When predicting the future, the how is as intriguing as the why
Every government sets goals and the better ones also set out the trajectories by which they will be met. When I did this for the UK government we didn’t have available an approach to target setting that factored in robust predictions of the future, although the approach of thinking in terms of systems and the techniques of foresight were being developed. For example, through the work of my then colleague, Geoff Mulgan.
One of the most exciting projects that I am involved with in my work at Pearson is applying the tools of prediction to better understand the skills, and skill combinations, that the US and UK economies will likely demand in 2030. The research is a partnership between Pearson, Nesta and Michael Osborne of the Oxford Martin School. Eventually our hope is that the insights we surface will provide a prompt to think about the future of learning. For example, if particular skill combinations are especially promising (the ability to problem solve combined with the ability to lead a team, say) what does this imply for the pedagogies deployed in our classrooms, or the way we approach learning in our universities?
There are three features of the methodology used in this project that I want to highlight and recommend.
The first is the recognition that the past has a lot to tell us about the future.
This is something that I’ve always found immensely valuable. For example, when I began my education reform work in the Punjab province in Pakistan one of the first things I did was to read about the country’s history so that I could understand the origins, and context, for the problems we were about to tackle. Similarly, when it comes to predicting the future of jobs, it is wise to acknowledge that occupational change is historically slow moving, and that the fears of technology induced joblessness are not a new phenomenon.
The second is the serious attention given to cataloguing a wide-set of the trends that will shape the future demand for skills.
As well as technological change and automation, we have mapped trends such as demographic change, globalisation, climate change, urbanisation and political uncertainty. This comprehensiveness in turn allows us to attend to how the trends will interact. For example, will an ageing society change the mix of jobs to favour the types of tasks that are less automatable — for instance, human empathy and caring?
We have now held two foresight exercises where experts have been invited to apply historical data, alongside the catalogue of trends, to predict the future demand for occupations as currently defined.
This allows the third feature I admire:
This analysis is anchored in the data, allowing quantifiable predictions to be generated.
It seems to me that something like this exercise needs to be built upon, extended, and the methods used the subject of more experimentation and testing. Other intriguing possibilities come-to-mind. Could something like this be institutionalised as a biennial exercise, a recurring scan of the future and an informed assessment of how well placed education is to prepare students for that? Or, how could foresight be linked to other data-sets, such as the OECD’s data on the learning achievements of 15 year olds across over 70 countries, in order to generate a view of the gap between current learner achievement and what is needed in the future?
I hope that the insights that this project surfaces will help with that — as importantly, I hope that its approach — the how — can be developed, innovated upon, and eventually become a standing function of what any good government should do: wisely plan for the future.
Find out more about this project.