“Big data” and politics. How did new technologies transform political environment in 2016?

It is hard to argue that in today’s world internet has become more than just another communication technology. We use it every day in order to catch up with news, to book our tickets, to communicate with friends, listen to music, and even earn a degree. As a result, internet with all its advantages and opportunities penetrated human life to its core. However, being such a powerful tool, internet can also be used as a dangerous weapon.

For instance, recently, the media reported that the unexpected results of “Brexit” and 2016 Presidential elections in the US were largely due to the actions of a marketing firm – “Cambridge Analytica”. These two campaigns allegedly relied on this firm to promote their political agendas by running sophisticated web-advertising and building a detailed database of potential voters. Another report claimed that during the last election campaign in the US, fake news related to the elections were more popular than most of the election-related factual news. Evidently, fake news stories were distributed via social media platforms. Finally, prior to November 8th, emails of the Democratic National Committee were leaked and published by Wikileaks, causing a major controversy about how hackers could possibly influence such major political events as presidential elections. How exactly does manipulation via advanced information technologies happen? Is Internet really becoming a major factor, influencing political processes around the world?

7

One of the popular news circulating around the Internet was that Trump called all Republicans “dumb” people; however, the news was completely fake

Often, political campaigns are quick to embrace new communication technologies. Back in 1960s, it was debated that the television was able to hugely influence the results of the United States presidential elections. In 2008 and 2012, Barack Obama’s presidential campaigns were famously technology-driven. They used social media platforms to spread political messages (big innovation for 2008!), spent millions for technological services; its IT-specialist built a number of applications to optimize management. In 2016, the campaign of Hillary Clinton relied on sophisticated technological approach – much of which was a continuation of Obama’s successful strategy. However, Clinton went even further: her tech team was described as “Silicon Valley Stars” – prosperous technicians who used their talents to optimize fundraising, agitation, and delivery of messages. Clinton’s tech team was also reported to use a database of voters “Narwhal” – a list of voters, which could be targeted by the campaign (basically those, who left their emails at sites of Obama and Clinton campaigns), matched with available data on every voter to get a detailed portrait of the targeted audience. And yet the usage of technologies by Trump and his team could have been more efficient, according to reports.

In short, Trump’s technological team (these hired “Cambridge Analytica” analysts) were reported to collect data (infamous “big data”) on almost every adult US citizen from various sources. This data was used to optimize marketing of the campaign. The company collected not only factual data on every citizen (age, gender, ethnic background, education, income, and similar), but also “psychographic” characteristics of each individual. As for the latter, each individual was assigned one or another “OCEAN”-profile (a 5-scale personality profile where openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism of each individual is measured) according to his or her consumption, lifestyle, political engagement and attitudes data. As for the data source, the company bought the data from several data inventories – like “DataTrust”.

How the data on attitudes and consumption was matched with each personality trail? Well, we do not know the precise procedure, but an example could be The Psychometrics Centre site developed by University of Cambridge, which uses likes from the Facebook personal pages to assess psychometric characteristics of each individual. For instance, the system gauges people who like music group “Queen” as more conservative and traditional, those who like sport teams and “Marvel” comics as more masculine, and those who like the movie “Big Lebowski” as more intelligent. Which is interesting, the more likes the system gets to evaluate the personality profile, the more the results of assessment resemble results of an undertaken psychometric test. Probably, Cambridge Analytica used a similar methodic to access personality profiles of individuals in their database (according to reports, co-founder of the campaign worked with inventor of method for “Psychometrics Centre”). Importantly, if the company used a similar method, they must have received incredibly precise profiles of each individual: CEO of the company claims to have “four to five thousands data points” on each individual in the United States.

picture1

The site of the “Psychometrics Centre” could estimate person’s personality traits, age, and gender based on his or her likes on Facebook.

2

3

4

The system gauges people who like music group “Queen” as more conservative and traditional, those who like sport teams and “Marvel” comics as more masculine, and those who like the movie “Harry Potter” as more intelligent.

Then, the company was able to identify the audience which should be targeted (for instance, conservative neurotic types who were still indecisive about candidates to choose in the next elections) and the message, which would perfectly resonate with such audience (for instance, gun rights advertising). Such method helped advertising to be more efficient for lower price in comparison with “blanket” advertising of other political campaigns, since only individuals which are most likely to response to the message of the campaign were targeted by Trump’s team. The company used different methods to deliver the message: online-advertisings, offline mails, and commercials in TV-programs most liked by the targeted audience. The message was adapted to each personality profile discovered in the preparatory research. Furthermore, the campaign also sent messages to would-be Hillary voters to overpersuade them from attending the elections. There, an important question arises. Were the US elections 2016 really “hacked”?

Is the result of US elections driven by more than usual personalized advertising or a number of other factors – Donald Trump’s celebrity status, enormous news coverage of his speeches, wide-spread anti-establishment sentiments among American voters? What we do know, however, that Trump’s team was not the only one to use sophisticated communication technologies in these elections. The main difference between technologies used by the two campaigns was that some messages were more personalized than others. We will never know if this was really an influential factor. There is no data of how well the personalized message of Trump’s campaign resonated with target audience in comparison with “blanket” advertising of Clinton’s campaign.

What we do know, however, that to resonate with certain audiences, the message should be believable. Despite any kind of marketing tools used, broad anti-establishment propaganda, protectionist policies, and anti-immigration emphasis – from peculiar messages to the campaign slogans – must have been persuasive by themselves. What we also know from the 2016 elections in US is that Internet could increase divide among people with different political attitudes rather than contribute to their mutual understanding. It seems that in social networks, such people consume different media products, look at different advertisings, and read different news. And political campaigns of 2016 only contributed to the differences in advertising.

What we also know, is that in a world dominated by the Internet and its products – the fast spreading of messages before their fact-checking, and enormous opportunities for surveillance – this is how any other political campaign in most of the countries will happen in the near future. It is obvious that there are many downsides of how politics is happening now.

However, Internet and “big data” also provide a number of opportunities for scientific research and even improvements in our life. For instance, Google was able to predict flu and dengue outbreaks relatively precisely using its own search data. Another project, “GDELT” monitors news from all over the world in 100 of languages to provide complex picture of what is happening in the world right now. Specifically, the project collects data on protests happening overall in the world, on humanitarian crises, responses to political reforms. It seems that there could be even more applications of such data. Isn’t that at least intriguing? What if in the near future necessary vaccines would be delivered to places where the outbreak is about to happen because there was a spike in search queries related to “headache”? What if the governments could optimize construction of roads, assigning more resources to repair of worst ones because there were more outraged tweets on quality of these roads? Despite shortcomings of the Internet and the big data political environment, there are major advantages that can be utilized to improve our lives in general.

Written by Zhamilya Mukasheva

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s