Proposals to tax robots have been debated by serious folks recently. The European Union considered but ultimately rejected the idea of taxing firms that use robots. And last week Quartz published an interview with Bill Gates in which he argues for a robot tax. These proposals follow a spate of recent articles on robots and automation, some of which argue there will be large job losses from robots and automation. These articles include one in the New Yorker, which profiled books by Martin Ford, Jerry Kaplan, Erik Brynjolfsson and Andrew McAfee, and David Autor’s article on workplace automation in Journal of Economic Perspectives, among others.
There’s no question that the potential increase in robots and automation requires policy makers to rethink fiscal policy for the 21stcentury (and other policy as well, such as education and retraining policy). But, based on the data we currently have, a tax on robots would be bad policy. Robot taxes would dissuade firms from investing in robots, which would lower economic growth, and, to the extent that robots complement labor in some cases, would lead to less hiring and lower wage growth.
What Is A Robot?
There are a number of problems with the idea of taxing robots. The first problem is definitional. In a manufacturing context, a robot typically refers to a robotic arm. As defined by ISO 8373, an industrial robot is an “automatically controlled, reprogrammable, multipurpose manipulator programmable in three or more axes, which may be either fixed in place or mobile for use in industrial automation applications.” But people also use the term “robot” to refer to software algorithms, including Amazon’s Alexa and Apple’s Siri, and many of us now routinely interact with chatbots, which have replaced human operators.
When we say we want to tax robots, do we want to tax all types of robots, including software robots and other forms of automation, or just robot arms? If we confine our definition of robot to a robot arm, then the robot tax ends up being a tax borne primarily by the manufacturing sector, and not by other sectors of the economy that will likely invest heavily in automation, including autonomous vehicles in trucking and transport, smart conveyor belts in warehouses, electronic checkouts in retail, etc. Manufacturing comprises about 9% of the U.S. economy on average, but up to 15% of employment in some states. Taxing investment in a handful of states based on an arbitrary definition of what does or does not comprise a “robot” does not seem to be good policy.
Robots Increase Economic Growth
Even if there were good ways to define a robot, there are good reasons not to tax the use of robots and automation. The Council of Economic Advisers’ 2016 Economic Report of the President cited research showing that robots can boost gross domestic product growth by 10% and labor productivity growth by over 15%, numbers similar to the impact of steam engines on British labor productivity growth in the 19thCentury.
Moreover, robots are complements to labor, not substitutes for labor, in many cases. For example, Amazon has been increasing the number of robots it uses in its fulfillment centers, but Amazon has also been increasing the number of humans it employs in its fulfillment centers. More broadly, as highlighted elsewhere, we need to do a better job collecting data on robots and how they are being used so that we can understand conditions under which robots are substitutes or complements. Therefore a robot tax would make it harder to achieve productivity growth and, to the extent that robots are complements to labor, a robot tax may perversely lead to fewer rather than more jobs.
Instead, Consider Proven Policies Such As EITC
Instead of a robot tax, policy makers should consider other policy options in response to automation. One idea with bipartisan support is to expand the Earned Income Tax Credit (EITC), which provides tax credits to working individuals who meet certain income thresholds, to a larger set of the population. If a person loses a job due to automation, globalization, or any other reason, the opportunity to benefit from the EITC provides an incentive to find another job. Research shows there are a number of other benefits of the EITC, including childhood development and poverty reduction.
Of course, new skills may be necessary for a new job, which is why vocational training and re-training programs are important. The United States spends less than 0.05% of its GDP on vocational training—less than most other OECD countries, a tenth of what Finland and Denmark spend, and a fifth of what Germany spends—and the U.S. number has been declining for the past couple of decades. At a minimum, funding for job training should be increased, and there are a range of options for how the training can be delivered and administered.
One recent idea that has gained popularity is a universal basic income (UBI). While UBI schemes vary in their specifics, the basic idea is to replace existing safety net programs with a single, unconditional cash transfer. While the apparent simplicity of a UBI scheme is appealing, a number of drawbacks have been noted, including that it does not reward work or provide the training needed to find higher paying jobs, and, depending on how it is structured, may leave many people worse off than they are today, particularly those at the lower end of the income distribution who currently rely on existing safety nets. Nevertheless, pilot studies on basic income schemes have been started, including in Finland, the Netherlands and Oakland, CA. These experiments will provide valuable data and, ultimately, insights for policy makers.
Thus, despite the legitimate concerns of its proponents, there are good reasons why there should not be a tax on robots. But Bill Gates is right to suggest the need for a larger debate about appropriate policies for addressing automation.