The Power of 50 Million Data Points
What role does Data actually played in 2016 presidential election?
By Ben Niu
November 2016, the results of the 2016 United States presidential election shocked the world. Donald Trump, the most unexpected candidates, winning a larger advantage of an electoral college ballot and has been elected as 45th president of the United States. Later on, the western news media report that the data analysis and cooperation with Cambridge Analytica was the secret weapon for Donald Trump. However, the Cambridge Analytica rejected that “ CA did not interfere with any Americans’ right to vote in the presidential election. What we did was just to increase the number of voters in the election.” Moreover, the New York Times said that the Cambridge Analytica use 50 million users’ personal data to play a decisive role in the election through a psychological and statistical model to target marketing voters. However, some research paper said, the effect of CA has brought to the election was only a few percents in all marketing strategies. What is the actual role does data analysis played in the 2016 presidential election?
Part 1 How and why CA and Trump did it.
In the fourth episode of the House of Cards, President Underwood encountered a strong challenge from Republican candidate William Conway during his second term.
Conway is handsome and his family is a military family. He is very popular among voters. However, the secret weapon that really makes him win the hearts of the people is not his personal charm, but his relationship with the search engine “Pollyhop”. With the help of Pollyhop, Conway’s campaign team digs in the voter data, analyzes their daily behaviors, and clearly understands what voters like, what they want, and what they want to hear, so that campaign and can be accurately placed. Conway’s positive image is everywhere, every policy and every sentence speak of the voters’ hearts, and the election of democracy has produced votes.
The big news that broke out in the real world has turned the story of “House of Cards” into a forward-looking political allegory.
During the 2016 US presidential election, there was a company similar to Pollyhop. It used an exclusive psychographic model to analyze user behavior which helped the Trump campaign team customize everything from a political standpoint to a campaign slogan, and accurately delivered digital advertising, which to some extent reversed Trump’s poor image and helped him win the election.
The company is from the UK and is called Cambridge Analytica. Facebook announced that it had closed the interface between CA and CA’s parent company, Strategic Communications Lab (SCL), and all the accounts of the two institutions on the entire platform.
Soon, the United States’ “New York Times” and the British “Guardian”, two authoritative media outlets published a document, exposed the scandal of Cambridge Analytica (hereinafter referred to as CA) illegally stealing data. Evidence from the two media sources through CA’s internal sources indicates that the company violated the Facebook Open Platform Agreement and obtained data on a total of more than 50 million Facebook users, most of whom were real American voters.
Although 50 million is really not much compared to Yahoo, which leaked the account information of 3 billion users. However, the data this time included the user’s name, gender, address, living status, political opinions, social relationships, etc. In the CA incident, those user features are sufficient to complete a portrait of a user. It is enough to rank among the largest data breaches in US history.
When a company has more than 50 million real people’s personal data and the ability to analyze it, its power is enough to subvert a society — if it is willing to.
In the past, only one company in the world has had this ability: Facebook. And now, there is one more company: CA. Facebook doesn’t have this will, but CA has it.
Here is an example of one voter professor Carroll ‘s psychological and statistical model.
This precise voter analysis can often play a decisive role in striving for swing tickets. In theory, the Trump campaign team will push ads to topics such as Trump’s trade protection policy, job creation, and the US middle-class crisis, or push it to voters who are close to him. “Hillary is trapped in the mail scandal” and “suspected with many men”.
It is conceivable that Professor Carroll, who has a very high moral sentiment score and is concerned about national debt, will vote for whom. Although he is “not likely to support the Republican Party,” these pushes are enough to sever his trust or fantasies about Trump’s political enemies.
It is no secret that the company uses psych statistical models to help campaign teams analyze voter preferences. Its official website clearly states: “We offer a range of services from predicting voter behavior to spot-based advertising,” and more explicit: “We will provide data to help users get votes and win elections.”
Through this model with personalized target marketing, CA could manipulate social and decision making bias of the cognitive bias and worldview backfire effect on voters.
Part 2 Really or not?
The story above exhibits how and why CA and Trump effect election through data. Meanwhile, on the other hand, Hillary relies heavily on the legacy of the first “social media president” Obama. She has a Democratic address list, works with leading-edge big data analysts from BlueLabs, and is backed by Google and DreamWorks. When Trump announced the appointment of CA in June 2016, the Hillary camp sneered. Will foreigners in custom suits know Americans better than Americans? (BlueLabs is a data analytics and technology company founded in Washington in 2013. Founders include senior members of the Obama data team. DreamWorks is an American film producer and publisher.) This also become the counterpoint of why Trump’s Data team did not have a significant effect on all marketing strategies. Because Hillary has stronger data analysis team and marketing channel.
Although the precise data on how much effects do Facebook advertising helped Trump is difficult to find. We can still find some side facts to prove it is true or not.
Pew Research Center is a nonpartisan fact tank that informs the public about the issues, attitudes, and trends shaping the world. My Expolotary Data Analysis will perform based on the data analysis provided by the Pew Research Center.
Facebook advertising was the major marketing channel of CA mentioned by news report. The Pew research center data report, the voter’s age demographics, shows 13% of voters aged 18–29, 30% of voters aged 30–49, 29% of voters aged 50–64, 27% of voters aged 65+, 33% of voters’ income is above 75K + and 37% of voters are college graduates. However, the Facebook demographics show a very different user’s age demographic from Pew Research center.
- 62% of online Seniors aged 65+ are on Facebook and 72% are between age 50–64.
- 88% of online users of age 18–29 are on Facebook, 84% of those 30–49.
- 82% of college graduates are on Facebook.
- 72% of online users of income more than $75K are on Facebook.
The difference may not directly explain the relationship between Facebook advertising effects and voters’ decision on presidential candidates. But the correlation score between two factors is small and not strong enough to say the Facebook advertising has a significant effect on those major demographics.
On another hand, although the psych statistical model with personalized target marketing could help CA and Trump manipulate social and decision making bias of the cognitive bias and worldview backfire effect on voters, he still has to, continually build his reputation as a “good president work for American.” Otherwise, he will be impeached. At the same time, in the era of this information explosion, a large amount of content touches people’s nerves every day. Perhaps, some kind of freshness will condense the attention of the moment, but when new things appear, people’s attention will shift. Over time, the eyeball attraction will be gradually forgotten.
In this regard, as early as 1964, Canadian scholar McLuhan once asserted that the greatest significance of the media to human society is not the specific information it carries as a carrier, but the human perception it brings as a “prolongation of the human body”. The world, the change in the way of grasping the world, and the impact on human social activities brought about by such changes are the “influence” brought about by the evolution of the media. The cognitive bias and backfire effects also change with times and depending on what major media told us what is the ‘truth’.
Reference
Ben Niu.(2019). Explore the change bought by Internet-based media through presidential election history of the United States
Burton, G. (2010). Media and society: Critical perspectives. Retrieved from https://ebookcentral.proquest.com/lib/ucb/reader.action?docID=771412
Jones, S., & Park, D. (2018). New Media & Society 20(1). New Media & Society, 20(1), 3–7. doi:10.1177/1461444817747062
Rosenberg, M. (2018, March 17). How Trump Consultants Exploited the Facebook Data of Millions. The New York Times. Retrieved March 15, 2019, from https://www.nytimes.com/2018/03/17/us/politics/cambridge-analytica-trump-campaign.html
