Correct me if I'm wrong, but I don't think you can reliably determine president-only votes for each candidate from this data set? You'd have to ignore the possibility of split-ticket voting.
You could figure out e.g. the minimum percentage of R house voters who would have to vote D president in order to eliminate the skew, and assess how reasonable that is?
It's bizarre the hate that was directed at people checking it when the data was sourced in the OP, in .CSV format. Anyone with a knowledge of how to do basic math operations in excel could check it for themselves - which is the point of providing sources in the first place.
Now we're in a situation where the president is tweeting extremely dubious information, at a time when all evidence of voter fraud is being called false, even the stuff that should be taken seriously (affidavits). Spreading fake news reduces the credibility of the real deal.
This is why it's important to actually check info before spreading it, even if you want to believe it. PedeInspector used flawed methodology, but provided all the data used, so anyone could have checked it for themselves. Unfortunately this spread out as a story before anyone (including GP) reviewed it.
Yeah it's hard to say if they messed it up or if they were trying to spray false info at first and then correct it for plausible deniability. According to their note and the internet archive, the info was updated almost immediately after they posted their article, but they didn't update the article until the next day. That and the obvious auto-generated content is some serious red flags for me.
Maybe some of these wards could be cross-checked against other timestamped media, since the county clerk apparently updated the information at 3:xx AM on Nov 4, before the article was posted?
As of now, the only site I could find that definitely reported 200% turnout is the article cited by the tweet author: https://mkecitywire.com/stories/564495243-updated-analysis-five-milwaukee-wards-report-89-turnout-in-2020-presidential-vote-biden-nets-146k-votes-in-city
Take a look at some of their other articles:
https://mkecitywire.com/stories/tag/126-politics
https://mkecitywire.com/stories/tag/9-business
https://mkecitywire.com/stories/tag/8-local-government
https://mkecitywire.com/stories/tag/10-real-estate
Each topic section is hundreds of auto-generated articles from the past ~1.5 years from publicly reported stats. Now look at the opinion section:
https://mkecitywire.com/stories/tag/52-opinion
Four articles, all in September or October of this year, all pertaining to politics. This looks like a fairly obvious opinion push site with a thin veneer of news plastered over it. If you look at the bottom of the page you can see that there are a bunch of identical sites in the network, purporting to be local news but all using the same content.
I'd be cautious about trusting "mkecitywire" as a source - check the other content on their subject pages. It's all auto-generated articles about donations to PACs, market stat reports, housing stats etc., with a few front-page articles stapled on the main page. All of the sites in their network are the same.
Source cited by the original article (data doesn't match): https://county.milwaukee.gov/EN/County-Clerk/Off-Nav/Election-Results/Election-Results-Fall-2020
EDIT: Closest web archive snapshots around when the article was posted. Article says that the original website changed numbers after the article was posted. Not enough info to conclusively know one way or the other.
This doesn't check out against the original source: https://county.milwaukee.gov/EN/County-Clerk/Off-Nav/Election-Results/Election-Results-Fall-2020
EDIT: Editorial note says the data was updated, checking web archive of info...
EDIT 2: Story was published some time between these two snaps, but can't find the data that the story was based on. Second snap was captured after the story was published, but has a listed update time twelve hours before the story was published.
If a lot of elections are around 73% it's extremely unlikely that a new one is at 90%.
This is an apples-to-oranges comparison. The numbers for previous years are based on votes cast vs. voting age population, the number for 2020 is calculated as votes cast vs. registered voters. If you want to compare on the same basis (assuming 98.5% of votes currently counted):
Voter turnout (voting age population):
2016: 67.34%
2020: ~73.8% (3,297,408/0.985/4,536,293)
Voter Turnout (registered voters)
2016: 84.4% (3,004,051/3,558,877)
2020: 90.8% (3,297,408/0.985/3,684,726)
On a comparable basis for voting age population, compared to 1948-2016, it's 1.58 standard deviations above the mean, which is nothing overly out of the ordinary, and there's nonrandom factors that influence turnout as well.
Sources:
https://elections.wi.gov/elections-voting/statistics/turnout
https://elections.wi.gov/node/7220
https://elections.wi.gov/publications/statistics/registered-voters-2016-november-2
The term was originally created as a contrast to the "extended family", which was the predominant type of household prior to about the 1960s.