Tuesday, October 21, 2008

The Polls Are Misleading, Not Dishonest - How Is That?

Oh, a few wild polls from known biased sources are suspect, perhaps sampling and phrasing questions so as to suit their hopes. But, most polls like Gallup, Rasmussen and Zogby for instance, are honest. But, they routinely underestimate the Republican showing in national elections. Just in the last one, Kerry was an average of 6 points ahead on the day of the election. He lost. Republicans always strongly out perform what the polls suggest. Why is that?

I was thinking about that, and a question occurred to me: Do pollsters over-sample in cities? City people are more likely to be too occupied to vote. And, rural populations are more likely to be animated by fears generated by national media coverage.

This year, Obama and his enthusiasts in the media have generated extraordinary fervor. And, there is a corollary enthusiasm among conservatives and in rural areas, which has gotten an extra kick from Sarah Palin (a lot of eggheads are dubious, but boy did McCain make the right pick).

My first prediction is that there will be a record turnout on both sides.

My second prediction is that like other Republicans, McCain will outperform what is reflected in the polls. Those red states where Obama leads or trails slightly? They’ll go to McCain. He’ll hold Ohio (if fraud is sorted out), North Carolina, Florida, Missouri, Colorado, and Nevada. If Obama doesn’t hold Minnesota, Iowa, Wisconsin, Michigan, Pennsylvania, and New Hampshire, he’s in definite trouble, and McCain may show close in any or all of them.

But assuming they both hold all those states, it may come down to Virginia, which has been trending decidedly purple, toward blue. Outside of Alexandria and the other DC suburbs, McCain will win big with a strong turnout. But, that suburban and heavily populated area has been turning strongly blue. Obama needs to generate a big turnout. But remember that in DC and its suburbs, McCain is hardly a stranger and known as a maverick (yes, I know the word has become overused, but that doesn’t change the meaning. Anyway, watch Virginia.

Possible surprises? First, I’ll go with New Mexico, which polls have not even been showing as a tossup, but solid blue. But demographically, it raises some questions. I think most of the so-called “Bradley effect” is accountable not to racism, but to misleading pollsters for the sake of perception. In any case, to whatever extent there is a Bradley effect; New Mexico is well-defined for it. It isn’t a big electoral number, but if the election is close…? But if Virginia is the decider, New Mexico may not make a difference. Secondly for surprises, I’d look at Pennsylvania. The polls have Obama in a comfortable lead and I still give it to him. But, those bitter Western Pennsylvanian clingers to guns and religion and “Joe the Plumber” types might make a statement (Murtha’s racist comment might help too). Pennsylvania may be close, but also most vulnerable among the blue states. And it’s elect, swingorally BIG. Basically, Obama doesn’t win without Pennsylvania.

The last big question is what will develop in disputes and challenges to the vote. This year, it seems exceptionally likely that Democrats will charge voter suppression AND Republicans will charge vote fraud. Maybe we’ll be spared by a decisive verdict. But it looks like it could be very close, and if it is, hold on to your hats. It might make Florida in 2000 look like a tea party.

5 comments:

Anonymous said...

You're the most optimistic McCain supporter I know. Even the more conservative polls are showing Obama with leads in almost every battleground state.

McCain supporters need to get to the polls, and vote quickly if any of that is to happen. Exit polls of early voters show Obama with a 58 to 40 lead among early voters.

Anonymous said...

Lee:

I beieve that. In fact, I'm surprised it isn't even a larger gap. The Obama camp knows that it has been riding a wave of sentiment which will likely only move away from them. And, they have been and still are pushing early voting, before there is opportunity to lose enthusiam or change minds.

Hey, I believe it's possible for it to be decisive, either way. But, I think it's more likely that it will be tight. In fact, my projections about the electoral map suggest tat Virginia could throw the election. I think the safest bet, which I still wouldn't make, is that whoever wins Virginia will win the election. I don't think most people understand how serious it could easily be if Obama wins. I'll be writing about that, shortly.

I hope I get responses quickly about the new date for Ecclesia. By the way, I haven't been feeling well. I fell on Friday night and hurt my back. It still was hampering me, yesterday. It still hurts but is better, today.

Anonymous said...

This is certainly not good news for McCain supporters. With Iowa and New Mexico already in the "solidly blue" column, and with the Kerry states from 2000 virtually locked up, Colorado is the only state Obama needs to clinch it.

http://www.guardian.co.uk/world/2008/oct/22/uselections2008-johnmccain-colorado-barackobama

Anonymous said...

McCain can win Colorado

I've just had a hunch for a long time that New Mexico could be a surprise.

If everything stacks up with each winning what they must, and that includes McCain coming from modest deficits in polls to win Missouri, Florida, and North Carolina as I think he can, Virginia is the big enchilada that I can't call. In that case, whoever wins Virginia wins the election.

Of course, if Obama wins Florida, Missouri, or Ohio (where McCain shows a small lead, it's over for McCain. Likewise if McCain wins Pennsylvania or MI, WI, MN, it's over for Obama.

Anonymous said...

Thought you might enjoy looking at this. It is a composite, showing virtually all of the polls, and does a cross analysis of them. Lots and lots of numbers.

http://www.fivethirtyeight.com/