2016 Annenberg-Oxford Media Policy Summer Institute alumnus Arthur D. Soto-Vásquez discusses the use of data analytics in political campaigns as well as key concerns about privacy and deliberative democracy stemming from its growth. The following post is based on a presentation given at the 2016 Annenberg-Oxford Media Policy Institute
On the evening of February 1st, following one of the longest electoral prologues in American history, Senator Ted Cruz of Texas was declared the victor of the 2016 United States Republican Iowa Caucus. While only securing slightly over a quarter of the vote, amongst a field of 12 candidates it was enough to gain a plurality. Cruz’s victory came as a bit of a shock to some as polls indicated Donald Trump was poised to win. It appears superior data analytics, and subsequent superior field organization, helped Cruz win the day (Hamburger 2016).
In late 2015, The Washington Post reported that the Cruz campaign organized a, “team of statisticians and behavioral psychologists who subscribe to the burgeoning practice of ‘psychographic targeting’ [and] built their own version of a Myers-Briggs personality test. The test data is supplemented by recent issue surveys, and together they are used to categorize supporters, who then receive specially tailored messages, phone calls and visits.” While the practice of micro-targeting voters is not new, the Cruz team’s efforts mark a deepened intensity of such practices.
The era of data analytics and computing has been heralded as a new age of knowledge. It is touted as an all-purpose tool to solve current and future issues. While industry and online commerce trace the clicks and purchases of online users to better market and sell products, political campaigns are finding data tools critical for winning close elections. Using data and predictive voter models, campaigns are able to identify factors from a voter’s past political and purchasing history to communicate a specific, tailored message (Kreiss 2012). These sophisticated messages are not only used for voter outreach, but also get out the vote (GOTV) mobilization and fundraising
For example, The Washington Post notes the Cruz campaign was able to identify at least 90 voters whose primary political concern was the statewide ban on fireworks. In addition, using psychological data, the Cruz campaign was also able to broadly identify two thematic objections to the ban, “fun-loving” and “libertarian.” Fun-lovers objected to the ban because they viewed fireworks as an essential component of family and community celebrations such at the 4th of July. Liberations philosophically objected to the government banning their sale. Using this data, the Cruz campaign was able to target unique messages to both groups.
Similarly, many other campaigns have utilized sophisticated data sets and predictive voter models, including the Obama for President campaigns in 2008 and 2012 (Kreiss 2012). In addition, a private cottage industry of firms has emerged to offer their services to campaigns willing to pay. American political data firms such as the Messina Group have also been involved in European politics, including the Remain campaign in the UK and the campaign for Rajoy in Spain (Mucha, Negre 2016, Behr 2016). The Cruz campaigns’ FEC filings emphasize their heavy reliance on digital data collection, analytics, and communication. As of their most recent FEC filing date, the Cruz campaign has paid Cambridge Analytics over $4.9 million since October of 2015. Cambridge Analytics, specializes in using,
“data modeling and psychographic profiling to grow audiences, identify key influencers, and connect with people in ways that move them to action. Our unique data sets and unparalleled modeling techniques help organizations across America build better relationships with their target audience across all media platforms”
In addition, the Cruz campaign has also paid over a $1 million for a voter list from Targeted Victory, another data firm. Targeted Victory advertises themselves as “audience specific, screen agnostic.” Overall expenditures to Targeted Victory total more than $2.5 million. This level of spending is also bipartisan, according to FEC filings through April, the Clinton and Sanders campaigns have spent over $40 million on data analytics.
This rise of sophisticated data analytics presents three major concerns and policy issues. First, there are significant privacy concerns regarding marrying commercial and political data to each other and making predictive behavioral models based on that data. Data is not always perfect, and the misapplication of it may leave voters (or entire communities) outside of the political process. Second, even with the accurate application of data, there are troubling implications for deliberative democracy. Instead of campaigns speaking to the broadest political audiences, data allows for campaigns to find the voters most sympathetic and persuadable to their message. In addition, campaigns with access to the most financial and technical knowledge resources are better positioned to utilize data analytics to their advantage. A final concern surrounds the redlining potential of political micro-targeting on minority communities. When political micro-targeting is based on past voting history, historically disenfranchised communities of color are less likely to be targeted by future campaigns. When campaigns, bound by limited resources, are faced with the difficult choice to either reach out to new potential voters or mobilize likely supporters – they will usually choose to make sure their likely voters get out and vote.
