Social Media Bullshit on the Facebook 'Peace' Page

The next session at AoIR 2017 starts with this year's AoIR Nancy Baym Book Award winner Nicholas John, whose focus here is on unfriending practices in the context of specific political events. There is limited information about unfriending as the platforms themselves do not provide a great deal of information about such practice.

However, facebook.com/peace offers data on Facebook ties across national divides (e.g. between Pakistan and India, or Palestine and Israel), and such data may potentially be valuable in this context. Unfortunately, though, the data provided by Facebook and the Stanford Peace Innovation Lab on this page is highly dubious, however.

The page operated in 2013, was mothballed for some time, and relaunched in February 2015; there was a substantial decline in numbers of friendships between Palestinian and Israeli users from June 2013 to February 2015; between November 2015 and early 2016 the number increased again by a factor of ten.

This is necessarily suspicious, even taking into account potential increases in the overall number of accounts. By Facebook's numbers, people would need to pick up some 8.5 friends across national divides per week (month?) – this is exceptionally unlikely. Independent surveys of Facebook friending behaviour also do not bear out these patterns.

Friending patterns also don't seem to follow other Facebook usage patterns in Israel, for instance: usage declines substantially around major religious holidays, for instance, yet friending appeared to continue. Nic queried Facebook about these discrepancies, and interestingly around the same the page stopped updating; it remains online, however, and continues to claim that the (now static) numbers represent new friendships over the past week.

Perhaps, then, the 'peace' page is an excellent example of social media bullshit. Perhaps the creators of the page are not concerned about the validity of these numbers at all; perhaps the numbers reported here since 2016 were massaged to 'appear' more impressive, in order to create the impression that Facebook contributes to world peace by connecting more people on a more personal basis.

The numbers claim to be real-time, high-volume, and accurate. They are created by Facebook, which should know the real numbers, and they draw on the semiotics of 'big data'. This raises real questions about how much we should trust Facebook in anything, and moves us from social media epistemology to social media agnotology: it highlights our continuing ignorance about some crucial, central aspects of social media.