It was always a lie to say that the American Dream was about a white picket fence. The white picket fence was merely a symbol of what people really wanted, which was the right to be boring and predictable. The American Dream was a dream about the monotony of life. People are not all the same, but they are all predictable. This has always been known, but until recently was of little significance. With the possible exception of those unlucky few whose names appeared on a CIA watchlist, people’s boring lives took place in privacy. The resources required to create individual profiles were too great, and the true patterns of people’s lives were therefore safe. But what once was a luxury has now been taken away.
Internet tracking and the use of machine learning models to profile users are relatively new, but the motivation for doing so is firmly rooted in years of economic literature. First degree price discrimination has always been a holy grail of sorts, and firms will do whatever they can to capture consumer surplus for themselves. As the costs of profiling users goes down, the viability of price discrimination goes up. And it’s not just a matter of price discrimination; firms can target people in other ways that they’ve never been able to before. Personalized news recommendations, targeted ads, flexible user interfaces that change based on revealed preferences—this is a world in which companies will use and abuse the predictability of people.
“But we get a better service”, they say as they sit passively in their box binging on Netflix. And they may be right. But there is a very real fear that the cost is too high, that people are losing control over their lives in ways that have not yet been appreciated. When a person chooses to read an article on a news site, that person unknowingly provides information that will be used to decide what content the person will be exposed to in the future. Everyone’s life is now path dependent, not only with regards to important matters like education and work, but also with respect to the mundane, like access to information, and the trivial, like entertainment.
Enter the GDPR, the European legislation that is meant to protect the rights of individuals in this domain. But what if it doesn’t work? What if May 25th comes and goes and everyone signs over their rights, or provides their consent? Is it still a win for freedom? The march of progress will go on as planned, and the position of Facebook and Google will be stronger than ever before.
On the face of it the requirements of the GDPR seem demanding. Companies need consent in order to collect the types of data they use to target people individually. That consent must be clear, and must be given for a specific purpose. In addition, providing consent cannot be a necessary condition for obtaining access to a service; no data paywalls allowed, so to speak.
But a different picture has emerged as companies prepare for the GDPR. Google is crafting a “consent policy” to give to publishes so website visitors can agree to be shown targeted ads. Facebook has sent out notices to users allowing them to manage their data settings. Here is an example of what that notice looks like.
The wording of this is subtly perverse, as this is not an “accept” or “reject” question. Instead, people are asked to accept or go elsewhere to “manage data settings”, which feels boring and burdensome. That clicking “accept” is optional is not at all obvious. Nonetheless, this probably complies with the letter of the law.
If a company is looking for a roadmap for how to collect consent effectively, Facebook has provided it. It wouldn’t be surprising if the wording and design of this system was itself the result of careful user testing. In the era of Big Data, even privacy options and notifications are seen as components to be optimized. This is how Facebook is winning the war against your privacy. In doing so, it is proving that there are limits to what the law can achieve.