Thread by @pbump

I am very skeptical of this, certainly in part due to my own predisposition on the subject. axios.com/russia-interfe…

But "25,000 cumulative retweets led to a 1-point jump for Trump a week later" is a remarkable bit of causation to imply. (The study authors don't seem to be implying it, FWIW.)

In part, too, this seems to be driven in part by a spike in retweets just as Trump's poll numbers were spiking ... after the GOP convention.

Also 25,000 retweets is a tiny, tiny number in the scale of the number of tweets each day. There were 75 million directly-election related tweets sent in 2016, excluding retweets. We're meant to believe 25,000 retweets was more than a drop in the ocean?

As I've written, there is no evidence that Russia's paid ads were highly targeted to influence the campaign. Did they maybe get lucky? Maybe! But, as I said, I'm skeptical.
washingtonpost.com/politics/2019/…