Comparing Early Voting Turnout 2008 vs. 2012

Comparing Turnout 2008 vs. 2012

We’re getting frequent requests for research on who is voting early in 2012 vs. those who voted early in 2008. My friend, colleague and data guru Mike Seigle was kind enough to research some comparisons for our Peach Pundit community.

A perfect comparison can’t be made because the laws permitting early voting have changed from 2008. Early voting was held for 45 days in 2008, but in 2012 early voting will only occur for 21 days.

However, percentages of turnout among various demographic groups can be compared.

Below you’ll see the turnout comparisons broken out by racial groups.

The Primary Finding: Essentially there has been a net +3.4% increase in the net number of white voters and nearly a 4% net decrease among black voters from 2008 numbers. Hispanic and Asian early voters in 2012 have voted in similar patterns to 2008.

Among white voters — comparing 2008 vs. 2012 early turnout as of 20 days before Election Day:
With 20 days left in the 2008election, 60.4% of voters were white.
With 20 days left in the 2012election, 63% of voters are white.

Among black voters — comparing 2008 vs. 2012 early turnout as of 20 days before Election Day:
With 20 days left in the 2008election, 34.9% of voters were black.
With 20 days left in the 2012election, 31.1% of voters are black.

Among Asian and Hispanic voters:
The early turnout percentages are essentially unchanged from 2008 to 2012 among both Asian and Hispanic voters (about half a percentage point each).


  1. Very misleading. Let’s go to the beginning of the week before advance voting started. With 23 days left in 2008, 479k ballots had already been cast and 37% of them were cast by black voters. This year, 59k had been cast and only 21% of them were cast by black voters. Of course in 2008 advance voting had been going on for ~ 22 days at that point, whereas this year only vote by mail had been going on.

    So let’s look first at just the vote by mail numbers, 2008 and 2012 side by side for the pre-advance voting period (day 23 or earlier).
    Total: 68k – 59k
    Black: 14k – 12k
    White: 51k – 42k.

    So as you can see for absentee voters only there were roughly 9,000 fewer total votes in 2012 than there were in 2008 and the difference comes pretty much entirely from white voters, though black voters are also down by a similar amount. The black % of mail voters who voted before advance voted started was 21.4% – in 2012 mail voters at this point were 21.5% black. So there is no difference.

    So we know that black voters are more likely than white voters to take advantage of in person early voting at least in 2008 when the final early vote turnout was > 34% black but the election day turnout was about 25% black and it averaged to 30% black. But what about in 2010? Minus the enthusiasm of Obama, black voters got off to a slow start – with 19 days left until the election they made up only 18% of early voters. But what happened in 2010? They surged thanks in large part to Saturday voting when they made up 55%+ of participants and the numbers continued to rise as election day approached. In 2010 black voters ended the advance voting period with 27% of all votes but a funny thing happened – they actually made up about 29% of voters on election day and pushed the overall total to above 28%. So unlike in 2008, 2010 was a rather even year with black voters (again turning out in record numbers compared to prior similar election cycles) splitting their turnout pretty evenly between advanced and early.

    So, taking that into account, we know that both black and white voters have turned out at much lower levels compared to 2008 because there haven’t been as many days of advance voting. But what about comparing to 2010? Well interesting that you would ask. What if I told you that compared to 2010, more than twice as many black voters have already voted as of day 19 (100k to 47k) but about the same number of whites have voted (202k to 204k, but I am missing race data on some newly registered people).

    So to sum it up a week ago before early voting started, blacks made up only 21% of voters. As of yesterday they make up > 31%. A huge gain for just one week of voting. Whites went from 73% to 63%, a similar drop. And compared to 2010, about the same number of whites have voted early but twice as many blacks have voted.

    And tomorrow, Saturday, DeKalb County and others will be open for Saturday early voting (next Saturday is mandatory statewide but counties like DeKalb can choose to be open all three Saturdays). I’ve been projecting the future totals based on what’s already come in each day and my projection for today’s numbers was off by about 1,000 total, 1,000 black, and 300 white. My projection says that after Saturday voting is booked, 463k voters will have voted, 144k black, 282k white for a black % of 32 and a white % of 62. That means we’ll start the day 15 week only about 225k lower total than we did in 2008 which is an impressive start for this year’s reduced timeframe. And we’ll be about 75% higher than the 2010 comparable period meaning all things considered, we’re on pace for a very similar year to 2008. And even if black voters finish the early period at 32% black instead of 35%, 2010 shows that turnout on election day can be quite good as well.

      • I’ll try to be concise. In 2008 early voting was a very big phenom in the black community and even though blacks only made up 29% of all registered voters, they made up 35% of early voters. On election day they only made up about 25% of all voters and it averaged out to 30%. In other words they voted at slightly higher % than the overall registered voter population.

