Prometheus 6   

1) Those blacks born in free states before 1865 were themselves typically no more than two generations removed from slavery, and very often less. The study therefore does not, and cannot, show that slavery didn't inflict lasting intergenerational harms and costs. It shows, in effect, that being the grandchild of a slave is equivalent, in some respects, to being the great- or the great-great grandchild of a slave, that certain effects level out. It does not show that being the descendant of a slave is equivalent to not being the descendant of a slave. This might mean that a reparations formula shouldn't differentiate between those blacks descended from slaves freed before the war and those slaves freed during and after the war. It doesn't show anything at all about the appropriateness of reparations in the first place.

2) Moreover, Sacerdote freely acknowledges that there is no 1920 convergence between blacks and whites on his measures. This could be, as he thinks, because of Jim Crow barriers. (Take it away, Professor Bernstein.) If so, that might mean that reparations need to be understood as reparations for slavery plus Jim Crow, not just for slavery; but what of that? Alternatively, the lack of convergence between blacks and whites could mean that slavery by itself inflicted harm that had enduring intergenerational effects on education, literacy, and occupation-- again, with a convergence over time between those whose ancestors were freed earlier and those whose ancestors were freed later, but with no convergence between descendants and non-descendants of slaves. I certainly favor the Jim Crow explanation. But neither explanation forms any part of an argument against reparations.

3) The items on which convergence occurs are certainly important ones. But Sacerdote has no good measure of wealth available. (He does have home ownership, but, as he explains, it's effectively worthless as a wealth measure, since home ownership in 1920 was heavily influenced by employment in agriculture.) And surely we would need this. Even if after two generations the descendants of slaves were as educated, as literate, and in the same professions as the non-descendants of slaves (which, remember, is not the finding of the paper), that would not mean that there was nothing to compensate. Wouldn't we still expect there to be a significant wealth difference between those whose families could not have accumulated more than two generations of capital and those whose families could? As is well-known, there is a very large wealth difference even today between blacks and whites today even after controlling for education and current income. If white wealth was accumulated with stolen black labor (note: I have no ability to judge the debate among economic historians about whether slavery was in fact economically beneficial for whites), then a convergence of literacy rates doesn't seem especially interesting or relevant to the reparations debate. It's not illiteracy that reparations are being demanded for.

I still see the Sacerdote paper as weakening the case for reparations. If we look at literacy, recently-freed slaves catch up to the earlier-freed slaves within about a generation.

This is consistent with two possible explanations:

1) each group of former slaves recovered from its slavery heritage, once freed (albeit still suffered under a more general burden of racism, which may be huge)

2) The earlier freed slaves made some progress, the later freed slaves then matched that same progress, both suffered about the same lasting amount from slavery.

Jacob points out, correctly, that 2) is possible. But 2) suggests that having a longer slave heritage, across previous generations, is not a bad thing for a later generation. This is not intuitive. If you believe "a slave heritage has persistent bad effects" you probably also should believe "having a longer slave heritage has especially bad persistent effects." But the data don't show the latter. And intuitively, think about it: there ought to be some long-term progress as you move away from a (relatively) short slave heritage.

So either a slavery heritage doesn't add much to the racism explanation (and other alternatives), or the persistent intergenerational costs of slavery kick in at a constant level, once your ancestors have seen a threshold level of enslavement in the past.

But in the meantime I am still more skeptical about the reparations case, and I would not have expected to find the Sacerdote result.