Bill Gates has stated in an interview that robots who take human jobs should pay taxes. This has some obvious attractions. Not only, as Gates says, will we be able to spend the money to finance jobs for which humans are particularly suited, such as caring for children or the elderly, but robots are also unlikely to complain about tax levels, they don’t use services financed by tax revenue such as education or the health services and they are most unlikely to salt away income and assets in a tax haven. What’s not to like?
Well, actually, you can’t tax robots any more than you can tax any other inanimate object – but Bill Gates’s suggestion does address some of today’s most important tax issues. What proportion of its tax revenue should a state raise from each of the three main tax bases; capital, labour and expenditure? And how can a state counteract tax avoidance by large companies and wealthy individuals?
Taxing robots would, in reality, be a tax on the capital employed by businesses in using them and might help to redress the long-term shift away from taxing capital. In 1981, the rate of corporation tax was 52%, although generous relief meant that the tax base was relatively narrow. This has now fallen to 20% and further reductions to 17% are planned by 2020.
By contrast, the principal expenditure tax, VAT, was originally set at 8% in 1973, but rose to 15% in 1979 and is now 20%. This means that individuals are contributing a larger proportion of tax revenue than previously through taxes on salaries and expenditure and businesses are contributing less through taxes on their profits – even though they make use of the UK’s transport, financial and legal infrastructure and benefit from the education and healthcare provided to their employees.
One of the justifications for not taxing capital is that companies do not bear the economic cost of taxation through lower returns to their shareholders, but pass it on to labour and consumers through lower wages and/or higher prices. This argument is contentious.
Edward Kleinbard, Professor of Law and Business at the University of Southern California, observes that “unseemly scuffles” can ensue when this topic is discussed at academic conferences. But whereas academics can have a good punch-up, agree to differ and then retire to the bar for drinks, Kleinbard cites several US government departments and committees, who must make policy based on their assumptions, estimate that capital bears between 75% and 95% of the economic cost of corporation tax, meaning that they can only shift a small part of it onto labour and consumers.
Furthermore, when US Uncut in April 2011 issued a hoax press release purporting to be from GE, stating that it would hand back a US$$3.2 billion tax refund as “contrition for past abuses”, the company’s market capitalisation very briefly fell by around US$3.5 billion.
On this evidence, shareholders behave as though they believe that capital bears the economic cost of corporation tax through reduced dividends. If they did not believe this, why would they have cared? On the evidence of Kleinbard and the hoax, the theory that capital does not bear the economic cost of tax would therefore appear to be a rationalising discourse put forward by those who benefit from lower taxes on capital.
Taxing robots might also help to counteract tax avoidance, because the tax would be calculated by taxing a notional salary paid to the robot, and the company would be allowed to deduct this notional payment for the purpose of corporation tax.
Tax avoidance by large multinationals typically operates by transferring taxable profits from where they economically arise to tax havens, where their presence is often no more than a brass plate on a wall and a mailbox, or even, in the case of Apple, to a company located in some mid-Atlantic limbo, whose profits were therefore not taxable in any tax jurisdiction. These companies pay the same rate of corporation tax as all other companies on their profits remaining in the UK or Ireland, but ensure that these are only a small fraction of their total profits.
In contrast, the robot tax, just like salaries, would be calculated on an amount notionally payable out of revenue and would be payable in the tax jurisdiction in which the robot was located. This would be where the revenue was economically generated and this location would be determined by economics rather than tax considerations.
Is it even possible?
Finally, Bill Gates puts forward the currently unfashionable view that governments have an important role to play in combating inequality. For the past 35 or 40 years the dominant view has been that this would be achieved through economic growth and should therefore be left to the private sector and the markets.
But Gates says that combating inequality will require large amounts of excess labour to be used to help those on lower incomes, that robots will free up this labour and that the impetus for the necessary changes must come from governments because business cannot or will not do this of their own accord.
Furthermore, he says that taxing robots will not discourage innovation. People are naturally anxious about the effects of such technology, but taxation is a better way of allaying these fears than the alternative of banning aspects of it.
Could taxation of robots ever happen? Certainly it could, but the $64,000 question is whether there is the political will to do it. It would take a major paradigm shift in our attitude towards taxation to see it as a possible force for good, rather than simply a dead weight and burden. However, in the 1960s and 1970s today’s attitude towards taxation would have been equally inconceivable. Never say never.