This paper questions if democratic elections are still even possible at all. The goal of this paper is to explain which methods Cambridge Analytica used to try to influence the 2016 presidential election. In that course, it aims to answer the question if these practices are a threat to democratic elections. The company in question was called Cambridge Analytica, a data analysis and political consulting company which used user data from Facebook and other social media services to manipulate voters' behaviour.
It is the 8th November 2016, day of the 58th presidential election in the United States of America. The republican Donald Trump and his vice president candidate Mike Pence are competing against the democrat Hillary Clinton and her vice president candidate Tim Kaine. According to surveys and polls prior to the election, Clinton has higher chance of winning and is expected to become the 45th president of the United States. A big surprise was revealed to the world that day when, for the fourth time in American history, a presidential candidate was elected without actually having the majority of citizens' votes. Donald Trump won the election and took office as the 45th president of the United States on January 20, 2017.
But what if this election was possibly manipulated? What if a single company had the power and the possibilities to completely analyse, predict and influence voters' behaviour? This might sound like an idea from a science fiction movie, but it is reality and it happened without the world even realising it.
After dealing with the case of Cambridge Analytica, one might think that electoral manipulation using data seems to be alarmingly easy in the digital age.
Table of Contents
1 Introduction
2 What Cambridge Analytica Was and What They Did
2.1 The Company´s History
2.2 Key Actors in the Trump Campaign
2.2.1 Alexander Nix
2.2.2 Aleksandr Kogan
2.2.3 Christopher Wylie
3 Cambridge Analytica´s Methodology
3.1 Microtargeting
3.2 Psychographics and Demographics
3.3 The OCEAN-Model (Big Five Personality Traits)
3.4 How Cambridge Analytica Practically Applied their Methodology in the Trump Campaign
4 A Threat to Democracy?
4.1 Definition Democracy
4.2 Discussing the Initial Question
5 Conclusion
6 Bibliography
7 Appendix
7.1 List of figures and videos
1 Introduction
It is the 8th November 2016, day of the 58th presidential election in the United States of America. The republican Donald Trump and his vice president candidate Mike Pence are competing against the democrat Hillary Clinton and her vice president candidate Tim Kaine. According to surveys and polls prior to the election, Clinton has higher chance of winning and is expected to become the 45th president of the United States. A big surprise was revealed to the world that day when, for the fourth time in American history, a presidential candidate was elected without actually having the majority of citizens´ votes. Donald Trump won the election and took office as the 45th president of the United States on January 20, 2017.
But what if this election was possibly manipulated? What if a single company had the power and the possibilities to completely analyse, predict and influence voters´ beha- viour? This might sound like an idea from a science fiction movie, but it is reality and it happened without the world even realising it. This being said, the company in ques- tion was called Cambridge Analytica, a data analysis and political consulting com- pany which used user data from Facebook and other social media services to manip- ulate voters´ behaviour.
After dealing with the case of Cambridge Analytica, one might think that electoral ma- nipulation using data seems to be alarmingly easy in the digital age. This leads to the question if democratic elections are still even possible at all. The goal of this paper is to explain which methods Cambridge Analytica used to try to influence the 2016 pres - idential election. In that course, it aims to answer the question if these practices are a threat to democratic elections.
2 What Cambridge Analytica Was and What They Did
2.1 The Company´s History
Cambridge Analytica was started as a subsidiary of the British data analysis firm SCL (Strategic Communication Laboratories) in 20131. At the time, SCL described them- selves as a company who specialised in “psychological warfare” and “influencing the behaviour of voters”. Cambridge Analytica specifically, was founded to manage SCL ´s operations in the United States and to engage in US-politics2. In order to test Cam- bridge Analytica´s operability, the firm engaged in the 2013 Virginia governor election by working for the republican Ken Cuccinelli as a first pilot project3. Even though Cuc- cinelli did not win, the American billionaire Robert Mercer saw the firm´s potential and decided to invest 15 million US-Dollars in Cambridge Analytica. With Mercer´s finan- cial support, the company was officially founded in 2014 by British business man Al - exander Nix, former Trump adviser Steve Bannon and SCL CEO Nigel Oakes45. Since then Cambridge Analytica managed election campaigns for multiple clients in more than 50 different countries. e.g. Afghanistan, Colombia, India, Libya and many more6. This being said, the company is most famous for their involvement in the 2016 US presidential election campaign which led to the “Facebook Cambridge Analytica Data Scandal”:
In March 2018, multiple different renowned newspapers such as the Guardian or the New York Times wrote headlines about Cambridge Analytica after whistleblower Christopher Wylie, a former employee at the firm, had claimed that Cambridge Ana- lytica harvested the data of up to 87 million Facebook users without their permission in order to use this data to create psychological profiles of US citizens and to manipu - late the presidential election in Trump´s favour 78. In addition to that Brittany Kaiser, another whistleblower and ex-employee of the company, claimed that Cambridge Analytica was also involved in the 2016 Brexit referendum by indirectly working for the UK Independence Party and influencing UK citizens to vote for Brexit. This accus- ation was denied by the UK´s Information Commissioner and has yet to be proven.9
After the data scandal and a following investigation by the UK government, Cam- bridge Analytica quickly voluntarily filed for bankruptcy and closed operations at their London headquarters on first May 2018.
