There’s a very good chance that Finland will become the first country to implement a nationwide Universal Basic Income (UBI). The country’s Social Insurance Institution (Kela) is studying proposals to replace most of its existing welfare state programs with a simple cash grant – 800 euros for every Finnish citizen, regardless of whether they work, regardless of whether they’re willing to work, and regardless of how wealthy they are or how much other income they earn.
I’ve argued in the past that a UBI has considerable attraction from a libertarian perspective, at least compared with more traditional welfare state programs. But here I want to focus on what many might regard as a puzzling argument that the Finnish government is making on behalf of it: they think a UBI will help reduce Finland’s serious problem of unemployment.
On its face, the claim that you can reduce unemployment by making welfare payments more generous might seem implausible. But there’s more to be said on behalf of this claim than one might expect.
To see why, we need to distinguish two different kinds of effect that a basic income might have on people’s willingness to work. The first effect, which economists call the “income effect,” leads people to work less as their income increases. All else being equal, leisure is a good that people tend to “buy” more of as their income increases. The more money you get from the government, the more likely you are to use some of this money to support yourself while you work fewer hours and relax. Therefore, because of the income effect, a basic income might lead to some increase in unemployment, just as most people assume it would.
However, a true UBI would have another, not-so-obvious effect. This effect, called the “substitution effect,” leads people to work more when the opportunity cost for leisure is increased. The more your after-tax earnings go up, the more expensive it is to take time off from work, and the more likely you are to substitute work for leisure. Because of the substitution effect, people might actually work more under a basic income.
The key is that, under Finland’s plan, the basic income would be replacing other welfare programs. Most of those other welfare programs are means-tested, which means that once recipients’ income passes a certain threshold, they lose the benefit. And that means that recipients can actually wind up financially worse off when they work more hours or get a raise. Means-tested welfare programs discourage work by increasing effective marginal tax rates, and thereby making work pay less than it otherwise would.
Because Finland’s UBI would be unconditional, there wouldn’t be any means-testing of the benefits. That means that if a Finnish citizen who received 800 euros from the UBI took a job and earned an additional 500 euros, she wouldn’t lose any of her original benefit. Whatever money she earns would be hers to keep. Of course, she’d still have to pay the normal income taxes on those earnings. But because her initial UBI benefit doesn’t go away, there’s less of a work disincentive under the UBI than there is under Finland’s existing welfare system.
Whether the UBI increases or decreases unemployment, therefore, depends on the relative magnitude of the income effect compared to the substitution effect. But as Ed Dolan has argued, for relatively modest UBIs of the sort Finland is considering, the income effect is likely to produce only a small work disincentive for those whose incomes are already relatively high, while the substitution effect is likely to produce a substantial increase in the incentive to work among low-income recipients. That means that the net effect on work incentives is likely to be positive.
The available empirical evidence appears to support this hypothesis. Opponents of a UBI often point to the famous Negative Income Tax (NIT) experiments conducted in the 1970s as evidence that a basic income would increase unemployment. But a NIT is not a UBI. Under the kind of NIT proposed by Milton Friedman, the benefits one receives from the program are a function of how far one’s income falls below a certain threshold. Because the size of one’s benefit diminishes with increases in one’s income, the NIT produces a much stronger work disincentive than a UBI. Moreover, as Dylan Matthews has reported, even the NIT experiments don’t show as much of an unemployment effect as is usually assumed. A UBI would almost certainly do better.
I have some serious concerns regarding the feasibility of a UBI, the most significant of which has to do with its very high cost. As Cato’s Michael Tanner has shown, there’s good reason to worry that the size of the grant provided by a UBI is either going to be too small to meet people’s basic needs, or too large to be affordable. Finland’s plan, at least in its current manifestation, seems to fall into that latter category.
Still, I’m happy to see that the proposal might now be tested in a rigorous way. As Dylan Matthews explains, the Finnish proposal is explicitly experimental in nature, and will, if adopted, start off on a relatively small scale in 2017, testing not just a full-scale UBI but also a more modest UBI and a NIT. Both opponents and supporters of a UBI should be grateful for the data that such an experiment will provide. One way or another, we’ll learn something!
Here’s a video I made for Learn Liberty on the benefits of a Universal Basic Income.