A universal basic income and a tax on robots were part of the platform which won Benoît Hamon the socialist nomination to be France’s next president. There was a time when he would have been lampooned internationally as the very caricature of the French left. But times have changed, and jokes about such policies would now be on those who make them. Robot taxes and UBI are being seriously debated across the political spectrum.
The most high-profile endorsement of robot taxes must surely be the one from Bill Gates, who expresses sympathy with the idea in a recent interview with Quartz’s Kevin Delaney. Gates seems to see it as one of many options for taxing the productivity increases that technology may bring — but that is already fairly radical: “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’s concern combines three realisations: that technological changes will make some jobs obsolete and may reduce living standards for people who depend on those jobs; that service work such as care will both be necessary and a place to employ the labour released by more efficient goods production; and that this sort of work will need to be paid for in part by taxation. Hence his overall view, which is surely right, that you need to “take that extra productivity and generate more taxes”, even though he’s agnostic about the best way to do this: “Exactly how you’d do it . . . it’s interesting for people to start talking about now.”
There are others, however, who are much less thoughtful and hurry to connect taxes on robots to help for those who lose their jobs to those robots. That could be seen as either a disincentive to automate too much or too fast (Bill Gates, too, says “you ought to be willing to . . . even slow down the speed” of automation) or a way to fund subsidies for the victims. One proposal to achieve the latter is, of course, UBI. (For an overview of the debate and experimentation around UBI, read Sonia Sodha’s excellent up-to-date report. I have previously argued that UBI is both affordable and desirable.) But as Leonid Bershidsky persuasively argues, while UBI itself is a good idea, taxing robots to pay for it may not be.
If the key idea is to tax the activities that increase labour productivity and therefore cause companies to shed workers, a special tax would need cast a much wider net than just robots. Mr Gates might, for example, reflect on how fewer secretaries, accountants, filing clerks and librarians are needed because of Windows and Microsoft Office. If robots are to be taxed because they make workers obsolete, presumably so should software, process innovation, and any other labour-saving change, but, at the same time, taxing labour-saving innovation is a means to have less of it. Surely that is not the right way to think about it.
A better question to ask is not what makes (some) workers obsolete, in order to tax it, but which activities give rise to what economists called “rent” — profitability over and above what it would take to keep an economic activity going. Robert Solow, one of the 20th century’s great economists, has recently advocated this perspective on what has happened in the US economy in the past 50 years or so, and in particular the changing shares of capital and labour in national income. He argues that there was always a large share of “rent” in private business, and that there used to be an implicit (and sometimes explicit) social contract protecting the split of this surplus between workers and business owners (and, we may add, senior management). That social contract has been eroded, with the effect that rent is now captured by those with the power to do so.
We should inquire, therefore, whether the march of the robots, and other labour-saving innovations, heralds a greater degree of rent derived from the economic activities they are used in. That may be so; it may also be that the answer varies between types of productivity advances. As the FT’s recent Technology and Society special report shows, artificial intelligence is as much a source of disruption as robotics — machine learning could presage the sort of scale economies that make large rents likely. In any case, the answer requires knowledge of much more than whether an innovation reduces the need for human workers; cost structure, ownership, and intellectual property protection may matter as much or more.
And as far as UBI goes, there is a deep tradition in seeing it as a remedy for excessive rent creation. Thomas Paine, who made the first fully-fledged proposal for UBI in his pamphlet Agrarian Justice, justified it by the existence of rent in the ownership of land — the fact that landowners, through no effort of their own, capture a part of the value due to development cultivation. Paine wrote: “Every proprietor, therefore, of cultivated lands, owes to the community ground rent (for I know of no better term to express the idea) for the land which he holds” because of this.
If there are rents arising from the adoption of robotics, AI, or other labour-saving technologies, Paine’s analysis applies to technological justice today as much as to agrarian justice in the late 18th century. So does his conclusion in favour of a tax on this rent to fund a universal payment to all citizens.
- For more on UBI, Cardiff Garcia and I discuss that and a host of other newly possible radical policy proposals in the latest edition of FTAlphaville’s Alphachat podcast.
- Kevin O’Rourke examines how Brexit will affect the Irish border: “It is logically incoherent to argue that if we remain in the EU and its customs union, and the North leaves both, there can be some special deal that will avoid the need for a customs frontier on the island.”
- German producer price inflation doubled from December to January, to 2.4 per cent year-on-year.