All of this writing and data analysis is a lot of work! So after more than five years of posting, I've finally launched a Patreon to help pay the bills.


The flimsy case for Russia's role in Sanders to Trump crossover vote

The Washington Post has just published an article about "the Russian effort to target Sanders supporters - and help elect Trump". Once you wade through about 1500 words of background, however, the substance is quite thin: the Post asked two researchers to "[examine] English-language tweets identified as coming from Russia". 

What did they find? 9000 tweets that used the word "Bernie," and "thousands of other tweets" that were allegedly "designed to appeal to his backers".

That's it. The article doesn't even indicate whether this was an exhaustive survey or whether they were just looking at a sample group, which means that we have no way of knowing the scale of the campaign. Taken at face value, however, these numbers are absolutely trivial. To say that they won Trump a single state, you would have to argue, for example, that three out of every five "Bernie" tweets flipped a Clinton voter. To say that they won him Michigan, Wisconsin, and Pennsylvania, you would have to argue that each "Bernie" tweet flipped Trump more than four votes. 

Here, we are back to arguing that Russia's ad campaign was, as Brian Feldman put it, "supernaturally effective and persuasive" - that it had an impact out of the reach of ordinary social media marketing campaigns by several orders of magnitude. It is entirely possible, of course, that the tweets discovered in the Post article were part of a much larger campaign - but it does nothing at all to establish this, leaving us with a handful of utterly meaningless numbers.

A final note: in passing, the Post cites an Ohio State University paper which argues that "about 4 percent of President Barack Obama's 2012 supporters were dissuaded from voting for Clinton in 2016 by belief in fake news stories." But to swing the election, Trump only needed to win about 2 percent of Sanders supporters.

In other words, all of this data is entirely consistent with a scenario in which Sanders supporters were twice as resistant to Russian propaganda as Obama supporters, but still swung the election. The article, then, tells us nothing about the scale of Russia's efforts to target Sanders supports, nothing about their effectiveness, and nothing about about how receptive Sanders supporters were. Why did they publish this, again?