The excellent Strange Maps produced a fascinating overlay of Obama votes in this month’s presidential election compared with cotton production in 1860. Not surprisingly there is a huge correlation: cotton in 1860 => slaves => black population now => Obama support.

The one thing that jumped out at me, after admiring the clarity of the general point, was that anomaly in southern Tennessee, where a concentration of cotton farms marked on the 1860 map translates into no correlation at all with the Obama votes in 2008.

The cotton concentration shown is mainly in Lawrence County (briefly the home of Davy Crockett, now the home of presidential hopeful Fred Thompson), with some spillover into Giles County; yet in the election, Lawrence voted for McCain over Obama by 66% to 32%, and Giles by 59% to 39%. (Lawrence has a 1.47% black population; Giles 11.80%.)

The commenters on the Strange Maps post put forward two theories to account for the lack of Obama votes in Lawrence and Giles Counties. First, that there were never many blacks there in the first place, because local agricultural practices were different: "Greg" says that "much of the cotton production in northern Alabama and southern Tennessee during the antebellum era was done on small family farms rather than large plantations. These poor white farmers couldn’t afford slaves, and resented having to compete with plantations that relied on slave labor." Sam Persons Parkes adds that "Lawrence County has always been about 99% white."

The second is that the black population has left. "Goateebird" suggests that it has to do with the Ku Klux Klan, which was actually founded in Pulaski, the seat of Giles County. "DG" decries the KKK explanation, and says instead that the black population of southern Tennessee fled en mass to Nashville (a visibly blue part of the map with no 1860 cotton production) at an early stage of the Civil War. I must say that myself I wondered if this might turn out to be a parallel to the ethnic cleansing phenomenon of sundown towns.

But after a bit more research, I am inclining towards the first theory. The 1860 map comes from Sam Bowers Hilliard’s Atlas of Antebellum Southern Agriculture, and can be found online here, an on-line publication of a book about the Savannah river by the National Park Service. That same page also has maps showing cotton production in 1820 and 1850, and most importantly slave population in 1860. Tennessee is represented as follows:

Tennessee cotton production, 1820

Tennessee cotton production, 1850

Tennessee cotton production, 1860

Tennessee enslaved population, 1860

That final map is pretty conclusive: rather than being kicked out by the KKK, or fleeing to Union-controlled territory, Lawrence County’s black cotton plantation slaves simply weren’t there in the first place. I have to say I find the sudden surge of cotton production in that area shown in the 1860 map a bit suspicious, and wonder just exactly how accurate it is to locate it in Lawrence County; I’m inclined to suspect that the real growth was a little further south and east around Decatur, Alabama.