What's Wrong With Bill Gates' Robot Tax

Imposing added costs on new technology will slow growth and won't help people displaced by automation.

Cute? He's after your job, pal.

Photographer: Ian Gavan/getty images

Microsoft Corp. founder Bill Gates made a splash in a recent interview, when he suggested that robots should be taxed in order to help humans keep their jobs:

Right now, the human worker who does, say, $50,000 worth of work in a factory, that income is taxed and you get income tax, social security tax, all those things. If a robot comes in to do the same thing, you’d think that we’d tax the robot at a similar level.

Gates is only one of many people in the tech world who have worried about automation and its threat to workers. YCombinator founder Sam Altman, for instance, is conducting an experiment with basic income -- a regular cash payment to all households. That policy has gained popularity across Silicon Valley, if not in the rest of the country.

It’s easy to see why the tech world is worried. The rise of machine learning has increased the fear that many humans could simply become obsolete -- for example, 3.5 million American truck drivers might soon find their jobs threatened by driverless trucks. Though in the past, technology usually complemented workers instead of replacing them, there’s no law of nature saying the technology of the future will work the same. A few economists even claim that cheap automation has already diverted income from workers to company owners.

Driverless Cars

The fear isn’t that all humans will become obsolete, but that automation will increase inequality among humans. Company owners and high-skilled workers -- people who tell machines what to do -- would be vastly enriched, while everyone else either works low-skilled jobs for meager wages or goes on welfare.

Another fear is that even if the mass of humanity ultimately does find new ways to add value by complementing new technology -- to “race with the machines,” as economist Erik Brynjolfsson puts it -- this transition could take a long time and hurt a lot of people. As Bloomberg View’s Tyler Cowen has noted, wages in Britain fell for four decades at the start of the Industrial Revolution. More recently, we’ve seen very slow and painful adjustment to the impact of globalization. If the machine learning revolution hurts workers for 40 years before ultimately helping them, it might be worth it to slow that revolution down and give them time to adjust.

The main argument against taxing the robots is that it might impede innovation. Growth in rich countries has slowed markedly in the past decade, suggesting that it’s getting harder and harder to find new ways of doing things. Stagnating productivity, combined with falling business investment, suggests that adoption of new technology is currently too slow rather than too fast -- the biggest problem right now isn’t too many robots, it’s too few. Taxing new technology, however it’s done, could make that slowdown worse.

The problem with Gates’ basic proposal is that it’s very hard to tell the difference between new technology that complements humans and new technology that replaces them. This is especially true over the long term. Power looms replaced human weavers back in the Industrial Revolution, but people eventually became more productive, by learning to operate those looms. If taxes had slowed the development of power looms, the eventual improvements would have come later.

This is a powerful argument against the taxation of automation. Gates is right to say that we should start thinking ahead of time about how to use policy to mitigate the disruptions of automation. But given the importance of sustaining innovation, we should look at alternative policies.

One idea is a wage subsidy for low-income workers. This basically puts a thumb on the scale in the human-robot struggle, by making human workers cheaper. The easiest way to do that is to cut payroll taxes, which disproportionally fall on low earners. That would mean paying for Medicare and even Social Security out of other revenue sources, such as higher income taxes on the rich or a value-added tax.

Another idea is to simply redistribute capital income more broadly. Income from capital gains, land rents and dividends now is highly concentrated among the wealthy. But policy could change that. One idea, suggested by economist Miles Kimball, is a sovereign-wealth fund. The government could use tax revenue to buy stocks and real estate, and distribute the profits to the populace. This would essentially redistribute some of the income produced by the robots, giving every citizen a stake in the new automation economy. The wealth fund could be split into many smaller funds, each with different managers, in order to prevent concentration of ownership.

So there are probably better ways than taxing robots to help humans avoid the harms of automation. Instead of slowing innovation, the government should think about taxing humans less and redistributing the income of robots more.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Noah Smith at

    To contact the editor responsible for this story:
    James Greiff at

    Before it's here, it's on the Bloomberg Terminal.