Friday, October 31, 2008
Based on those exit polls, we can pretty much call the race already for Obama in Washington and Oregon (no surprise) and also in the supposed battleground states of New Mexico and Colorado that were carried by Bush in 2004.
Here's the data (Note: all our data on early voting--from more than 50 polls so far--can be found HERE, including links to the polls discussed below):
Washington: Survey USA concluded a poll Oct. 28 in which 54% of the sample reported having already voted. Obama leads that group by a huge 59%-37% margin. This is no surprise, but it basically says the race is over in Washington.
Oregon: Similar deal here. Public Policy Polling (PPP) did a survey here in which 59% said they'd already voted. Obama leads among that group by an astounding 64%-35%. McCain would have to win more thatn 75% of the remaining votes to overcome that deficit.
Likewise, in the hard fought Oregon senate race, Democratic challenger Jeff Merkley is poised to knock off incumbent Republican Gordon Smith. Merkley leads in the early voting by 59%-37%.
While Oregon and Washington are not surprising--although the margins are huge--Colorado and New Mexico were supposed to be close. It appears they aren't.
In Colorado, we now have three separate polls showing that roughly half to two-thirds of voters there have already cast their ballots, with Obama having a big lead:
PPP--65% already voted--Obama 58%, McCain 41%
Marist--44% already voted--Obama 59%, McCain 41%
Roper--59% already voted--Obama 57%, McCain 34%
These polls are quite consistent and suggest that Obama has all but already captured this key state. (Likewise, Democratic Senate candidate Mark Udall leads for the open Senate seat in Colorado, blowing away his challenger by 60%-38% among early voters in the PPP poll.)
Finally, we have New Mexico, where early voting has also been on a torrid pace. In New Mexico, PPP just took a poll in which 56% of the sample had already voted, and Obama's lead was an astounding 64%-36%. This is consistent with an earlier Survey USA poll, in which only 10% of the sample had yet voted, but in which Obama led by 60%-37%.
Also in NM, the other Udall--Tom--has a huge lead over his challenger in the race for the open senate seat there, leading by 64%-34% among early voters.
By the way, the percentages in these polls who say they've already voted are pretty consistent with data on Prof. McDonald's early voting website, where he calculates the early voting as a percentage of 2004 votes. As of the latest posting, those percentages for these states (if available) were as follows:
New Mexico 62%
Washington: roughly 50%
Thursday, October 30, 2008
(Note: Updates on 11/4: Survey USA poll of Florida; Harris, Rasmussen national polls.)
(Note: Updates on 11/3 include: PPP polls from FL, GA, Indiana, Missouri, Montana, NC, OH, PA, Nevada, VA: Quinnipiac polls from FL and OH; Survey USA from GA, OH, WA; National polls from ABC/Wash. Post, NBC/WSJ, FOX & Hotline)
We've put them all here--we now have more than 80 early voting exit polls, with more showing up every day.
Note that for our purposes "early voting" is any process--absentee ballot, mail voting, in-person unrestricted early voting--whereby a voter may cast his/her ballot before election day. In 2004, roughly 20% of voters nationwide voted early. Experts predict an even higher percentage this year--as much as 25-33%.
We've gotten some of our demographic information from George Mason University Prof. McDonald's excellent blog on the topic, which you can also check out for additional information on other states. According to McDonald, a majority of early voters in 2004 were Republicans. This year, the trend is going the other way, especially in battleground states.
Here's what we have so far, which is updated daily, divided between national polls and state polls. For each poll, we give you the pollster, with a link to the pollster's data, the dates the survey was taken, and the percentage of the overall survey sample that reported having voted early. We then give you the preferences of the EARLY VOTERS in that poll. We've tried to keep them in chronological order, with the most recent poll listed first.
