Nate Silver has a blog about some interesting data, a poll that puts Barack Obama in a tie with Mitt Romney in Arizona.
The difference between that poll and other polls conducted in Arizona? This one conducted polls in English and Spanish both. According to this poll, Barack Obama looks ready to take more than 70% of the Latino vote.
There are some caveats. First, Silver says that the polling firm in question is awesome, that it does good work, and that it's methodology is sound, but even good polling firms will sometimes have polls that are outliers. Even great pollsters will sometimes have polls that are way out of the ballpark, which is why you have to look at every single poll in a state.
Those other polls in Arizona, by the way, do polling only in English.
Second, the fundamentals in Arizona, the way it tends to vote, makes it pretty likely that the state will be in the R column on election day. He implies that while the republicans might still win the state, it will be far more of a squeaker than state republicans will be comfortable with, which is one of the reasons that right wingers in Arizona have been having a total meltdown for the last few years about brown people.
Silver does point out that Spanish-Speaking populations tend to vote heavily Democrat, and that those populations are consistently underpolled. This might account for as much as 1-5 percentage points in polling in those states with large Latino populations.
What I'm about to say is very important. DO NOT do what the freepers do and go around adding 5 points to every poll. That way lies madness, and that's not how math works. This only applies in states with large Spanish-Speaking populations like Florida, California, Arizona, Texas, New Mexico, and states with smaller Spanish-Speaking populations like Virginia.
The freepers have this whole conspiracy theory process where they go around un-skewing polls, and create this fantasy world where Romney's going to win in a landslide just like Reagan did that one time.
Don't be like them, because they're wrong. Let's not be the other face of the internet crazy coin, let's be reasonable, responsible, users of math.
In order to develop a mathematical picture of the voting situation, we'll need more polling than we have. I've written it before, and I'll write it again, one or two polls are not the sum of all electoral data. You have to look at every poll. You have to compare the national polls to the state polls, and the state polls to the national polls. Even Rasmussen can tell us things about the direction of the race, though their polls aren't very good at telling us where the race stands. That is, Rasmussen is at least consistent with its bias, so we can learn about the race by comparing it with other polls.
By looking at all of the data together, you get a pretty clear picture. When it comes to measuring the Spanish-Speaking vote, we're going to need more polling from houses that do bilingual polling before we can draw a very clear picture. Still, here's what Nate Silver found when he adjusted for Spanish Speakers:
He then says that in past elections, Spanish Speakers haven't been that likely to cast ballots, and that he thinks his model is still more accurate.
I did some of my own research to compare states with large Spanish-Speaking populations in 2008. Here's the polling and actual results.
Nevada:
1 Barack Obama:53.2% John McCain:42.4% Margin:10.8
2 Barack Obama:51% John McCain:47% Margin:4
3 Barack Obama:50.7% John McCain:42.9% Margin:7.8
Actual Result:
Barack Obama:55.2% John McCain:42.7% Margin:12.5
New Mexico:
1 Barack Obama:52% John McCain:45% Margin:7
2 Barack Obama:51% John McCain:43% Margin:8
3 Barack Obama:58% John McCain:41% Margin:17
Actual Result:
Barack Obama:56.9% John McCain:41.8% Margin: 15.1
Colorado
1 Barack Obama:51 Mitt Romney:47 Margin:4
2 Barack Obama:52 Mitt Romney:45 Margin:7
3 Barack Obama:54 Mitt Romney:44 Margin:10
Actual Result:
Barack Obama 53.7 John McCain 44.7 Margin: 10
Florida
1 Barack Obama:49 John McCain:50 Margin:-1
2 Barack Obama:50 John McCain:47 Margin:3
3 Barack Obama:49 John McCain:48 Margin:1
Actual Result:
Barack Obama:51 John McCain:48.2 Margin:2.8
We notice a few things here. First, the most accurate pollster here is PPP, the guys currently in that awesome DailyKos SEIU partnership. I don't know if they conduct Bilingual polls. (Anyone?)
Second, most of the pollsters are off. The larger the Spanish-Speaking population, the further off they are. In Florida, it's a few percentage points, in other states with large Spanish-Speaking populations, the numbers tend to skew farther right.
So it's pretty clear that polling which doesn't take into account the Spanish speaking population is going to skew to the right. Republicans know this, which is why they're pushing through as many voter-ID laws as they can. They're attempting to bring back Jim Crow because they've seen these numbers, and they know the score.
Once more with that warning. Don't try to un-skew the polls yourself, like the freepers do. Instead, we should encourage pollsters, including PPP if they don't already, to start doing bilingual polling in states with large Latino populations, just to make sure we have the most accurate picture. Furthermore, we should work on GOTV efforts with Latinos in places like Colorado, Nevada, New Mexico, Virginia, Arizona, Florida, and even Texas.
That's all for this Election, but looking at 2014, and 2016, there's a lot that can be done to grow the Democratic party.
This data means that over time we can turn Texas and Arizona into battleground states and put Florida into a lean-D column, if we focus our energies.
If we can break the guaranteed lock the republicans have on Texas, and make them fight for every single vote in the Lone Star State, we can draw funding and attention away from other seats. Even if we don't win Texas (turning Texas blue is possible, but will take some time) we can make winning Texas a Pyrrhic victory for the Republicans. If Texas starts grey on the board rather than Red, the Republicans will end up with a much, much harder time getting to the white house.
That's where we need to focus our attention after the election. It's in the southwest and Florida where we can really grow the Democratic party. And how sweet would it be if we made sure that the Democrats who came from Florida and the Southwest were more progressives like Floridan Alan Grayson and Arizona's Raul Grijalva?
We can make that happen.
But first, we need to make this election happen.
GOTV.