Other approaches lead to much bigger sums. One calculates the difference between what slaves were given by way of maintenance, and what free workers were paid.
An estimate puts that at roughly $4trn in today's money (19% of GDP),
once you account for the financial returns that could have been made if the money had been paid on time.
But some argue that slaves held down the wages of free workers, meaning that the true value of slaves' lost wages is higher.
Mr Darity and Ms Mullen say that the difference in mean net wealth between white and black households ($795,000 in 2016)
is the "most robust indicator of the cumulative economic effects of white supremacy". That points to reparations of nearly $8trn, or 37% of GDP.
(The authors suggest that this should be partly financed by printing money, something that will make most wonks queasy.)
Another area of disagreement concerns the form that reparations should take.
Mr Darity and Ms Mullen argue that for "both symbolic and substantive reasons, an effective programme of restitution must include direct payments".
But cash transfers may do less to reduce inequality than their supporters hope.
Research on inheritances, for instance, suggests that most heirs consume their windfall within a few years (purchases of cars are especially popular).
A sizeable part of the income gap between black and white Americans reflects differences in education levels; big one-off payments alone cannot alter that.
And research by Mr Darity and Dania Francis of the University of Massachusetts Boston finds that
reparation payments could increase non-black incomes relative to black ones, if the spending thus facilitated flowed largely to non-black-owned firms.
"Our paper points to the need to improve the infrastructure of black-owned businesses and banking
so that dollars from reparations can flow into black communities," says Ms Francis.
To this end, some economists argue that reparations should fund training and education programmes, or subsidise business lending.
Others point to "baby bonds", which would be targeted at poor children and help them pay for university or to start up a business.
Naomi Zewde of the City University of New York finds that baby bonds could substantially reduce racial wealth gaps among young people.
Reparation payments could be spent in other ways.
Money paid out to Japanese exinternees has been used to fund academic chairs and historical archives.
Reparations from Germany pay for food and medicine for Holocaust survivors.
But before America can widen support for reparations, it will have to debate what works.