Get the DealBook newsletter to make sense of major business and policy headlines — and the power-brokers who shape them.
__________
If it’s true that robots are coming for our jobs, developing countries could have a much tougher time coping with the shock.
A new study by the economists Lukas Schlogl and Andy Sumner of King’s College London, written for the Washington think tank Center for Global Development, suggests that there is still a need to consider how developing countries will cope with the rise of automation. Most strategies to help workers displaced by robots, some of which look promising, have so far been devised for developed nations and may not translate to the developing world, they argue.
Much research has gone into investigating what kinds of work might be automatable, and predicting when machines will take over tasks performed by humans. Forecasts vary greatly, but there is a consensus that routine tasks that don’t require emotional intelligence, complex human reasoning, or creativity will gradually be filled by robots and artificial intelligence. Reports by the McKinsey Global Institute and the World Bank both suggest that agricultural and industrial sectors have higher potential for automation than service sector jobs, which typically require creative thinking or face-to-face interaction.
That poses particular concern for developing countries, where there are typically larger pools of unskilled labor working in agriculture or simple manufacturing roles.
In the developed world, much of the discussion about how to address the rise of automation has focused on 1) how to constrain it and 2) how to cope with the fallout.
Governments could constrain automation through policy measures such as disincentivization — by, say, taxing the use of robots — or reducing the cost of human labor through tax breaks or cutting minimum wages.
The problem is that disincentevizing automation could just push it into other nations. Developing countries are likely to feel unable to implement those kinds of constraints, for fear of companies or even entire sectors relocating to regions where the use of robots isn’t penalized.
Mr. Schlogl told DealBook the impact of automation could be felt acutely by sectors that are labor-intensive in developing countries, irrespective of where robots are put to use. Cheap goods can be built by machines in one location and distributed worldwide; a single country could mechanize agriculture and flood a subcontinent with affordable produce.
As for cutting labor costs, it’s unclear exactly how much lower wages can get in many developing countries before they become unethical.
The second way for governments to cope with the disruption caused by automation is to implement strategies that deal with the effects. That includes retraining workers to give them skills required for future jobs, or providing citizens with a basic income to make up for stagnant wages or job losses.
Retraining would be hard to implement in developing nations, argue Mr. Schlogl and Mr. Sumner, because they typically have limited education sectors in place through which to deliver it. And a universal basic income, they argue, would be hard to finance in developing countries, as it assumes the existence of a highly productive services sector from which to siphon money, which is often missing in such economies.
There are some potential routes forward. One idea suggested by Mr. Schlogl and Mr. Sumner is a so-called global universal basic income — administered internationally, and paid for in developing countries via aid. This plan, says Mr. Schlogl, “has the advantage of being only politically impossible.”
Another might be for developing countries to build out labor-intensive sectors that look set to be resistant to automation over the coming decades — such as social care, education, health care, tourism or infrastructure construction. But this is a risky approach, requiring large upfront investment without a guarantee of protection from automation in the long-term.
That makes the pair’s ultimate conclusion — that “we need to ask different policy and research questions” about how automation will affect the developing world — seem justifiable.
