Community Development Field

Election Polling, Big Data, and Movement Building

There is a data geek Internet flame war going on between Sam Wang of the Princeton Election Consortium and Nate Silver of over 2014 election projections. Evidence of the confrontation can be found here, here, and here. My research interests (such as they are) tend more toward demographic analysis, and I am far from […]

There is a data geek Internet flame war going on between Sam Wang of the Princeton Election Consortium and Nate Silver of over 2014 election projections. Evidence of the confrontation can be found here, here, and here. My research interests (such as they are) tend more toward demographic analysis, and I am far from an election-polling expert, but I find this confrontation fascinating.

On the surface, the rivalry seems to be about methodology. Wang aggregates only polls, while Silver adjusts polling data with other factors (e.g., economic trends). Wang is open source. Silver’s process is proprietary.

But this fight is about more than methodology; it is a fight for legitimacy in a shifting social media space, which says something larger about how technology is aiding and abetting some pretty big socio-cultural shifts right now.
The first shift is the increased prevalence of big data to drive decisions big and small. This is Google using your Internet search history to narrowcast advertising directly to you. This is NSA computers sorting through phone records to tag potential terrorists. This is HMOs making decisions about protocols for treatment options. This is the increased push for metric-driven social policy.

Statistical and social science-based decision making is nothing new, but computers have made the collection and processing of the data much more efficient. Over and above increased technological capacity, there is an increased belief in quantitative data and data-driven analytics. Within popular culture—TV cop shows for example—it’s how the gritty detective has given ground to the geeky methodologist.

In this vein, it seems fitting to me that Silver first came to national prominence as a baseball analyst and that is currently hosted by ESPN. As popularized by the film Moneyball, new-fangled statistical analysis has reshaped how baseball teams are constructed and how people talk and think about baseball. But baseball and TV shows are not the only places where this is happening. Advertising, Wall Street, social policy, education, etc. are all increasingly the province of data geeks (mostly male, mostly white, but with a smattering of Asian Americans).

The second big shift is around the ways in which technology has disrupted traditional lines of communication and authority. We’re very near the point where practically anyone with a laptop and an Internet connection can do what Wang and Silver do, and not just in terms of creating a website and running a social media campaign, but also in terms of analyzing the data. The Internet means that, more than ever before, anyone can become a pop star, an artist, a pundit (a blogger!), and a data analyst. Anyone can be a political prognosticator, but perception about quality and relevance of analysis still matters—perhaps even more now. In an environment where there are fewer filters and anyone can say anything about everything, the ability to generate buzz and have name recognition to drive clicks is a precious commodity. In this brave new maximum access social media world, Silver is good at what he does, and seems to understand the value of building, promoting, and defending a personal brand (the bet with Joe Scarborough during the 2012 elections is an example).

But, putting these factors together—the ascendency of big data thinking, the punctured equilibrium of pundit space—and you get an online skirmish between two people who have far more in common than their conflict would indicate. Wang publicly wonders if this wasn’t a situation better solved in private, over a couple of beers; but that’s not how things are done anymore. The conflict is a form of performance art in service of online content generation (a performance that Wang participates in with such enthusiasm and dedication that I question his sincerity in really wanting to solve it privately). It’s micro cast drama for Internet political junkies, a virtual Roman arena where spreadsheets and tweets are the weapons of choice.

It’s also a competition where we can pretend that the scoreboard is real-life election results.

Who cares whether most of us scoreboard watchers misuse and misunderstand (sometimes willfully) the underlying math? If someone who says a coin flip is 50-50 is not proven wrong when the coin comes up heads, and if someone says a coin flip is 60-40 is not likewise proven right when the coin comes up tails. But the underlying stats—boring things like standard deviations and sample sizes—don’t matter to the general public. We want to see competition. We want a winner and a loser. Intensifying political scoreboard watching in this way—making political discourse more about data analytics, more about the horse race and less about the issues—is dangerous when it’s pushed through a media to a public that doesn’t take the time to explain or understand the stats. It sets up a dynamic where, like sports fans, we consume the stats passively while the real game is played on a field that we can’t touch.

Meanwhile, outside of pundit space, where the real candidate and issue campaigns are run, data analytics is also reshaping the ground game of politics in the trenches. Like in online advertising, political campaigns use increasingly sophisticated databases and other technological tools to target mailings, phone calls, door knocking, events, social media, etc. to maximize “touches,” to fundraise and to get their bases out to vote. Like with our emerging social policy regime, tasks are more directly tied to specific, measurable outcomes. Message and medium are tested; everything is more narrowcast, more parsed and targeted and geared toward short-term results. In politics this often results in pandering to multiple bases separately yet simultaneously. It means messaging is more segregated and knee-jerk, more around emotions (like anger and fear) that more consistently inspire immediate responses. It puts us on a path towards an even more efficient version of the politics that we already have.

So, herein lies our challenge. Against this backdrop, how do we build compelling campaigns—and not necessarily campaigns in the traditional single-issue/candidate sense of the word—that have heart and vision? How do we use, but not become controlled, by data? How do we have an awareness of the urgency of the short-term but keep our eyes on the long-term prize? How do we keep bending the arc toward justice while we’re increasingly focused on the ground right in front of our feet?

In this once-again crazy election season, as we fixate on polls and election results and start handicapping the 2016 elections, I’d like to say that now—as much as ever—we need big, ambitious, values-based, long-term movement building.

(Photo credit, Flickr user Lynn Friedman, CC BY-NC)

Related Articles