This is my assessment of the Presidential race as it stands today. Below are the latest numbers from the daily tracking polls (yesterday's numbers are in parentheses):
Research 2000 (10/17-10/19): Obama 50 - McCain 42 (Obama 50 - McCain 43)
Rasmussen Reports (10/17-10/19): Obama 50 - McCain 46 (Obama 51 - McCain 45)
Zogby (10/17-10/19): Obama 50 - McCain 44 (Obama 48 - McCain 45)
Hotline (10/17-10/19): Obama 47 - McCain 42 (Obama 48 - McCain 41)
Gallup - Traditional (10/17-10/19): Obama 50 - McCain 45 (Obama 49 - McCain 46)
Gallup - Expanded (10/17-10/19): Obama 52 - McCain 43 (Obama 51 - McCain 44)
Tipp (10/15-10/19): Obama 47 - McCain 41 (Obama 47 - McCain 42)
Battleground (10/13-10/19): Obama 49 - McCain 45 (Obama 49 - McCain 45)
There are many divergent opinions from pollsters about party identification and that explains the disparity you are seeing in some of the results. For example, some pollsters are using conservative models like Zogby, employing party identification percentages from 2004 that favor Republicans. Others, like the Daily Kos/Research 2000 tracking poll, are optimistically projecting a huge Democratic turnout, as has been predicted in some quarters.
Today's average of all the tracking polls is: Obama 49.375 - McCain 43.5. Yesterday's average was Obama 49.125 - McCain 43.875.
The 2004 Election Map:
With just two weeks left, and McCain running approximately 6 points behind in the national polls, Obama is also enjoying very significant leads in all the states he must win: New Hampshire, Pennsylvania, Michigan, Wisconsin, Iowa, Virgina, and New Mexico. Right now, state polls are lagging indicators of the momentum shifts in the race and that won't change until the daily tracking polls begin in the battleground states closer to the election. But, what we can discern is that for McCain to emerge victorious he will need all of the momentum to be running his favor from now until election day. Any day in which Obama is able to stabilize his lead, or expand upon it, makes it much less likely that John McCain will follow George Bush as the forty fourth President of the United States.
Based on my average, McCain needs to recover 5.875 points against Obama's lead if he is to level the race and stand any chance of emerging victorious. With the race that close it's likely that undecideds will break for McCain and he will take Florida, Ohio, Missouri, North Carolina, Nevada, and Colorado. That will put him within striking distance of the blue states where, presuming New Mexico and Iowa are out of reach, he'll have to win one of Michigan, Pennsylvania, Minnesota, or Wisconsin.
For me, the first problem with this unlikely scenario is Virginia. I think the Democratic candidate for Senate, Mark Warner (pictured right), a very popular former governor, is making this an incredibly difficult state for McCain to win. Warner currently leads his Republican opponent by 20+ points and I can't imagine a scenario where Obama doesn't benefit from this tremendously on election day. Secondly, Obama's leads in those critical blue states that he must hold are very big. Pollster.com puts Obama's average lead in Pennsylvania at 14 points, Michigan 10 points, Wisconsin 8 points, and Minnesota 7 points. Those are big leads for McCain to overcome in the space of two weeks. Obama's lead in the 2004 George Bush states Iowa and New Mexico are 12 and 7 points respectively, and those spreads have been too stable throughout this election season to change now. That means McCain must win one of those four blue states to get to the magic 270 number in the electoral college.
14 days of polling remains. That means that McCain must recover an average gain on Obama's lead of 0.42 points a day. McCain actually lost ground today, ceding 0.625 in the spread towards Obama. Anymore polling days like this will make it next to impossible for anything other than an Obama victory.
However, there are some caveats to take into consideration. Firstly, Obama's support is notoriously soft. We saw that against Hillary in Texas, Ohio, and Pennsyvania, where the electorate actually moved away from Obama and towards Hillary in the final days of polling. We also saw it post Republican convention where his lead went from +8 against McCain to -6 in the space of a weekend. Additionally, throughout this entire Presidential race up until the economic crisis began in September, Obama's leads have fluctuated wildly from +15 to -6. The electorate's trust of Obama is subject to a rapid erosion of support given the right set of circumstances.
Secondly, undecideds typically broke against Obama in the primaries. We saw this in New Hampshire, and in many other states throughout the battle with Hillary. If voters still have reservations at this point about siding with Obama then the probability is that they will resort to their fears and concerns, as opposed to their hope and desire for change, even in this political climate. Also in the 2006 mid-term elections, we saw a similar trend in favor of Republicans and against the Democrats. It has been forgotten, because of the innumerable tight victories Democrats were able to eventually win, but the Democratic lead going into that election should have, in reality, resulted in far more emphatic margins. The United States, ultimately, is a center right nation and these natural proclivities amongst the electorate will always work against Democrats in the closing days of most political races, in my opinion.
Taking this into account I think it's fair to estimate that undecideds in the last 3 days will predominantly break to McCain by a margin of 65% to 35%. What this means is that if Obama is leading 50-46 heading into the weekend before election day, we will probably be looking at a final result of something like 51.4 - 48.6 in favor of Obama. The difference undecideds can make is negligible unless McCain is able to draw level.
The Intrade Predication Market gives McCain a 15% chance of victory and I think this sounds about right. However, news cycles like yesterday or today are crushing for McCain. The Colin Powell endorsement not only elevated Obama's standing on national security issues, but Powell also launched into a stinging attack on the McCain campaign, the selection of Palin as a candidate for VP, and on the Republican party for its conduct in general (video below). In actual fact, Powell went much further than endorsing Obama as was expected. Instead, he turned into a fully fledged advocate providing some of the most effective arguments for Obama's candidacy that we have heard this election season.
McCain has only one chance, and that is to pivot away from the negative campaigning, to positively evoke a sense of America's greatness being founded in Conservative ideals. He has to acknowledge that George Bush has failed, but explain that the failure is as a result of George Bush, and not conservatism. This is his only hope.
Bush ran a brutally negative and effective campaign against Kerry in 2004, but he did two things differently than McCain.
Firstly, he started early. When McCain was running ads about Paris Hilton and Britney Spears, Republicans had already defined Kerry as a flip flopper and a politician with no integrity. McCain simply started the "guilt by association" attacks too late, after Obama had already been introduced and defined positively in the minds of too many.
Secondly, Bush did hardly any of the dirty work himself. He came very close to publicly condemning the swift boat ads while surreptitiously co-ordinating their smears. George Bush pivoted from the attacks of surrogates by simply presenting a positive message about the greatness and strength of America . McCain, by contrast, has become defined by the Ayers attacks, or the lipstick on a pig reference that Obama made. He has taken too much responsibility for the work of his campaign.
Ultimately, the momentum is with Obama right now and not with McCain. The Powell endorsement, preceded by center right endorsements from newspapers nationwide, have provided the platform for an Obama landslide. Victory for Obama can be guaranteed, not by playing it safe, but by truly winning the hearts and minds of the American center ground, offering himself as a President for all, and not just a President of the few.
My final suggestion for following the race over the next fourteen days is to follow the Gallup tracking polls. They have been the greatest early indicator of momentum shifts throughout this political season, with Rasmussen, Zogby, and Battleground always playing catch up. The margin might be greater or smaller, but right now it's all about the trend. No trend for McCain equals no hope for his campaign. Over the past two days Obama has improved his spread by 4 points in the Gallup poll. If this isn't halted within the next couple of days then we will be heading for a landslide of Reagan-esque proportions.