Harris Poll (10/30-11/3) (29% of sample)
Colorado Early Voting
Public Policy Polling (Oct. 28-30) (65% of sample)
Obama 59% (+18)
Roper (Oct. 22-26) (59% of sample)
Obama 57 (+23)
Georgia Early Voting
Public Policy Polling (10/31-11/2) (57% of sample)
2008 early vote as % of 2004 total vote: 61%
Roper (Oct. 22-26) (43.5%)
Obama 45% (+7)
Early votes: 4,272,280
2008 early vote as % of 2004 total vote: 56%
Public Policy Polling (10/31-11/2)(22% of sample)
Research 2000 (Oct. 27-29) (16% of sample)
Louisiana Early Voting
Total casting early votes: 266,880
Black: 36.3% (Note: African-Americans make up 30.4% of La. registered voters)
Early voters as % of 2004 election: 13.6%
Maine Early Voting
Mississippi Early Voting
Nevada Early Voting
Public Policy Polling (10/28-10/30) (56%)
Survey USA (10/30-11/2) (57% of sample)
NORTH CAROLINA EARLY VOTING DEMOGRAPHIC DATA (thru 10/31/08)
2008 early vote as % of 2004 total vote: 73%
2008 early vote as % of all registered voters: 42%
Note: Early voting in NC has now closed. This appears to be the final data.
(In the hard fought Senate race, Democratic challenger Jeff Merkley leads incumbent Republican Gordon Smith in the early voting by 51%-41%)
Washington Early Voting
The race for the White House has tightened significantly -- with Barack Obama now ahead of John McCain by three percentage points -- according to a FOX News poll released Thursday.
Despite pundit discussion of a "narrowing" of polls, the fact is that Obama leads in every national poll, and has for several weeks now. He also leads in most battleground states--indeed, he's turned new states into battlegrounds.
Furthermore, between 15-20% of Americans have already voted, and Obama leads among those voters.
We've collected data from polls that include breakdowns of early voters' preferences--about 50 of them so far--in one place HERE.
This morning we've updated our comprehensive early voting compilation with the following: national poll data from an ABC-Wash. Post showing Obama leading 59%-40% among early voters. We've also updated our demographic data on early voters from Georgia, Louisiana and North Carolina. In North Carolina, where early voting already accounts for more than 50% the number of voters who cast ballots in 2004, the demographic data is showing some regression to the mean, i.e., the outsize percentages of Democrats and African-American voters are slowly declining--but are still quite high.
We expect that by day's end there will be some additional data, and we'll do another update then.
Wednesday, October 29, 2008
Judy Feder, a Democrat, is running against incumbent GOP Representative Frank Wolf. We gave Feder some money and have generally been supportive.
Earlier this week, however, a Feder "tracker"--a staffer who follows the opposition with a video camera hoping to catch a gaffe (like George Allen's infamous "macaca" comment)--got into a scuffle with an elderly Wolf aide, who whacked 'em with a cane.
The Feder campaign is trying to make this into something good for Feder.
Sorry, it's not. Feder's tracker was WAY out of line. The tracker was following Wolf, and his wife, and yelling questions. That is not a tracker's job. A tracker records what's happening, but should not insert himself/herself into things (now, if the opposition candidate decides to single out the tracker, as George Allen did, then that's different).
Feder's tracker was, frankly, being obnoxious.
Frank Wolf is actually a quite moderate congressman who has served a district that is also moderate. There has to be some level of civil discourse in these campaigns.
Feder should fire her tracker and apologize to Wolf for the behavior.
The Harris interactive poll, conducted Oct. 20-27, found that 12% of its sample had already voted. (This comports with other data: GMU Professor MacDonald estimates on his quite useful blog that we're up to 13.4% nationally; an ABC/Wash. Post poll from 10/25-10/28 had 11% of its sample reporting they had voted already; and a Diageo/Hotline poll from 10/25-10/28 had 17% of its sample already voting.)
Of the early voters in the Harris poll, Obama led by 51%-42%.
Gallup has also reported on early voters. In its sample, as of Oct. 28, Gallup found that 18% had already voted, and they went for Obama by 53%-43%. Note: in the Gallup poll, this does not mean early voters are disproportionately voting for Obama, because the margin was about the same for those who said they still planned to vote.
The other national polls did not report who the early voters in their samples actually voted for.
For complete results of early voting information from more than 45 polls, covering most of the battleground states, click HERE.
A new poll shows McCain leading Obama by only 46%-44%, and is consistent with at least two other recent polls showing serious slippage by McCain at home.
That's close enough to make it all depend on turnout.
Will McCain send Sarah Palin to Phoenix to help him out?
For twice daily updated comprehensive list of states with early voting data, including results of more than 30 polls that include the preferences of voters who have ALREADY VOTED in this election, click HERE.
At 3:00 pm today we updated our data with results from more than 10 new polls with breakouts of early voting preferences.
