What happens when you give people free money with no strings attached?
The question is of enormous importance to America’s safety-net policy, given widespread proposals for a “universal basic income” or a “child allowance”— which would give unconditional cash payments to all adults or all parents.
A massive scientific experiment has now provided the most compelling answers to date.
Those answers pretty much align with common sense: When people get free money, they work a bit less and play a bit more.
And while they spend more on health care, they don’t see enduring health improvements as a result.
This study won’t end the movement for no-strings cash, of course, but it certainly validates some common criticisms of these policies.
The experiment included 3,000 lower-income Americans aged 21 to 40, drawn from 19 counties in the Dallas and Chicago areas, with an average household income of just under $30,000.
Two-thirds of these folks were randomly assigned to the control group and paid $50 a month for their continued participation. The other 1,000 got $1,000 every month for three years.
That works out to $1 million a month, and $36 million over the whole time period, funded by private donations.
Researchers collected detailed data on what happened to the participants and published some key findings Monday in a pair of working papers through the National Bureau for Economic Research.
The first paper focuses on employment results, a major focus of the UBI and child-allowance debates — despite the limited and contradictory evidence on the question, much of it from small or dated experiments.
Some advocates have even raised their hopes that a UBI could boost employment if recipients use the money, say, to get training or to pay for transportation.
The upshot of this experiment, though, is a reduction in work effort that, as the authors put it, “does not appear offset by other productive activities.”
Labor-force participation fell by two percentage points, and spells of unemployment grew about a month longer.
Participants and their partners each cut their working hours by more than one hour per week, on average.
Overall, giving someone $12,000 a year reduced their individual income by about $1,500, and also reduced partners’ income, for an overall reduction of more than 20 cents per dollar received. Scaled up nationwide, these would be substantial effects.
The authors also collected data on what individuals did with their extra time: Leisure increased by about as much as work went down.
Recipients did not meaningfully increase “time spent on childcare, exercising, searching for a job, or time spent on self-improvement.”
However, the reduction in work was strongly concentrated among non-parents, implying that child allowances might have less impact than universal payments to all adults.
As far as investing in human capital, no detectable improvement occurred in recipients’ job quality (including amenities), though younger participants may have gotten a bit more education.
There also wasn’t much growth in savings; recipients went through “nearly all” of the money they received, either by reducing work or by increasing consumption.
The second paper, which shares some authors with the first, looks at health: Recipients spent about $20 per month more on medical care, went to the hospital more, and enjoyed short-term improvements in stress levels and food security.
Yet the transfers did not seem to enhance overall physical health and only temporarily improved mental health.
As the authors summarize, “More targeted interventions may be more effective at reducing health inequality between high- and low-income individuals.”
Of course, large as this experiment was, it’s just one study, testing a specific type of payment. Significantly larger or smaller payments would almost certainly lead to different outcomes.
Permanent payments would also have different effects — in this experiment, the employment reduction got stronger with time until close to the end, when recipients knew that they could not count on the money much longer.
The experiment was also limited to lower-income Americans, those making at most three times the poverty level.
Presumably, giving money to people making more money would have smaller effects on work. However, if “universal” benefits were funded through higher taxes, many higher earners would also face a new work disincentive.
At any rate, these results have — at great expense — confirmed what always seemed likely: Unconditional cash transfers reduce work while benefiting recipients in generally short-term ways.
That finding may not end the universal basic income debate, but it deserves to be a big part of any discussion going forward.
Robert VerBruggen is a fellow at the Manhattan Institute. Adapted from City Journal.