills and tasks as captured by the US Department for Labour’s O*Net surveys.
By combining historical analysis, qualitative foresight and quantitative machine learning techniques in this novel way, we hope to present more accurate predictions about future jobs and skills demand than would be the case if we relied on any one of the approaches alone.
The potential importance of our research should be clear. Consider one popular estimate that 65 per cent of children entering primary school in 2016 will by the time they are economically active (in 15 or so years) work in completely new jobs that do not exist today. This makes it all the more important that we set learning priorities for young people today that are grounded in a rigorous assessment of what skills will be required of them when they enter the workforce.
Hasan Bakhshi is Senior Director, Creative Economy and Data Analytics, Nesta, and Project Director of the Employment in 2030: Skills, Competencies and the Implications for Learning study.
Image credit: Flazingo Photos via Flickr, CC 2.0