Here are the highlights:
Ohio: two new polls, on top of two yesterday, give us a pretty good picture of Ohio early voters. In a Roper poll, Obama led 56%-28% among early voters, who made up nearly 30% of the sample. The margin was a bit smaller in a Quinnipiac poll, but still heavily favored Obama, 57%-41%. These results are consistent with yesterday's polls. For all the Ohio results, including links to each poll, click HERE.
Florida: two new polls here as well. In the Roper poll, Obama leads among early voters by 45%-38%. In that poll, more than 40% of the sample reported having already voted. We do have concerns about the large number of respondents who refused to answer as to who they voted for, however (14%). In the Quinnipiac poll, Obama leads by a larger margin, 58%-34%. Quinnipiac, however, did not say what percentage of its respondents had already voted. The Quinnipiac result is much more favorable for Obama than other polls of Florida--some of which have had McCain leading the early voting--so we view it as a possible outlier. For all the Florida results so far, as well as some demographic information (where the trend is in favor of Obama as more early voters vote in-person), click HERE.
Colorado: In Colorado, early voting already accounts for as much as 40-50% of the expected vote. In the Roper poll, 59% of the sample (!) reported already voting, and they went for Obama by 57%-34%. It this poll is accurate, then it may be all but over in Colorado, as McCain would have to win more than 60% of the remaining vote to pull out a win! More details here.
Nevada: Another state where as much as half the electorate may have already voted. In the Roper poll, 51% of respondents said they'd already voted and they went heavily for Obama, by 56%-33%. Again, that would make for a huge mountain for McCain to climb with remaining voters. HERE for details.
North Carolina: This is a battleground state thought to be pretty close. We now have quite a few exit polls that confirm what the demographic data show on early voters--they heavily favor Obama. In NC, it will all come down to overall turnout and turnout among African-Americans. In the Roper poll, 39% of the survey had already voted--consistent with statistics reported by the State Board of Elections--and Obama leads among those early voters by 59%-33%. This is consistent with two other polls. Details HERE.
We also have new data (from Survey USA) for the state of Washington, where more than half of voters have already cast their ballots. As expected, Obama is leading by a wide margin, 59%-37%.
And, we have additional data from Roper for New Hampshire, Pennsylvania and Virginia--all states with only traditional absentee balloting, so with smaller samples of voters who've already voted. In each of those states, Obama has a huge lead over McCain among those absentee voters. Details here.
Finally, we've updated demographic data from some southern states showing a huge African-American turnout among early voters and suggesting that Obama may well take Georgia if the trend continues.
Tuesday, October 28, 2008
Today's news is as follows:
The Pew Center has conducted a national poll of 1500 respondents during Oct. 23-26 and found that 15% of its sample had already voted. The poll did not break out presidential preferences for those who have voted, however.
Survey USA issued a new poll of Ohio voters. Ohio has very active early voting going on. The results are quite interesting. Two weeks ago, SUSA conducted a similar survey. In it, 12% of respondents reported having already voted, and they heavily favored Obama, by 57%-39%.
We wondered if that was just early enthusiasm and whether the numbers would level out over time. They haven't. The new survey, done through Oct. 27, includes 22% of respondents who have already voted. Among them, Obama still has a commanding lead, 56%-37%, almost the same as two weeks ago.
We also wondered if this wasn't just Survey USA's methodology. Fortunately, we found a second poll, this afternoon, from LA Times/Bloomberg, which has very similar results. In that poll, 21% of respondents report having already voted, and they went for Obama by a wide margin, 57%-35%.
This means that with one in five Ohio voters having already gone to the polls, McCain has some major ground to make up in the remaining 80% of voters.
LA Times/Bloomberg also did a poll in Florida, which found that 14% had already voted there. Unlike almost all our other states, however, Obama is not leading among early voters in the Sunshine State. Instead, McCain leads by 49%-45%.
We also missed a poll yesterday with North Carolina early voting data, and have included it today. Public Policy Polling conducted a survey that ended Oct. 26. In it, fully one-third of respondents had already voted, and they've gone heavily for Obama, 63%-36%. (In the Senate race, Democratic challenger Kay Hagan also leads in early voting over incumbent Elizabeth Dole, 61%-35%.)
