Within the Facebook corporation, there are entire divisions dedicated to the analysis of the connections and communications of users, using unrestricted access to the entire Facebook dataset. If you have a Facebook page for your cause or organization, you can use Facebook Insights to see a far more limited breakdown of information about the people who like or read the posts on your page. Individual people who use Facebook and don’t have a Page cannot access any accumulated information about the people who are their Facebook “friends” (unless they visit each of their friends’ pages in turn) or about Facebook users in general.
For people who maintain a Facebook page, however, there is a curious back-door method for getting fairly broad information about Facebook users. When we log in to our Irregular Times Facebook page, for instance, we see this solicitation, inviting us to create an advertisement on Facebook. By clicking on the solicitation (or finding the “Create an Ad” option under the “Build Audience” tab at the top), we’re taken to a screen on which we can build an advertisement — and in so doing find out a surprising amount about trends among Facebook users.
Don’t get me wrong — Irregular Times has never advertised on Facebook and has no intention of buying any Facebook advertisements. We don’t advertise generally speaking, and the price Facebook asks for every time someone clicks on an advertisement — around $2.00 — is far more costly than any benefit we’d reap from someone’s eyeballs. The entities that have so much money to burn that they’re willing to pay two bucks for every time someone clicks an advertisement are almost always inhuman entities — corporations like WalMart that sell things or corporations like The Common Sense Coalition and Americans Elect that are hawking ideas.
But here’s the trick: in the act of building up a price quote for a hypothetical ad, Facebook provides counts of all sorts of people who use its service.
For instance, Facebook tells me that there are 446,660 people who have liked Occupy Wall Street or one of the other Occupy pages on the Internet. Such numbers are already public. But if I’m on Facebook’s build-an-advertisement page and looking for a market to target, Facebook also tells me that 242,240 of these people — 54.2% of them — are college graduates. That information is not otherwise public.
If I act as if I’m interested in purchasing an advertisement, I can further compare that information to other movements and parties in the United States. It turns out that the Occupy movement has the highest proportion of college graduates among its fans among all the movements and parties listed below:
53.4% of those who like the Green Party Facebook page are college graduates
52.3% of those who like the Democratic Party Facebook page are college graduates
48.1% of those who like the Libertarian Party Facebook page are college graduates
41.5% of those who like the Republican Party Facebook page are college graduates
39.7% of those who like the largest Tea Party Facebook page are college graduates
38.3% of those who like the No Labels Facebook page are college graduates
34.0% of those who like the Americans Elect Facebook page are college graduates
That’s multivariate data, and that’s where data we can sneak from Facebook begins to get interesting.
It’s worth it to pause and ask whether any old person should be able to obtain such information (and more, as I’ll describe later) willy-nilly by following the path to Facebook’s advertising widget for page holders. In fact I invite your ethical musings on the subject. What’s worth noting is that this information has been available to corporations and the wealthy for quite some time; this backdoor method directs the flood of “big data” more democratically.