Universal Basic Income (UBI) is back, like a space zombie in a science fiction movie, resurrected from policy oblivion, hungry for the attention of policymakers: brains!
Andrew Yang, whose “Yang Gang” zealotry briefly rocked the Democratic presidential nomination in 2020 by promoting a “Freedom Dividend” to protect workers from automation — $1,000 a month for every American adult — is again the main bearer of the bug: offering a UBI to save the nation when robots eat all our jobs.
Yang hopes this time the chat GPT will help ground his argument: If artificial intelligence really does make human labor redundant, as many denizens of the tech bubble in Silicon Valley hope, then society will need something other than jobs for all of us to make ends meet.
Yet while the warning is true, the recipe still fails. If there was a super-human intelligence to do all the jobs, we would need something big and new to spread the wealth around. But a UBI, as envisioned by its current cheerleaders, does not begin to address the real challenges of an economy that has outgrown human labor.
Ask a truck driver (Yang was concerned about truck drivers) to live on $1,000 per month. A two-parent, two-child family on the “Freedom Dividend” would be in deep water, living on 25% less than they need to get over the poverty line.
The bill, which would provide every adult with a guaranteed income of $53,000 per year, equal to the earnings of the average American worker, would raise more than $14 trillion, approximately 45% of the United States’ gross domestic product (GDP). Good luck to any politician running on a platform to finance this brave new world.
To put this in perspective, since 1980, the first year of the Organization for Economic Co-operation and Development publishes that data, public social spending in the United States – which includes health, pensions, disability, unemployment insurance and all that – has never hit 25% of GDP. Indeed, since the 1960s, total tax revenues raised by all levels of government have never reached 30% of GDP.
And this doesn’t even consider how challenging redistribution will become when AI eliminates all labor income, which generates most tax revenue today.
Yang suggested financing his “freedom dividend” with a value-added tax. It’s a tax on consumption that the US doesn’t use but funds a large portion of Europe’s welfare states. It has advantages: It can raise a lot of money, because it is easy to collect at store checkouts, and it does not reduce incentives to work and invest, as income taxes do. But it seems a bit ridiculous to propose a world without work in which most people’s livelihoods depend on taxes on the things they buy.
If it meets the high expectations of its investors, the AI-powered economy will be fundamentally different from what we know, with the cost of machines replacing human labor falling below the cost of human subsistence. Nobel economist Vasily Leontief’s comment about horses comes to mind: “The role of humans as the most important factor of production will diminish in the same way as the role of horses in agricultural production was first diminished and then eliminated by the advent of tractors.”
Perhaps we can keep humanity alive through redistribution. Machines that don’t require workers can produce huge volumes, so it may be easier to fund a future UBI.
Given that there will be no workers, taxes will have to be raised on something else: carbon emissions, perhaps, or other things producing bad externalities, or land, which can be taxed without discouraging production. But this world would probably need to impose substantial taxes on robot owners.
And it will raise new questions about power: Who will determine how much everyone gets? More likely it will be a select gang of tech oligarchs who own the machines. In an economy where labor’s share of income has fallen to zero, the owners of capital ultimately reap all the benefits.
To quote Eric Brynjolfsson, an economist who runs the Digital Economy Lab at Stanford University: In this world, most of us will be “indeterminately dependent on the decisions of those in control of the technology.” Society would risk “getting stuck in an equilibrium where those without power have no way to improve their outcomes”.
UBI has features that will prove valuable in an AI-driven future. It removes the work requirements that often come with well-being, a desirable characteristic when human work has no meaning. But it fails to address key challenges, particularly the huge inherent inequality brought about by an AI economy, which could demand a redistribution not of income but of capital ownership across robots.
Problematically, UBI does not even meet the challenge of the present. America’s current problem is not zero employment, but a large number of service jobs that do not provide livable wages. However, a universal advantage is an exceptionally expensive tool to fix it. Wage subsidy would work much better. What do we think about improving the design of the Earned Income Tax Credit, signed by President Gerald Ford in 1975?
Less work – such as fewer working hours – does not require a new paradigm. Australians already work 20% less than Americans; Danes and Finns work 24% less. Spaniards work an average of two-thirds more hours per day than Americans; the French only as much as 62%; Italians about half. These countries are not dependent on UBI, just half-decent social safety nets. The US might try to do so before it tries to reconfigure its welfare state.