Our other news is all updates of demographic data from those Southern states that report such data on early voting, including Georgia, Florida, North Carolina and Louisiana. Check our comprehensive guide for the details, but the bottom line is this:
In GA, LA and NC, blacks are voting at well above their registration percentages. In NC, early voting turnout has been HUGE--as much as 40% of the state's voters (1.4 million) have already voted, and blacks account for 28% of those voters (versus 21% of registered voters). We thought the trend might change as more voters went to the polls, but it has not changed much. It appears the Obama machine is really getting the minority voters to the polls in NC.
Georgia is similar. In Louisiana, the number of early voters so far is smaller, but the trend is similar, although not quite as dramatic as NC and GA. In Florida, Democrats are outnumbering Republicans at in-person early voting stations.
Needless to say, if Obama takes Ohio, Florida, Georgia and NC, this election is over.
Monday, October 27, 2008
We've now done our update. Here's what's new: we updated Florida, Georgia and North Carolina early voting statistics, which show a continued torrid pace of in-person early voting in all three states. In NC and GA, as many as a third of voters have already cast their ballots with another few days of early voting left to go.
In NC and GA, blacks continue to vote well above their registration percentages. In both those states, women far outnumber men in early voting. In Florida, Democrats are gradually stretching out a lead over Republicans as early voting continues apace.
We could only find one poll today with data broken out on voters who'd already cast a ballot, and that was for Virginia. In the Old Dominion, 9% of the Survey USA poll sample had already voted, and they went for Obama by 67%-30%.
UPDATE: We found a Survey USA poll of Oregon voters. In the SUSA sample, 50% of Oregon voters had already voted (in OR everyone votes by mail). Obama had a huge lead among those who had already voted, 62%-34%. That's virtually impossible for McCain to overcome. Maybe the networks should go ahead and call Oregon for Obama.
Interestingly, in the hotly contested Oregon Senate race, Democratic challenger Jeff Merkley also has a large--but not as large as Obama's--lead amongst those who've already voted in his effort to unseat GOP incumbent Gordon Smith: 51%-41%.
We expect to see a lot more tomorrow, so we'll update a couple times.
Sunday, October 26, 2008
In today's Washington Post, Marc Fisher-who we usually find quite reasonable--argues against Maryland's early voting proposition. His reasoning lacks coherence or rationality.
Fisher seems to be wallowing in some kind of nostalgia.
His first argument is what if something happens after early voting starts to change people's minds, like the economic collapse of late. This is pretty ridiculous--after all, what if something happens after election day to change our minds? We'd say that a lot of people wish'd, after election day, that they hadn't voted for W. Bush. The economic collapse could've happened in December; or 9/11 could've happened before the 2000 elections. Any election date brings with it the risk--indeed the certainty--that events after it could change voters' feelings.
On the other hand, MOST voters would not change their minds absent something awfully dramatic. The economic collapse MAY have shifted 5% of the electorate to Obama. Or maybe not--maybe it was McCain's Palin pick, his awful campaign and Obama's strong debate performances. In any event, the economic collapse hardly tells us much about the candidates themselves.
Fisher's second--and apparently, to him, more significant--argument is that voting should be a "communal experience." Voting "is a statement we make about ourselves, to ourselves, but also to each other. It is how we say, 'I am part of something larger, and my voice matters, and so does yours.' When we chip away at that communal experience, we diminish democracy."
We read this over and over, trying to see if we missed something. Nope, it's meaningless dribble. First, there have long been at least some people who vote absentee, usually at least 5-10% of the voting population. Is there something wrong with these folks that there votes aren't as valuable as those who share the "communal experience?"
Second, in most early voting states they have in-person early voting at regular polling sites and you get to wait in a line with your fellow voters. You just don't have to wait as long and you can go when it is more convenient. So it is communal, at least communal enough.
But some states, such as Oregon, have ALL mail-in voting, so it isn't communal. Is that so horrible? Commenters on Fisher's article made some excellent points about mail-in voting, including that such voters, who have their ballots in advance, often call campaigns and ask questions, and they can also research issues and be familiar with other issues, such as the wording of propositions and bond referenda. Turnout in Oregon is generally higher than other jurisdictions, certainly than Virginia.
In any event, Fisher's "communal experience" is a bunch of hogwash. In Virginia, polls close at 7:00 pm. Have you ever tried to get around town in evening rush hour traffic, after taking kids to their activities and so forth, in time to vote before 7:00?
