With apologies to our many friends and posters outside the United States... it's time for another one of these! Culture war thread rules apply, and you are permitted to openly advocate for or against an issue or candidate on the ballot (if you clearly identify which ballot, and can do so without knocking down any strawmen along the way). "Small-scale" questions and answers are also permitted if you refrain from shitposting or being otherwise insulting to others here. Please keep the spirit of the law--this is a discussion forum!--carefully in mind.
If you're a U.S. citizen with voting rights, your polling place can reportedly be located here.
If you're still researching issues, Ballotpedia is usually reasonably helpful.
Any other reasonably neutral election resources you'd like me to add to this notification, I'm happy to add.
Jump in the discussion.
No email address required.
Notes -
A Tuesday Morning Black Pill: What's the Margin for Error on an Election?
TLDR: Pennsylvania is the most likely deciding state for this election. In the last two elections running, both featuring Trump, the deciding margin of votes has been so small, that in any other context in which untrained people filled out a form, we would expect more than that quantity of people to fill out the form wrong. Have a nice day.
NOTE: This entire post is premised on the idea that the results are more-or-less actual results.
Greetings from the swing township of the swing county in the swing region of the swing state. What's been on my mind lately, but which I can't find anywhere: what's the rate of people filling out their ballots incorrectly? What's the error rate? Not spoiling their ballot or otherwise destroying it so that it doesn't count, but voting for the wrong person. The voting equivalent of walking to the fridge, pouring yourself a glass of milk, then putting the glass in the fridge and carrying the jug to the breakfast table. I'm looking for the people who walk in saying "I plan to vote for [A]", walk in and mistakenly fill out their ballot for [B}, then turn it in and walk out thinking "I voted for [A]!"
Most business studies and data analysis texts put the base assumed error rate on manual data entry at around 1%, though that feels like a round number bias more than a legitimate estimate. And, of course, election day is just about the worst case scenario for manual data entry! Even the best voters only fill out a ballot twice a year, more commonly it's somewhere between every two and every four, and for many voters it will be even longer. Millions of voters will be filling out this particular ballot for the first time, either because they have never voted before, or because the ballot layout has changed since they last voted. Voters will often be waiting in line for hours before they have the opportunity to vote, often at the end of the day after work. Mistakes will be made.
But it's impossible to measure under actual conditions, as it's impossible to distinguish voting by error from voting secretly (except under particularly egregious conditions, like tens of thousands of liberal Palm Beach Jews mysteriously voting for Pat Buchanan in 2000). It's totally permissible, to return to my platonic ideal scenario, to walk in wearing a "WOMEN'S RIGHTS ARE HUMAN RIGHTS" T shirt and vote straight Republican once your wife isn't looking over your shoulder. America considers that not just permissible, but your right, and an important part of the system. The votes are immediately anonymized anyway.
But I've been unable to find anyone talking about plain voter error, any studies done under laboratory conditions to determine a likely percentage. Because, for the Cathedral construed most broadly, this idea is just too scary, it's too existentially dreadful: for some closeness of election, the result doesn't actually reflect anything but luck. A 1% error rate, which is the bare minimum and it is likely higher, would quite frequently swing an election as close as several swing states have been in recent elections. Our political parties have optimized us right out of democracy, the choice is so narrow that there is no real choice occurring. The result will likely reflect the will of the people only in the most tenuous sense.
It doesn’t make sense to look at and functions better as part of a more broad “error rate” (somewhat organic inherent noise). First of all in aggregate an error like this is typically symmetric, ie does not “bias” results. Second, on a basic level it’s like asking how many people fill out their gender wrong to what they intended? I’m imagining quite low. I honestly don’t think it’s ever happened to me at least. But if you want to ask the related question about how many people mistake one name for another, well then you’re taking about name ID, which is studied, and better treated as its own thing. Or if you look at bias towards the first name appearing on the ballot, again that is studied and is its own thing.
Edit: However I should add that a pretty close analogue is the debate about whether an online poll should or should not provide a “back” button. I don’t have any literature on hand, but some say a back button can provide greater accuracy, but IIRC most people use the back button to rethink an answer as opposed to fixing an error, although one potential trade off is that this can lead to more bias (some claim and I agree) because “motivation” to click the back button can vary, potentially/in theory, by response. But most post election polls are not this type anyways and pollsters are more worried about deliberate lies to pollsters because they are inherently biased in a statistic sense (again are plausibly non symmetric)
More options
Context Copy link
More options
Context Copy link