Republicans Try to Bridge the Big Data Gap
At hackathons and in the cubicle-less offices of Silicon Valley firms, Republican recruiters are sidling up to programmers and cautiously sounding them out to see how they feel about the NSA spying program and the Obamacare rollout. They’re headhunting for Para Bellum Labs, the GOP’s attempt to answer the computed power of President Obama’s Organizing for America. The organization is meant to take advantage of a start-up spirit, nurturing small, agile political projects that can scale up to national races.
The new initiative takes its name from the Latin motto, Si vis pacem, para bellum—If you seek peace, prepare for war. But it might have been more apropos to say, Si vis vitam, para bellum—prepare for war, if you want to live. In 2012, the GOP’s get out the vote project, Project Orca, blew up in the Romney campaigns face. On Election Day, the webpage went down and stayed down, unprepared to handle the load of field volunteers checking voter turnout and following up with tardy voters.
Meanwhile, the Obama team’s outreach rolled along smoothly, messages tailored to individual voters through Project Narwhal, a massive data mining initiative similar to the algorithms and A/B testing that online advertisers use to maximize the chance that you’ll click through on an ad. The Obama tech team didn’t just work on what message to send, but what medium to send it through as well. At the peak of the campaign, the Obama team was carefully looking for undervalued, highly-targeted media buys. Their big data machine meant that the Democrats were spending, on average, $72 less per TV spot than the Romney campaign.
The Para Bellum team may be able to help the GOP staunch the bleeding, but they’re unlikely to deliver the same kind of advantage as Project Narwhal did for the Democrats. The Obama campaign benefited from being the first mover in the political data movement. By using big data, they were able to segment voters and media markets into smaller demographic slices and to identify opportunities that had been systematically mispriced.
The Democrats’ playbook resembled the sabermetric approach of Billy Beane, the Oakland A’s coach profiled in Moneyball. By taking a finer-grained approach to a market (of persuadable voters, in the Democrats case, of free agents and draftees, in Beane’s), the quants were able to steal a march on their opponents.
But Billy Beane’s Oakland A’s only benefited from his Moneyball strategies for a few seasons before the rest of the league caught up, and his advantage didn’t last long enough to reach the World Series. His data-intensive approach only helped as long as players were systematically mispriced. Once the other general managers wised up, the market for players settled into a new equilibrium and the richer teams were able to outbid their rivals for better players.
The Para Bellum initiative can blunt the advantage of the Organizing for America team, but it will be difficult to outstrip them. As both parties get better at leveraging big data, they’ll find it easier and easier to spend their resources efficiently on the undecideds who are most likely to become supporters, and the supporters who are the most likely to become campaign contributors.
Big data and algorithms mean that future political campaigns will be conducted in a world of better slicing and better pricing. But better pricing, in this case, means more efficient pricing, not cheaper pricing. If Para Bellum Labs succeeds in its mission, the Republicans won’t be leaving money on the table for Democratic programmers anymore. But instead of reaping an advantage of their own, the GOP coders will just have shifted the electoral market into a new equilibrium.
Just as the segmentation of the electoral college allows both parties to better target their outreach, to the only voters whose votes could make a difference, the data crunchers on both sides will let the parties fight fiercely over an ever smaller set of voters.