Georgia Election Analysis III

I did some analysis of the Presidential election in the state of Georgia in two installments. Today I’ll look at Fulton county where I live as well as the specific precincts in Alpharetta, Milton and South Atlanta where our team worked as volunteers for the Obama campaign.

Year D R Turnout %Turnout
2000 57.53% 39.68% 264,276 65.06%
2004 59.35% 39.99% 330,791 74.15%
2008 67.07% 32.09% 405,531 73.11%

There was a 15.3% swing towards Obama in Fulton county compared to Kerry in 2004. The turnout, based on the number of registered voters, stayed constant from 2004 to 2008. However, if we use the voting age population estimates (VAP) for Fulton county, we get the following turnout rates:

Year Turnout based on VAP
2000 42.30%
2004 48.00%
2008 52.72%

This shows a bigger turnout in 2008. One reason for the discrepancy between the turnout in the first table and this one is that the Obama campaign focused a lot on voter registration this year and thus got more people registered.

Also, please note that the VAP turnout estimate is probably lower than the real turnout which should be calculated as a proportion of the voting eligible population.

Out of the 3,924,440 votes cast for President in Georgia this year, 2,084,179 (or 53.11%) were cast during advance/early voting or by absentee mail-in. In Fulton county, 184,240 votes (45.42%) were cast early, absentee or provisional out of a total of 405,628 votes cast. This is very unusual for Georgia and Fulton as can be seen in the table below for Fulton county. (This shows the numbers for the 2004 general election for President, the 2006 election for Governor, the 2008 Presidential Primary on Super Tuesday and the 2008 general Presidential election.)

Year Absentee/Early/Provisional votes
2004 Prez 12.68%
2006 Gov 11.76%
2008 Prez Primary 7.65%
2008 Prez 45.42%

In the 2008 general election, the Obama campaign tried to get everyone to vote early. The effect of this can be seen in the early voting numbers in Fulton county where Obama had a 49.5% lead in early voting compared to a 22.9% lead on polling day itself.

The large numbers of early voters have complicated my precinct level analysis. The precinct level data does not include early or absentee votes which are listed separately as one per county. Still let’s see what we can conclude for our precincts.

For the Get Out the Vote (GOTV) campaign in October and then the first four days of November, we were based in the South Atlanta precincts, 12E1, 12J and 12T, which are just north of Atlanta’s Hartsfield-Jackson Airport. These precincts are heavily Democratic (more than 90%) but there are a lot of sporadic voters there. So the main task was to get the vote out.

Election 12E1 12J 12T
Voted %Turnout Voted %Turnout Voted %Turnout
2004 518 49.5% 1024 59.7% 566 62.3%
2006 224 22.1% 473 29.2% 581 35.8%
2008 Primary 280 27.5% 564 34.6% 609 39.1%
2008 747 55.1% 1,360 67.3% 1,232 68.0%

Thus, turnout as a percentage of registered voters increased from 57.4% to 64.3% while actual number of people who voted in these three precincts increased from 2,108 to 3,339, a 58% increase in the number of voters. Compare this 58% increase to a 22.6% increase over the whole of Fulton county and a 19.6% increase over all of Georgia. I am actually pleased at these numbers and I am sure all of our Obama team would be too.

Let’s now look at the precincts in Alpharetta and Milton where we made the most calls, starting in August, for identifying voters, persuading them and then getting them out to vote. The table below shows the Republican advantage over Democrats. For the 2008 Presidential Primary, I added the votes for all the Republican Presidential candidates together and did the same for the Democrats. Let’s take the example of ML01A in the 2008 general election, McCain got 73.5% and Obama got 25.3%, which means a Republican advantage of 73.5-25.3=48.2%.

Precinct 2006 2008 Primary 2008
AP07B +66.0% +33.2% +40.9%
ML01A +64.0% +34.4% +48.2%
ML02A +67.4% +40.8% +62.6%
ML02B +68.0% +45.2% +57.5%
ML03 +66.9% +45.7% +59.5%
ML04 +65.2% +38.5% +56.5%
ML05 +48.9% +14.9% +37.7%
ML06 +29.9% -11.4% +8.2%
ML07 +65.1% +41.8% +52.0%

I couldn’t find the data for the 2004 Presidential election for the Milton (MLxx) precincts but AP07B had a 42.9% Republican advantage in 2004.

Before I opine on those results, let’s look at the turnout (as a percentage of registered voters) in these precincts:

Precinct 2008 Turnout
AP07B 44.6%
ML01A 44.6%
ML02A 55.3%
ML02B 55.3%
ML03 52.5%
ML04 44.7%
ML05 44.2%
ML06 41.9%
ML07 54.2%

We already know that about half of Georgia and Fulton county voters voted early. The early voters are not listed in the last two tables for the Alpharetta/Milton precincts because they are not reported that way by Fulton county. Looking at the election day turnout for Milton, it seems that fewer than half of the voters voted early.

Because we only have the vote breakdown by party for those who voted in person at their polling location on November 4 and a significant number of voters had voted early in 2008, we cannot really say what the margin between McCain and Obama was in Alpharetta and Milton. There is no reason to believe that the early voters had the same partisan distribution as those on election day. In fact, there is reason to believe that the early voters were more likely to be Obama supporters. The Obama campaign had been working hard asking people to vote early. Also, in the whole of Fulton county, Obama had a 49.5% advantage in early voting compared to a 22.9% advantage on polling day. Thus, the very large Republican margin in the Milton precincts in 2008 is most likely not correct. I can say with certainty that McCain won all those precincts except ML06 but probably with a somewhat smaller margin. Even then, Republicans had a huge advantage here in North Fulton. Unfortunately, we don’t have any way of finding out whether our team was able to reduce that Republican advantage or not.

Next: A look at the Senate runoff between Jim Martin and Saxby Chambliss.

By Zack

Dad, gadget guy, bookworm, political animal, global nomad, cyclist, hiker, tennis player, photographer

7 comments

  1. what do D and R stand for in the first table? Is it the turn out among the registered Democrats and Republicans, or the percent of people who voted for the candidate of the designated party? In the latter case, the total is more than 100%!

  2. It’s the percentage of votes that the Democratic or Republican candidate got. And you are right, it summed up to more than 100%. This is what happens when I use the candidate votes from the certified numbers and total votes from another place on the Secretary of State’s website which does not show official, certified numbers. I have fixed it now.

  3. Zach,
    Did you do any analysis by precinct in Atlanta Council District 1 to see what the impact of the Obama vols was on U. S. Senate runoff turnout as compared to the”average” turnout for the rest of the state?
    Riki

Comments are closed.