Anywho, North Carolina is a fun state, isn't it? Now that's what I call election fraud! They don't mess around with ineligible voters showing up to multiple polling places to cast multiple votes, or any of that nonsense that the GOP claims to happen as a justification for voter ID laws, which don't do much anyway. You might have seen a reference to this paper lately, showing that voter ID laws are basically nothingburgers, or whatever weird spelling Dear Leader might use, and you know what? I think I kind of posted about that. As an example, here. Voter fraud, of the kind in GOP ghost stories, is basically phantasmal, and the laws they write have minimal effects. That doesn't mean you should ignore them, but this stuff isn't the main event.
The main event is when some crook goes around collecting absentee ballots, filling them out fraudulently, tossing the ones they don't like, etc. We've known all along that, to the degree that there would be fraud, it would be in the absentee ballot system where voter ID laws can't do squat. You just need to keep close tabs on the operatives. Otherwise, you get the crooks in North Carolina, and for the purposes of this post, I'll leave it to others to summarize just how bad it was. WOW! It's almost like they looked at the history of fraud in New Orleans, Chicago, and the other champions of corruption and said, "hey! We want a shot at the title! We'll dance like a butterfly, sting like a bee!" Yeah, I hate sports, but even I know that quote. Then again, he'd probably have his ballot stolen in North Carolina. Zing!
Point being, that was some crazy stuff. Massive, massive fraud. Who won? Uh... We didn't have an election in that House race. It just didn't happen. So, do-over. And that leads to the broader question of when we do this. This isn't something we do. Why now? Because the counting of ballots just couldn't be done, given the fraud. As an interesting point of comparison, though, consider the 2000 presidential election.
Calm down. Breathe. And remember, wouldn't you give a hell of a lot to trade Trump for Bush? I would. A lot.
So, what happened? Who won? The final vote tally in Florida put Bush ahead of Gore by 537 votes. Was that because of fraud? No. Was there fraud? No. Bush eventually told his DoJ to go on a hunt for voter fraud, they couldn't find any, so he fired a bunch of US attorneys, and that was what counted as a major national scandal a decade ago, but Bush's lead over Gore wasn't because of fraud. To our knowledge, there wasn't anyone going around the state, fraudulently filling out absentee ballots, and pulling what I think we should now call "the North Carolina maneuver."
You may also recall a bunch of stuff about "dimpled chads" and "hanging chads," and if you weren't there, you just picture hanging out with some WASP-y dude named "Chad," who had dimples. And for some reason, you have the general impression that he was pounding "skis" with PJ and Squee, or something. What a douchebag...
Somehow, this made its way to the Supreme Court, which made one of the most cowardly rulings in all of US history-- Bush v. Gore. A 5-4 majority halted a three-county recount, leaving Bush the official winner, and said in the ruling that the ruling shouldn't ever be used as precedent for any future case.
That's how you know it's total bullshit. If you know that your ruling is based purely on partisan preferences at the time and don't want your partisan ruling to bite you in the ass later on, this is what you do. Cowards. The five conservative members of the Court tried to come up with a way to write a ruling that would favor Bush but wouldn't give Democrats any future benefits, couldn't do it, and Bush v. Gore was the result.
Here's what you don't remember, though. A bunch of journalistic organizations decided to put together a project to see what would have happened if the three-county recount hadn't been halted. Bush still would have won.
And Gore wasn't asking for a do-over. Just a recount, in three counties, under a specific standard. Rather different from North Carolina. Interesting.
Yet, one of the things we know about the 2000 election is that the structure of the ballot in Palm Beach County handed the election to Bush. We don't even need to get into that nitwit, Ralph Nader. Palm Beach County, under the administration of a Democrat named Theresa LePore, used the "butterfly ballot," which caused a bunch of careless people to vote for Pat Buchanan rather than Al Gore. How do we know this? Simple. If you look statewide in Florida, there was a very clean, linear relationship between how any given voting bloc voted between Bush and Buchanan, which makes sense because Buchanan was a hardcore racist isolationist. Basically, a proto-Trump. A constituency with more Republicans was a constituency with more Buchananites. Tell me what proportion of a geographic unit voted for Bush, and I'll tell you almost exactly the vote share Buchanan got there. Except for Palm Beach County. Buchanan got way, way, way, way, way more votes there than can be explained by anything other than the butterfly ballot. A bunch of voters punched the second hole thinking they were voting for the second name on the left, but the second name on the left was associated with the third hole, because the second hole was for the first name on the right. The second hole was associated with Buchanan.
A stupid ballot confused enough sloppy, careless voters to hand the election to George W. Bush.
Who won North Carolina's House race? I dunno. Who won the 2000 presidential election? Al Gore.
