May 18, 2017

"A robot tax is not expedient"

					Dr. Susanne Bieller

Dr. Susanne Bieller, Project Manager at the European robotics association EUnited Robotics: "We should tax profits and not equipment that presumably contributes to increased productivity."
Source: EUnited Robotics

In February 2017, the EU Parliament met to examine the effects of robotics and artificial intelligence on society and the job market. A robot tax that had been proposed earlier was rejected, but liability issues and mandatory insurance are still being discussed. In an interview, Dr. Susanne Bieller, Project Manager at the European robotics association EUnited Robotics, explains why a robot tax is not expedient and why there is no reason to be afraid of autonomous robots.

Dr. Bieller, you have been following the robot debate for quite a while. Why is the debate coming to a boil, particularly now?

Bieller: This time the reason was an EU-funded project, the results of which were presented before Parliament. Generally speaking, dealing with the issue of robotics law and seeing if legislation takes current requirements into account is a good thing. Besides consumer protection, it is also important to review regulations that might prevent the use of robotics for desirable purposes such as in medical technology. In that respect, the EU Parliament's initiative is welcome.

But as EUnited Robotics, you categorically reject the robot tax proposed in the draft, which is supposed to offset possible negative effects on the job market. Why?

Bieller: The recommended robot tax simply focuses on the wrong thing. It has been suggested that robots take away jobs and that social systems will not be able to absorb that in the long term. Even if multiple studies come to that conclusion, it still doesn't prove that jobs are being lost. We assume that exactly the opposite is happening. That is also evident when you look at strongly robotized branches of industry where, so far, employment appears to be increasing instead of decreasing.

So robots create jobs?

Bieller: Yes, in the long term, they do. If a company automates, it can in some cases initially lead to lost jobs on the assembly line. But overall, productivity increases, and increased output creates new jobs, often in other divisions such as sales or marketing. The end result is that more jobs are created.

Is there a concrete example of that in actual practice?

Bieller: Yes. That theory has been corroborated by the numbers in the automotive industry. It has undergone extensive automation during the past few years and created huge numbers of jobs at the same time. And you have to keep in mind which jobs were automated: Take the painting lines in the automotive industry, for example. That work is performed almost exclusively by robots because one can no longer expect an employee to perform such an unpleasant task involving toxic materials.

Still, a number of studies are causing a stir because they predict that about half of all jobs are in danger due to ongoing technological progress.

Bieller: You have to keep in mind that these studies are based on a model and simplify the real world accordingly. For example, as a rule these studies only count which jobs might be eliminated in the future. But they ignore the new jobs that are created. In other words, many studies do not take the net effect into account. And many of them assume that entire occupations or job descriptions will be eliminated. But in most cases, an employee carries out a number of different tasks—some of them can be automated, but some of them can't.

Does anyone disagree?

Bieller: Yes, McKinsey Global Institute assumes that it will not be possible to completely automate more than 90 percent of jobs in the future.

Still, some pretty clever people are in favor of a robot tax. For example, Robert James Shiller, winner of the Nobel Prize for Economics and professor at Yale University, feels that robotization warrants government intervention.

Bieller: The crux of the matter is, how do we define 'robot'? What is a robot in the actual sense of the word and what is software or artificial intelligence, for example? Computers and software automation have already had a massive influence on the working world. But that doesn't mean that computers and software have caused massive unemployment. Jobs have changed: A secretary today performs completely different tasks than twenty years ago. Inequality can have a number of causes. But the best way to overcome it is through education and healthy economic development.

But education costs money. Wouldn't a robot tax be a good way to pay for it?

Bieller: No, we should tax profits and not the working tools that we use because they might contribute to increased productivity. That is also what the Federation of German Trade Unions (DGB) says. In that respect, we agree with the trade unions that introducing a robot tax would do more harm than good. Besides, practical implementation would also prove to be very difficult. That is why the first recommendation of the European Parliament was that companies should estimate what share of profits can be attributed to automation. How in the world are companies supposed to calculate that? That would be an enormous bureaucratic burden!

Dr. Bieller, thank you for talking to us.