In a recent interview with Quartz, Bill Gates floated an idea for mitigating the impact of automation technology on human jobs: tax the robots. Governments could use the revenue, he argued, to fund social safety-net programs and support jobs that still require human empathy and understanding.
But Harvard economist and former US Treasury secretary Lawrence Summers calls the suggestion “profoundly misguided.” In a Washington Post op-ed published March 5, he laid out his reasons for his skepticism.
Taxing technologies is a slippery slope
Summers argues that there are a number of technologies that already make it easier for a human to work, or to work more effectively—for instance, Microsoft Word.
Kiosks replace human work by allowing customers to serve themselves, but don’t involve artificial intelligence or robots. Vaccines prevent disease, so you might argue they destroy jobs in medicine.
Where is the line between which technologies destroy jobs and which do not? Summers argues that even if the government could create a precise definition, tax authorities like the IRS would be unlikely to effectively administer one.
True innovation shouldn’t be stifled
Automation technology has the potential to replace work that accounts for $14.6 trillion in wages globally, according to a McKinsey analysis.
But along with its potential to dislocate workers, new technology can bring with it improvements in goods and services. As Summers notes, we already benefit from technology that has given the world more efficient ways to book travel or more precision in delicate surgeries. AI currently under development promises to do the same. Automated cars, for instance, are meant to be safer than human drivers.
A tax on robots—a version of which was proposed and rejected by EU officials in February—risks squeezing innovation out of the system. As Summers writes, “A sufficiently high tax on robots would prevent them from being produced.”
Summers seems to agree with Microsoft co-founder Gates that the job destruction caused by automation argues for governments to invest in more education and retraining programs. He also suggests that direct government employment programs and wage subsidies for some groups of affected workers would be welcome.
But a tax?
If anything, Summers might endorse the exact opposite.
Between “emulation and competition, innovators capture only a small part of the benefit of their innovation,” Summers writes. “It follows that there is as much a case for subsidizing as taxing types of capital that embody innovation.”