In 2008, one of the best posts in the time around the November election was by Ed from Gin and Tacos, a whip smart blog that I link here from time to time. It’s about the idea of constructed explanations, or what is created by the public and the media from events where there is inadequate data for a more objective explanation.
In the context of elections it’s very apt. Think about it this way. Modern society is feedback-driven, whether it’s a form asking you how your hotel experience was, or a text box that opens up when you ask to unsubscribe from an email list. It’s easier than ever to tell a business what you thought about whatever it is that they do. Far beyond the era of hotlines, it’s something you have to avoid these days.
So it seems odd that a ballot, despite being part of an immensely important process, has nothing to provide context to what is marked. Why did this person vote for Proposition 23 but not 25? Wouldn’t they want both? Don’t know. They might as well be cryptic runes from a thousand years ago.
What emerges then, is a guessing game about a huge, complex event. There is a ton of potential data to collect, but very little is; it remains in the mind of each individual voter. Exit polls are notoriously inaccurate and don’t take a representative sample. In any recent United States presidential election, you would have a pretty decent idea of what Ohio or Florida voters did – including important data like their key issues and what influenced their vote.
The hundred million plus who live in safe states? Not likely to meet a data collector. If you’re trying to create a large-scale political narrative, the map looks like a crappy cellular network. Key places are covered, but most is a black hole. When it comes to voters in states like California or Oklahoma, media explanations fall on stereotypes more than anything.
In the 21st century the only other major post-vote data source are online polls, which measure the most politically engaged slice of the electorate. Voters who keep to themselves are a question mark. When the numbers come in, the contours of the results may lie with them and the subtle, small reasons many of them showed up to vote, and what they ultimately voted for.
With the EU elections going on, and the US midterms approaching in a few months, narratives will be constructed well in advance, then paired with polling. If they line up well with the results, they are accepted as gospel. This is problematic, because there are many reasons a party wins or doesn’t win. Is the narrative that X Party won, or is it that they only won by that amount? In a context like the EU elections, where are supporters moving among the various parties? Did turnout bolster certain parties, and should it be considered high, low, or normal given the circumstances?
2012. Mitt Romney wasn’t a good communicator. It was a bad year to be a Republican. The Tea Party dragged the ticket down. Obama’s campaign was run very well. Or maybe just better than Romney’s. Or maybe they cocked things up and got lucky with a weak candidate. These are all estimations because you’re looking at numbers and assigning agency and motives to them. But just like the Man in the Moon isn’t a real face, just something that resembles a face, sometimes numbers resemble a narrative.
The advantage for the media is that it’s hard to call bullshit. And as any detour into cable news can show you, the narrative factory – the myth-making, if you will, goes beyond being a part of the business.
It is the business, now.