In fact, according to the CDC, about 90 people die every day from vehicular accidents, so that comparison is basically just unfair. Sometimes, then, I pick weirder ways to die, like lightning strikes. Here's a NOAA paper from earlier this year by John Jensenius on death by lightening strikes, showing 352 people over the 10-year period from 2006 through 2016.
The total number of people who have individually swung a presidential election over that period? Zero. You can make a case, if you go back to 2000, that five people did it: the five-to-four ruling in Bush v. Gore, but even that is wrong because if the Florida recount had been conducted under the procedure Gore was requesting, Bush still probably would have won. (Bush did not legitimately win-- the butterfly ballot in Palm Beach County gave him the election. There were other requests Gore could have made, but Gore's specific request in that case was for a specific recount in three counties under a specific standard, and... I'm getting off-topic here.)
Senate elections? House elections? One vote doesn't really swing these things. You really are more likely to be killed trying to get to the polling station than you are to swing a national election.
Then, there's local stuff. Smaller electorates. Here's how the probabilities work, in short. The reason you aren't going to be "pivotal," in voting jargon, is that it only happens when everybody else is within exactly one vote of each other. In the US Supreme Court, one Justice is often pivotal. Rulings can easily be five-four. Why? There aren't many of them. In the Senate, it happens, but not as often. There are 100 of 'em. With the filibuster pivot (such as it remains), getting to 60 matters. Splits of 60-40 happen. When they do, everyone who is on the 60 side is casting a "pivotal" vote. If they either flipped, or didn't show up, the outcome would be different.
In the House of Representatives, it isn't as common. When the House is fully-seated, there are 435 members. That means someone is "pivotal" when the chamber is split 218-217. That happens, but not very often, and not as often as the 60-40 splits in the Senate, and nowhere near as often as 5-4 SCOTUS rulings.
The more people there are, the less likely there is to be a divide close enough for one person to be pivotal.
And when you aren't pivotal, you don't matter. Deal with it. There are just over 326,469,000, according to the Census estimates as of this morning. You are one person, and you don't fucking matter. Get over yourself, you precious, little snowflake.
I don't matter either. I'm just some asshole, ranting over my morning coffee.
Coffee. Coffee matters.
Anyway, the larger the voting body, the lower the probability of a split close enough for you to be "pivotal," and once that body gets large enough, you are more likely to die on the way to the polling place than cast a pivotal vote.
So, hey, do you pay attention to state or local elections?
(Admission: I pay... rather less attention than I should.)
Do you... vote in state and local elections? There is an interesting phenomenon that we call "ballot roll-off." Even the people who do vote sometimes just vote for the top-of-the-ticket races, like the presidential race, and leave the ballot blank for the state and local stuff. And, when there isn't a presidential election, turnout is way lower.
The irony here is that this messes with the rationality/irrationality of voting. It means you get much smaller voting groups, and the probability of being pivotal goes up. It is still absurdly low, but you aren't quite in lightening strike territory anymore, and the absurdity of the jokes I can make goes down.
Which brings me to Virginia's lower state house. The race in one district-- Yancey versus Simonds-- is tied at 11, 608 votes after the latest recount! And, to make this shit even crazier, if Simonds wins, that brings the chamber to a tie, and the Democrats get effective control of the whole legislature! Holy fucking shit!
What happens now? They actually, truly, draw lots! This election will actually be decided by random fucking chance! Don't you fucking competitive elections goo-goo morons love that?! This won't be decided by the merits of the candidates, their platforms, voter deliberation, or any of that stupid, fucking nonsense, nor will control of the House of Delegates. No, something far more important for democracy will determine everything. Random fucking chance!
Behold! Democracy! The true arbiter of all that is good and holy! Random fucking chance!
Yes, I am making fun of this stupid, fucking process. Why? Because I wrote a whole book about the stupidity of having this kind of thing be decided by random chance rather than actual deliberative processes. Are you comfortable with control of a legislature being decided by random chance? No? Then fuck competitive elections. Welcome to the dark side, right along with me and my occasional co-author, Tom Brunell (Trump's possible pick to head the Census!).
Anyway, let's go through pivotality here. They're going to draw lots. Someone will lose, and that party will lose control of the chamber. The 11,608 people who voted for the candidate who drew the short straw will have voted pointlessly because the outcome would have been the same regardless of whether or not they voted. The 11,608 people who voted for the candidate who drew the long straw will have determined control, not just of that seat, but of the chamber! That's actually 11,608 pivotal voters!
That's a fuckload, right? Yes, but over how many contests, in how many years? Numerators and denominators. If you want to know your probability of being in this situation, you need a denominator too. We've been running elections in this country for a couple of centuries now, and the thing about random fucking chance is that if you do things often enough, weird shit happens.
Like now! Holy shit, right?!
Remember, though, what I call "the paradox of news." This is news because it is so weird and new and different.
Remember, too, that this is what happens at the state and local level, where turnout is way lower! 11,608. That's an order of magnitude lower than competitive elections at the level of a US House election, to say nothing of a presidential election. And people ignore this shit! State politics matter!
[Cough, cough... me reminding myself to pay attention to this shit...]
And there's something else going on here. Back when James Comey was handing the 2016 election to Donny-boy, I wrote a lot about the "minimax regret rule." It goes as follows. You make choices, not to maximize your expected utility, but to minimize the maximum regret you might feel.
There were people in that Virginia district who didn't vote. Some of them thought the way that I tell my students to think. Don't vote. It won't affect the outcome. It is irrational to vote because you are more likely to be killed on the way to the polling place than you are to cast a pivotal vote.
And mathematically, that is true! The math on that is still right. Deal with it.
But, so, Candidate A wins the drawing of straws. Think about how fucking shitty a citizen who would have voted for Candidate B will feel after that drawing of straws, having decided not to vote based on the calculation of expected utility.
Pretty fucking shitty, right? Not only would that voter have swung the district, that voter would have swung the whole fucking legislature.
That's a metric fuckton of regret right there. If you want to minimize your maximum amount of regret, you vote! It is, technically, irrational to vote because you really are more likely to be killed on the way to the polling place than you are to cast a pivotal vote. Those CDC statistics are right, as is my math. (OK, it isn't my math-- I stole these analogies from people who have been making these jokes for so long that I don't even know where they originated anymore.)
The minimax regret rule is not true rationality, but some people are going to be feeling it real soon.
Does anyone else feel like there's an infinite improbability drive operating in America right now? It would explain so much.
Just sayin'...