Of these concerns, privacy issues are especially potent. Political campaigns are utilizing voter data combined with data from other sources, including commercial databases, as well as user provided and extracted data. Campaigns can extract data not only through user provided information such as age, gender and location but also through browser enabled technologies. For example, campaign websites typically will track which website a user came from (google, Facebook, a blog) to review their advertising strategies.
As such, these current practices generate serious questions about privacy such as, should campaigns be allowed to use commercial data to reach out to voters? Many campaigns have privacy and user agreements, however many are out of sight and hard to find. In addition, opting out of data collection equates to opting out of use. Thus another significant policy question emerges, should campaigns be required to provide an “opt out” choice to site visitors?
“your mobile device’s unique ID number, your mobile device’s geographic location while the app is actively running, your computer’s IP address, technical information about your computer or mobile device (such as type of device, web browser or operating system), your preferences and settings (time zone, language, privacy preferences, product preferences, etc.), the URL of the last web page you visited before coming to one of our sites, the buttons, controls and ads you clicked on (if any), how long you used our website or app and which services and features you used, and the online or offline status of Cruz Crew.”
The campaign also notes they reserve the right to share a voter’s data with service providers, ideologically or politically aligned organizations and analytics companies.
Political leaders often have divergent opinions regarding surveillance policy at massive scale, such as the data monitoring done by the National Security Agency. Cruz, who has been critical in the past of the widespread data monitoring and tracking of American citizens by the National Security Agency, has a campaign apparatus which collects much more personal data about its supporters and potential supporters. Should we doubt potential leader’s commitment to privacy when their campaigns engage is similar invasive techniques?
“Data is the new oil!” So how can researchers understand this developing phenomena? The data regime is a rich area of study, and here are some of the questions guiding my future work. First, how can researchers conceptualize data in new ways? Can we think of data as digital labor and/or social currency? Second, what does this mean for privacy in general? Are we past the point of no return? Finally, how does data about a community change its identity? What does this mean for growing minority communities as they politically “wake up?”
Behr, R. (2016, July 5). How Remain Failed: The Inside Story of a Doomed Campaign. The Guardian, UK. http://www.theguardian.com/politics/2016/jul/05/how-remain-failed-inside-story-doomed-campaign
Hamburger, T. (2015, December 13). Cruz campaign credits psychological data and analytics for its rising success. The Washington Post, USA. https://www.washingtonpost.com/politics/cruz-campaign-credits-psychological-data-and-analytics-for-its-rising-success/2015/12/13/4cb0baf8-9dc5-11e5-bce4-708fe33e3288_story.html
Karpf, D. (2010). Online Political Mobilization from the Advocacy Group’s Perspective: Looking Beyond Clicktivism. Policy & Internet, 2(4), 7-41. doi:10.2202/1944-2866.1098
Kreiss, D. (2012). Yes We Can (Profile You) A Brief Primer on Campaigns and Political Data. The Stanford Law Review, 64, 70-74.
Kreiss, D. (2015). Digital campaigning. In S. Coleman & D. Freelon (Eds.), Handbook of digital politics (pp. 118-135). Cheltenham, UK: Edward Elgar.
Mucha, M. Negre, J. (2016, July 3). La Gurú de San Francisco que hizo Ganar las Glecciones a Mariano Rajoy. El Mundo, Spain.
Arthur D. Soto-Vásquez is a proud native of El Chuco Town – otherwise known as El Paso, Texas. His academic, professional, and civic endeavors focus on the political socialization of Latino populations in the United States. Arthur is currently a Doctoral student at American University School of Communication in Washington, DC. His research focuses on the social and racial integration of U.S. Latinos into American democracy through political communication efforts by official and non-official actors mediated through digital technologies. Politically, Arthur has worked on the communications team for the Center for Public Policy Priorities. During the 84th Legislative session in Austin, Texas, Arthur helped organize opposition to SB 1819, a repeal of the Texas DREAM Act and was successful. He has also served on numerous progressive political campaigns including serving as the Deputy Campaign Director of the Hector H. Lopez for Mayor of El Paso in 2013. Civically, Arthur has served on several nonprofit boards, including Texas Student Media and The National Hispanic Institute. Arthur also began his primary civic engagement with the National Hispanic Institute eight years ago and has served at numerous leadership experiences every year since, including one-term as the President of the Collegiate Leadership Network in 2012. Arthur received his M.A. in Media Studies from the Department of Radio-Television-Film at the University of Texas at Austin in 2015 and his B.A. in Political Science from St. Edward’s University in 2012.