And what if it rains--or snows as is the case in some northern climes--on election day? What's so communal about standing in line (or rushing to a polling place in such weather) in a pouring rain?
Finally, we have a totally irrational system of voting on Tuesdays, of all days. Most democracies vote on a Saturday or a Sunday and have it as a national holiday. We vote on Tuesdays, in November, because our founders thought it most convenient for an agrarian society. It was after the fall harvest, and it gave voters Monday to get to the polls and Wednesday to get home. Fisher could at least advocate for moving election day to some more convenient day!
Fisher has some other arguments--such as that because early voters are more motivated, more committed, or older, or younger, or whatever, it gives campaigns a chance to parse voters into different categories and treat them differently.
PUHLEEZE. They do it already.
How 'bout this. Let's have one day to vote, on a Saturday, and let's prohibit all campaign activities until two weeks before the election, so that we don't have to suffer from months and months (or years and years) of endless campaigning.
In any event, voters have spoken with their feet: in jurisdictions with early voting, it is immensely popular. Yet, for those traditionalists, such as Fisher (does he have a cell phone or email yet?) who like the communal experience of voting on election day, the opportunity still exists.
Saturday, October 25, 2008
We've come across no new state polls that include separate data on voters who have already cast their ballots. We did add an exit poll from El Paso, Texas of voters only in that city who had voted at an early voting site.
We've updated NC and Florida demographic information. In Florida, early voting is continuing at a torrid pace. With each day's update of information, Democrats are stretching their margin over Republicans in the early voting. This reflects that absentee voting, which started earlier, was dominated by Republicans, but that in-person early voting, which is going on now, has been tilted toward Democrats.
Just to give you one indication about North Carolina and African-American enthusiasm in that state: so far, nearly 25% of all black voters registered in NC have already voted. That's incredibly high. (For whites, about 18% of registered have early voted.)
Our correspondent from South Carolina reports that when he voted today in Charleston, 26,000 voters had already been to the polls there. That's well above the record 16,000 in Charleston in the last election. SC does not have "no-excuse" early voting like NC and Georgia, but it does have some broad absentee voter exceptions, including one for older voters.
Friday, October 24, 2008
First, we updated our comprehensive guide to early voting--see here. No new state surveys revealing early voting preferences. We did update demographic data for Florida, Georgia and NC--very strong turnout for early voting in all three. Data still favors Obama.
Gallup did a national poll that included early voters and found that Obama had a slight lead among the nine percent of voters who said they'd already voted.
Second, in the RealClearPolitics moving average of all presidential polls, Obama closed today with a lead of 7.9 points over McCain, which is his highest yet in the RCP average. Since October 13, Obama has not been below 5 points in this running cumulative average of the various polls.
Need we say that with 11 days to go, that's quite a huge margin?
As we indicated last week, the polls that include cellphones give Obama about an additional three point lead.
Thursday, October 23, 2008
For all the results, click here.
We'll do another update late this afternoon, after additional polling data is released during the day.
UPDATE: As of 6:30 pm we've added data from a series of Big 10 Battleground polls of Illinois, Indiana, Iowa, Minnesota, Michigan, Ohio and Wisconsin. The numbers of "already voted" in these polls were VERY small, so the data is not too useful. In the aggregate, however, it tells us that Obama is probably leading the early voting throughout the midwest. (In Minnesota, McCain led, but that was based on three voters in the poll sample, two for McCain, one for Obama).
We've also updated Florida, Georgia and North Carolina demographic information.
Wednesday, October 22, 2008
Since when do protesters show up at election sites and heckle people? This is like some awful throwback to the early days of civil rights marches, with redneck whites out protesting and yelling. (In fact, the local white sheriff, just like the old days, was standing there and said he didn't see anything wrong with it.)
WHAT IS WRONG WITH THESE PEOPLE!
For more on this, go to Christina Bellantoni's blog at the Washington Times.
No longer. In today's Mason-Dixon poll of Virginia likely voters, taken over Oct. 20-21, Obama has a small--but statistically insignificant--lead, 47%-45%. In the last Mason-Dixon poll of Commonwealth voters, at the beginning of October, McCain led 48%-45%.
We note that the M-D poll sample is 50/50 between men and women. Men somewhat favored McCain, whereas women somewhat favored Obama. Recent patterns in Virginia--and throughout the nation--are that women tend to turnout and vote in greater proportions than men. If that's true, the poll probably underestimates Obama's support.
