There is a vivid discussion right now whether or not Facebook played a crucial part in making Trump. It’s funny to me that the very same company that not too long ago faced enormous criticism for allegedly being biased towards liberal news, is now responsible for Trump. Apparently Facebook can have its cake and eat it too.
Okay, on a more serious note: I don’t deny either the filter bubble effect or that Facebook is good at spreading fake news. But I think it’s unlikely that Trumps electorate consists of mostly moronic folks who believe that every link found on Facebook leads to a credible news source regardless of how crappy the page they arrive at looks like. Sure, these people might exist but believing that the majority of Trump voters falls in that category sounds almost disrespectful to me. After all, we are talking about more than a quarter of the US population eligible for voting.
Also, I assume that there were fake stories about Trump as well which we are blinking at when blaming Facebook for Trump – a traditional case of survivorship bias. (Okay, given the education gap between both electorates maybe Trump voters would be more inclined to believe fakes, but still)
Most importantly though, I think this whole debate is dangerous. It reduces the problems – of which there are plenty, I addressed some of them here – and looks for an easy scapegoat while distracting us from the thorough analysis that is due. Yes, there certainly is an issue with our societies being increasingly fragmented and residing in filter bubbles. But blaming it all on Facebook is just too lazy.
Also, it might be plain wrong. If you really want to find out what role Facebook played in Trump happening, I suggest you look at the following questions and find the data to answer them:
- How many of Trump’s voters are actually heavy Facebook users or fall into the category of people who primarily get their news from Facebook? My guess is that his voters in general are less tech-savvy heavy-users than tech-savvy heavy-users who discuss such matters on the internet assume.
- What media did Trump voters consume the most? Isn’t it reasonable to assume this population (“on average, older, whiter and less-educated than the rest of the U.S”) is tilted towards TV and conservative news channels?
- As how trustworthy do people regard a random article found on Facebook? Clicking on something is not the same as believing it is true. If you can find that data, weigh it so it fits the Trump electorate and the numbers from the former questions.
Let’s have a discussion when you can answer them.
I do mean that sincerely, I’d be really interested in those numbers. A good starting point might be the raw data of this study by the highly regarded Pew Research Center.
Rough Notes 