It is now well known that Cambridge Analytica, the British firm at the centre of the data breach controversy, was working with users’ details obtained from Facebook. What is not clear is exactly how the research firm influenced voters in the American presidential election and other campaigns.
Over the last decade, Facebook and Twitter have become important platforms for political campaigns and exchange. But the two popular social media websites have also been exploited to air a lot more than rhetoric. In fact, it appears now that a lot of the political material available on these platforms are propaganda that helped in influencing the voting behaviour of many users.
But how can data acquired from Facebook help shape political campaigns and reap electoral dividend?
When an individual interacts with Facebook over a period of time, the platform gathers enough information about her likes and dislikes. It is not only aware of the kinds of people the individual follows, but also the types of news sources that she prefers, and the range of actions and reactions the posts from these sources elicit from the user. All these data, when combined, help in putting together a picture of the user’s political affiliation and inclinations.
Facebook has an advertisement preferences page (https://www.facebook.com/ads/preferences/edit/) that helps the platform to serve up advertisements according to user preferences. Based on an individual’s likes and dislikes, the social network lists topics, people and interests, which it thinks are good enough to push ads to the user.
However, Cambridge Analytica did not gather data from Facebook in this way. The firm allegedly took the help of Dr Aleksandr Kogan, a psychology professor at Cambridge University. Kogan designed an application, thisisyourdigitallife, to gather information about users. While details of the app are not known, given that it was a personality quiz, the questions are likely to have been framed in a manner that would draw out details about the quiz-taker’s political leanings and related aspects. The acquired data would then provide raw material for a detailed profile of the potential voter.
With the availability of advertisements of all kinds, Facebook allows a good degree of targeting. Usually, the targeting is based on age, location and interests. But with the kind of data Cambridge Analytica had, it allegedly targeted voters in many American states, who had not made up their minds, and had a high probability of moving. For instance, a group of people disgruntled with the economic slowdown and the resultant unemployment could be targeted with the pro-Donald Trump campaign. Pushing a certain line of narrative to these voters over a period of time may have helped them in making their choices. For example, if it involved individuals who lost their jobs in the 2008 American recession, stories blaming Democrats for the situation could be targeted at them. Targeting users with a particular line of narrative has often had the desired effect.
An important caveat: Facebook does allow users to change their preferences through the advertisement preferences page. The preferences that are found on this page are most likely to have been added based on browsing patterns, and not the intent of the user.