Should robots pay taxes?
It may sound strange, but a number of prominent people have been asking this question lately. As fears about the impact of automation grow, calls for a “robot tax” are gaining momentum. Earlier this month, the European parliament considered one for the EU. Benoît Hamon, the French Socialist party presidential candidate who is often described as his country’s Bernie Sanders, has put a robot tax in his platform. Even Bill Gates recently endorsed the idea.
The proposals vary, but they share a common premise. As machines and algorithms get smarter, they’ll replace a widening share of the workforce. A robot tax could raise revenue to retrain those displaced workers, or supply them with a basic income.
The good news is that the robot apocalypse hasn’t arrived just yet. Despite a steady stream of alarming headlines about clever computers gobbling up our jobs, the economic data suggests that automation isn’t happening on a large scale. The bad news is that if it does, it will produce a level of inequality that will make present-day America look like an egalitarian utopia by comparison.
The real threat posed by robots isn’t that they will become evil and kill us all, which is what keeps Elon Musk up at night – it’s that they will amplify economic disparities to such an extreme that life will become, quite literally, unlivable for the vast majority. A robot tax may or may not be a useful policy tool for averting this scenario. But it’s a good starting point for an important conversation. Mass automation presents a serious political problem – one that demands a serious political solution.
Automation isn’t new. In the late 16th century, an English inventor developed a knitting machine known as the stocking frame. By hand, workers averaged 100 stitches per minute; with the stocking frame, they averaged 1,000. This is the basic pattern, repeated through centuries: as technology improves, it reduces the amount of labor required to produce a certain number of goods.
So far, however, this phenomenon hasn’t produced extreme unemployment. That’s because automation can create jobs as well as destroy them. One recent example is bank tellers: ATMs began to appear in the 1970s, but the total number of tellers has actually grown since then. As ATMs made it cheaper to run a branch, banks opened more branches, leading to more tellers overall. The job description has changed –today’s tellers spend more time selling financial services than dispensing cash – but the jobs are still there.
What’s different this time is the possibility that technology will become so sophisticated that there won’t be anything left for humans to do. What if your ATM could not only give you a hundred bucks, but sell you an adjustable-rate mortgage? While the current rhetoric around artificial intelligence is overhyped, there have been meaningful advances over the past several years. And it’s not inconceivable that much bigger breakthroughs are on the horizon. Instead of merely transforming work, technology might begin to eliminate it. Instead of making it possible to create more wealth with less labor, automation might make it possible to create more wealth without labor.
What’s so bad about wealth without labor? It depends on who owns the wealth. Under capitalism, wages are how workers receive a portion of what they produce. That portion has always been small, relative to the rewards that flow to the owners of capital. And over the past several decades, it’s gotten smaller: the share of the national income that goes to wages has been steadily shrinking, while the share that goes to capital has been growing. Technology has made workers more productive, but the profits have trickled up, not down. Productivity increased by 80.4% between 1973 and 2011, but the real hourly compensation of the median worker went up by only 10.7%.
As bad as this is, mass automation threatens to make it much worse. If you think inequality is a problem now, imagine a world where the rich can get richer all by themselves. Capital liberated from labor means not merely the end of work, but the end of the wage. And without the wage, workers lose their only access to wealth – not to mention their only means of survival. They also lose their primary source of social power. So long as workers control the point of production, they can shut it down. The strike is still the most effective weapon workers have, even if they rarely use it any more. A fully automated economy would make them not just redundant, but powerless.
Meanwhile, robotic capital would enable elites to completely secede from society. From private jets to private islands, the rich already devote a great deal of time and expense to insulating themselves from other people. But even the best fortified luxury bunker is tethered to the outside world, so long as capital needs labor to reproduce itself. Mass automation would make it possible to sever this link. Equipped with an infinite supply of workerless wealth, elites could seal themselves off in a gated paradise, leaving the unemployed masses to rot.
If that scenario isn’t bleak enough, consider the possibility that mass automation could lead not only to the impoverishment of working people, but to their annihilation. In his book Four Futures, Peter Frase speculates that the economically redundant hordes outside the gates would only be tolerated for so long. After all, they might get restless – and that’s a lot of possible pitchforks. “What happens if the masses are dangerous but are no longer a working class, and hence of no value to the rulers?” Frase writes. “Someone will eventually get the idea that it would be better to get rid of them.” He gives this future an appropriately frightening name: “exterminism”, a world defined by the “genocidal war of the rich against the poor”.
These dystopias may sound like science fiction, but they’re perfectly plausible given our current trajectory. The technology around robotics and artificial intelligence will continue to improve – but without substantive political change, the outcome will range from bad to apocalyptic for most people. That’s why the recent rumblings about a robot tax are worth taking seriously. They offer an opportunity to develop the political response to mass automation now, before it’s too late.
When I asked the prominent leftwing thinker Matt Bruenig for his thoughts, he explained that whatever we do, we shouldn’t try to discourage automation. “The problem with robots is not the manufacturing and application of them – that’s actually good for productivity,” he told me. “The problem is that they are owned by the wealthy, which means that the income that flows to the robots go out to a small slice of wealthy people.”
Job-killing robots are good, in other words, so long as the prosperity they produce is widely distributed. An Oxfam report released earlier this year revealed that the eight richest men in the world own as much wealth as half the human race. Imagine what those numbers will look like if automation accelerates. At some point, a handful of billionaires could control close to one hundred percent of society’s wealth. Then, perhaps, the idea that wealth should be owned by the many, rather than monopolized by the few, won’t seem so radical, and we can undertake a bit of sorely needed redistribution – before robot capitalism kills us all.