Still, a man hears what he wants to hear
And disregards the rest. -- Paul Simon, The Boxer
I'm sorry that I didn't get this out yesterday but just as I was about to start on it, my Intarwebs went down. That's probably a good thing because all I could think of yesterday was that I was completely blindsided, had no idea how it happened and now we're completely doomed.
In the time since I think I discovered how I managed to be so completely wrong.
A few very wise pundits often say that the only real poll that matters is the one held on election day. All of the other ones are various different forms of guessing, some of them better than others.
What threw me (and a few people much better at prognosticating than your humble blogger) was a seeming disparity in the numbers coming out of the polls. The major polling organizations were asking the right questions but the cumulative answers didn't make sense, so we picked the answers we liked.
And that's where I screwed up.
All the way up until election day the polls were telling us that in the swing states, Obama had a slight lead. The national polls showed Romney up slightly and the national Demonrat/Republican split was about R+2 where in 2010 it was R+1 and in 2008 it was D+6.
The swing state polls were showing Obama and Romney neck and neck. We looked at that, looked at the changing split and decided that one of them had to be wrong. Given the huge amount of people turning out to see Romney and Ryan, it had to be the swing state polls that were wrong and on election day, Obama was going down by 5 or 6 points.
But the swing state polls weren't wrong.
In other states, like NC and TX and UT, Romney won those states by bigger margins than even Bush in 2004. In the swing states, Romney didn't turn out the numbers McCain did. Obama didn't turn out the voters he turned out in 2004 either but he eked out a slight lead, just like the polls said he would.
What the hell happened?
Well, a couple of things. None of them (save one) was a deal breaker. It took a combination of factors to get us to where we ended up.
1) Sandy. Even the national polls showed Romney by 5 or so before Sandy. After Sandy he started to slide a bit. It wasn't much but the pictures of Obama and NJ Governor Chris Christie and the gushing by Christie that made the SCOAMF look all presidential hurt. When the guy that's been one of your harshest critics says you're doing a good job it helps you. It was not enough to change the election all by itself, but it had an effect.
2) Sandy again. Not as a political event but as a Black Swan news event. Romney's strategy in the closing days of the election was about getting his message out, clear and strong and consistent. The storm pushed politics off of the front page and stopped people talking about the election. The steady surge hit a bump and instead of recovering and continuing up it just kind of sat there. Not for long. Again, not enough to change the election, but a bit.
3) Romney's strategy. In the final weeks running up to the election, Romney avoided any conversation that could be bad for him. He did almost no interviews and stuck to his stump speeches. Generally, this is wise. Many of us said that the election was Romney's to lose and doing something stupid would certainly do that. But the Obama campaign and the media were throwing bombs about Romney's fund raising after the storm, women in binders, 47%, rape, and a host of other attacks. Romney stuck to his message which is generally good, but didn't respond to any of the attacks, which is generally bad but better than responding badly, which would be worse. He played it safe right into the loser's column.
4) Romney himself. This is where we have hard numbers. Compared to 2008, Romney gained some ground and lost some ground. With evangelicals, Romney gained ground. Not a lot, but some. So the new meme that evangelicals didn't turn out for Romney is wrong. He gained ground with Catholics. Not as much as with evangelicals, but still a gain. Romney lost ground with women, hispanics and the 18-24 crowd (demographics that matter) and gays ( a demographic that isn't statistically significant) but none of them in number significantly large to make a difference in the election. Demographics for Obama were also down, mostly where we said they would be but not as large as we thought and not nearly as much in swing states as nation-wide. But the one demographic that was enough by itself to sink Romney was the generic Republican voter. Down by a significant percentage.
But wait! Turnout was up! Long lines at the polls!
Yep. The numbers for Johnston County tell an interesting tale. In Johnston County, NC the winner of the race for the NC Commissioner of Agriculture, Steve Troxler, received 50,615 votes. The three Republican county commissioners that were running un-opposed netted 50,417, 50,409, and 50,813.
Romney received 48,202.
Somewhere around 2000 Johnston County Republicans voted for local or state-level candidates and then did not vote for Romney.
That's 4% of your base, those R's that the polls indicated were there but weren't showing up in the Obama/Romney split. And it's not like they lied to the pollsters. These were certainly likely voters. They turned out, they waited in line. And they voted. But not for Mitt Romney. In a normal election cycle that's enough to beat you right there.
Why? That's a subject for another post. For now let's suffice to say that Romney was a better candidate than everyone hoped but not quite good enough.
-- But the final thing is something that was much more effective than the pundits thought it would be. --
5) The Obama campaign had a new tool they've been putting together since 2007 called Operation Narwhal. Here's how it worked.
Back in 2008, the 13 million people that signed up for Obama campaign alerts got generic, milquetoast, hopey-changey messages about The Lightbringer. Signing up asked for an email address and your zip code. Nothing odd about that, right? But since then, the Obama campaign has been cross-referencing "other databases" to tease out more specific information about what might motivate a voter or the opposite. And I don't know which databases. I suspect that I don't want to know.
This election the messages sent out to targeted swing states were targeted more specifically than ever before. A young, white, female college student would get a message saying how Mitt Romeny was going to take away her contraception. All young college students would be told how Romney was going to reverse the part of Obamacare that guaranteed their student loans. Gays would be told about how JugEars now supported gay marriage. African Americans would *not* see the messages about gay marriage but one more crafted to their interests. The same for the other demographic targets. If you were poor you were told how rich Romney is. If you were a professional you'd be told about how Obama saved Wall Street and the banks. If you were a auto industry worker you'd hear about how Obama saved the car industry and Romney would have let it go bankrupt. Those kind of voters don't go to rallies. They don't stand in the cold and the rain to see their candidate because they don't really have a candidate. They have an enemy.
Does that "revenge" comment Obama dropped in that speech six days before the election make more sense now? It was the signal for his army of mindless zombies to rise up and strike.
In effect, the Obama campaign turned out an army of single-issue voters in the swing states. All of them knowing almost nothing about any issues other than the one they were programmed to be concerned about. The Great Uniter divided his voters like they've never been divided before while at the same time giving them all the same target. The only thing they had in common was anger about part of Romney's ideas.
Of course, once they start to realize that Obama really doesn't care about their issues it should be entertaining. But then he won't be up for re-election again will he? Obama kamikazed his supporters into Romney. That's the kind of dick move that you can't use more than once but once was enough.
It was just enough to push him over the top.
So, where do we go from here? I've got a few ideas. I'll put together another post in a few days about that.