A robot tax won’t fix our public funding crisis


The threats posed by automation are now a common topic for futurologists.

The reality is that automation has been quietly progressing for some time, and while we are a long way off a Skynet-style singularity, the development of artificial intelligence means robots are not only taking our current jobs, they are going to take more.

No one seems to know how many – predictions vary from a doom-laden 47 per cent to a highly optimistic five per cent – but most hover around a third.

Read more: Questioning AI ethics doesn’t make you a gloomy Luddite

So far, fears have coalesced around loss of income to now-redundant workers, but there is another issue that has yet to be properly examined: the effect on tax revenues.

In the UK, as in most countries, a large slice of government revenue comes from taxes on workers’ wages. Together, income tax and National Insurance contributions make up around 44 per cent of the complete UK tax take.

Corporation tax, by contrast, makes up just eight per cent. So even if automation means companies see revenues grow and generate more capacity to pay corporation tax – and more taxable dividends for shareholders – it may not be enough to compensate for the losses to employment income.

This would have a significant impact on vital public spending areas like healthcare and pensions – already under pressure – and would come at a time when redundant workers who have lost their incomes need far greater support. Income inequality, which has been falling, would skyrocket.

Governments would have to face even harder choices than they do today.

So should we tax robots? It is a concept that has been mooted in more than one jurisdiction. Last year, South Korea was the first country to attempt something like a robot tax, although in practice it was simply a limit to the incentives available for investment in automated machines. Diverse figures from Bill Gates to Jeremy Corbyn have suggested that robots should be taxed.

The idea has some obvious attractions. The proposition is that either the installation of robots or the profits companies enjoy thanks to savings in displaced labour could be taxed, and the money used to retrain workers or finance an expansion in socially important but hard-to-automate sectors like healthcare, education, or social care.

The problem is how it would work in practice. You cannot tax a machine as you can a worker, since the burden falls on the company that has acquired the robot rather than the employee.

Essentially, it would be a tax on the capital employed by those companies. This might lead them pass the costs on to both labour and consumers, through lower wages and higher prices.

But the real danger is that it would come at a terrible time for the British economy.

A 2016 report by the Centre for Policy Studies suggested that Britain’s problem is not too many robots, but too few. In the UK, there are currently 71 robots for every 10,000 employers in the manufacturing sector. By contrast, in Germany there are 300.

Increased automation would be a shot in the arm for efficiency and productivity, two areas where the UK lags behind many competitors.

There is broad consensus that the UK needs to invest more in infrastructure in order to improve growth, and automation is a key part of this.

Slowing the deployment of robots into business or making it much more expensive to use them would both increase costs and hinder productivity. The likely result would be more expensive goods and services, further draining remaining worker incomes.

Ultimately, a “robot tax” looks to be a short-sighted, knee-jerk response to the deeper systemic problem of how to organise and fund a society where public finances are shrinking while demands increase.

We need non-ideological strategies to rethink our labour landscape and tax base in the long term, not quick fixes that will ultimately make the underlying problem worse.

Read more: No more dodging hard questions, we need a conversation about AI

Original Article