Several times in the last few weeks, I’ve seen claims made about “the south.” Things like polling data that reported regional cross-tabs, political analysts claiming something distinctive about the region and the 2010 election, or just a friend conjecturing about some regional cultural phenomenon. And, as usual, none of the sources defined what they meant by “the south.” Which makes it pretty hard to assess the claim. Of course, part of the problem is that’s there’s not really any utterly stable definition; it depends what you are talking about it.
This is more extreme than it first appears. One would think it’s not all that controversial to include Georgia in your definition of the south. But how often do you hear, “well, Atlanta’s not really in the south” at some point when the conversation regards about urban modernization? Or “yeah, but North Carolina’s not really in the south,” when the issue is the relative proportion of prestigious universities in the south. And while such comments are usually the downstream effects of uncomfortable counter-examples to some “the south is culturally backwards” thesis, they can run the opposite direction. “The South is just so much more beautiful than the North. And the people are so much nicer.” Um, yeah, but have you been to Houston? “Well, Texas isn’t the real south.”
But aside from all that pap, the problem remains. You’ve got a social science question to address. You need to define “the south.” How do you do it. From the political science research I’ve encountered, there are basically five options:
1) Go with the Confederacy. States that left the union in 1860 or 1861 (11 States): Virginia, North Carolina, South Carolina, Georgia, Florida, Alabama, Mississippi, Louisiana, Texas, Tennessee, Arkansas
2) Go with the slave states. State land with legal slavery as of March 4, 1861 (16 States): Confederacy + Missouri, Kentucky, West Virginia, Maryland, and Delaware.
3) Go with the segregation states. States with required segregation in public schools at the time of the Brown decision in 1954 (17 states): Slave states + Oklahoma.
4) Go with the VRA states, minus the obvious exceptions. States with at least some territory covered by the “preclearance” requriements of Section 5 of the Voting Rights Act of 1965, minus states that cannot be in the south by any reasonable definition (10 States): Texas, Louisiana, Mississippi, Alabama, Georgia, Florida, North Carolina, South Carolina, Virginia, Arizona(?). This leaves out the VRA states of Alaska, California, South Dakota, Michigan, New Hampshire, New York.
5) Just dump Texas. Either Confederacy or Slave States, minus the Lone Star (10 or 15 states).
It’s not clear to me why people choose one definition over another, or whether one is, ceteris parabis, better than the rest. Certainly there are circumstances that dictate ones choice — you can’t study how section 5 of the VRA was implemented in states where it wasn’t in effect, so any implication use of southern in regard to the VRA necessarily excludes Arkansas, Tennessee, Kentucky, etc. On the other hand, I’m not convinced that people don’t manipulate the definition in order to cook their data; the “south is backward” thesis is empirically aided greatly if you can squeeze the appalachia poverty of West Virginia, Kentucky, and eastern Tennessee into the south and out of the midwest. Similarly, excluding Texas can solve a lot of messiness if you want to really drive home the economic depression of the region. And I don’t mean to highlight the “backward south” thesis. One great thing about the maleability of the definition is that you can take any social science finding that might plausibly have a southern aspect to it, and then choose from a menu of defensible “souths,” working backwards to pick the one that best shows off your data.
Of course, there are a number of sociology options, as well, ranging from things like historical patterns of public opinon about different policies to subscriptions to bassmasster magainze or places with less than a 5% chance of a white Christmas. Or a Soviet professor’s vision of the future breakup of the union. Just go browse strange maps for a while. Something even weirder will pop up.
And then there’s my personal favorite, the soft drink map. States with dominating usage of “Coke” as a generic for soft drink (8 states): Texas, Arkansas, Louisiana, Mississippi, Alabama, Georgia, South Carolina, Tennessee.
