Every four years, the Pew Research Center, a vital resource for the kind of work we do around here, releases its splendid Political Typology study, in which it slices up the American electorate into a series of subgroups (libertarians and liberals, but also postmoderns and disaffecteds, etc.) differentiated by their ideological beliefs and viewed through their lens of their demographic composition, e.g., some groups are more devoutly religious or older or whiter than others, etc. Studies of this kind can be illuminating and entertaining. Yet for political professionals hoping to not just to understand the electorate but to influence the behavior of its individual members, slicing up the electorate into subgroups isn’t actually very useful. Think of the electorate as a constellation, in which each star represents an individual voter. We can draw circles around particular clusters of stars, and some of these clusters might really be pretty distinct from the other clusters. But the more stars we try to capture in our circles, the more we will find that the clusters get blurry around the edges; they bleed into each other, to the point where the boundaries start to seem pretty arbitrary. And more to the point, clusters that tell us about one dimension, like ideological views or how enthusiastic you are about spamming your friends with political propaganda on Facebook (not to sound too judgmental), might not give us useful information about another dimension. A grandmother living in Anaheim really wants to protect Social Security while her granddaughter in Modesto believes that the elderly should be forced to wander an irradiated wasteland in search of water and food. Yet they might both be reliable GOP voters, and they might be susceptible to the same kind of guilt-trip tactics when it comes to voting in off-year elections, e.g., some young person with a clipboard comes by and points out that everyone else in the neighborhood voted. So for one action (making an ideological appeal to solicit a donation) you might have to make a different pitch to the grandmother and the granddaughter. But for another action (get them to vote), you’d actually be making the same pitch.
All of this is to say that telling a political professional that they ought to focus on NASCAR dads is actually worse than useless, because it confuses the issue. Every intervention initiated by a campaign has to be considered not in terms of its generalized impact on ill-defined yet rigid subgroups (“our campaign is really going after waitress moms, and when we run ads bashing Nancy Pelosi, they like us more”). Rather, each intervention has to be assessed in terms of its specific, measurable effects on individual behavior, with a close eye on bang for buck. The “rules” we identify in one campaign (“door-knocking is always better than Facebook ads”) won’t necessarily apply in another; indeed, they don’t necessarily apply to all voters in a given election, so you might be leaving money on the table by treating all voters as though they adhere to such a generalized rule. Successful campaigns will be the campaigns that fully embrace a strategy of resegmenting the voter base around specific behaviors that will contribute to victory (raising donations, seeing to it that sympathetic eligible voters are registered, persuading voters to move towards your candidate, persuading voters who are already persuaded to move towards your candidate to persuade others in their social network, getting your voters to the polls, etc.), and keeping close tabs on cost-benefit tradeoffs. (I write a bit about this subject in my new Reuters Opinion column.)
This might sound a bit nihilistic, particularly coming from someone who thinks (or tries to think) deeply about the policies we as a country should embrace, and the goals that we as conservatives should be trying to achieve. But campaigns are not where these larger questions get settled. Rather, campaigns are where the coalitions we have, armed with the policies they’ve embraced, compete to motivate (or demotivate) committed voters, and to persuade the small universe of voters who participate in the process and yet are willing to keep an open mind about which team to support. This is a zero-sum competition, and it pays to be data-driven and willing to discard what doesn’t work for what does instead of sentimental and stubborn.