With more than 20 million Americans having already voted as of October 30th (more than 16% of the 2004 total), and very large percents of the total voting population having voted in certain critical swing states, there is a good indication that Obama is doing very well. This analysis looks at selected swing states, including Colorado, Florida, Nevada, New Mexico, North Carolina, Ohio, and Virginia.
Assuming that Obama hangs on to New Mexico and the other states that either Kerry or Gore won, including Iowa, Pennsylvania, and New Hampshire, all which are currently polling for Obama with double digit or nearly double digit leads, Obama needs just one additional of the above states to win the election.
Early voting information for the purpose of this analysis is taken from
The United States Election Project – 2008 Early Voting Statistics.
Keep in mind a couple of things: First, the early voting statistics are not compiled according to whom people have actually voted for, although in most of these states the statistics are compiled by party affiliation. In assessing Party affiliation, the absolute percents of who voted from which party are not what is most important. Rather, it is the percents of who voted from which party
relative to the voter composition of the state and recent polling information. For example, if a state is composed of 40% Democratic voters and 60% Republican voters, and yet during the time that it exhibited that composition was polling about even for Obama and McCain, that means that if 50% of the early voters are Democrats that would be a very good sign for Obama, because it would mean that Democrats are performing above what is predicted by their state composition.
Secondly, although the early voting statistics are not exit polls, regular “pre-election” polls taken after early voting has commenced may be considered at to be at least partially exit polls. That is because polls taken after early voting has started will obviously include some people who have already voted, roughly in proportion to the percentage of voters who have voted early.
ColoradoAs of October 30th, the number of Coloradoans who have voted early already account for 60% of the total 2004 Colorado vote (Only 48% of Coloradoans voted early in 2004).
Broken down by Party, 38.6% or Colorado’s early voters have been Democrats, compared to 36.7% Republicans. That may not seem like much of a difference. But keep in mind that Colorado has been a Republican state. Even though Democrats demonstrated four times the increase in registered voters between January and September 30th, 2008, by the end of September, registered Republicans still outnumbered registered Democrats by nearly 5%, and by October 22nd they still outnumbered Democrats by 1.1%.
And despite that, Obama has been
outpolling McCain in Colorado by an average of 6.5% in the last week of October. Furthermore, you can see from the graph (scroll down a little) that Obama’s lead over McCain has been increasing steadily throughout October.
Colorado summary: Early voting and current polling trends both favor Obama in Colorado. Since more than 60% of Coloradoans have already voted (judging by 2004 figures), current polls are more like exit polls than they are like pre-election polls.
FloridaThe number of Floridians who have voted early so far account for 44% of the total 2004 Florida vote (compared to only 36% of Floridian voters who voted early in 2004).
By Party, 45.6% of early voters have been Democrats, compared to 38.2% who have been Republicans. That means that Democratic early voters in Florida outnumber Republican early voters by 19% (45.6% divided by 38.2%). Statewide, voters who are registered Democratic outnumber those who are registered Republican by 16%, as of October 6th. So Democrats are slightly outperforming Republicans in early voting in Florida, compared to their statewide composition.
Obama has risen in the polls from about even with McCain at the beginning of October to a lead of 3.5% in an
average of 8 polls over the past week or so.
Florida summary: As in Colorado, early voting and current polling trends favor Obama in Florida. Since about 44% of Floridians have already voted (judging by 2004 figures), current polls are similar to exit polls to a significant extent.
NevadaClark County accounted for 66% of Nevada’s votes in 2004, and it was the only county in Nevada that voted for John Kerry, giving him a margin of about 5%. As of October 13th, 2008, there were 43% more registered Democrats in Clark County than registered Republicans. But among early voters, there were 73% more registered Democrats in Clark County than registered Republicans (52.5% Democrats, 30.4% Republicans). And 63.5% of Clark County voters (using 2004 as a standard) had already voted as of October 30th, 2008.
Obama has risen in the polls from about even with McCain at the beginning of October to a lead of 7% in an
average of 6 polls over the past week or so.
Nevada summary: In Clark County, by far Nevada’s largest county, early voting favors Democrats substantially, compared to the voter composition of the county. Polling trends in Nevada as a whole also favor Obama substantially. Since almost two thirds of Clark County voters have already voted, and Clark County accounts for almost two thirds of Nevada’s voters, current Nevada polls are as much or more like exit polls as they are like pre-election polls.
New MexicoNew Mexicans who have already voted in 2008 account for 55% of the 2004 New Mexican vote.
By Party, there were 64% more registered Democrats who voted early than Republicans who voted early, as of October 30th (53.9% to 32.8%). That compares with a statewide excess of Democrats in New Mexico of 57%.
Polling in New Mexico has not been as prolific as in other swing states, but still, Obama appears to holding about a
7 point lead, with a 10 point lead in the only recent poll.
