Determining a course of action will not be easy for governments as businesses accelerate their use of AI and robotics — whether physical robots in the factory or robotic process automation software in the office.
On one hand are complex and critical tax questions. If job losses cut into countries’ income tax revenues, how would governments fill the gap in their treasuries?
What will be the public cost of continually reskilling populations for an ever-increasing-tech world, as well as providing adequate support for people put out of work by robotics and AI?
On the other hand, governments want to encourage technological innovation, and putting taxes or regulation on new technologies at an early stage could hinder their development.
Some academics and governments are already conceiving and experimenting with targeted tax and wage solutions. Even the most fundamental tax precepts could be rewritten, according to Charles Davis, EY Global Tax Lead Analyst in London.
For example, effective corporate income tax rates tend to be lower in many countries than the effective tax rate (i.e., income and social security taxes) that employees pay on their wages.
So an increase in corporate profits and reduction in employee costs because of the use of robotics would produce less tax revenue for the government as employees are taxed at a higher rate than businesses.
“The balance between the two could become a really tough question, over time, because — if value creation moves from being delivered by humans to robots — you’re talking about quite a big potential hit to government revenue,“ Davis says.
Many suggestions, few answers
In an interview with Quartz in early 2017, Microsoft Co-founder Bill Gates caused a stir by suggesting the creation of a “robot tax” — basically, taxing a robot at a similar level to a human worker performing the same task. One way could be by taxing profits generated by the labor-saving efficiency of robotics and AI, Gates said in the interview.
During the past year, proposed tax solutions have only proliferated. For example, taxing capital investment in automation could be an alternative to taxing profits.
In another twist, a robot tax could be paired with minimum income payments for everyone, since a robot tax alone would have to be extremely high to address societal goals, according to research by Sergio Rebelo, a professor at the Kellogg School of Management at Northwestern University, and Pedro Teles, a professor at the Universidade Católica Portuguesa.
There is growing discussion about the merits of a universal basic income in the AI age, providing all individuals with a minimum amount paid regularly and unconditionally. Proponents argue that the approach could actually incentivize entrepreneurship, volunteerism and other socially desirable outcomes.
One World Economic Forum (WEF) panelist this year broadly suggested eliminating corporate tax breaks and subsidies to support a universal basic income for all, regardless of work status.
A pilot of the basic income concept is planned for Stockton, CA, in 2018. Finland is expected to publish the results of its pilot at the end of 2018.
Other alternatives include wage insurance, which could provide incentives for workers with old-economy skills to take a pay cut in the short term as they get on-the-job training for 21st-century positions, according to a white paper co-written by Kellogg School Professor David A. Besanko.