George W. Bush took the oath of office. We're re-running that North Carolina House race.
Kinda weird, right?
What's going on here? The answer is that our assessment of the 2000 presidential race, while damn-near indisputable, is based on statistical analysis. In fact, statistical analysis done by my grad school advisor, Henry Brady (truth in advertising). The North Carolina situation is based on specific cases of ballot fraud. Our system takes into consideration specific cases of ballot fraud, but not statistical analysis.
So, imagine this. A bunch of political scientists with our fancy, highfalutin' math, walk into the courtroom and challenge the election results of 2000 by saying that we are as certain as we can be that Al Gore won, despite the formal tally, 'cuz math. Why doesn't that count for anything? Basically, the legal system doesn't even have the capacity to respond to math.
Here's a bullshit quote that you have probably heard, and heard misattributed: "There are three kinds of lies: Lies, damned lies, and statistics." (Benjamin Disraeli, not Mark Twain) This is a bullshit line, used by people who don't understand statistics as an excuse to a) not study statistics, and b) disregard any statistical argument.
Frequently, although not always... lawyers.
The North Carolina case is easier for people who don't study math to wrap their brains around. We don't know who won, we know fraud happened, so just have a do-over. A non-math person can get that. Florida 2000?
You can sort of think of this in terms of two kinds of math. There's counting, which sounds simple, and there's statistical analysis, which a) people don't study, and b) people use a bullshit quote to disregard.
Counting to three is easy. Counting to ten is easy. Counting up into the thousands, without error, is hard. The reason that recounts exist is because errors occur. Hand recounts incorporate human error. Voting systems have error rates. Florida, in 2000, was using one of the worst systems--- punch cards. Along with Matthew Jarvis and John McNulty, I have written about this. There are some nasty racial effects here! Yes, it's measurable.
For some reason, though, people think of counting, particularly human-performed counting, as true and reliable, in contrast with statistical analysis, which the statistically-uneducated think of as being some sort of untrustworthy, mystical process akin to astrology, except that lots of people are stupid enough to believe astrology.
Look, Gore won, but we inaugurated Bush because the courts don't care about statistical analysis and the legal system isn't even really set up to care. We have no idea who won in North Carolina, but we're having a do-over because instead of statistical analysis, we can see the individual process of the ballot fraud. That's the point here, and it is a weird thing about how the electoral system responds to problems.
The question is whether or not we translate the votes and the voting intentions into outcomes. We knew what the intent was in 2000. More so than in North Carolina. This is weird.
So... what about 2016? I just set this up as a sort of statisticians-versus-lawyers kind of dichotomy. Here's what we know about 2016. Contrary to Trump's constant lying about everything ever, Russia hacked the DNC and leaked the emails through WikiLeaks to try to help Trump win, along with an extensive social media disinformation campaign. Trump's campaign and surrogates coordinated in the effort. Roger Stone and Jerome Corsi coordinated directly with WikiLeaks, while Trump's campaign manager, Paul Manafort, who had a long history of working for the Russians, coordinated directly with Russia through a variety of means, including providing detailed polling data to the Russians to help with Russia's efforts to sway the election. All of this is public information, along with much more, keeping in mind that everyone associated with Trump has been caught lying about how deep in bed they are with the Russians. Including the little-big man himself.
What if none of this had happened? We have no idea. What if Comey hadn't "reopened" the FBI investigation of Clinton two weeks before the election? I have stated here, repeatedly, that my interpretation of the polling data is that this is what flipped the election from Clinton's glide path to victory to a Trump victory. I didn't originally think that the announcement could have a big enough effect, but watching the polls afterwards, they moved, and Trump had narrow enough margins in enough states that I think this is what did it.
In my opinion as a political scientist, Comey flipped the 2016 election.
If you read around, though, there is a debate about this. Kathleen Hall Jamieson, for example, is more confident that Russia did it. There's a debate among political scientists about what happened in 2016.
2000? Not so much. The butterfly did it. That's even the title of the most important paper. Did Trump win "legitimately?" Um... What does that mean?
Is it illegitimate to be swayed by Comey? To be swayed by Russian lies? Is the electorate ever responding rationally to truthful information? I'm pretty cynical, and like I remind people, the default pattern in presidential elections is DDRRDDRR, which is why one of the best predictive models is Alan Abramowitz's "Time for a Change" model, so what's legitimate if it's all just silliness anyway?
In contrast to either 2000 or the North Carolina mess, there is a lively debate in political science about what happened in 2016, and all of this points to the general weirdness about when we accept an election result and when we don't.
Do I have a real answer here? No. That was just a ramble. Sometimes I don't have real answers. I just ramble.