New Mexico summary: As in Colorado, Florida, and Nevada, both early voting and polling trends favor Obama in New Mexico. Also, as in those other states, recent polls in New Mexico largely represent exit polls.
North CarolinaNorth Carolinians who have already voted in 2008 account for 58.5% of the 2004 North Carolina vote.
By Party, there were 78% more registered Democrats who voted early than Republicans who voted early, as of October 30th (52.6% to 29.6%). That compares with a statewide excess of Democrats in North Carolina of 40%.
Obama has risen in the polls from about even with McCain at the beginning of October to a lead of 2.6% in an
average of 7 polls over the past week or so.
North Carolina summary: As in all the other states discussed so far, both early voting and polling trends favor Obama in North Carolina. Also, as in those other states, recent polls in North Carolina are more similar to exit polls than to pre-election polls, since nearly 60% of North Carolinians have already voted (using 2004 numbers as a standard).
OhioCuyahoga and Franklin Counties are two of the most populous counties in Ohio, together accounting for 21% of the state’s votes in 2004. They were also the two states that gave John Kerry his greatest vote margins in 2004, with 67% and 54% of their vote, respectively. And they were also the only two counties in Ohio whose early voting statistics are recorded, whose early voters in 2008 accounted for more than 30% of the 2004 county vote (Early voters accounted for only 8% and 12% of the total in 2004). Early voting statistics are not presented by party affiliation in Ohio.
Obama has risen from about even with McCain at the beginning of October to almost a 6 point lead in an
average of 9 polls in the past week.
Ohio summary: Although early voting statistics are not presented by county in Ohio, the two counties that gave John Kerry his largest pluralities in 2004 are the ones that have exhibited the highest rate of early voting so far. Polling trends also strongly favor Obama over the past month. Current polling in those counties reflects exit polls to a substantial extent, since about 30% of their population have already voted.
VirginiaThere is a paucity of early voting statistics for Virginia, compared to the other states discussed in this post. The only information presented is that early voters from Fairfax County in 2008 have accounted for 17% of the 2004 Fairfax County vote, compared to 10.5% early voters from Fairfax County in 2004. Fairfax County accounted for John Kerry’s second largest vote margin over George W. Bush in 2004.
Obama has risen from about even with McCain at the end of September to a 6.5 point lead in an
average of 8 polls in the past week.
Overall summary of early voting in critical swing statesThe critical swing states noted in this post all have a great deal in common:
1) They all voted for George W. Bush in 2004, and except for New Mexico they also all voted for Bush in 2000 as well.
2) Any one of them will give Obama an Electoral College victory on Election Day if he holds onto the states that Kerry won in 2004, plus New Mexico.
3) Early voting has accounted for 45% or more of the total 2004 vote in all these states except Ohio and Virginia.
4) Early voting has already exceeded the 2004 vote percentage in all the states except possibly Virginia, for which comparative data is not available.
5) Democrats are over-represented among early voters in each state, compared to their composition of the electorate for the whole state – except possibly for Virginia and Ohio, for which Party breakdown in not available (and in Ohio, the relatively large early voting percentage in the two largest Democratic counties suggest that Democrats are over-represented in that state as well).
6) Obama has risen in the polls in all seven states from about even with McCain (except in Colorado and New Mexico) in late September or early October to a significant lead in the last week of October. Although the lead is not above the margin of error in all polls, the
combined lead for all polls in the last week is above the margin of error in each state.
7) Obama’s rising poll numbers in each of these states has occurred in conjunction with relatively large increases in Democratic voter registration (except possibly in Ohio and Virginia, for which party affiliation is not officially recorded) and with large numbers of early voters, approaching 50% or more in every state except Ohio and Virginia.
What does this all mean?The disproportionate percentage of registered Democrats among early voters, in all of these states with the possible exception of Virginia (and maybe Ohio), is very encouraging. It is of course possible that early voters are not representative of states’ voters as a whole. If that is the case, Election Day could see a relatively greater Republican turnout than what we’ve seen so far in early voting. But since so large a percentage of voters have already voted in these states, the comparative Republican turnout would have to be quite large to turn the election around in these states.
Polling trends in all of these states show substantial gains for Obama over the past month and substantial leads in all of them, above the margin of error when the polls of the past week are combined. Furthermore, compared to past years, a reversal of the trend large enough to change the results of the election will need to be quite large, since so many voters have already voted.
None of this should provide reason for complacency, of course. Neither our country nor the world can afford another four years of Republican rule, especially with Republicans like John McCain and Sarah Palin. That could literally mean the end of world civilization as we know it. And furthermore, it is very important to achieve the largest Obama and Congressional vote margin possible in order to get our country back on track.
But given the current situation, if John McCain was to somehow pull out a “victory”, that should certainly be considered very suspicious, so say the least, unless accompanied by a sudden large plunge in Obama’s poll numbers. It would be time to get the lawyers heavily involved.