Obama and the African-American vote in Georgia.

Using President Bush's margin of victory in Georgia in 04 (roughly 550,000 votes) as a baseline, and if we control for the state's population growth over the last four years (9.2%), we can project that Obama will need to close a 600,000 vote gap in November. (This projection assumes that the demographic and political identification breakdown of the increased population is similar to the demographics of the voting population in 04.)

Much has been made of the Obama campaign's voter registration drive, specifically how they are targeting African-Americans. Can the Obama campaign close the gap in Georgia solely through registering and motivating AA voters to come out in November?

The number of AA voters in the 2004 General Election was 820,101. Of that total 88%, or 721,689 AA voters, voted for Kerry and 12%, or 98,412, voted for Bush. In the 2008 Primary, 96.6% of AA primary voters voted Democratic and I would expect the percentage of AA voters voting for Obama would be closer to this 97% than the 88% that voted for Kerry. For these projections I am assuming that 95% of AA voters in November will vote for Obama. If we were to retroactively apply this 95% figure to the 2004 General, Obama running as the Democratic nominee would've received 779,095 votes or 57,407 more votes than Kerry. Additionally, these votes are migrating from the Republican candidate to Obama so they count double when applying them to the 600,000 deficit. So as a result of our prediction that Obama will receive 95% of the AA vote versus Kerry's 88%, the deficit reduces to 485,186. Based on these calculations, 1,264,282 AA voters will need to cast votes for him in November (95% of AA vote in 2004 + 485,186 additional votes), for Obama to close the gap on the strength of AA voters alone.

The most recent voter registration numbers in Georgia (6/1/08) indicate that there are 1,314,755 registered AA voters. This number is 50,473 registered voters more than the Obama campaign's target number, so it closes the rest of the deficit, right? Well, no. The chance that all AA registered voters will actually show up at the polls is slim, but what percentage will? In 2004, 70% of AA registered voters turned out. If 70% turn out in November, the Obama campaign will only reach their target number (1,264,282) if 1,901,176 AAs are registered. Is it possible to register 600,000 new AA voters by November? Recent voter registration trends indicate no. In March, 41,084 new AA voters were registered. In April, that number dipped to 18,574, and in May, it dropped further to 12,840. While these numbers are impressive they fall well short of getting the total number of AA registered active voters to 1.9 million.

A 70% AA active registered voter turnout seems a bit conservative to me. Certainly the Obama candidacy will boost AA turnout to at least 80%, which was the rate at which non-AA active registered voters turned out in 04. If 80% of AA active registered voters turn out to the polls in November, the Obama campaign would need 1,663,529 AAs registered to vote in November to reach his target number of votes (1,264,282). Is 1.66 million registered AAs plausible? There are five months until the election. If the Obama campaign registers 30,000 new AA voters per month (less than March, more than April and May respectively), that would total 150,000 additional AA registered voters. These 150,000 additional voters would bring the AA registered voter total to 1,464,755, still 200,000 voters short.

How many registered voters would the Obama campaign need if 90% of AA registered voters turned out in November - a longshot scenario for sure? He would need 1,478,692 registered voters, or about 13,000 more than he would have if 30,000 new voters were registered each month between now and November.

Keep in mind, this analysis is solely focused on the AA vote. It doesn't consider the possibility of Evangelicals sitting out (as compared to the number that voted for Bush), the Bob Barr/Ron Paul effect (Georgia had the highest percentage of Libertarian voters in 2004 when compared to the other states considered "in play" this election cycle, Barr represented Georgia's 7th Congressional District in Congress for 8 years, and Ron Paul took 3% of the primary vote.), or general disillusionment with the Republican brand. By the same token, this analysis also doesn't account for any Appalachia effect benefitting McCain, or any of the 10 commenters on hillaryis44.com that may live in Georgia.

Realistically turnout among AA active registered voters will probably be near 80%. If the Obama campaign were able to register 30,000 voters per month, then he would still need to close a gap somewhere between 150,000 and 200,000 voters. Keep an eye out for the June registration numbers in Georgia, usually reported here a few days after the month ends. As the summer numbers come in we should have a better idea of whether the Obama voter registration drive is a force to be reckoned with or merely another talking point. Perhaps the aggregate of the above-listed "effects" would close anything within that range. Either way, AA voters have an opportunity to swing a state significantly closer toward Barack Obama, and, possibly, with truly historic voter registration and turnout, could push him over the top.

The Protocol does not discriminate based on race, color or national origin, pursuant to Title VI of the Civil Rights Act of 1964 (42 U.S.C. 2000d.) and Bylaw XII of Protocol Inception Act (III.ii PIA 2008).