Whether you’re for or against universal basic income, your data is no good

Come and get it.
Come and get it.
Image: Stefan Bohrer / Wikimedia
We may earn a commission from links on this page.

The Swiss just voted down a proposal to give every citizen a guaranteed minimum payment known as a universal basic income (UBI), but around the world, the debate is only getting started.

Some Libertarians see it as a chance to dismantle the welfare bureaucracy. Silicon Valley investors say it’ll be a lifeline for workers displaced by the coming robot revolution. And economists dismiss it as a costly fantasy and a disincentive to work.

Amid all the strongly held views about UBI, it’s important to remember it all rests on rather spotty data. Researchers reviewing the science in 1993 were simultaneously claiming “few adverse effects” and “an avalanche of negative results.” It’s not that different today.

GiveDirectly, a charity that gives unconditional cash to the poor, wanted to see what research to date had to say. It found eight programs that had tested how versions of universal basic incomes affected people’s well-being over the long term by covering their basic expenses. But the glaring problem with the experiments was that none met the three fundamental criteria for a viable universal basic income program: that it be universal (for everyone in a geography); basic (cover minimum living expenses); and long-term (over 10 years or more). Instead, the data was a series of experimental payment schemes around the world over the last 50 years that, while extensive, didn’t tick all the boxes.

To get a better answer to the question of whether UBI really works, GiveDirectly is running its own pilot. It’s raising private and crowdfunded money for its own $30 million pilot giving regular cash payments to thousands of extremely poor households in East Africa.

“We expect some of the most interesting results within the first year or two just by making a long-term commitment,” wrote Ian Bassin, the chief operating officer of GiveDirectly in the United States. “Once we’ve guaranteed an income to someone for 10 years, how does their behavior change immediately: do they go back to school? Start a business? Spend more time raising their kids?”

Critics charge that a universal income is too expensive, eliminates a force that ”organizes people’s lives,” encourages sloth (pdf), and dissolves community bonds. Bassin challenged these findings. ”Cash grants to the poor have arguably the strongest evidence base for anti-poverty policies,” wrote Bassin in an email (links are his). “People don’t get drunk and they don’t stop working.”

Another problem, however, is the evidence depends so heavily on context. What works in Kenya does not necessary apply to veterans on disability payments in the United States. Most studies extrapolate from existing welfare programs or labor data, rather than strict UBI experiments. You can see a comprehensive list of related studies in this Reddit group.

Better answers should be forthcoming in the next few years. At least five more pilots are underway or planned in Oakland, OntarioThe NetherlandsGermany and Finland.

The image above was taken by Stefan Bohrer and shared under a Creative Commons license on Wikicommons.