The impact of automated technologies on the workplace has been a topic that I’ve covered numerous times, not least via a recent study led by MIT’s Eric Brynjolfsson, which argued that the ‘machines are taking our jobs’ narrative at the moment is excessively simplistic and the reality is that some tasks will be automated, but it will be incredibly rare for entire jobs to be taken by machines.
This more nuanced perspective is something I’ve advocated in all of my articles on this topic, with a general theme that education is a priority, both in terms of developing new skills to work effectively alongside machines, and to help people adapt if their profession is impacted by new autonomous technologies.
It’s on this general topic that a new report from the Center for Global Development think tank focuses. As their name suggests however, rather than focusing on how society should respond to automation in the developed world, it instead explores the situation in the developing world.
The general crux of their argument is that the developing world tends to have lower skilled jobs, and that these are in theory easier to automate than higher skilled jobs. This basic theory seems sensible enough, and rather than predicting how many job losses this may cause, they thankfully veer instead towards predicting that this will result in de-industrialization and stagnant wage growth.
In a development context, this might cause a rise in inequality and challenges in reducing poverty levels. In turn, the authors suggest this might place the social contract in a country under strain, or alternatively hinder the evolution of more inclusive social contracts. Therefore, they focus their attention on how countries should respond to this challenge.
They argue that automation is likely to result in a number of key policy challenges. They suggest that attempts to upskill the population are likely to be insufficient given the widespread potential for disruption, and that there needs therefore to be a broader range of policy initiatives, ranging from infrastructure construction to social, education and healthcare provision.
Despite suggesting that we need to do more than upskilling the population however, it will remain a central tranche of government policy as it will help to move people away from the routine skills that are easier to automate.
The authors suggest however that the approach has a number of challenges, not least in understanding what skills are likely to be immune to automation and in overcoming the barriers to upskilling that I’ve touched on in previous articles.
“Competition with currently available technology increasingly seems to require a tertiary education which is still very rare throughout the developing world,” the authors say. “Given that even advanced industrialized countries are struggling to keep their labor forces competitive, the success of a skills development strategy alone remains questionable.”
An alternative strategy is to increase the safety net offered, typically offered via unemployment insurance, wage subsidies or other safety nets. This is an approach that aims to address the inequalities that automation may cause. It requires a high level of profit across society however so that income can be redistributed.
Putting off the inevitable
The authors believe that neither of these strategies have been deployed in developing countries thus far, with most instead choosing to invest in already highly labor-intensive sectors. It’s a profound case of kicking the can down the road rather than preparing for the inevitable.
A more appropriate response is to try and develop the skills and sectors in areas that are more resistant to the influence of automation.
“Such an automation resistant sector could, for example, involve the social, education and health-care sectors, and some forms of tourism, and infrastructure construction which are generally considered resilient despite increasing service automation,” the authors say.
The challenge in implementing such a strategy is that these sectors tend to require high skill levels, whilst many of them aren’t hugely value-adding, so it is not an easy area to grow.
Ultimately however, they argue that focusing excessively on employment when discussing automation renders the debate far too narrow to be effective. They ignore the broader questions about the potential impact of the digital revolution on structural change, and how these questions can be satisfactorily answered.
Whilst the authors don’t offer any solutions of their own, they do offer a number of areas that they believe warrant further investigation. These revolve around three core areas:
- How, when and why does productivity growth translate into employment growth?
- How does automation impact different countries in different ways?
- What political consequences might there be of changes to employment due to automation?
Overall, it’s a paper that poses more questions than answers, and in that sense is common with many other papers produced by governments and think tanks in the past few years. It would be pleasing to see some level of action and experimentation as a result of these papers, but for the time being at least, we will have to wait for that.