The Robot Tax.
There has been an increase in calls for the taxation of companies that use robots to replace employees. Bill Gates made the argument in 2017. He resurfaced this idea recently. Senator Bernie Sanders is currently touting the idea. In 2024, Brookings also made a case for a robot tax. I worry about a robot tax, but not for the obvious enlightened self-interest reasons that the robotics industry might raise. My concerns are fundamental and lie at the roots of the concept.
The problems with a robot tax remind me of a set of Russian Matryoshka dolls. For those who have forgotten the childhood toy, Matryoshka dolls are nestled one inside another. You open the first doll, you find another; you open the second doll, you find another, etc. I have a friend whose aunt would give the children Matryoshka dolls at Christmas. The last doll always contained money.
Reasons To Have A Robot Tax.
There are many positive aspects to a robot tax. It is the implementation that is challenging. Contrary to some beliefs, a robot tax is not like a set of Matryoshka dolls, where a pot of money is found in the last doll.
While the goal of a robot tax has been discussed in general, there are some obvious positive implications. I will address four: 1. The tax could make up for lost tax revenue if robots replace employees. 2. The tax could be utilized to retrain those who have lost their jobs. 3. The tax could provide a subsidy for those who have lost their jobs and cannot find work. 4. Finally, the tax may slow the adoption of robots in the workplace. Each of these pluses has a set of negatives within it.
Replacing Federal Income: The First Matryoshka Doll.
Let's address the need for the federal government to replace lost revenue. The government will lose revenue if employees and employers do not pay taxes on employee income. One critical area is Social Security: it will lose payments necessary to provide benefits to current recipients. However, there is the potential for an offset. If robots reduce the cost of operations, the income for a corporation should increase. This increased income will be subject to taxation and could replace some of the taxes not collected from employers and employees. (I realize the corporate and personal tax rates vary and would have an impact.)
However, let's assume that the taxes on increased profits will not meet the financial shortfall. How would the robot tax be calculated? Since Senator Sanders has said that he wants the tax to make it expensive for companies to replace workers, the tax should be designed to provide financial equilibrium.
An Equilibrium Equation.
A simple equilibrium equation, I will call RET, might be set up as follows:
COR + RET ≥ COE
RET ≥ COE - COR
COR = Cost of a Robot: RET = Robot Equality Tax: COE = Cost of Employee(s). This assumes that the COR will be less than the COE, which is what proponents of the robot tax believe. (State taxes are ignored, but could be added by taking a percentage of the RET.)
This type of equation would require the expertise of cost accountants, economists, and actuaries, among others. A few of the challenges might include estimating the mechanical life span of a robot, the cost of maintaining it, the number of employees a robot would replace, and the procedures for corporations to calculate the RET. Obviously, these factors would vary by industry and production output. I doubt that the government will bother with the niceties. Instead, the government will impose a tax that is not highly correlated with the above equation but will provide another source of revenue.
Another challenge is whether the tax would be a one-time charge in the tax year the robot is purchased or begins operations, or if it would be an annual tax charged as long as the robot is in operation. Most taxes, once put in place, rarely go away. It is doubtful this would be a tax imposed only when a robot is purchased or first enters the workforce.
Retraining Workers: The Second Matryoshka Doll.
Assume you have four workers at a fast-food chain. Also, assume one robot could replace all four workers; a robot tax could be used to retrain the workers. However, as robots take over more and more positions, this becomes a do-loop. For example, assume a fast-food worker is retrained to be a brick mason. It will only be a matter of time before a robot replaces the brick mason. Then the brick mason must be retrained again. Let's assume the brick mason becomes an accounting auditor. It will only be a matter of time before the worker needs retraining, as AI software, computers, and robots can replace the accountant. Realistically, accountants, professors, and even doctors could be replaced.
Those proposing a robot tax assume you can retrain a fast-food worker for another position. While this is possible, it is not guaranteed. Programs to retrain workers have not always been successful. Many have said that the solution lies in creating millions of new jobs to provide opportunities for those who are displaced. This is addressed below.
Replacing Lost Income: The Third Matryoshka Doll.
There is no doubt that robots will replace workers. What is critical is how quickly workers will be replaced and whether they can be retrained. From an ethical and humane perspective, society probably has a responsibility to assist displaced workers overwhelmed by a technological tsunami, a problem they did not create. The robot tax dollars could provide income while displaced workers are being retrained. If they cannot be retrained, the funds could also offer permanent income. Our society is replete with these types of safety nets. Funding for this safety net should come from corporate robot taxes. Of course, the government is willing to borrow trillions of dollars for other programs. Surely, the government would go into debt a few more trillion to provide the safety net if the robot tax is not adequate.
Slowing The Adoption Of Robots: The Fourth Matryoshka Doll.
If the government decides to increase taxes above what it would take to have equilibrium, then companies would hesitate to add robotic initiatives. Protecting employees in this manner is a short-term Trojan horse victory for those trying to slow the implementation of robotics. As robots improve in capabilities and costs decline, companies will eventually adopt them. The reasons are apparent. While you might have an increased cost because of a heavy tax, robots usually do not take sick leave or family leave, fail to show up for work, have varying rates of productivity, or need paid vacation: costs that would be hard to include in the calculation of the tax. At some point, factors like these will force companies to use more robots. Besides, taxes are deductible as a business expense unless they are treated as federal income taxes. This deduction will help offset the cost of adding robots.
If companies adopt the use of robots and the governmental taxes are onerous, companies will increase prices. The price increases will produce inflationary pressure, which is the reverse of what should happen when robot usage is increased.
Some Additional Perspective.
Job Opportunities. Suppose 75,000,000 workers out of a United States labor force of approximately 175,000,000 workers are replaced, where will they go? If the 75,000,000 are retrained for the other 100,000,000 jobs, will there be a glut of employees for the remaining jobs? What about new workers? Will they all be qualified for the 100,00,000 jobs?
More importantly, if new jobs are created, will they be filled by people or will AI/robots take these over? Assuming that new jobs will solve the problem for those who are displaced may be like jousting at windmills. Simply put, the supply of jobs human beings can fill as a percentage of the population that wants to work will probably decline.
A Broader Tax. I am all for a robot tax. However, it seems that the concept of a robot tax is a lot easier to discuss than to implement. I fear that the tax will be a SWAG (a wild guess). Some people, replaced by robots, cannot be retrained, and a significant number of these will end up on the robot tax welfare rolls. It seems that considering a robot tax is not a serious way to address technological challenges. As AI improves and its operations become more like those of a human brain, it will replace jobs that do not require physical assets like robots. It is not just jobs replaced by robots that should concern us. The increase in the use of all forms of technology may be problematic. We should be looking at a technology tax, not just a robot tax.
While some predict a doomsday scenario for workers, do not despair. Humans will maintain control for a few more years.
If you haven't read the blog describing the "However View," click here.
