In this particular case, cotton production in 2008 is irrelevant, thanks to slavery being ended and agriculture largely mechanized — it has no relevance to population distributions today. But cotton production in 1860 does, because poverty and hostility made it unlikely that southern ex-slaves and their families would redistribute themselves around the South. The ones that migrated, and there were many, left the South altogether, especially in the early 20th century as industry opened up jobs in the urban north, and destroyed jobs in the agricultural south.

It is this chain of reasoning that is being tested by the graph above, and while not perfect, the correlation is pretty good.

So where, in this analysis, is the racism, or the non-objectivity? Is it wrong to even ask questions about a group of people who identify themselves as belonging to a particular race? Is it wrong to look at a map of electoral results and wonder why?

We live in a country and culture (like everyone in the world) with codes, signifiers, and symbols. Of course, these are open to interpretation; but I think a reasonable interpretation one can make from this visualization is a nice, euphemistic “poor, black people voted for Barack Obama.” I think this visualization is especially loaded, considering the sets of data are pulled from two different time periods not comparing the same thing. Cotton picking in 1860 vs. cotton picking in 2008? Helpful! But this one? Not so helpful.

Is it racist? Very possible that some would find it so and some would not. But don’t make the argument that data is data and it’s objective. That is simply unsophisticated thinking.

Actually, it’s “if you can get your county to have grown more cotton 150 years ago…”

But if you could hop in your time machine and do that, it might work, since it would have changed the geographical distribution of slave labor then, and thus black voters now.