Finland's Basic Income Experiment

On October 31, Finland's public media organization Yle reported that the Finnish welfare institution Kela had begun the initial stages of putting together a Universal Basic Income (UBI) proposal. In the last week, an enormous number of US publications suddenly picked up on this development and produced a deluge of totally inaccurate stories about it (CNN, Forbes, Quartz, HuffPo, and many others). This then set off a mini-storm of takes based upon the false reports. Overall, it was a mess.

Finally, yesterday, Vox's Dylan Matthews pubished an accurately reported story about Finland's nascent UBI plans. The short of it is that there isn't actually a plan yet. The individuals charged with coming up with a plan will, in the next year, produce a report reviewing the results of UBI-type programs used elsewhere in the world. Then, they will eventually come up with a concrete proposal for a limited UBI experiment that will provide benefits to a small fraction of the population. Yet to be decided is how high the experimental UBI payments will be and what, if any, benefits those payments would replace.

The main mistake most of the media coverage made was assuming that the UBI would replace all other benefits. There is no way, even on the most extreme option, that the UBI would replace all other Finnish welfare benefits. But it would probaby replace some. So one of the important questions in implementing a UBI design is how to harmonize it with an existing welfare scheme.

When thinking about what welfare benefits a UBI could or should replace, it's helpful to distinguish between four different benefit types:

  • Services. These are benefits that do not pay out income, but instead provide specific services in-kind. Examples include child care, elder care, disabled care, health care, and education.
  • Flat Benefits. These are benefits that pay out an equal welfare income amount to every eligible recipient. Alternatively, you can understand these as income benefits that are paid out to people regardless what other income they have or their prior earnings. Examples include child allowances, student living grants (in some countries), and minimum benefit levels for old-age, survivor's, and disability pension. A basic income is also this kind of benefit.
  • Means-Tested Benefits. These are benefits that pay out more welfare income to those with lower current incomes than those with higher current incomes. So, for instance, someone whose current income is $10,000 might get $2,000 of benefits, someone whose current income is $15,000 might get $1,000 of benefits, and someone whose current income is $20,000 or more would get no benefits. Examples include housing allowances.
  • Earnings-Related Benefits. These are benefits that pay out more welfare income to those with a stronger earnings record than those with a weaker earnings record. So, for instance, someone who has earned $50,000/year for the last 10 years might get $30,000 of benefits (60% earnings replacement) while someone who has earned $30,000/year for the last 10 eyears would get $25,000 of benefits (83% earnings replacement). Examples include the above-minimum benefit levels for old-age, survivor's, and disability pension, as well as unemployment and paid leave benefits.

There is no indication that Finland's UBI planners intend to replace any of the services. Nor would that make much sense. Care and education benefits serve a function beyond providing extra disposable income to those who would otherwise pay for care and education out of pocket. Specifically, those benefits serve certain labor market functions, either by facilitating higher employment levels (child, elder, and disabled care) or by increasing worker quality and innovation (health care and education).

It is conceivable that some flat benefits could be replaced. Child allowances would have to stay. But the minimum benefit levels for old-age, survivor's, and disability insurance could concievably be folded into a UBI scheme.

It's also conceivable that some or all means-tested benefits could be replaced. If everyone received some kind of housing allowance for instance (folded into the UBI and child allowance to ensure that it was responsive to family size as the housing allowance currently is), then you could probaby kill that benefit as well. The same is true of other means-tested benefits.

Earnings-related benefits are where the interesting benefit replacement decisions will be made. Given that one of the points of the UBI proposal seems to be facilitating higher employment levels, earnings-related working-age benefits like unemployment and disability would seem like obvious targets for replacement. A UBI would cushion working-age people who find themselves unemployed and disabled (as unemployment and disability benefit already do). But, because the UBI wouldn't go away when they became re-employed, it wouldn't create the possibility of discouraging them from accepting work.

So, all in all, I would expect the UBI to leave services untouched, to make certain flat and means-tested benefits obsolete, and to replace certain targeted earnings-related working-age benefits. Whether you can come up with a UBI amount that would make the numbers all work out while also not putting an unacceptably high amount of hardship on disabled and unemployed people is, in my mind, the biggest practical challenge.