Universal basic income (UBI) is back, like a space zombie in a sci-fi movie, resurrected from policy oblivion, hungry for policymakers’ attention: brains!
Andrew Yang, whose “Yang Gang” enthusiasm briefly shook up the Democratic presidential nomination in 2020 promoting a “Freedom Dividend” to save workers from automation – $1,000 a month for every American adult – is again the main carrier of the bug: offering UBI to save the nation when robots eat all our jobs.
This time Chat GPT, Yang hopes, will help his argument land: if artificial intelligence truly makes human labor redundant, as so many citizens of the tech bubble in Silicon Valley expect, society will need something other than employment for all of us to make ends meet.
Yet while the warning rings true, the prescription still falls flat. We will need something big and new to spread money around if some super-human intelligence comes for all the jobs. But a UBI, as contemplated by its current cheerleaders, does not start to address the real challenges of an economy that has moved past human labor.
Ask a truck driver (Yang was worried about truck drivers) to live on $1,000 a month. A two-parent, two-kid family on the “Freedom Dividend” would be pretty deep under water, living on 25% less than needed to poke through the poverty line.
The bill to provide every adult a guaranteed income worth, say, $53,000 per year, equivalent to the median earnings of American workers, would add up to over $14tn, about 45% of the United States’ gross domestic product (GDP). Good luck to the politician running on a platform to fund this brave new world.
To put it in perspective, since 1980, the first year for which the Organisation for Economic Co-operation and Development publishes that data, public social spending in the United States – covering health, pensions, disability, unemployment insurance and all that – has never hit 25% of the GDP. Indeed, since the 1960s, the aggregate tax revenue raised by all levels of government has never reached 30% of GDP.
And this doesn’t even consider how challenging redistribution will become once AI kills all labor income, which today generates most tax revenue.
Yang suggested funding his “Freedom Dividend” with a value added tax. This is a tax on consumption that the US does not use but funds a big chunk of Europe’s welfare states. It has merits: It can raise a lot of money, because it is easy to collect at the store checkout, and it does not sap incentives to work and invest, as income taxes do. But it seems a bit ridiculous to propose a world without work in which the livelihoods of most people are funded with a tax on what they buy.
If it meets its investors’ lofty expectations, the AI-powered economy will be radically different from what we know, driving the cost of machines that substitute for human labor below the cost of human subsistence. Nobel economist Wassily Leontief’s observation about horses comes to mind: “the role of humans as the most important factor of production is bound to diminish in the same way that the role of horses in agricultural production was first diminished and then eliminated by the introduction of tractors.”
Maybe we can keep humanity alive via redistribution. Machines that don’t require workers could produce enormous amounts of output, so it might be easy to raise the money for the UBIs of the future.
Given there would be no workers, taxes would have to be raised on something else: carbon emissions, perhaps, or other stuff producing bad externalities, or land, which can be taxed without discouraging production. But this world would likely require substantial taxation of the owners of the robots.
And this would raise new questions about power: Who would determine how much everybody gets? More than likely it would be the select gang of tech oligarchs who own the machines. In an economy in which the labor share of income has gone to zero, the owners of capital end up reaping it all.
To quote economist Erik Brynjolfsson, who runs the digital economy lab at Stanford University: In this world, most of us “would depend precariously on the decisions of those in control of the technology.” Society would risk “being trapped in an equilibrium where those without power have no way to improve their outcomes”.
UBI has features that would prove valuable in an AI-driven future. It does away with the work requirements that often come with welfare, a desirable feature when human work makes no sense. But it fails to address key challenges, notably the enormous built-in inequality that the AI economy would bring about, which might demand redistributing not income but capital ownership in the robots themselves.
Problematically, UBI does not meet the challenge of the present either. America’s current quandary is not zero employment but a large footprint of service jobs that do not provide a living wage. A universal benefit is an extraordinarily expensive tool to fix that, though. A wage subsidy would do much better. How about we improve the design of the earned income tax credit, signed into law by president, Gerald Ford, in 1975?
Less work – as in fewer working hours – does not necessarily require a new paradigm. Australians work 20% less than Americans already; Danes and Finns work 24% less. Spaniards work two-thirds as many hours per day as Americans, on average; the French only 62% as much; Italians about half. These countries don’t rely on UBI, just on a halfway decent social safety net. Before the US tries to reconfigure its welfare state, it might just try that.
