Saturday, July 26, 2008

Polls and Predictions

Bro KY,

I notice the pundits are trying out a new narrative that goes something like "Obama have had a fantastic week and McCain could not look more pitiful by comparison" - what else can you say about the one about showing up in a german restaurant in Ohio to counter Obama's speech before 200,000 people in Berlin - "but that is not translating into the polls".

Like you I check the RCP Latest Polls. (After all, great minds think alike) And I must say that the trends are clearly in Obama favour and an Obama sweep is on the way this November.

First of all, although the daily national tracking polls tend to fluctuate between a tie and +6 for Obama, the average seems to be steady around +4 for Obama. For McCain's numbers, I see a floor of around 39; a ceiling of around 44 and average of about 42. For Obama, the floor seems to be around 44 ceiling of 50 and average of 46-47.

If you look at the battleground polls, where it all matters, the situation on the ground is even better for Obama. He leads solidly in most states that Democrats won in '04 except MI where he has a small lead despite not campaining during the primaries. He is also placing many previously red states in competition - he looks solid in IO, NM, NV, OH - good chances in CO, MO & FL and has decent chances in VA and IN. Obama wins when he takes the first 4, I can see a sweep when he takes the next 3 and possibly a landslide if he takes the next two. That is despite a 4% lead nationally.

I also see other factors that boosts his margins. This time the Democrats has a more aggressive, better funded and better organized ground strategy.

On my street, I have already been approached 3 times by Democratic Party canvassers trying to register voters - unfortunately as a non-American I am not eligible - 2x by Obama volunteers giving out pamphlets. Taking a leaf from your book, next time I see one I will ask them how their efforts are going. As a result, voter registration is apparently at record level in many states. With statistics showing that 35% of African-Americans and 45% of hispanics not registered to vote, this is a huge opportunity for Obama especially in swing states in the Southwest. I see voter registration as giving at least +1% or +2%% advantage to Obama and given that on average 50% of people vote in the general elections, the impact can potentally be 5% in some states.

That same ground organization will super-charge the Democrat's get-out-the-vote efforts. Obama's strategy against HRC relied extensively on his experience as a grass-root organizer and he is applying that same technique to build a nationwide network of volunteers. As I mentioned earlier, the high costs of gasoline places the ability to identify voters and getting them to the voting stations an important factor. For the less-enthusiastic supporters who may decide to save money instead of driving to vote (as is common for semi-rural communities in the Southwest, Midwest and South) - or poorer voters who does not have a car, this will have a major impact on voter turn-out. The advantage for Obama is that by having fought all those primary countests and having led the voter registration drive in all these states, he has a head start in being organized locally, of knowing where his voters live and the organization to put them into play. Given that historically the Democrats have been sadly lagging compared to the GOP's network on the Christian-right, this should add another 1% to 2% to Obama's numbers.

Finally, voter enthusiasm amongst Obama supporters are 4-times higher than for his opponent. This has the effect of higher voter turn-out for his supporters and having a larger pool of potential volunteers to leverage upon. Given that the Christian-right appears to be less than enthusiastic for McCain, the total impact of voter enthusiasm gives Obama yet another 1% advantage.

You add it up, and that is an in-built 5% advantage (possibly more) even if for a the improbable 50:50 situation. Hence, I am convinced a +4% poll advantage is potentially worth even more in practice.

Obama is now working to lift his poll upwards from the 50% ceiling. As you pointed out, he has the courage to confront his challenges and embrace perceived shortcomings. He is focused on lessening the 'fears' and concerns about his background and his capability. I saw on a poll that 58% of people has a +ive impression of Obama meaning he can potentially improve his poll ceiling on the merits i.e. by persuading them of his ability or policies. (Conversely the 35% negatives are probably no longer open-minded to persuasion). The trip to Mideast and Europe should help with this group of voters. Another opportunity are the almost 80% of Americans who disagree with the way the country is going; that is potentially another 30% in-the-play if they can be persuaded that McCain is McSame. Even if they do not vote Obama they will be less likely to vote for the otherside.

Polls tend to be backward looking rather than forward looking, I believe it will take a month or so for any sets of polling to present a clearer picture. Lets keep checking the RCP.

2 comments:

View from NY said...

Since posting, I discovered an intruiging site called - www.fivethirtyeight.com - which does analysis of polls using statistical, historical accuracy and demographic factors. The person running it based it on systems that analyses baseball scores! Enjoy the geek fest!

View from HK said...

the lastest from RCP is that it is a TIE at Florida.

As we are having a high expectation of Obama winning the election, we must not forget the reality of politics at the end is performance, he has to deliver.

Our favorite, Ma Ying-Jeou, after winning the election, has failed to live up with the expectation and he has since seen his favorable rating dropped more than
30%. He is probably the only politician who won a landslide without enjoying a honeymoon.