Former Harvard University President and U.S. Treasury Secretary recently demolished the idea of a robot tax and also issued a ringing endorsement of an activist government role to manage the impact of artificial intelligence on the labor market and on society.
On the negative side, he asks why pick on robots? Lots of technologies save labor - from kiosks at airports to automatic teller machines to industrial robots. Why not tax all of them? Some technologies also enhance labor, increasing the effectiveness of humans at their jobs. AI systems do both. Do we tax both? If just the labor-destroying technologies, how can we tell from the outside which is which? AI also makes products better not just cheaper. Cars were a better product than horses, and autonomous cars will be safer than cars driven by humans. Why do we want to tax improved products? Besides, why do we want to tax, and thereby impede, developments that are making us collectively wealthier? Robot labor increases output – which we should then intelligently and fairly distribute to make everyone better off. Making us collectively poorer does no one any good.
All that said, Summers recognizes that artificial intelligence is creating extraordinary policy challenges. It is contributing to unemployment, wage stagnation and inequality, while paradoxically not doing enough to pull us out of our productivity slump.
How should governments deal with these challenges? Here, he is a full-throated believer in an activist government:
They likely will need to take a more explicit role in ensuring full employment than has been the practice in the US. Among other things, this will mean major reforms of education and retraining systems, consideration of targeted wage subsidies for groups with particularly severe employment problems, major investments in infrastructure and, possibly, direct public employment programmes.
Needless to say, this is a substantial government reform agenda. But if we correctly insist that measures to increase the size of the economic pie should be encouraged, not impeded through a robot tax, then we must also insist on measures to make sure the resulting increase in community wealth improves life for all of us. We found out what happened to public commitment to an integrated world economy when political elites failed to compensate those who lost ground from trade. The same could happen with artificial intelligence. And a robot tax will not be the worst idea to come along if we fail to take seriously the task of sharing the wealth the robots will help create.