Don't believe the hype about Big Data in 2016
Political campaigns are known for being a step behind when it comes to innovation.
If you've ever watched a bunch of cookie-cutter campaign ads and wondered why they look like they were produced by a couple of college students who just learned how to use Final Cut Pro — and not, say, Madison Avenue execs capable of creating heart-wrenching 30-second film masterpieces — you're not alone. But if there's one place where campaigns are supposed to be utilizing the latest in techno-whizbangery, it's in their exploitation of Big Data. With the information tools now at their disposal, they can microtarget voters down to the depths of their very souls.
This vision of highly sophisticated, algorithm-driven campaigns isn't completely inaccurate. But it's missing an important piece, the piece that is supposed to make it all seem either exciting or sinister, depending on your perspective. That missing piece is persuasion. In short, campaigns know how to find you in ways they never did before. But once they've found you, they haven't gotten any better at winning you over.
In a front-page story on Monday, The Washington Post reported on the data operation that supposedly may have propelled Ted Cruz to his current position near the front of the Republican pack:
Cruz has largely built his program out of his Houston headquarters, where a team of statisticians and behavioral psychologists who subscribe to the burgeoning practice of “psychographic targeting” built their own version of a Myers-Briggs personality test. The test data is supplemented by recent issue surveys, and together they are used to categorize supporters, who then receive specially tailored messages, phone calls, and visits. Micro-targeting of voters has been around for well over a decade, but the Cruz operation has deepened the intensity of the effort and the use of psychological data. [The Washington Post]
There's no real way to demarcate the arrival of microtargeting, because for decades campaigns have targeted people with whatever information they could. Thirty years ago they had little apart from registration data, but that could tell you a voter's gender, age, neighborhood (which gives a good approximation of their income), and how often they voted. As time went on, the companies specializing in list management added more and more data points, like magazine subscriptions, to their databanks, giving a finer-grained portrait of each voter.
And today there are more data points than ever, since we leave such an informational trail behind us as we go about our lives. So it's true that if you're a Republican voter in Iowa, Ted Cruz's campaign knows many things about you, enough to come up with some kind of profile that may or may not actually describe your personality. The problem, though, isn't that they know less about you than they think — though that could be the case, since they're dealing in probabilities, i.e. someone with your profile is probably the kind of person who would care about these issues and have these opinions. The problem is what happens after they've identified you.
That's because once they've employed all that 21st century technology to decide who you are and what you believe, the tools at their disposal to change your vote are pretty primitive. They might have a volunteer call you on the phone and read from script 16A instead of 16B or 16C. They might send you a flyer — one tailored to you, of course, but still a flyer in your mailbox that you'll probably glance at before tossing in the trash. Or they might even put an ad in your Facebook feed, after they've carefully mapped your social network. And you'll pay as much attention to it as you do to the rest of the ads you see on Facebook. They may have fancy new tools of analysis, but their tools of persuasion are still pretty mundane.
And, of course, those tools are dependent in large part on voter ignorance to have an impact, just like most campaign communication. If the Cruz campaign has discovered that you have a concealed weapon permit and sends you an email saying, "Did you know that Ted Cruz is a fervent supporter of the Second Amendment?", the only way you're going to say, "Wow, I think Ted Cruz is my guy!" is if you're unaware that every Republican candidate is a fervent supporter of the Second Amendment.
Maybe you are unaware, in which case that email might do its job. But this is a difference in degree, not in kind, from what candidates have always done: tailored their message to whomever they're talking to. The candidate will emphasize his military hawkishness when he's at the VFW hall, or talk about health care on a visit to a hospital, or blurt out "Who let the dogs out? Hoo! Hoo!" upon finding himself amidst a group of African-American teens. Those appeals are the same kind of probability assessment: These are the type of people who probably care about this sort of thing, so that's what I'll talk to them about. But it doesn't mean they'll change their minds.
Fortunately, persuasion always has its limits. A well-run data operation can help a campaign keep in touch with its voters efficiently, and figure out what kinds of emails are better at bringing in donations and getting people to volunteer, and help the campaign spread its message as widely as possible. But microtargeting can't turn Ted Cruz into a likeable human, or make Ben Carson knowledgeable about policy, or give Jeb Bush charisma. In the end, the candidacy depends on the man or woman whose face is on all those flyers and web pages and ads.
So there isn't that much to be afraid of, even if it creeps you out that Cruz is all up in your Facebook.