2.2 Key Actors in the Trump Campaign
Their work for Trump´s election campaign and the following data scandal made Cam- bridge Analytica famous. To understand how exactly the company managed to influ- ence the election, one first has to get an overview of the most important figures:
2.2.1 Alexander Nix
Alexander James Ashburner Nix (see image at 7.1.1), descending from the aristo- cratic Ashburner Nix family, is a British financial- and data analyst who was Cam- bridge Analytica´s CEO from 2014 – 201810. According to statements from former em- ployee Brittany Kaiser, Nix was a key player in Trump´s campaign because he nego- tiated contracts and strategies for the firm´s work with Trump´s election team 11. After the numerous headlines concerning the Facebook data scandal and a secret filming investigation by British news agency “Channel 4 News”, Nix was suspended as Cam- bridge Analytica´s CEO in March 2018. The secretly recorded video shows Nix boast- ing about the company´s practices and him taking credit for influencing the 2016 presidential election1213.
2.2.2 Aleksandr Kogan
Aleksandr Kogan (see image at 7.1.2) is a Moldovan-American data scientist who worked as a professor and lecturer at Cambridge University in 201214. In 2014, Kogan developed an app named “This Is Your Digital Life” which let Facebook users take personality tests for them to receive a small amount of money in return. “This Is Your Digital Life” was used by approximately 270.000 volunteers who agreed on their data being used for Kogan´s academic research at Cambridge University. What most of them did not know was that they also agreed to the app harvesting the data of all the people on the person´s Facebook friendlist using the app15. This “snowball effect” led to a total of up to 87 million individuals having their personal data mined by the app. Kogan later violated his conditions of only using the data for academic purposes by selling the data set to Cambridge Analytica16. This data set was used as the basis for Cambridge Analytica´s data algorithms and analyses which were later used for their work in the Trump campaign17. Therefore one can say that the firm´s practices would not have been possible to perform without Kogan´s work.
2.2.3 Christopher Wylie
Christopher Wylie (see image at 7.1.3) is a Canadian data-scientist and consulter who worked for Cambridge Analytica from 2013 – 2014. Even though he left school at the age of 16 without a graduation, he later studied law at the London School of Economics and taught himself how to code. When being offered a job at Cambridge Analytica, CEO Nix promised him total freedom to experiment with new ideas in the field of influencing elections trough the use of data. Highly gifted Wylie was excited to accept the offer because he saw an opportunity to test out an idea he had thought about in early 201318: Using psychographic data to create psychological profiles of in- dividuals and accurately identify voter target audiences (see 3.2). According to former Cambridge Analytica employees, he also came up with the concept of using microtar- geting (see 3.1) in the firm´s election campaigns19. To get to the point, Wylie´s ideas were the fundamentals for Cambridge Analytica´s methodology they applied in the Trump election campaign. According to own statements by Wylie, he never had an idea how dangerous these tools would be at the time he developed them. Because of his guilty conscience, he later acted as a whistleblower by disclosing information about Cambridge Analytica to various newspapers and exposing the Facebook data scandal.
3 Cambridge Analytica´s Methodology
After taking a closer look at Cambridge Analytica´s history and the most important actors in their work for Trump´s election campaign, it is really important to focus on the methods the firm used to try to influence the presidential election in favour of the republicans. Understanding the company´s practices is the most decisive factor in understanding why Cambridge Analytica could be a potential threat to democracy.
