• by Michael Cornfield , Op-Ed Contributor, October 21, 2014

Recent columns and comments in this space reflect competitions among political consultants and media outlets for election spending dollars. Use branded channels, Sean Cunningham and, indirectly, Michael Beach argued. Broadcast TV reigns supreme, Abby Auerbach responded. Don’t forget geo-targeted online video, said Roy Moskowitz.

A casual reader following this lively and factual tussle might be left with the impression that television ads make a big difference in elections. Allow me to put on my political scientist beanie and push back on that notion.

Institutional and socioeconomic conditions, parties, candidates, field organizations, and big donors decide elections, roughly in that descending order. Television has occasional and marginal effects after these factors have spoken, as it were.

Institutional conditions refers to the powers of incumbency and the life cycle of the presidency: year-six elections often go against the party of the president. The impact of the economy on voters and the roll of the dice that is big news also matters. Ebola is today the riveting story many thought Obamacare would be a year ago, and there are two weeks left for the virus scare to recede or explode. 

Parties are weaker in the U.S. than elsewhere and at their historic peaks, but they still recruit candidates, articulate national themes, and allocate crucial resources, including the insider sense of who has a chance to win. Candidates embody their party’s themes and their own locally crafted messages with varying levels of skill and energy. In the field, staffers and volunteers communicate peer-to-peer with voters, and political influence is very much a retail business. Big donors, when they spend unmatched by the other side in particular contests, can have a big impact. 

These factors largely determine the size and composition of the electorate, the people who turn out to vote, and with that, the outcomes of most elections. Television advertising ranks at the top of the second tier, a stimulus tool that campaigners can use to expand and contract the electorate by slivers. It’s meaningful in so far as it plays well off the big factors. But for the most part,TV’s power gets understandably over-rated by its purveyors, with an assist from the press, which enjoys reviewing television ads and giving a few of them a second jolt of exposure.

As for radio, display graphics, direct mail, web advertising, debates, and press coverage, they take a back seat to TV. Smart campaigns, of course, make sure to coordinate across media. And there is an X factor in the form of scandals, as seen to a degree in this year’s Montana Senate race, where appointed incumbent John Walsh, already in trouble, had to quit because of a plagiarism discovery in his past.

When it comes to determining how to spend campaign dollars, I agree wholeheartedly with the professionals cited above that big data analytics are changing the game. All resource allocation choices, from how to raise more money to where to put ads and who should knock on which doors, can now be subjected to testing and modeling. And they should be, because data-driven optimization achieves more for the buck. And they are being data-driven, increasingly, up and down the ballot. Someday soon, data will crowd television at the top of the second tier of determining factors.

In this regard there is some news: the RNC boosted its digital efforts significantly in the 2014 cycle. More on that next time.