Politics in the age of big data
One thing allegedly struck Mark Zuckerberg when he was building a prototype of Facebook as a Harvard University undergraduate student. This was the readiness with which people shared aspects of their private lives online in exchange for the chance to multiply their social connections. Such pervasive gullibility supposedly shocked the Facebook founder.
Two billion Facebook subscribers later, this astounding insight into human nature continues to be validated. People do not seem to care that anyone might exploit or misuse the “harmless” information they post about themselves online. Maybe they cannot imagine that anyone would have the capability to aggregate such information on a mind-boggling scale, and transform it into data that can be used to manipulate human behavior in ways never before seen.
Zuckerberg’s realization stands in contrast to the widespread notion that modern educated individuals are more deeply private, more protective of information about themselves, and less keen to share intimate details of their lives with people outside their narrow circle of friends.
As we now know, that narrow circle has expanded into a global social network encompassing billions of people around the world. Surpassing all expectations, Facebook has become the first virtual global village in a real sense. The word “village” is appropriate because it conjures images of the traditional community where secrets are seldom kept, and the lives of neighbors and acquaintances are treated as open books.
When people freely share opinions, beliefs, fears, likes and dislikes, and photos of themselves on Facebook — as though they were doing this with a childhood friend or a next-door neighbor — they must be unaware that they are, in fact, baring themselves before a global audience. Would they sign up for a Facebook account if they knew this? Probably not.
Still, I am inclined to believe that, in general, it doesn’t take much to get people to shed their apprehensions and inhibitions. Not knowing the potential uses of these information-gathering machines, as nearly every popular website today indeed has become, they might regard it as delusional to imagine that anyone might take any interest in them. To be fair, it is not just Facebook that has this capability.
An advertisement from Avast, the antivirus company that sells all kinds of virtual shields, claims that as much as 75 percent of all the websites we regularly visit keep track of what we do online, and store this for future use. These websites construct images or profiles of the kinds of persons we are. Though we may believe that these images are far from being faithful representations of who we are, we would miss the point if we thought this was nothing but a useless expenditure of effort.
Big data miners are seldom concerned with getting an accurate representation of who we “really” are. Their aim is more specific: to anticipate our desires so as to get us to buy or sign up for something, or to induce us to act in a particular way.
The attempt to grasp how voters think and perceive their milieu for the purpose of shaping their electoral behavior is, of course, as old as politics itself. Political strategists and tacticians do this all the time. Survey firms, doubling as political consultants, formulate campaign lines and messages for their clients based on observed preferences, fears, anxieties, and hopes.
What is probably new is this: that, unlike the opinion survey outfits we know, there are today research and consultancy organizations that are capable of processing data from millions of individuals in order to arrive at sharper (or more “nuanced”) definitions of voter or consumer types. But, more frightening is their ability to extract data — not from the responses of informants to survey questions but from the observed online behavior of people. One of the basic limitations of conventional survey research has been that findings are typically drawn from the responses of people to questions or issues that may not be salient to them. Big data analytics dodges this limitation.
Facebook knew the potential uses of its platform in electoral campaigns, and, indeed, its people actively promoted these in workshops they gave to campaign strategists. For Zuckerberg and his associates, bringing electoral discourse to Facebook would not only increase their traffic, it was also good for democracy.
Clearly, this was a naïve view. There were people who saw Facebook’s uses beyond these civic-minded intentions. One of them was Alexander Nix, CEO of Cambridge Analytica, who said: “If you know the personality of the people you’re targeting, you can nuance your messaging to resonate more effectively with those key groups.” His firm today stands accused of improperly using data it had obtained from Facebook on false pretenses in order to craft campaign software for their clients, including some from the Philippines. Facebook itself is accused of the unauthorized sharing of users’ accounts with Cambridge Analytica, including those of around 1,175,870 Filipino users.
The issues against Cambridge Analytica and Facebook have mainly centered on breaches of privacy. I think these issues pale in comparison to what has become Facebook’s biggest danger — the intensification of bigotry and partisan resentments resulting from the micro-targeted manipulation of Facebook users’ media feeds. Just take a look at the normalization of hate speech on social media. Politics in the age of big data preys upon desires, hopes, and fears, that often lie at the level of the unconscious. That is what makes it insidious and, quite often, deadly.
