Read on Twitter
Ok this is kind of too sad to be funny. One of the things going around the internet right now is this theory that 100's of thousands of dead people voted in Pennsylvania. Here's an example:
So I downloaded the csv they pointed to, and ran a quick program. Turns out, a completely reasonable number of people that age requested mail ballots. 128,628. Not "over a million"...
I also ran a quick histogram, which looks like an entirely normal distribution of voters. Some outliers? Sure. But nothing to indicate systematic problems over likely data entry errors. Also, they use 1800 as an anonymous birthday for people like Domestic Violence survivors.
Finally, there are some improbable data points, but the likelihood that these are data entry errors given that this is on 3MM total voters seems pretty high. Also, speaks to a low-materiality problem.
