In 1865, toward the end of the Civil War, Union Army General William Tecumseh Sherman promised slaves that they’d receive 40 acres and a mule. Land was even set aside, but the promise was recanted by President Andrew Johnson. Ever since, the issue of reparations has come up many times, often fiercely debated. Although most Americans generally don’t support reparations, according to University of Connecticut researcher Thomas Craemer, it matters greatly how the question is worded, who would get reparations and in what form. For example, the idea of reparations paid in educational benefits are more popular than others, Craemer says.

On the other hand, one of the cases often made against reparations is that it'd be impractically difficult to calculate how to fairly take and give so many years after the fact. But in a new paper, published in the journal Social Science Quarterly, Craemer makes the case that there are other examples of historical reparations paid many decades later after “damages” were incurred. He also has come up with what he says is the most economically sound estimate to date of what reparations could cost: between $5.9 trillion and $14.2 trillion.

Craemer came up with those figures by tabulating how many hours all slaves—men, women and children—worked in the United States from when the country was officially established in 1776 until 1865, when slavery was officially abolished. He multiplied the amount of time they worked by average wage prices at the time, and then a compounding interest rate of 3 percent per year (more than making up for inflation). There is a range because the amount of time worked isn’t a hard figure.

Previous estimates of reparations have ranged from around $36 billion to $10 trillion (in 2009 dollars), Craemer says. Those calculations mostly looked at wealth created by slaves as opposed to services provided, resulting in underestimates. Craemer believes that “the economic assumptions underlying [his method] are more sound” than those used in previous papers.

The paper also illustrates several historical examples in which reparations were paid, many decades later, despite being initially unpopular—showing that repayment of age-old claims is not without precedent.

One example is the case of “French spoliation claims.” During “a seven-year period from 1793 to 1800…France attacked American ships in retaliation for the United States’s neutral stance in the war between France and Britain,” Craemer writes. Though France refused to pay for the damage they caused to people’s property during the war, hundreds of Americans sued their own federal government anyway, arguing that the hostilities, and damages, were ultimately the responsibility of the United States. They claimed that France was retaliating for the United States’s failure to pay back the European country for the support it gave to the U.S. during the Revolutionary War. By 1910, after decades of debate, the United States eventually agreed to pay back $1 million in claims ($38 million in today’s dollars) to hundreds of its own citizens.

Robert Westley, a professor at Tulane University who wasn’t involved in the paper, says that this and other examples can be used to refute arguments that slavery reparations would necessarily be too difficult to figure out. The French spoliation claims and others “were made and demanded over many generations,” he says. “Somehow problems of proof were not insurmountable in those cases, and shouldn't be in the case of the United States with slavery.”

Craemer also makes a case in the paper that the stimulus packages paid for by the government after the 2008 financial collapse were unpopular, but thought to be necessary at the time. If reparations were perceived as necessary for righting a past wrong, as some argue, then perhaps the public would back them. Craemer’s assumption, of course, is that the government would pay for reparations, since it allowed slavery to exist, and slaveholders were, after all, acting in accordance with the law at the time.

Part of Craemer’s interest in the issue of reparations derives from his upbringing in Germany, a country that agreed to pay reparations to Jewish victims of the Nazis. As of 2012, Germany had paid $89 billion in such compensation.

“I grew up with this guilt complex about the Holocaust, and I remember kind of feeling good that my country paid reparations,” Craemer says. When the policy was first put forward, there was a lot of resistance, he says. He sees some parallels with slavery reparations in the U.S., and thinks that such a move has the potential to help race relations.

“Reparations will never bring one life back, and it’s totally inadequate to the terror of the [past], but having a meaningful symbol of reparations is a good thing, not just for recipients but for the people who provide it,” he says.

Both Craemer and Westley point out that the number is just an estimate and not set in stone.

“To me the important point is not the number, but the dialogue” it starts, Westley says, “a negotiation about [righting] this historical wrong.”