As far as we can tell, M-D only calls landline phones. Again, that will underestimate Obama's support to some degree, because younger people are much more likely to live in households with no landline, using only cellphones.
By the same token, the M-D poll is useful, because we ought to think of Virginia as a close race and make sure to get everyone to the polls by election day.
The demographics of the early voters continue to favor Obama (as well as Democratic senatorial candidate Kay Hagan and gubernatorial candidate Bev Perdue).
African-Americans, who make up about 20% of NC registered voters, account for 29% of early voters in the state.
Here's the latest data, as of the end of day yesterday (Tuesday, October 21) from the NC Board of Elections:
Early votes: 629,266
Percent white: 67.13%
Percent black: 29.04% (Note: blacks are 20% of NC registered voters)
Percent Democrat: 56.41%
Percent Republican: 27.08%
Percent Women: 56.2%
Percent Men: 43.0%
2008 early vote as % of 2004 total vote: 17.97%
For a comprehensive look at early voting data around the country, click here.
In the 2004 presidential election, 3.3 million Georgians voted. That means that fully 25% of the number of ballots cast in Georgian in 2004 have already been cast. We predict a considerably higher total turnout in Georgia this year, but the number of early voters is still quite impressive.
The demographics of this early voting tsunami continue to favor Obama, with a very high proportion of African-American and female voters, as shown below.
Polls of Georgia give McCain a small lead, but if those polls have underweighted blacks, then McCain's lead disappears.
Here's the data, through yesterday, from Georgia on early voting:
GEORGIA EARLY VOTING (thru 10/21/08)
Early votes: 825,923
Percent white: 60.67%
Percent black: 35.57% (Blacks make up 29% of Ga. registered voters)
Percent women: 56.1%
Percent men: 41.1%
2008 early vote as % of 2004 total vote: 25.01%
Tuesday, October 21, 2008
First, Survey USA has updated it's NC sample, with polling through Oct. 20. In the new SUSA poll, 14% of respondents reported that they had already voted. Of those, Obama leads handily by a margin of 59%-36%, or 23 points.
That's a bit lower than in the initial SUSA poll a couple weeks ago, where only 5% had already voted and Obama led by 65%-31%.
The Survey USA poll matches up pretty well with what the NC Board of Elections is reporting (see below), with about 14% of likely voters having already cast ballots. Since about 56% of those voters are Democrats, Obama is also picking up some independent voters, which is good.
A second poll was released by NCCivitas, in which 4% of the sample volunteered that they had already voted (this is from a personal communication with the poll's spokesperson). Again, Obama had a large lead amongst those voters, 64%-32%.
By all accounts, the demographics of the early voting in NC have heavily favored Democrats, so Obama's margin is no surprise in these polls. Here's the latest demographic info from NC:
NORTH CAROLINA EARLY VOTING (as of 10/21/08)
Early votes: 481,468
Percent white: 68.3%
Percent black: 28.8% (Note: blacks are 20% of NC registered voters)
Percent Democrat: 56.11%
Percent Republican: 27.39
2008 early vote as % of 2004 total vote: 13.75%
The question is whether he can keep it up as early voting proceeds, or whether the lead will gradually diminish. The polls show overall that NC is a virtual tie, but as SUSA notes, the outcome may hinge heavily on turnout amongst African-Americans. If they turnout above historical rates, NC could go for Obama, elect a new Democratic Senator and have a Democratic Governor.
Notably, Obama is doing slightly better in early voting that his downticket Democratic running mates, Kay Hagan for Senate and Bev Perdue for Governor. Hagan's early voting margin over incumbent Senator Liddy Dole is 56%-37%, while Perdue's margin is 54%-41%.
In Wyoming, 14% of the SUSA sample reported already voting. Of those, 55% voted for McCain, 41% for Obama. That's still more favorable to Obama than those "likely voters" who haven't yet voted, who went for McCain by a margin of 58%-37%.
Monday, October 20, 2008
Note 1: For our purposes, "early voting" is any process--absentee ballot, mail voting, in-person unrestricted early voting--whereby a voter may cast his/her ballot before election day. In 2004, roughly 20% of voters nationwide voted early. Experts predict an even higher percentage this year--as much as 25-33%.
