20th Century policies may not be enough for 21st Century digital disruption

March 31, 2017

     By Duncan Green     

It’s often a good sign when you rock up at a conference and hardly know anyone there. That was my experience at a

Sango Patekile Holomisa, Minister of Labour, South Africa, opening the conference

Sango Patekile Holomisa, Minister of Labour, South Africa, opening the conference

recent, rather grandiosely-named, ‘Digital Development Summit’, hosted by IDS, Nesta and the Web Foundation, which clearly got people’s attention – the places were fully booked within a day of going live. Participants were diverse: developing country ministers, donor officials, tech company execs, AI pioneers, and civil society types like me.

The topic was ‘the future of work in the digital age’ (see the IDS background paper for more details), and I got to listen to a day of presentations, taking in both the substance and the mood of the discussion. This topic, more than most others, attracts both tiggers and eeyores (for Winnie the Pooh fans – optimists and pessimists, if you’re not). The tiggers bounce around telling stories of amazing start-ups in African slums, or how drones are about to start delivering aid; the eeyores sigh and say ‘every driver in America is about to lose their job to driverless vehicles, and developing country jobs are even more at risk’. Both were represented in roughly equal numbers (and some people seemed to move from one to the other at a startling rate). tiggereeyoreWho’s right?

The pessimists have to show why this time is different – people have been predicting the end of jobs since the Luddites, and yet new sectors and jobs have emerged as fast or faster than those being destroyed. The most cogent argument was that automation, artificial intelligence and ubiquitous algorithms are different because so many sectors are being simultaneously affected, and because the pace of job destruction is outstripping the pace of skills upgrading.

The optimists, on the other hand, have to show that their technologically-driven visions can address some of the fundamental divides that we know are being deepened by the digital revolution. And there is also a need to address the fact that developing countries are most prone to the impacts of automation, and least able to build resilience to shocks and stresses.

I got my favourite task – randomly reacting on the final panel of the day to what we had heard from the actual experts. Here’s a potted summary of what I said:

Reducing vulnerability versus enhancing agency: The opening speakers talked about the need to keep people’s agency at the centre of the debate. There was a great exercise (which I missed, but which participants raved about) which used ideas from human-centred design and scenario thinking to get participants to think about how different kinds of jobs might be affected by technology, using this 2×2 matrix.Ben R 3


While it’s clearly not an either-or, the discussions tended to focus on reducing vulnerabilities, rather than enhancing adaptive capacities. Having identified different groups of workers who would be most (negatively) affected by technology, the debate was on what could be done to reduce exposure through measures like social protection, education, and so on. There were also some ideas generated about enhancing adaptive capacity through digital technologies: for example, how digital tools could be used to enhance collective action and bargaining for digital jobs; or used to expand access to education opportunities. But these seemed to be in the minority.

What’s the critical juncture? While there is no unanimous agreement on whether robots and AI will destroy or create jobs overall, what is clear is that the current wave of technologies – from mobile phones to uberisation of the marketplace – is contributing to falling wages, living standards and job security. Do we need some kind of global wake-up call to take account of this hidden reality? I suggested that this we need a digital equivalent of the 2008 financial crisis in order to see the real impacts and trigger genuine policy engagement and change. Until then the discussions about regulation – whether voluntary, soft or hard – will remain largely theoretical.

Compensate digital ‘losers’ from the profits of digital ‘winners’: there are many ways of thinking about this, see my previous post, but most would involve some kind of taxation and compensation. Along with my suggestion for digital transaction taxes, robot taxes got a lot of air time, as did novel approaches to social protection such as Universal Basic Income.

Monkey to robot20th century solutions for 21st century problems? But if even half of the predictions of the eeyores are correct, these kinds of measures hardly feel up to the seismic challenges under way. What is the point in circling the wagons around shrinking pockets of paid, non-automatable jobs, if they are all destined for the scrapheap in the end?  It may be that something far deeper is changing, such as a definitive sundering of the links between consumers and producers (previously both people, now a world with many human consumers but very few human producers).

Much of the discussion of this new and emerging challenge seemed to be based on 20th century ideas of social, government and corporate policy – social protection, education, taxation, regulation and so on.

But we could be facing a tipping point, the moment that will finally overwhelm the fraying edifice of traditional economics, opening the way for a return to a ‘moral economy’ based on  broader understanding of ‘value’, which includes culture, leisure and care at its centre, rather than as an afterthought. After all, even if robots pick the kids up from school, aspects of the care economy are likely to persist, and could well end up at the heart of a post-automation division of labour. We do the loving, the singing and the sport, while the robots get on with the making. Sounds good to me (but then you haven’t heard me sing).