I was musing about how to implement the Georgist idea of replacing all taxes with a single tax on land value, even if you don't have the authority to do anything about the other taxes.
Eg think a local government that can raise a land tax, but can't muck about with federal income taxes; or even a private community of people banding together. They can buy land as a cooperative/corporation, and lease it out to individuals (ala collecting land tax), but don't have any proper taxing authority.
For convenience sake, I will assume a city in the following.
Now, there's an interesting result in economics that given capital and labour mobility, All Taxes eventually Come Out of land Rent (ATCOR), and all subsidies, benefits from infrastructure etc eventually go towards land rent. (See https://en.wikipedia.org/wiki/Henry_George_theorem)
The plausibility (even if not correctness) of this claim is easy to see: real estate next to a train station or a good school is more expensive. Real estate in an area of high taxation is cheaper. (You can see that nicely on the border between Switzerland and Germany, where it's entirely feasible to live in one country and commute to the other for work.)
My first idea was thus: everybody records the federal taxes paid on economic activity within the city. If ATCOR is true, land values (and thus land taxes collected) will increase until the refund is fully covered. People pay higher land rents, but don't care about federal taxes any more. Especially, not about the marginal tax rate.
Of course, this system might be gamed. So I added some bureaucracy: instead of direct refunds, give people tax credits that they can trade and use in a combination with cash for paying the land tax. (Or pass on to their landlord, grocery shop etc for them to pay land taxes with. Or give to grandma as a gift.) Also, you might only count one dollar of federal taxes paid as much as 80 cents of cash for paying land taxes (to preserve some private incentives for lowering your federal taxes); and do something in case you get more credits handed over in the beginning of scheme than you are actually taking in land tax. And you'd also want to be very careful about auditing people's claim about how much federal tax they paid.
The impact on marginal tax rates remain: their disincentives are blunted. Earning an extra dollar more means an extra dollar more in your pocket, as long as you don't have to directly or indirectly use more land to earn that extra dollar.
Where we might turn a profit is that effectively untaxing economic activity might make the economy of the city boom, thus increasing the value of the land beyond the status quo---even after taking the discounts/refunds we give into account.
(There's a bunch more stuff to explain, eg how to assess the land values. For the sake of argument, please take the Georgist ideas for granted.)
Now to re-introduce the basic income: the city could give out some credits per resident for free. Someone who only uses a very small amount of land might come out net-positive (via trading their credits on eg a centralised exchange).
Of course, if everyone tried to loaf of their basic income, the whole system would collapse. On average, people need to still generate economic activity that keep the value of the land high.
But in some sense that's a very familiar argument against universal basic income.
I said you might turn a profit on the whole scheme. That requires two assumptions: first that the basic scheme without basic income creates enough of a boom to make it worthwhile for a private investor to set up; second that providing a basic income would further encourage economic activity. (An argument that some basic income proponents already make.)
Please critique the above sketch. I suggest a two-pronged approach: if you want to show that the approach is impractical, please steelman it as much as possible before you shoot it down. (Ie please try to shoot down the best possible argument you can think of, even giving all the concessions about what might or might not work in practice you can.) This is so that we can get a definitive verdict.
If you are on the other side, and think the scheme might succeed: please point out all weaknesses (and try to repair them, if possible).
I am fairly confident that a scheme like the above can succeed for a city. The land tax plus credits scheme will bring in revenue at less deadweight loss than any other taxes.
I am less confident that adding basic income to the scheme will increase its net profits. But a city might do it anyway for humanitarian reasons.
(If you believe that a universal basic income should be the only welfare benefit necessary, any other welfare benefits would be subtracted from your tax credits. This single-benefit-scheme is a nice symmetry to the way we can emulate a single tax. I am not sure, whether that's a good idea, or even legally possible.)