| Hotline published the memo here. I put it into a spreadsheet for easy viewing here. Obviously these numbers change daily and by the time you read this it more accurately reflects the state of the race as of a couple days ago so please analyze accordingly.
As expected it varies by state and they only have the additional useful information available for a few states, but on first glance things look very promising for Barack.
In North Carolina the percentages of Dem/Rep/Ind by registration that have early voted are 56%/27%/17% and the raw total is 520k Dem vs. 255k Rep for about a difference of 265k, we're doubling them on raw turnout. That's great, but what's really worth getting excited about are the numbers of new and infrequent voters. Here the Dems have turned out 162k to just 87k for the GOP for a difference of 75k voters. In 2004 Kerry lost North Carolina to Bush by about 435k votes. Assuming party ID only remained constant we've already made up at least 75k votes during early voting. With party ID moving in favor of the Democrats all across the country, the actual gain for Obama is probably higher, but it's difficult to say how much.
If you think that's fun to look at wait until you look at the Nevada numbers. The Dem/Rep/Ind percentages are 53%/32%/15% with a raw total Dem/Rep of 107k to 64k for a 42k advantage. In new and infrequent voters the Dems are up 35k to 16k for a difference of about 19k votes. In 2004 Kerry lost to Bush by 21.5k votes. Assuming party ID remained constant the Dems have almost made up the difference that put Bush over the top in Nevada in 2004. This is a very good sign.
Iowa is as exciting as Nevada. The Dem/Rep/Ind percentages are 50%/29%/20% with a raw total Dem/Rep of 133k to 77k for a 56k advantage. In new and infrequent voters the Dems are up 34k to 17k for a difference of a little under 17k. In 2004 Kerry lost Iowa to Bush by 10k votes. These early vote numbers combined with current Iowa polling imply that Iowa could very likely be an Obama state.
Those are the only states that they provided info on new and infrequent voters, but as you can see by clicking the links above they do give raw totals for several other states, and the Democratic enthusiasm is showing up in the early vote totals. The two states where the Republicans are showing enough strength to stay neck and neck with the Democrats are Florida and Colorado. In 2004 the Republicans had a significant advantage in early voting in Colorado and a slight advantage in Florida and both states are now essentially even. We'll have to keep an eye on those.
Anyway, these numbers are suggestive of good news for Obama, at least in these few states anyway. But it's important to note that while they suggest an advantage for Obama, there's not enough information to know that for sure. Different subsets of the electorate behave in different ways, and if the demographics of the types of people who are more likely to take advantage of early voting are not normally distributed among the parties it could just mean that one group votes earlier than another. Also, we're making assumptions that the Dem early voters are Obama voters and the GOP early voters are McCain voters. That's a reasonable and rational assumption, but it's just an assumption and there's no way to know for sure if the small minority in each party that are not supporting their party's nominee are disproportionately voting early or not voting at all. I could go on and on the but the point is that there's any number of unknowns that make it impossible to know whether or not these numbers are indicative of the outcome or if it's just fun reading and interesting trivia. The only safe conclusion that you can draw is that right now in Nevada, Iowa and North Carolina the Obama campaign is beating the McCain campaign at getting their run universe to cast early ballots.
On a related note I just want to give a huge thanks to Alan Pirtle, Chairman of the Illinois Democratic County Chairman's association and the Durbin campaign for what they are doing to build the VAN in Illinois and retain useful voter data. In the 02 Governor's race we spent hundreds of thousands of dollars to build a useful voterfile and ID tons of independent voters, and once the campaign was over I put all the data on a CD that went into storage and I haven't seen it since. We looked for it in 06 but never found it so we did the same thing all over again without the benefit of any of the information we had in the race before. And 06 was the first cycle for early voting in Illinois and we put a huge statewide effort into turning out dropoff voters early. We were able to get daily data from Chicago and Cook County to check our effectiveness, but the rest of the counties weren't really technologically advanced enough yet to get us that data. And we never really did a post-election analysis to see how effective our program was. There's just too much good information lost or not followed up on from election to election when you don't have a central body focused on data collection and data analysis from cycle to cycle. What these guys are doing with the VAN could have a huge benefit for the next couple cycles. |