Note 2: according to George Mason University Prof. McDonald's excellent blog on the topic, in 2004 a majority of early voters were Republicans. This year, the trend is going the other way, especially in battleground states.
Georgia Early Voting
Obama 52% (+6)
GEORGIA EARLY VOTING DATA (thru 10/29/08)
Early votes: 1,572,293
Percent white: 60.6%
Percent black: 35.1% (Blacks make up 29% of Ga. registered voters)
Percent women: 56.3%
Percent men: 40.5%
Florida Early Voting
Roper (Oct. 22-26) (43.5%)
Quinnipiac (Oct. 22-26) (?% of sample)
Obama 58% (+24)
LA Times/Bloomberg (Oct. 25-27) (21% of sample)
McCain 49% (+4)
(as of 10/16) (13% of SUSA sample)
McCain 53% (+8)
Quinnipiac (10/16-10/21) (?% of sample)
Obama 48% (+4)
St. Petersburg Times poll: no data, but text describing findings (poll taken Oct. 20-22) says "McCain and Obama were tied among people who already had voted."
FLORIDA EARLY VOTING DATA (thru 10/29/08)
Early votes: 2,872,459 (includes absentee votes)
Percent Democrat: 46.0% (Note: the data is trending more Democratic with each update)
Percent Republican: 38.2%
2008 early vote as % of 2004 total vote: 37.6%
Illinois Early Voting
Big 10 Battleground (10/19-10/22) (2.1% of sample)
Obama 83% (+67)
Indiana Early Voting
Big 10 Battleground (10/19-10/22) (1.9% of sample)
Obama 100% (+100)
Iowa Early Voting
Obama 65% (+34)
Maine Early Voting
Obama 61% (+27)
Michigan Early Voting
Big 10 Battleground (10/19-10/22) (1% of sample)
Obama 60% (+20)
Minnesota Early Voting
Big 10 Battleground (10/19-10/22) (.6% of sample)
McCain 66% (+33)
New Hampshire Early Voting
New Mexico Early Voting
Obama 60% (+23)
North Carolina Early Voting
Public Policy Polling (as of 10/26) (33% of survey)
Obama 63% (+27)
(as of 10/20) (14% of SUSA sample)
Obama 59% (+23)
(as of 10/20) (4% of CIVITAS sample (volunteered))
Obama 64% (+32)
NORTH CAROLINA EARLY VOTING DATA (thru 10/29/08)
Early votes: 1,847,860
Percent white: 69.0%
Percent black: 27.1% (Note: blacks are 21% of NC registered voters)
Percent Democrat: 53.3%
Percent Republican: 29.2
Percent Women: 56.4%
Percent Men: 42.7%
2008 early vote as % of 2004 total vote: 52%
2008 early vote as % of all registered voters: 30%
Note: the data, while still quite favorable to Obama, is trending the other way. On Oct. 23, blacks made up 28.7% of the early vote; now they make up 27.1% of the early vote. The percentage of Democrats has similarly declined. However, with as much as 40-50% of all voters having voted, African-Americans still make up a disproportionate number of voters, good news for Obama.
Ohio Early Voting
SUSA (as of 10/13) (12% of sample)
Obama 57% (+18)
Quinnipiac (10/16-10/21) (?% of sample)
Obama 71% (+51)
Big 10 Battleground (10/19-10/22) (2.7% of sample)
Obama 66% (+33)
Oregon Early Voting
SUSA (10/25) (50% of sample)
Obama 62% (+28)
(In the hard fought Senate race, Democratic challenger Jeff Merkley leads incumbent Republican Gordon Smith in the early voting by 51%-41%)
Pennsylvania Early Voting
Roper (Oct. 22-26) (12% of sample)
Obama 65 (+43)
Texas Early Voting
El Paso Times (10/24) (100% of sample) [Note: this is a poll only of El Paso voters, taken as they left early voting sites.]
Obama 69% (+46)
Virginia Early Voting
Roper (Oct. 22-26) (12.5% of sample)
Obama 67% (+44)
SUSA (10/26) (9% of sample)
Wisconsin Early Voting
SUSA (as of 10/19) (11% of sample)
Obama 52% (+15)
Big 10 Battleground (10/19-10/22) (1.3% of sample)
Obama 63% (+26)
Wyoming Early Voting
(as of 10/20) (14% of SUSA sample)
McCain 55% (+14)