No, Bill Gates, We Don’t Need a Robot Tax
Bill Gates made waves in a recent interview when he said he believed robots should be taxed, saying, “You ought to be willing to raise the tax level and even slow down the speed of that adoption” in order to smooth the transition for people that would be displaced and bolster programs and processes to retrain them.
These concerns are no longer abstract, as European Union legislators recently considered but ultimately rejected a similar proposal. Evidence does not support the fear that the labor market is currently undergoing a massive disruption, and ultimately an ill-defined “robot tax” could stifle innovation, lower growth and reduce living standards.
Gates is the most noticeable example of a growing concern that in the coming decades, a swath of occupations is at risk of being automated. Some earlier studies estimated that 47 percent of occupations would be at risk of being automated within the next decade or two. Lower wage jobs faced elevated risks, with 83 percent of jobs with median wages below $20 dollars. These estimates have sparked fears of lower-skill and lower-wage workers being shut out of the evolving labor market. Once they have been displaced, some say, they could fall farther behind and find it harder to re-enter the labor force.
What some of these studies did not fully account for is that in most cases automation is not so much geared towards replacing all of the duties associated with occupations, but with specific tasks. Many occupations previously categorized as “at risk” have at least some tasks that would be difficult to automate. This suggests that at least some of these occupations will evolve, along with changing technologies. A blunt “robot tax” would likely be unable to differentiate between those machines that replaced and those that complemented human workers, which could choke of an important source of new, often less physically demanding, jobs. Using a task-based approach that allows for more automation to be complementary in nature, a recent working paper from the OECD estimates that a much-less-alarming nine percent of people are at high risk of automation.
As rigorous as are these projections, they carry with them a degree of uncertainty inherent in trying to project the evolution of technologies that have not yet emerged, so it is also worth seeing if there are signs that this dynamic is already taking place. A corresponding increase in labor productivity should accompany a surge in automation and a shift away from low-skilled labor. That is not what has occurred happened. According to the Bureau of Labor Statistics, productivity has grown at an average rate of 1.1 percent from the end of 2007 to the third quarter of 2016, far below than the 2.3 percent rate from 1947 to 2007 or the 2.7 percent rate from 2001 to 2007.
Source: Bureau of Labor Statistics, “Labor Productivity Growth Since the Great Recession,” February 24, 2017.
Lackluster productivity growth in most developed countries would seem to suggest that, if anything, the pace of automation and introduction of new technologies is too slow, not something that is far outpacing society’s ability to cope. Policymakers should be seeking to encourage more innovation to boost economic growth, not deter it.
Concerns about unprecedented upheaval in the labor market, and what that means for income taxes, discount how much the labor market has evolved. In 1870 farm workers were just over half of the labor force, and their share declined drastically to 8.1 percent by 1960 and continued in the years since. Over this period, the landscape of the American economy and labor market changed fundamentally, but by 1960 the country had attained levels of prosperity that would have been difficult to imagine a century before. If the equivalent of a robot tax had been imposed in 1870, discouraging new innovations in the interest of minimizing disruption, everyone today would be much worse off.
Source: Output, Employment, and Productivity in the United States after 1800, Table 2, NBER Book Series Studies in Income and Wealth.
The advent of personal computers, the Internet, industrial robots, and many other innovations have introduced new instances of automation, but individual income tax revenue as a share of GDP has remained relatively steady, equivalent to 8.4 percent of GDP last year, above its long term average of 7.9 percent.
Source: Congressional Budget Office, “Historical Budget Tables, January 2017,” Table 2. Revenues, by Major Source, Since 1967.
The core concerns that lead Gates to his “robot tax” do not seem to hold. Sluggish productivity growth does not indicate that there is currently a wave of disruption in the labor market due to quicker adoption of new technologies. In addition, previous transformations of the labor market have led to a more prosperous country and have raised living standards without eroding the tax base.
Any form of robot tax would bring a host of practical complications. What, exactly, would qualify as a robot subject to the tax? A mechanical loom? Would Alexa and Siri run afoul of the new tax? What about Microsoft Office? These concerns aside, the biggest problem with the proposed tax is that it would discourage all sorts of future innovations, many of which could bring about tangible increases in the standard of living for the very people Gates is most worried about. The calls for a “robot tax” are misguided, and raging against the machines could end up making things worse by curbing innovation and hindering economic growth.
Charles Hughes is a policy analyst at the Manhattan Institute. Follow him on twitter @CharlesHHughes.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, e21 delivers a short email that includes e21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the e21 Morning Ebrief.