[efaref]

Functionally they're equivalent. In a UBI/negative-income-tax schenariom things like food stamps also shouldn't be cut off at a simple threshold, but instead factored in at your marginal rate. If your marginal rate is always your marginal rate, then it's always beneficial to take a higher-paying job.

The problem with means testing is that it effectively raises the marginal rate, often to astronomical percentages. For example, the withdrawal of a $1000 benefit because of an additional $1 that pushes you over a means testing threshold means your marginal tax rate on that dollar is 100,000%. Of course the rational thing to do is to not earn that dollar, or indeed any of the next $1250 that it would take to get you back to where you started (assuming a normal marginal rate of 20%).

[dragonwriter]

> In a UBI/negative-income-tax schenariom things like food stamps also shouldn't be cut off at a simple threshold, but instead factored in at your marginal rate.

In a mature UBI system [0], food stamps as use-restricted means-tested welfare wouldn't exist at all. (Food stamps in their original purpose as a system of agricultural subsidies might.) Their function would be rolled into the single benefit payment.

[0] transitional proposals often keep other means-tested programs during the transition, scaled back as UBI scales up.

[nottorp]

In a UBI scenario food stamps shouldn't exist, i think? The idea is to replace most benefits with just the basic income and save money not tracking those benefits.

[efaref]

That's the idea. In practice you need to be careful as a failure mode of introducing UBI is it being used as an excuse to cut benefits without an adequate replacement.

For example, a reasonable stepping stone would be for people to get their food stamps by default, but then be able to "pay" some of the taxes on any income they earn using those food stamps. This means food stamps are being reduced gradually and at the marginal tax rate, so it's fairer than just falling off a cliff when you hit a threshold.