This chart indicates that 13% of Obama's voters in 2008 say they are going to support Mitt Romney in 2012. The poll finds that 84% report they are voting for Obama and 3% are backing others or haven't made up their minds.
Obama's problem is that he has almost no upside on his votes from four years ago but he is vulnerable in a number of states that he carried last time. Also, it does not take a lot of voters to switch sides or stay home to hurt Obama.In fact, at that time I pointed out four states, North Carolina, Indiana, Florida and Ohio, were all carried by Obama by less than 5% of the vote in 2008. It takes only one out of every 40 voters who voted for Obama in 2008 to switch to Romney swing the result in these states if everything else is equal.
I also cited 4 other states, Virginia, Colorado, Iowa and New Hampshire, that were decided by less than a 10% margin in 2008. It only takes one out of every 20 voters in these states to swing these states to Romney.
Where are we today? In every one of these states Romney is leading in the polls or is slightly behind and within the margin of error of the polling sample. In addition, the polls show that he is very close in Pennsylvania (10.35% margin in 2008), Minnesota (10.24%) and Wisconsin (13.90%).
Let's look at the overall math of the national popular vote because I think it is very interesting.
In 2008, Obama garnered 69.5 million votes to McCain's 60 million. A winning margin of 9.5 million votes. Head to head this was 53.7% for Obama and 46.3% for McCain-a margin of a little over 7 points.
In 2004, Bush won with 62 million votes to Kerry's 59 million votes.
Therefore, Obama got 10 million more votes than Kerry and McCain got 2 million less than Bush. It is easy to see why Obama won.
Also note that the total vote count increased for the two major candidates from 121.0 million votes in 2004 to 129.5 million votes in 2008. Simply stated, it was as if Obama got the votes of every new voter between 2004 and 2008 and also took Bush some votes away on top of it.
By contrast, in 2000, only 101.5 million votes were cast for Bush and Gore. Therefore, compared to a population increase of 9.7% between 2000 and 2010, votes cast increased by about 25%. This would seem to put to rest the concern that any voter suppression has been going on. Many more people have been voting each year over the last several Presidential elections.
Let's look at how the math plays out using the 2008 national popular vote totals and the Washington Post/ABCNews poll assuming 13% of Obama voters do vote for Romney.
| Obama | McCain/Romney | |
| 2008 Results Baseline Popular Vote | 69.5 | 60.0 |
| 13% of Obama voters switch to Romney | -9.0 | 9.0 |
| 60.5 | 69.0 | |
| 2012 Projection A | 46.7% | 53.3% |
| 5% of Romney voters switch to Obama | 3.0 | -3.0 |
| 63.5 | 66.0 | |
| 2012 Projection B | 49.0% | 51.0% |
| 1% of Obama voters stay home | -0.7 | |
| 62.8 | 66.0 | |
| 2012 Projection C | 48.8% | 51.2% |
| 1% more Republican voters go to the polls | 0.6 | |
| 62.8 | 66.6 | |
| 2012 Projection D | 48.5% | 51.5% |
If 13% of Obama voters in 2008 switch to Romney and everything else stays the same (the total vote turnout, same voter energy on both sides etc.) Romney would win the popular vote by 53.3% to 46.7% (Projection A) in a head to head match-up. Voters that switch carry the most leverage in a political race as they subtract one and add one. A 1% switch away from one candidate to the other actually changes the overall margin by 2%.
Bear in mind that the numbers above does not include third party votes so if we compare to the polls or the final results we must reduce the total vote percentage for those votes. For this purpose I will assume that number is 1% and assume that total is borne equally by each candidate-or .5% off of each candidate's final vote percentage. This would equate to 52.8% for Romney and 46.2% for Obama in the final results in Projection A-a 6.6% margin. Less than Obama's margin in 2008 but a very comfortable win.
When I wrote about the race last June I cited an analysis by Wayne Allyn Root who is a political oddsmaker (and Libertarian Party Vice Presidential candidate in 2008) who approached the Presidential race from a simple premise-he did not think there was one person in the land that voted for McCain four years ago who would vote for Obama in 2012. On the other hand, he believed there were millions of angry, disillusioned voters filled with buyer's remorse.
He cited various voting blocs and asked if it was reasonable to believe whether Obama could expect any of them to deliver more votes for him in 2012 than in 2008. Black voters? Hispanic voters? Jewish voters? Youth voters? Catholic voters? Small business owners? Blue collar working class whites? The polls numbers above support this view.
However, is Root's premise correct that there are no McCain voters who will now turn around and vote for Romney? There are apparently some out there. The best number I have been able to find in the polling is that this number is about 5%.
If we assume this number is correct, the race tightens to a 2% margin (51%-49%) for Romney (Projection B).
Of course, all of this assumes everything stays the same as 2008 as far as turnout. What happens if Obama's loyal base turnout degrades by 1% but there is no improvement in the Republican base?
Projection C layers this assumption on and it shows Romney's margin increasing to 2.4% (51.2%-48.8%).
Many expect that Democrat turnout will lag 2008 and Projection C models this without assuming any increase in turnout of Republican loyalists. Remember from the numbers above that McCain had 2 million fewer voters than Bush while Obama ended up with 10 million more votes than Kerry. Some of this was undoubtedly due to Bush votes moving to Obama. However, it also resulted from many new voters turning out for Obama (minorities and younger voters in particular). Some number of Republicans also undoubtedly sat out the election either because they did not like McCain/Palin or the pre-election polls and hype for Obama simply discouraged them from bothering to vote.
Therefore, what happens if we further assume that the McCain vote increases by just 1% due to more Republicans taking an interest this year? That additional wrinkle added to the other assumptions expands the margin for Romney to 3% (51.5%-48.5%). (Projection D)
Keep in mind that I am also not taking any of the 3% of Obama voters from 2008 who say that they are either undecided or are going to back a third party candidate. My numbers above assume they stay with Obama. This could be an additional benefit for Romney on Tuesday.
The polls this year show an incredibly close race. There seems to be a poll for every result. A large reason for this is that all polls have a margin of error. The margin of error results because, after all, the poll is a sample of the population. The more voters in the statistical sample, the smaller the margin of error. Fewer voters in the sample, a larger margin of error.
The Washington Post ABC tracking poll cited above has a margin of error of plus or minus three percentage points. It shows the race as Romney 49% and Obama 48% on November 1. This means that Romney could actually be as high as 52% and as low as 46%. Obama could be as high as 51% and as low as 45%. Rasmussen's poll today is 48%-48% with a 2.5% margin of error. Gallup usually surveys a larger sample of likely voters and has a margin of error of 2%. Gallup suspended polling for a few days because of Superstorm Sandy but it was Romney 51% and Obama 46% as of Monday.
You will notice my analysis above matches up pretty well with the overall polling data. Almost any way you look at the data we will probably have a final result that will be within a margin of error of plus or minus 3%. My math suggest the same thing. It seems unlikely that either candidate will win by more than a 3 percentage point margin in the popular vote. If anyone can do that it will be Romney based on the math.
Get ready for an interesting Tuesday. +/- 3 Points will decide it. You can count on it if any of this poll data is right.
No comments:
Post a Comment