I stopped reading David Brooks a while ago, when he decided it was his mission not to provide thoughtful commentary on current events but instead to produce one column after another that read like the Cliff Notes for a high school "character education" class. But his column today is worthy of comment, because it gets at something that I've been thinking about for a while, raises an important issue about how campaigns are conducted today, and still manages to be utterly wrong.
As we all know, campaigns have become increasingly sophisticated at targeting voters. Back in the stone age when I worked on campaigns, they had little more information to work with than what was on your voter registration card. They knew your age, your gender, your address, and a couple of other data points, and with some creativity they could infer other things about you (for instance, one campaign I worked on sent a mailer about gay issues to any household with two people over 30 of the same gender living together). But today it's possible, by aggregating multiple sources of information, for campaigns to have hundreds of data points on any given voter. Here's why Brooks thinks this is a problem:
That's because the data-driven style of politics is built on a questionable philosophy and a set of dubious assumptions. Data-driven politics is built on a philosophy you might call Impersonalism. This is the belief that what matters in politics is the reaction of populations and not the idiosyncratic judgment, moral character or creativity of individuals.
Data-driven politics assumes that demography is destiny, that the electorate is not best seen as a group of free-thinking citizens but as a collection of demographic slices. This method assumes that mobilization is more important than persuasion; that it is more important to target your likely supporters than to try to reframe debates or persuade the whole country.
I'm not sure when he thinks it was that campaigns appealed to the "idiosyncratic judgment, moral character, or creativity" of each individual voter. Candidates have always thought about different groups and how to persuade and mobilize them, it's just that now they can be more precise about it. The danger, I suppose, is that campaigns will micro-target with different messages, in effect telling different people what they want to hear. Which is no different from what candidates have always done. There wasn't some halcyon time decades ago when a candidate would give exactly the same speech at the chamber of commerce, the union hall, and the ladies' sewing circle.
But are they really pandering in dramatically different ways to different groups? Are candidates telling some people, in a quiet email or text message, that they'll be pro-choice while telling others they'll be pro-life? Of course not. They're talking about different issues to different people, in the hopes of convincing them they share their priorities and concerrns, just like they always have. Let's check back in with Brooks:
Candidates who have overrelied on these techniques have been hurt by them. One victim was Mitt Romney, who ran for president not as himself, but as a thin slice of himself. Another victim was President Obama. His 2012 campaign was legendary from an analytic point of view, and, of course, it was victorious. But it lacked a policy agenda and produced no mandate. Without a compelling agenda, the administration has projected an image of reactive drift and lost public confidence.
That's why Mitt Romney lost—because he was too reliant on microtargeting? And not, say, because he was a largely inept candidate leading a party with an unpopular agenda? And the problems Barack Obama has had in his second term have been because he "lacked a policy agenda and produced no mandate," presumably also because of microtargeting? Give me a break. Obama had an ample policy agenda. The places it hasn't been enacted aren't because he didn't offer enough of a universal, sweeping vision. It's because there's this thing called "Congress" which has been something of a problem.
But Brooks sees the slaves of microtargeting failing everywhere:
The other victims include the Democratic senators in red states. Winning in a state that the other party dominates is a personal enterprise. It requires an ineffable individual connection with voters. It requires an idiosyncratic approach to issues. By eclipsing individual quirks with generic messages, the data-driven style deprives outnumbered candidates of precisely what they need to survive. The Democrat Alison Lundergan Grimes could have made a real run at Senator Mitch McConnell in Kentucky if she'd been a little more creative.
Oh, so that's why Alison Lundergan Grimes is probably going to fall short by three or four points tonight—her "generic messages" and "data-driven style" have deprived her of the "ineffable individual connection to voters" she might otherwise have built up. It couldn't be that she's running as a Democrat in a state that Barack Obama lost by 23 points, could it? Nah. If only she had a more "idiosyncratic approach to issues," that wouldn't have mattered, and instead of overperforming what a Democrat should be expected to achieve in Kentucky, she'd be winning by 10 points.
I'm all for imagination and grand visions, and of course it's possible to get mired in the weeds of precisely honed messages and forget to tell voters an inspiring story of why they should elect you. But Brooks is taking utterly predictable political outcomes, like Democrats losing in deeply conservative states, and blaming them on developments in campaign technology that they have nothing to do with.
Here's an actual problem I see with the way campaigns use all this data. They drill deep down into every voter, knowing an incredible amount about them, and what's the result of all this knowledge? What's the extraordinary technique of persuasion that emerges at the other side? It's...a flyer. Or an email telling them to get out and vote. Or an ad, aimed at a carefully selected group of viewers, saying, "My opponent: bad for our state, bad for America." In other words, it's the same old crap as when campaigns had only the barest sense of who voters were.
If you want to assign blame for that, you're going to have to go a lot farther back.