More electoral analysis

I cover polls a lot on this blog. This is because it is the only refutation to wingnut assertions.
Currently, Obama is leading McCain on Gallup by 48 to 42 in the national polls. Obama has led for almost two months now, proving that McCain's negative ads are only speaking to his base and not to the general public. Independent voters, the group that is supposed to make all of the difference to McCain's chances, still break evenly among the two candidates (Obama 43%, McCain 41%). Usually, effective negative campaigns drag down the victim's "favourables". In this case however, it is McCain's favourables that have dropped for the past month, with a current virtual tie (Obama 58%, McCain 56%).

On Rasmussen, Obama has been between 43 and 47% since early June. He's been between 44 and 45% for the past two weeks. In spite of the attacks, Obama's support seems stable among likely voters. McCain has been between 40 and 44% since early June, and in that same range for the past two weeks. There was a tie for two days in early August.

Both Rasmussen and fivethirtyeight.com give Obama the win in electoral vote projections.

One commenter on fivethirtyeight.com broke the electoral vote situation down this way:

What we are talking about, though, is that Obama has a ton of winning maps, while McCain has very few. Kerry+IA/CO/NM is a win, and he's up in all of those. Same with IA/CO/NV. Now, those fairly likely scenarios aside, if he wins Kerry+IA (which he should) then a win pretty much anywhere else, like OH, VA, NC, FL, IN puts him in extremely good shape. That's why the 50-state strategy is such a better idea than the campaigns Gore and Kerry ran, and why people do think it's going to be different this time. It might not be, but the odds are a lot better at least.

And that is essence of this election, and what makes it so very different from 2000 and 2004. The Republicans would love to make this a battle over Florida and Ohio, but the Obama campaign is too smart for that. It's fairly certain that the 2 or 3 state "battleground" approach is the route Hillary would taken as well.
Republican support is so weak that states that once were solid red are now within reach for he Democrats. And the idiotic "battleground" approach has left several states with little in the way of organisation. Those mired in the old paradigm see the "50 state strategy" as a waste of effort, but Obama is doing the work that the Party needs done by organising these putatively "red" states. Even if none of them break for Obama, the Party has fantastic data for House and Senate races, as well as future Presidential races. However, the odds are that a certain number of these states will "flip", and then Obama wins. McCain has only one or two "paths" to victory.

Meanwhile, the McCain campaign is targeting only 19 states. Among those 19 are very safe states for Obama: Maine, New Jersey, Minnesota, and Washington. And with the abbreviated Republican primary season, the GOP was cheated out of a Party's best chance of updating and expanding their lists of contacts. They are left with the old lists from 2004 in many cases. Thus, we are seeing the McCain campaign rely more on ads than usual.

The GOP won in 2000 and 2004 by conducting massive registration drives and bringing in millions of new voters. This year, however, the numbers netted by the GOP in nearly every state fell far below those of the Democrats. In Indiana alone, the Democrats registered more voters during the primary than voted for Kerry in the previous election.

Conventional wisdom says that 5% is the maximum lead that can be overcome by a ground campaign. Right now, the polling shows Obama with a 5 point lead in Wisconsin, and an eight point lead in Pennsylvania. New Mexico is projected to go to Obama by 5.5%, with Iowa about the same. Ohio and Virginia are still evenly split. McCain leads in Florida by a little over 3%. Obama is projected to win by a bit over 2% in Colorado.

The issue is that a lot of these polls use the "likely voter" model to skew their data. The Obama campaign has registered hundreds of thousands of new voters who will not show up as "likely voters". We are left to believe that none of these people will vote if we accept the likely voter screen on the polls. Another issue is the weighting of the polls in party identification.

The Republicans seem stuck in a rut, obsessing over Obama's inability to gain a decisive advantage. They seem to forget that McCain has failed to score any advantage. More than that:

The primary "underperforming" that's going on in this election is John McCain. Where Bush consistently polled in the 45-47% range, on average, all year in 2004 in swing states, McCain polled all summer in the 40-42% range, on average. In an electoral system where +/- 3% in national polling gives rise to +/- 350+ EV landslides, that's a huge ground that needs to be made up. If you want to see the difference, go to electoral-vote.com and click on states like IA, PA, WI, MN, MI, OH, VA. McCain is generally polling 3-5% less than Bush across the board.

With numbers like this, relying on past scenarios for a Republican win is dangerous.
Other polls show McCain's support among evangelicals is 9% weaker than Bush's 2004 numbers. And this time around, evangelicals are less of a monolithic bloc. In Texas, McCain is polling 11% weaker than Bush in 2004. Republican states such as Nevada, Montana, North and South Dakota, and North Carolina are all within 5%. McCain has immense defensive work to do that Bush never did. And of those five, he only has ground campaigns targeted for Nevada and North Carolina. That is 9 electoral votes he has given up on, or at least decided to rest on his laurels in.

McCain is also relying on a national campaign with national issues. Obama is targeting states based on local issues. The objective is to win states, after all.

The jingoism the Bush campaign employed is now yielding diminishing returns. Iraq is regarded as a mistake; whether we "win" or not is unimportant in the big picture as far as using it as an issue to rally around. And, sure, McCain is a war hero. Everyone knows he's a war hero already, though. The people likely to vote based on that are already on his side, and he's still behind Obama. He's not going to become any more of a war hero in the coming months. Likewise, McCain has hit a ceiling on his "favourables". Everyone already knows him. Everyone likely to like him already does. He can only go down or stay steady.

Lastly, his ad campaign to tear down Obama has not only met with no perceptible results in the polls, it says nothing about him. Putting down Obama does not make McCain look any better, and these ads show no contrast between McCain and Obama. At best, they make McCain the default alternative to Obama, which is very weak tea.