        In 2010, their early vote % was 27 and their election day was 29 for an average of 28 – again about the same as their overall makeup of the electorate but slightly lower.

        Mark argues that this year because blacks currently make up 31% of all voters instead of 34% that black enthusiasm is somehow down. But you can’t compare 2008 to 2012 at this point because at this point (ie the 3rd friday before election day) there had been 28 days of early voting in 2008 vs just 4 now and that is the primary way that black advance voters vote – not by absentee ballot.

        And even if black voters end the early voting period at only 31% of all voters, that doesn’t spell doom for the general election because in 2010, the second year of big early voting, blacks actually voted in heavier numbers on election day.

        • Chris, I didn’t see Mark “argue that black enthusiasm [was] somehow down.” I saw him present some statistics. You however, jumped to a conclusion in a manner that would have made Felix Baumgartner proud.

          • Dave Bearse says:

            Mark was uncharacteristically sloppy with the statement: “The Primary Finding: Essentially there has been a net +3.4% increase in the net number of white voters and nearly a 4% net decrease among black voters from 2008 numbers” directly states black enthusiam is down. After that Mark’s following supporting statements clearly imply that black enthusiam is down relative to whites.

  2. The Democratic vote is simply less intense in its support of Obama. It’s a pretty common finding among polling firms (…and soccer moms nationwide).

    This is from Politico’s poll released this Monday morning (Politico is Democrat leaning in its angle of news, the poll was conducted by the bipartisan pollster Battleground):

    “Two weeks from Election Day, the GOP nominee also continues to maintain a potentially pivotal advantage in intensity among his supporters. Seventy-two percent of those who support Obama say they are “extremely likely” to vote, compared to 80 percent who back Romney. Among this group, Romney leads Obama by 7 points, 52 to 45 percent.”

    Don’t blame the messenger. Blame the debater.

    • The Last Democrat in Georgia says:

      Prediction: Willard “Mitt” Romney wins the General Election by a tally of 53% to Obama’s 47%, carrying 32 states and 289 electoral votes and becomes the 45th President of the United States.

      • Calypso says:

        While I agree with your desire regarding the outcome of the election, I hope you wrote this prediction in pencil and your eraser is in a nearby desk drawer.

        • The Last Democrat in Georgia says:

          You’re right…I should have typed that Romney will win the General Election by a tally of AT LEAST 53-47 while carrying AT LEAST 32 states and winning AT LEAST 289 electoral votes enroute to becoming the 45th President of these United States of America as when all is said and done, Romney is likely to earn over 300 electoral votes.

          Though the winning margin does not matter just as long as we can get that Chairman Mao clone out of the White House so that this nation can finally proceed with some serious economic growth for once and for all without being stuck-up by hordes of moocher advocates at every turn.

    • In Georgia, where race is a pretty direct proxy for partisan support at this point, blacks are 32% of early voters. Yes, that’s down slightly from 2008 when the rules were entirely different, so it’s very hard to make a direct comparison. But they only make up 29% of all voters.

      So far, we have over 400,000 actual datapoints about what people have done. I’ll take that over polls of what people say they will do – notoriously bad at predicting actual behavior on this stuff.

      • Doug Deal says:

        It might be a faulty assumption that blacks will again vote as high as 95% for Obama and the Democrats. It might be closer to 90%, which is closer to the normal result. Elections are all about margin and 95% is a 90% margin, while 90% is an 80% margin (10% swing). This means that even with 32% turnout, the President would be losing 3.2%. Nationwide, this would result in about a 1.2% difference in margin. To put this in perspective, Obama won by about 7%.

        • Great points – though if you look at the precinct results for South DeKalb County, I think the % in 2008 was closer to 99% here, honestly.

          My overall point is that in Georgia, where race gives us a good look at what’s going on, black enthusiasm is down at most 2% compared to 2008 – and that’s with a direct comparison. But if the same thing is going on in other states (and some polls say it is and some polls paint a rosier picture for Republicans) if Obama had a bigger % to start with, he’ll probably be fine.

            • Well right now Democrats don’t really get any independent or conservative white voters in this state and Republicans don’t really get any independent or conservative black voters, even if absent race they might re-align.

              I’m not sure who a re-alignment would be better for – although I’d point out that if you’re getting the supermajority of the bigger group, be careful what you wish for.

              But agreed.

  3. KudzuDave says:

    “(Politico is Democrat leaning in its angle of news, the poll was conducted by the bipartisan pollster Battleground)” I have never heard Politico described as a “Democratic leaning”. Wait I forgot, if you are not a shill news organization for the GOP, then you must be “Democratic leaning”?

Comments are closed.