People have asked me why I believe that Romney will win, when all the polls seem to be going increasingly against him. First of all, as I have pointed out several times, the majority of polls are skewed to unrealistic turnout models. But the more telling reason is Obama's approval ratings, or the lack thereof.
I wrote about this and the importance of turnout in elections awhile back where I pointed out:
Republicans will show up at the polls, perhaps in unprecedented numbers. But what about Democrats and others who voted for Obama? Will they show up? It is very doubtful that they will show up and vote in anywhere near the numbers that they did in 2008. I can say this without much hesitation simply by looking at Obama's approval ratings.My contention is that when a President's approval rating drops among people who supported him, it must manifest itself somewhere. It does not mean that everyone who voted for Obama who does not approve of his job performance is going to vote for Romney...though some will. It also doesn't mean that his supporters who don't approve of the job he is doing just won't vote...though some won't. What it does mean is that as Obama loses voters, which he has it draws Romney closer.
So let's begin with a couple of safe assumptions. A minuscule number of McCain voters will switch and vote for Obama, I doubt few people would argue with that. But there are quite a few Republican voters who sat out 2008 that will not sit out 2012. Party registration numbers and common sense points this out. As I have pointed out several times, 2008 was the lowest percentage of Republicans as a percentage of the electorate in decades 32%. The historic norm is 35% and I suspect that is probably a minimum to be expected in 2012. So there is that for the Republican side of the equation, however the real problem for Obama is on the Democratic side.
For purposes of this discussion I will just use ethnic polling I could use other demographics but since ethnicity is the area which the media likes to focus on let's focus on it. But let's begin with this other safe assumption, that across all ethnic groups a certain percentage of people voted for Obama simply because he was the first black Presidential candidate and he will never have that advantage again, after all there is "only one first time".
The media and the left like to say that being a black candidate hurt him among the "racist" white population, but the facts do no bear that out. Here are the percentages of the white vote for every Democratic candidate starting with Carter in 1980
So How did the first Black candidate do with the "racist whites" in America, Obama received 43% of the white vote. More than any other Democratic candidate accept Bubba in 1996. The majority of whites in America are not prejudiced against black candidates, they are prejudiced against Democrats.
What has changed is that whites are not as large a demographic as they were in 1980 when they were 88% of the electorate compared to 74% in 2008. But whites still make up the largest ethnic demographic in the electorate and this is why that 40% threshold is so important to Obama. Without 40% of the white vote Obama's path to victory becomes extremely difficult.
So we look to the Gallup to see how he is doing among the various ethnic groups. It is my contention that the approval ratings are a far better for forecasting the election than the horse race polls. This is why I have simply plugged in Obama's approval ratings into the actual turn out ratios from 2008. Here are the turn out ratios for ethnic groups in 2008.
Now we take Obama's most recent week approval ratings and pug it in.
Whites 74% x 39% = 28.86% of the vote
Blacks 13% x 88% = 11.44% of the vote
Hisp. 9% x 62% = 5.58% of the vote
Other 4% x 60% = 2.40% of the vote
Giving Obama a total of 48.28% of the vote. If you take out the normal 3% for third party candidates, Romney still wins the popular vote in a squeaker. The Gallup poll has another category which is simply White and Non White which shows virtually the same thing. Obama has a 75% approval rating among non-white voters 26% x 75% = 19.5 plus 28.6% (the white vote)=48.1%,
Why do I feel confident in this way of calculating the election results? Well as I said disapproval has to go somewhere and every disaffected voter that voted for Obama last time that votes for Romney is the equivalent of two votes for Romney and everyone that doesn't vote is like one vote for Romney in the scheme of things. These two factors I believe make Obama's approval ratings a better forecast matrix than regular polls.
But it is even worse for Obama.
It is well known that there are three categories of polls, those among "all adults" those of "registered voters" and polls of "likely voters". Republicans do better in each of these narrowing fields of voters, Meaning a poll of "Likely voters will favor Republicans far more than that of one of "registered voters" and even more so than of "likely voters". Why does this matter? The figures above of Obama's approval ratings are of "all adults". The chances are very good that the approval ratings above overestimate Obama's actual approval among the people who will actually vote.
I also do not see blacks turning out for Obama in the numbers they did in 2008, probably closer to the 11% that is usual rather than the 13% he received. I suspect that the Hispanic vote will increase marginally to compensate for this with the "non-white" vote, but even that would not help since Obama will receive a lesser share of the Hispanic vote than from blacks.
Now is entirely possible that this won't matter. That the election will be so close that it will all come down to those "swing states" and like Bush in 2000 Obama will win the electoral vote while losing the popular vote, boy won't that be fun? But my bet is that it will be a wave election and in the end Romney will handily.
But if you are feeling bad about the polls, just keep an eye on the white vote and white approval numbers, if they stay below 40% nationally Obama remains in trouble.