The Problem with Polls

Yesterday I went over Oklahoma's (atrocious) poll numbers. However, there are several reasons why polls may be unusually hard to get accurate this cycle. First the negative.


The Bradley Effect.
Named after Tom Bradley, former mayor of L.A. When he ran for Governor of California in 1982 polls had him ahead, but he ended up losing. The theory behind this is that when an African-American is running, people may not want to look like bigots to the cute young pollster asking them questions, so many of them lie. This idea got resurrected this cycle when Hillary came from behind to win New Hampshire.

Even though this theory is eye-catching, it is actually quite controversial among pollsters. They looked hard for a Bradley Effect in later races in various states, and could find no evidence of it. Today is not 1982, and many pollsters claim that the effect has diminished over time. Some even claim it never existed in the first place.

GOP Voter Suppression
When you take a bunch of people who either have no moral scruples in the first place, or feel that they are fighting on the side of God, and put them in a situation where it looks like they are losing, it isn't tough to imagine them doing some unsavory and/or downright illegal things. In other words, they will try to cheat. I'm not saying all Republicans are like this. However, it only takes a few. This is a party that destroyed America's anti-torture principles, which date back to the founding struggle of this country, simply for some imagined expedience. Don't think those same folks will blanch one bit at taking the vote away from some poor people.

However, the GOP is, if nothing else, a party that respects authority. Their best efforts are always directed top-down, and frankly they aren't going to waste them on what looks to be one of their safer states. I expect to see a lot of shenanigans in the next 3 weeks, but I expect them mostly to be directed at blue and swing states. That doesn't mean we won't get some rogues doing caging, but I don't expect it to have a significant effect. I hope I'm right.

Now for the positive:

New Registrations
There have been a huge surge in voter registrations all across the country this cycle, and Oklahoma is no exception. The vast majority of these new registrations are for Democrats.

Unlikely Voters
The polls at this point are all weighted to "likely voters". That means that if you aren't in a demographic group that has historically voted in Oklahoma in past elections, they won't count your answers in the polls as strongly. However, this is definitively not a normal election. People who haven't voted in ages are expected to come out. Since Republicans are generally reliable voters, this effect can only skew Democratic.

Reverse Bradley Effect
Pollsters have actually noticed that Barak Obama has a kind of Bradley Effect in reverse in any state that has a significant African American community. Oklahoma's is about 8%. We are also about 8% native-american and %7 hispanic, both communities that are trending overwhelmingly Obama's way this cycle.

Ground Game
The Democratic get-out-the-vote ground efforts have always been rather shabby compared to what our opponents had to offer. We are doing our best to change that in this state. We still aren't anywhere near where we should be, but we are getting better. This is going to be the big X-factor in this campaign.

This is also where you come in. Don't just sit on your butt this year and exepect your one vote to work its magic. If you love Oklahoma, get out and fight for it. Democratic HQ is at 3930 East 31 Street, and I'm but an email away.

Debate Tonight
The last debate is tonight. That means the last debate watch party at the Circle Cinema Quad this evening starting at 7. See you there.