Nate Silver has recently pointed out that a surge in the black vote could radically change this election, well outside what his or anyone else's models would predict. Today he further points out there is evidence the surge is actually occuring. This led me to ask "what percentage increase in the black vote would actually result in flipping Louisiana, Alabama, Mississippi, Georgia and South Carolina?"
Calculations below the fold.
Take a look at the chart I've constructed below (most data available from FEC and Census Buruea, the rest from the Joint Center for Political and Economic Studies) comparing total state population, percentage of that population that is black, and then calculating the total number of black votes based on the percentage of the black population that voted.
State State Pop. % Black 2004 Black Vote/%Voting
LA 4,287,768 31.7 788,349/58
AL 4,599,030 26.3 653,154/54
MS 2,910,540 37.1 539,905/50*
GA 9,963,941 29.9 1,459,816/49
SC 4,321,248 29.0 877,212/50
*Could not find data on MS, chose 50% as a conservative estimate based on other Southern states voting percentages.
Based on this data we can start playing with SCENARIOS in which we increase the percentage of the black vote to see what impact it would have on the voting number, ceterus paribus (not changing anything else, like accounting for Katrina, increases in Democratic Voter registrations, or declines in GOP identification). This chart compares number of votes the Democratic Party lost each state by to my estimates of the increase in votes represented by a 10%, 20%,30% and even 40%* increases in the black vote.
*I quit calculating at a win, and I am in no way prognisticating a 90% voter turnout, just investigating it.
State VotesLostBy +10% +20% +30% +40%
LA 281,870 135,921 271,844 407,766
AL 482,461 120,954 241,908 362,862 483,817
MS 214,894 107,981 215,962
GA 548,105 327,714 595,844
SC 276,725 125,308 250,624 375,940
This "back of the envelope math" demonstrates that Mississippi, Georgia and even South Carolina are all very close to being through 20-25% increases in voting rates in the black population alone. And that percentage increase brings them well within the projected voter turnouts of other US populations (65%-75% depending on who I was reading this morning). Further, Louisiana would probably be in this category but for the changes in population brought about by Katrina. (Alabama is, well, still mighty might unlikely.) Even if we there aren't any "Southern Surprises" in the offing this election, this points to inroads that we will be making and can capitalize on in future elections. We have the bones of an infrastructure, now we can put some flesh on it.
This is by no means a scientific method; it is more akin to reading chicken entrails with arithmetic. But, it does give us some benchmarks, some metrics, with which to measure voter turnout in the South, to assess if we are witnessing a wave or even a tsunami. But all useless labels aside, even current observations and anecdotal evidence indicate that one of the sleeping giants of the US electorate has finally awakened. And good morning to him!
Thanks for reading, and if you have any professional or more precise knowledge of election statistics, please chime in the comments on what I did wrong here!
-SPX