Yes, they did a bunch of shit. They hacked the DNC, and a lot of other stuff. Yes, they were actively working against Clinton, and on behalf of Trump. None of this has ever really been a question. Yes, they had lines of communication with Papadapoulos, Don Jr., Flynn and others. Yup. All true.
I have regularly stated, though, that I haven't seen clear indications that they flipped the election, and the indictments touch on this, but let's address why this is such a hard question.
James Comey's pre-election announcement was easy to assess. You have what we call, in social science, a quasi-experimental design. An exogenous event occurs, and you can do a pre- and post- measure. The polls before Comey's announcement were very different from the polls after, by much more than Trump's margins in enough states to swing the election, so shut up about that "Clinton didn't go to Wisconsin" shit. I'm really sick of that. Comey flipped the election. That is measurable with a pre- and post-announcement assessment of polls.
What else is, in principle, measurable? In principle, we can actually count up candidate visits to states, and look for patterns of statistical association between that and vote shares, controlling for factors like state partisanship and the other important stuff. If Clinton spends proportionately more time in Florida, and visits matter, then she should do better in Florida. If not, candidate visits don't matter. That kind of thing is measurable, but it must be generalizable rather than an ad hoc rule for Wisconsin because Trumpists want to make statements about Wisconsin. (Can you tell how sick I am of that Wisconsin bullshit?)
We can look at spending. We can look at aggregate spending across years. Does the candidate who spends more money, overall, defeat the candidate who spends less money, controlling for other factors, like the state of the economy, Abramowitz's "time for a change" variable, and so forth? We can look at spending across states within years and ask whether or not the candidate who spends more within a state does better, controlling for the state's political leanings.
All of this is based on the same kind of logic. Find a baseline, and ask whether or not we systematically observe deviations from that baseline associated with the causal factors we are discussing.
That's how science works.
So, um... how do we do that with the Russian hacking?
Well, we could look at the timing of the release of the DNC hack emails. Here's RealClearPolitics's favorable/unfavorables for Hillary Clinton. Can you pick out the date? It was right before the Democratic Convention. Did it hurt her? Not that I could tell.
Then, there's all the trolling. Here's the problem. How do you measure the effect of trolling? Is there variation across states in levels of trolling? Variation across counties? Can we conduct surveys and ask, in the surveys, "so, um, hey. How much of your FaceBook feed was a bunch of Russian trolling?"
You see the problem. Without a way to measure the variation, our method breaks down.
Did the DNC hacks move the poll numbers? I just didn't see it. Russian trolling? That's... hard. I don't know how to measure that effect.
Here's how this stuff could have mattered. Maybe it helped shift media attention, and that helped Trump by distracting the press away from all of the very real shit surrounding Trump. Donald Trump is and has always been a con artist. See: Trump University. The press, though, spent a lot of the campaign discussing bullshit pseudo-scandals like Clinton's email thing. Throw in Comey, and maybe things might have been different.
That's a lot of conditionals, and that's why I'm uncomfortable attributing outcomes to Russian trolling. It's just too hard to draw the causal connection.
Comey? Nope. He gave the White House to Trump, and the Russians did engage in meddling. The effect of that meddling? I dunno. That doesn't legally exonerate them, and Trump is full of shit on "collusion." His campaign absolutely did collude. Flynn, Don Jr., Papadapoulos...
My point is merely that, from a social science perspective, assessing the impact of Russian trolling is extremely difficult, and I am not convinced that it affected the outcome of the 2016 election.
Comey? Pin the blame on him, if you want to blame anything other than just American gullibility.