3.1 Microtargeting
The term microtargeting describes the targeted addressing of individuals with the goal of influencing the individuals´ view of something in the initiator´s favour20. In the context of election campaigns, it describes the data based targeting of individual voters or bigger groups of voters with the goal of convincing them to vote for the initi - ator´s party. Potential voters can be targeted through multiple different ways, for ex- ample through direct mail, television ads or billboards. But in the age of digitalisation, the most common method is to reach the target audience through personalised ad- vertisements on social platforms such as Facebook or Instagram. Since there are many different factors that can influence the outcome of an election, the actual effect- iveness of microtargeting is not empirically provable21.
3.2 Psychographics and Demographics
Demographic data is factual data based on certain attributes such as age, gender, ethnicity or income22. Only until a few years ago, companies mostly used demo- graphic data for their advertising. The issue with demographic data is, that even in narrowed down groups in society, there still is a big difference between individuals. For example, take a group of American, middle-aged white men who have a similar income. From the perspective of demographic data they would belong to the same target audience even though they might be very different individuals with different opinions and interests. This leads to the conclusion that traditional advertising meth- ods based on demographic data such as billboards or TV-ads might not be that ef- fective. They reach a big audience but a major part of that audience probably just does not care about them because the advertisements are not sufficiently adapted to their interests.
To solve this problem, Cambridge Analytica established a new, more effective and ac- curate method of defining and reaching target audiences: the use of psychographic data.
Psychographic data, in contrast to demographic data, does not only focus on factual information, it focuses on character traits, interests and the overall personality of an individual23. Psychographics allow marketers to gain an insight on why exactly a per- son made a decision, for example to vote for a certain party. According to Alexander Nix, advertising campaigns based on psychographic data are significantly more ef- fective because ´it is personality that drives behaviour´24, not your race or age for ex- ample.
3.3 The OCEAN-Model (Big Five Personality Traits)
When talking about psychographics, it is almost impossible to not encounter the OCEAN-Model. After over 50 years of research by various personality scientists, American Psychologist Lewis Goldberg presented the model in 199325. The OCEAN model is the leading personality model in behavioural psychology and it represents the concept that personality differences between individuals can be traced to the varying degrees of five central personality traits. These “big five personality traits” are o penness (1), c onscientiousness (2), e xtroversion (3), a greeableness (4) and n eur- oticism (5)(see example 7.1.5)26. In other words the five personality traits could be described as how open an individual is to new experiences (1), how much someone values organization and planning in their life (2), how social a person is (3), how much compassion they have (4) and how emotional or impulsive an individual is (5)2728.
3.4 How Cambridge Analytica Practically Applied their Methodology in the Trump Campaign
With the data Aleksandr Kogan´s App had harvested (see 2.2.2) and additional per- sonality quizzes voluntary American citizens had taken, Cambridge Analytica had a lot of raw data to work with when they accepted the offer to work for Trump´s presid- ential campaign.2728
Using the psychographic- and demographic data they had available, the firm now proceeded to divide individuals into different categories according to the varying de- grees of the five different personality traits in the OCEAN-model (see example 7.1.7)29. This leads to the assumption that, on the basis of the psychological profiles the firm had created, Cambridge Analytica now had the possibility to create a wide variety of different advertisements, precisely tailored to the individuals´ personality traits (see example 7.1.8)30.
In the course of Trump´s election campaign, this meant that Cambridge Analytica would determine different target groups based on the individuals´ political views. For instance, they determined a group of people who, according to the evaluated data, were very likely to vote for Clinton, a group of people who were likely to vote for Trump and a group of individuals who had not clearly decided yet who they want to vote for, “the persuadeable”31. In comparison with the whole US-population, this group only represented a fairly small amount of people but that did not matter be- cause big elections like the presidential election are often decided by only small per- centage differences. Therefore, influencing this small group of individuals would already lead to a victory for Trump. Cambridge Analytica went ahead and created dif- ferent advertisements that presented presented Clinton in a very bad light in order to convince the target audience to vote for Trump32. Applying the concept of microtarget- ing, the firm then displayed the specifically tailored advertisements to the individuals on different social media platforms such as Facebook or Instagram 33.
[...]
