The Politics of Test-and-Learn

by Reihan Salam

Sasha Issenberg’s article on the roots of President Obama’s reelection victory is well worth reading:

Concern that the technical supremacy of Rove and his crew would ensure the Democrats’ future as a minority party drove consultants who usually competed with one another to collaborate on previously unimaginable research projects. Major donors like George Soros decided not to focus their funding on campaigns to win single elections, as they had in the hopes of beating Bush in 2004, but instead to seed institutions committed to learning how to run better campaigns. Liberals, generally in awe of the success that Republicans had during the 1980s and 1990s in building a think-tank and media infrastructure to disseminate conservative ideas, responded by building a vast left-wing campaign research culture through groups like the Analyst Institute (devoted to scientific experimentation), Catalist (a common voter-data resource), and the New Organizing Institute (improved field tactics). 

With an eager pool of academic collaborators in political science, behavioral psychology, and economics linking up with curious political operatives and hacks, the left has birthed an unexpected subculture. It now contains a full-fledged electioneering intelligentsia, focused on integrating large-scale survey research with randomized experimental methods to isolateparticular populations that can be moved by political contact.

“There is not much of a commitment to that type of research on the right,” says Daron Shaw, a University of Texas at Austin political scientist who worked on both of George W. Bush’s presidential campaigns. “There is no real understanding of the experimental stuff.”

If Republicans brought consumer data into politics during Bush’s re-election, Democrats are mastering the techniques that give campaigns the ability to understand what actually moves voters. As a result, Democrats are beginning to engage a wider set of questions about what exactly a campaign is capable of accomplishing in an election year: not just how to modify nonvoters’ behavior to get them to the polls, but what exactly can change someone’s mindoutside of the artificial confines of a focus group. [Emphasis added]

Back in May, after reading an earlier Issenberg account of the origins of the Analyst Institute, I expressed my concern that this data-driven approach would win President Obama the election. I also recall a Republican campaign operative for having derided the Obama campaign’s strategy of throwing dozens of messages at the wall and seeing what sticks, not realizing that this was part of a very deliberate, and very shrewd, test-and-learn strategy.

What Republicans need most is Jim Manzi, but my sense is that he has bigger fish to fry. 

The Agenda

NRO’s domestic-policy blog, by Reihan Salam.