That isn't a very strong argument. Assuming the basic income and associated taxation is calculated to approximate the same level of redistribution of wealth as the existing welfare system, the amount a person making a given level of income would have to have their taxes increased by in order to balance the budget is only equal to the amount they themselves would receive from the basic income but would not have qualified for in a means tested welfare system. For high income earners that amount is negligible because the basic income constitutes such a small proportion of their total income, and even then it has no effect on their effective tax rate (i.e. net transfers with the government; taxes paid less basic income), only that small change to the marginal rate needed to keep the effective rate the same.

And in particular what is really happening is that the disincentive (which is created by any taxation for any purpose) is being spread more evenly throughout the population, which all else equal can be expected to have less effect on behaviour than concentrating the effect on a subgroup which will then be much more compelled to react to it.

Moreover, how the taxes are collected is separate from the basic income. If we find that it is for some strange reason better to charge a higher marginal rate to lower income people than higher income people then we can have a basic income and yet still do that.

Agreed, which is why I object to your statement that: We already know how to design a program that won't discourage work. It's not that basic income discourages work more or less, but that it evens out the marginal tax rate. This might be a good or a bad thing, but your original post contains no argument either way. You run a very strong risk of the reader misinterpreting what you are saying, and believing that there will be a uniform improvement in incentives across the whole population.

which all else equal can be expected to have less effect on behaviour than concentrating the effect on a subgroup which will then be much more compelled to react to it.

I think the main argument behind the current system is that low income people need the money, and therefore are less affected by the marginal tax rate.

Moreover, how the taxes are collected is separate from the basic income. If we find that it is for some strange reason better to charge a higher marginal rate to lower income people than higher income people then we can have a basic income and yet still do that.

Either "basic income" should be used to denote a flat tax plus basic income, or a varying marginal tax plus basic income. Your first post implies the former to the exclusion of the latter. I think you are changing the definition because it's important to you that basic income "win", but really there are two separate issues, which are negative taxation vs welfare, and fixed vs varying effective marginal tax rates. Each of these issues should be decided independently on their own merits.

Because the marginal tax rate is a separate thing. You can raise the money however you like. Tax the rich, tax the poor, tax securities derivatives, print lots of money, it's a separate consideration. What a basic income does is dramatically simplify the welfare system so that when you balance the other side of the ledger you can better see what you're looking at.

It may then also make sense to simplify the tax system and impose a uniform rate on everyone (because the level of redistribution of wealth can now be trivially set by adjusting the amount of the basic income), but there is no inherent reason why any other method of raising the money could not be used instead. Even in that case a basic income is preferable to the existing system of ad hoc means tested welfare programs.

> I think the main argument behind the current system is that low income people need the money, and therefore are less affected by the marginal tax rate.

There are three points here.

First, if these people need the money so badly, is it really good policy to be imposing high marginal tax rates on them, regardless of whether they'll bear it because they have to?

Second, the people we currently take benefits away from are not the lowest income people (who do qualify for the programs), they're the low-middle income people who only just barely don't. If the true consideration is to collect from people who will most be forced to work anyway then shouldn't we be sticking it to the even poorer people who have even less choice? (Of course this runs right back into the first point.)

And finally, if these people need the money so badly then preventing them from earning as much by working legitimate jobs that report their income will cause more to turn to off-the-books income like dealing drugs and stealing in order to make ends meat. Presumably that isn't the work we want to maintain an incentive for.

See my other post for answers to the rest.

Of course it does, because it's in the context of receiving the benefit. Since receiving the benefit is unconditional, it doesn't cost you money/benefits to take a job or work more hours, unlike with means tested programs which do exactly that.

How the government comes up with the money to pay out the benefit is a separate question because you can do it in any number of ways. Imagine the government pays for the basic income by conquering foreign lands and seizing their resources. Unambiguously no disincentive to work for all the citizens.

Obviously in the real world the money is going to come from taxes. But taxes are separate. All transfer programs require taxes, on top of the effects created by the program itself. Because the method of taxation that has zero effect on the incentive to work is the mirror image of the basic income: Everyone pays the same fixed dollar amount in taxes regardless of income. Using that method of taxation to fund a basic income converts it from a transfer payment to a no-op.

So if your point is that you cannot have a transfer payment without creating a disincentive to work for someone, that is correct, because that is what a transfer payment is: In order to take from those with more and give to those with less, you have to condition at least one of the taking or the giving on the existence of some threshold level of wealth.

But the difference between the program and the tax that pays for it is still quote important, because the people they affect are not necessarily the same people. For example, suppose you want to fund some transfer program using a tax exclusively paid by the top 1% of income earners. Everyone in the bottom 99% therefore does not pay the tax at all and is not directly affected by it. But now the separate choice comes between whether the program is a traditional means tested welfare program or a basic income. The means tested program will provide a disincentive for the 99% to work because the government will claw back some significant portion of the money they earn; the basic income will not, even inclusive of its funding source, because it is not funded by them.

Also, there are some taxes where the disincentive effect is beneficial. Basic income funded by a carbon tax seems like a good combination.

At any income level you can make arguments why their work output would be more/less sensitive to taxation. We are talking about a difference in a difference (how much one group changes their work in response to tax, vs another) and not absolutes, which complicates matters. E.g. poor people need the money more, so maybe they are less responsive to marginal tax rates.

Also, there are some taxes where the disincentive effect is beneficial. Basic income funded by a carbon tax seems like a good combination.

Pollution taxes are a good idea for their own sake, but they will only cover a tiny percentage of basic income or welfare.