1 Kaiser: Targeted, 2019, p.12
2 cf. Sturgis, Sue: Cambridge Analytica´s Reach into Southern Politics, in:facingsouth.org , 17th March 2018
3 cf. Rosenberg, Confessore, Cadwalladr: How Trump Consultants Exploited the Facebook Data of Millions, in: The New York Times, 19th March 2019
4 cf. Kaiser: Targeted, 2019, p.108
5 cf. Paytoncular: Cambridge Analytica Main Players: Where Are They Now?, in: women in technology, 19th December 2019
6 cf. Kaiser: Targeted , 2019, p.10
7 cf. Cadwalladr and Graham-Harrison: Revealed: 50 million Facebook profiles harvested for Cambridge Analytica in major data breach, 2018
8 cf. The Guardian: Cambridge Analytica whistleblower: ´We spent $1m harvesting millions of Facebook profiles´, 2018
9 cf. BBC News: Cambridge Analytica ´not involved´ in Brexit referendum says the watchdog , in: BBC News, 7th October 2020
10 cf. McKee: Alexander Nix, Cambridge Analytica CEO, suspended after data scandal, in: the Guard- ian, 25th March 2018
11 cf. The Guardian: Brittany Kaiser, former Cambridge Analytica director: 'I voted for Bernie', in: youtube.com, 23rd March 2018
12 cf. Dwilson: Alexander Nix: 5 Fast Facts You Need to Know, in: Heavy.com, 20th March 2018
13 cf. McKee: Alexander Nix, Cambridge Analytica CEO, suspended after data scandal, in: the Guard- ian, 25th March 2018
14 cf. Statement from the University of Cambridge about Dr Aleksandr Kogan: in: University of Cambridge, 23rd March 2018
15 cf. Weaver: Facebook scandal: I am being used as scapegoat – academic who mined data, in: the Guardian, 25th March 2018,
16 cf. Rosenberg, Confessore, Cadwalladr: How Trump Consultants Exploited the Facebook Data of Millions, in: The New York Times, 19th March 2019
17 cf. Cadwalladr: ‘I made Steve Bannon’s psychological warfare tool’: meet the data war whis- tleblower, in: the Guardian, 24th July 2019
18 cf. Cadwalladr: ‘I made Steve Bannon’s psychological warfare tool’: meet the data war whis- tleblower, in: the Guardian, 24th July 2019
19 cf. Cadwalladr: The great British Brexit robbery: how our democracy was hijacked, in: the Guardian, 13th July 2021
20 cf . Jungherr: Datengestützte Verfahren im Wahlkampf , 3rd March, 2017, p. 4
21 cf. Kolany-Raiser, Barbara and Radtke, Tristan: Microtargeting – Gezielte Wähleransprache im Wahlkampf , in:ABIDA- Dossier, January 2018, p.2
22 cf. RyteWiki: Demographic Data, in: Ryte, 2019
23 cf. CBInsights: What is Psychograhpics? Understanding the Tech that Threatens Elections , 6th May 2020
24 Alexander Nix at 2016 Concordia Summit, minute 2:52
25 cf. Belyh, Anastasia: The Big Five Personality Traits Model (OCEAN Model), in: Cleverism, 25 th September 2019
26 cf. Gerlitz and Schupp: Zur Erhebung der Big-Five-basierten persoenlichkeitsmerkmale im SOEP, 2005, p.2
27 cf. Alexander Nix at 2016 Concordia Summit, minute 3:15 – 3:40
28 cf. CBInsights: What is Psychograhpics? Understanding the Tech that Threatens Elections , 6th May 2020
29 cf. Cambridge Analyitca: Data Driven Behavior Change Presentation - Cambridge Analytica, in: documentcloud.org, date unknown, p. 18
30 cf. Cambridge Analyitca: Data Driven Behavior Change Presentation - Cambridge Analytica, in: documentcloud.org, date unknown, p. 19
31 cf. cf. Amer, Karim and Noujaim, Jehane: The Great Hack, in: Netflix, 26th January 2019 (first publication), minute 41:10 – 42:10
32 cf. Amer, Karim and Noujaim, Jehane: The Great Hack, in: Netflix, 26th January 2019 (first publica- tion), 1:14:28 – 1:16:08
33 cf. Amer, Karim and Noujaim, Jehane: The Great Hack, in: Netflix, 26th January 2019 (first publica- tion), minute 42:15 – 42:36
- Arbeit zitieren
- Anonym,, 2022, Election Manipulation in the USA. Using the Example of Cambridge Analytica, München, GRIN Verlag, https://www.grin.com/document/1322832
-
-
-
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen.