Lots of folks who are sympathetic to the idea of a basic income guarantee, and many folks who aren’t, have questions about the details of what such a policy would look like. Would a basic income be given only to adults, or should children get it too? Is eligibility limited only to citizens? Would it be distributed to families or to individuals? And, perhaps most importantly, how much would it cost?

Most discussions of the basic income – my own included – are heavy on theoretical justification and light on the details of how it would be implemented. But there’s a new paper by two specialists in tax law – Miranda Fleischer, a colleague of mine at the University of San Diego, and Daniel Hemel of the University of Chicago, that gives these questions the attention they deserve.

The paper, titled, “The Architecture of a Basic Income,” performs two valuable services. First, it helpfully surveys the range of options available in the design of a basic income, and the way in which those options are supported or opposed by various justificatory frameworks such as resource egalitarianism, welfarism, and libertarianism. Second, for each of the questions of design it considers (such as size of the grant, eligibility, frequency of payment, etc.) it tells us what answer the authors think is best justified, and why.

The authors wind up advocating a $500/month basic income payed to every American citizen or permanent resident, regardless of age, income, or assets, to be distributed through the Social Security System and funded by the consolidation of some (but not all) existing transfer programs and a modest increase in taxes.

$500 a month isn’t a lot. It’s certainly not going to eliminate poverty. But it does line up with the Census Bureau’s definition of “deep poverty.” And if you think the point of a basic income is not to substitute for work and voluntary charity but to supplement them for those who are, temporarily, unable to provide for themselves or be adequately provided for by others, then a low dollar value is a feature not a bug.

A $500/month value also keeps the cost within reasonable limits. According to Fleischer and Hemel’s calculations, the total cost of their basic income program would be approximately $1.8 trillion. That’s a sizable amount, but a good chunk of it could be covered by rolling over SNAP, TANF, unemployment insurance, Section 8 housing, the EITC and the Child Tax Credit. Much more could be recouped if we were willing (as the authors argue we should not be) to roll over Social Security benefits as well. The rest could be funded by tax increases. Those tax increases would not be negligible, but some of them would simply recollect the cash being distributed by the basic income and therefore not constitute a true cost, and the net result would still keep our spending-to-GPD ratios below the level of Sweden and Norway – countries which some of us on this blog think are doing things pretty well.

Along the way, there are lots of helpful discussions and asides. For instance, why favor a Universal Basic Income that gives money to everyone rather than a Negative Income Tax that gives money only to those in need? The authors’ answer to that question makes for a long blockquote, but it’s so well done that I’m going to reproduce the whole thing anyways:

While the differences between a UBI and the existing welfare state are significant, the differences between a UBI and a so-called “negative income tax” are not. Moreover, some of the distinctions drawn between different UBI variants—for example, a UBI with a phaseout and a UBI that never phases out—turn out to be alternative ways of accounting for the same flow of funds.

To illustrate: Imagine a UBI of $6000 per year ($500 per month) that extends to all individuals regardless of income, coupled with an income tax of 20 percent on all non-UBI income. Thus:

— A, who earns $0, receives a UBI of $6000 and pays no income tax, for a net transfer from the government of $6000;

— B, who earns $30,000, receives a UBI of $6000 and pays income tax of $6000, for a net transfer to/from the government of zero; and

— C, who earns $60,000, receives a UBI of $6000 and pays income tax of $12,000, for a net transfer to the government of $6000.

Now imagine that instead of the above-described system, we implement a UBI of $6000 that phases out at 20 percent rate starting with one’s first dollar of non-UBI income, and that we couple that with a tax of 20 percent on non-UBI income over $30,000. Thus:

— A, who earns $0, receives a UBI of $6000 and pays no income tax, for a net transfer from the government of $6000;

— B, who earns $30,000, receives no UBI (because her UBI has fully phased out) and pays no income tax, for a net transfer to/from the government of zero; and

— C, who earns $60,000, receives no UBI (again, on account of the phaseout) and pays income tax of $6000, for a net transfer to the government of $6000.

The first example involved a UBI with no formal phaseout; the second UBI phased out explicitly. But as should be evident, the first and second arrangements are identical in economic terms.

Now imagine that instead of a UBI, we implement an arrangement involving a “negative income tax.” Specifically, the income tax is 20% x (income – $30,000), with individuals whose tax is negative (i.e., whose pre-tax income is less than $30,000) receiving a payment from the government of 20 percent times the difference between $30,000 and their pre-tax income. Thus:

— A, who earns $0, faces an income tax of 20% x ($0 – $30,000), which is to say, receives a net transfer from the government of $6000;

— B, who earns $30,000, owes nothing and is owed nothing; and

— C, who earns $60,000, pays a tax of 20% x ($60,000 – $30,000) = $6000.

As should be apparent, this third regime is the exact same in economic terms—resulting in the exact same net transfers between individuals and the government—as the first and second.

That is, the same arrangement can be described as a UBI with no phaseout, or as a UBI with a phaseout, or a negative income tax. The various framings may affect the political popularity of a UBI proposal, but they do nothing to change its essential features.

Read the whole paper for yourself, on SSRN.