Earlier this year I wrote a piece on my fear that Big Data was going to change the way elections are managed. Briefly I argued that elections are going to start to use an unholy alliance of big data and behavioral targeting to socially engineer elections. True elections have in some since always been about organizing and motivating people to get to the polls, but the emerging digital landscape substantially alters the playing field, in a way that ought to make us concerned about whether or not this is healthy for a democracy. I won’t restate my whole argument here, you can read it yourself, and read what someone working in the industry had to say in response. Then you can also read Zeynep Tufekci’s piece on why we should “Beware of the Smart Campaign.”
While prior to the election there were a few articles talking about digital tools, micro-targeting and politics, post election cycle there has been a deluge of articles explaining the inner-workings of this new electoral battleground. I will have more thoughts on this later, a longer piece. But in the meantime I thought I would just give a run-down of all the pieces that have been written on this matter lately, a sort of weekend reading list on Big Data politics.
I think it is easy to argue that this is just more of the same, that we have always had this type of public manipulation in politics. But I also think this argument is wrong. It fails to recognize the degree of difference here that is producing a difference of kind, and more importantly how a refinement of behavioral targeting techniques coupled with big data is producing a world that is dangerous to democracy. Early on many argued that the digital network would enable more people to participate in a grass roots politics, but it is pretty clear now that the opposite force is just as powerful: using the digital network to engineer the public.
It’s all too easy to point out Big Bird, Binders Full of Women, and Bayonets, to talk about how the Romney campaign and the Obama campaign used Facebook or Youtube to get its message out, but to pay attention only to the level of content I think is a mistake. As McLuhan says the content is the juicy meat meant to distract the guard dog. What happened below the level of content here is the bigger story.
Pro Publica Primer. For a while Pro Publica has been following the story of Big Data and politics. Following the election they posted this article which is a good place to start on what was going on, highlighting what we know, and what we have yet to learn about how the Obama team leveraged data. The article talks about the all important “persuasion score” that the campaign used to decide who to target, and the “persuasion” phone call which is likened to a drug dose, “only effective for so long.” If you only read one piece this is the one I would pick.
Frontline: The Digital Factor in Election 2012. Produced prior to the election this Frontline video attempts to understand what types of data are out there and how campaigns are using it to target voters. Aristotle one of the companies involved in this data targeting claims to have 500 points of data on voters, and explains how they “exploit” this data. As the CEO of Aristotle says the campaigns aren’t worried about your privacy they are concerned about winning the election. This video is a good baseline for what is going on, but really it only scratches the surface about the dangers.
**Obama’s Data Crunchers: **This article at Time’s Swampland was one of the first I saw most election that gave an inside look at what the digital team for Obama was doing. Obviously they had access during the election, pre-wrote the article, and just held it until after the election was over. There is some interesting information here, like the analytics department was five times the size as 2008, the “chief scientist” was someone who had experience in maximizing supermarket sales, how efficiently they raised money, how they used safe states like California as test beds before rolling something out to a battleground state, and ran 66,000 simulations every night.
**Behavorial Targeting Brought to You by Academics: **This article at the NYTimes illustrates how a group of academics, dubbed the “dream team” helped Obama to carefully craft his message for maximum effect. The Obama team took advice from these behavioral scientists. Unfortunately the Obama team won’t release information on the details, and the academics signed a non-disclosure agreement. But the details the Times has are telling, for example how precisely to word a message in order to make a not-likely voter feel guilty and turn out to vote.
**Optimizing the Message. **A large part of the story here is not just that behavioral targeting or persuading voters is happening, but that the Obama campaign maximized efficiency, to get this done. In other words they relied on data to tell them how to most effectively and efficiently persuade voters and raise revenue. Mother Jones describes how the Obama campaign was able to spend less per add, and raise more per email than the Romney effort.
Romney Fail. One of the more telling stories to come out about this election cycle, was how far behind the Romney campaign was in terms of its digital strategy. Romney tried to have a real time update of people’s voting behaviors. This didn’t go so well. Tellingly it was similar to the system that the Obama team realized 4 years ago didn’t work. But the larger picture here is how the rhetoric post-election has been how envious the Republicans are of the Democrats advantage and how they will look to close the gap. For those interested in the ORCA fail you should check out the blog post at Ace of Spades HQ, which was by someone who tried to volunteer, and was the first one I saw post election. Also Ars Technica (which in general has had some of the better inside looks at tech and politics) covered the Romney fail.
The Technical Side: For those interested in the more technical side of the Obama advantage GigaOM described some of the details of the operation, but the ArsTechnica piece the best rundown of the specifics of Obama’s tech team. This article describes in detail project Narwhal and how the Obama campaign built the infrastructure from the ground up in order to maintain a dominate advantage. The article discusses everything from design philosophy to which tools they used.
The Republican Disadvantage. One of the consistent narratives post election has been disappointment within the Republican campaign about the inability to compete. Republicans are being openly envious about the Obama campaign, in many cases vowing not to get beat in 2016.
**The Minds behind the Tech. **Alexis Madrigal’s piece at The Atlantic profiles Harper Reed, the chief technology officer for the Obama campaign. While I am generally skeptical of any history which focuses on the contributions of one individual, (if it wasn’t Reed it would be someone else), Madrigal’s piece demonstrates something more, the way that a culture of technology influenced the Obama campaign. The argument here is one that is pretty consistent across post election coverage: Nerds are now key to winning elections.
Will the Republicans Catch Up?. As Democrats celebrate, Republicans work hard to close this gap. Nancy Scola writes a piece that should be comforting to most Democrats on how the Republicans will have a hard time countering the data advantage. I am not sure I agree, in fact nothing about the technology here is particularly aligned with party politics, if anything I would say that engineering public opinion in this matter lends itself to a right leaning ideology.
The Long History. If you are interested in the long history of this read The Victory Lab by Sasha Issenberg, in which he discusses how campaigns have increasingly leveraged social science techniques for competitive advantage.
The Democratic History. If you want a sense of how the Democrats and particularly Obama developed an advantage read Daniel Kreiss’s book, Taking Our Country Back, in which he traces the history of the digital campaign from Dean to Obama. In particular the book demonstrates how it was the Dean loss that lead to building infrastructure for the Obama win.