A basic income could be universal OR conditional

A UBI true to its name would be unconditional and have no means test for eligibility. It would be given to every individual, regardless of their own or their family income. A related, but quite distinct, policy would be a conditional basic income program, or a Negative Income Tax (NIT), as it was named by University of Chicago economist Milton Friedman in his 1962 book Capitalism and Freedom. Under a conditional basic income scheme, the government would provide every individual a guaranteed income level, or a stipend, which would gradually be reduced as their earned income increased. This type of scheme was considered by the Nixon administration and evaluated in randomized controlled trial (RCT) social experiments in the 1960s and 1970s (discussed below).

We view a UBI to be related to a negative income tax, albeit much more expensive and with a very “leaky bucket,” to use the metaphor coined by Brookings scholar Arthur Okun (1975). to describe the “socio-economic leakages” of redistributive policy that arise from administrative costs, reduced savings or investment, or work disincentives. One of our main objections to a UBI is that by making the payment universal and unconditional, the government would be paying a lot of money to well off families, which would not serve any useful redistributive purpose but would divert public funds away from programs that could usefully advance human capital development and economic opportunity for truly needy Americans.

Here we outline key elements of some specific UBI proposals that have been put forward by UBI advocates in the past few years. As this discussion makes clear, many people who support UBI are actually arguing for a conditional basic income, not a universal basic income. We also offer some calculations showing how much it would cost the government to provide conditional basic income payments. These calculations highlight both the large expense of such schemes and the trade-off implicit in a conditional basic income scheme in terms of giving more money to fewer, lower income people or less money to more people, including those with fairly high levels of income. For simplicity, the calculations assume no behavioral response. That said, incorporating behavioral responses is likely to make the UBI proposals even more expensive, both directly through negative income effects on labor supply and indirectly through the distortionary costs of the additional taxes necessary to pay for the program.