Is taxing robots a useful idea?

Dear Mr Bill Gates

Since the end of August last year I have been researching tax and treasury issues in a Belgian context. Just like you, I am not a tax expert, but just like you, I was wondering how we should address the problems we will face in the next three decades, when employment will decrease by up to 50% according to a study published by Carl Benedikt Frey from Oxford Martin School en Michael A. Osborne from Oxford University.

The idea you just proposed to tax companies that replace a part of their workforce by computers or robots might seem valuable at first glance. But when you think about it there are some problems.

Tax policies should be fair

There is an increasing feeling in society that tax policies are not fair. Whether that is true is not a topic to discuss now, but when talking about new tax policies we at least should be prudent to propose new tax regulations that are perceived as fair.

When you propose to tax companies that are replacing employees with automation, you forget to consider what to do with start-up companies, companies that do not replace employees by computers or robots, but just don’t need any employees (or at least not many relative to their revenue).

What will be the start date for measuring the replacement and how will you measure this at all? Facebook for example had a revenue of 17.93 billion dollars in 2015 and at that moment a workforce of 12,691 FTEs, bringing Facebooks revenue at 1,412,812.23 dollars per FTE. On the other hand we have General Electric with a workforce of 333,000 FTEs for a revenue of 117.00 billion dollars, a mere 351,352.35 dollars per FTE.

You state that companies need to be taxed for the replaced workforce but that personal tax still should also be paid. When that is the outcome, companies from the ‘old economy’ that try to keep pace with evolution are punished twice, while new companies don’t pay tax at all.

Aside from the fact that this seems to be unfair, it is also far from the sustainable solution we have to find to address the dramatic decreases in employment. It is not a solution that will fund our government in the long run, because we will end up with more digitally born companies and therefor companies that do not pay tax.

How will you measure replacement

Measuring replacement will be hard. You propose to tax the robot that replaces an employee responsible for 50K of revenue at the same level. But what happens when, due to robotization or automation, the revenue of that company grows? And what happens when the worker isn’t replaced at all? What happens when the company, instead of buying robots or computers to replace workforce, is renting computing capacity in the cloud?

Companies will often have very good reasons to rely on cloud computing solutions. Reasons such as efficiency: they only pay for the used capacity and technology; the computing capacity will always run on the latest technology and the most advanced machine learning algorithms.

Or will governments outsource their tax collection to companies such as Amazon, Google and well yes … Microsoft?

And what about other tax inequalities and tax regulations?

There’s more than meets the eye. You only address the problems we‘re facing when employment is decreasing. But when thinking and talking about new tax regulations, wouldn’t it be nice to solve all the flaws in current tax policies?

What about instrumental tax regulations that cause society a lot of problems?fla I think about, for instance, company cars that are used to pay untaxed wages.

It’s not the time for quick fixes as the one you propose. It’s time for deep thinking and sustainable solutions.