Good morning! I have had very little sleep. So now is a good time to do some #Math in #Public. Thread 1/ https://twitter.com/KimStrassel/status/1324093624611532800

Strassel's is a point that I've seen raised across RW Twitter: turnout in Wisconsin is unnaturally high, as measured by the ratio of votes cast to registered voters on 11/1.

The state of WI *explicitly* does not do this. Why? They have election-day registration.

https://elections.wi.gov/elections-voting/statistics/turnout

This, some (including Strassel) might argue, does not explain the gap. EDR can't be that high, right? Actually according to this 2013 report on possibly dumping it, EDR is a) generally around 10 - 15% and b) EDR rises as turnout rises (naturally)

https://elections.wi.gov/sites/elections.wi.gov/files/publication/65/final_edr_report_02_18_2013_pdf_86368.pdf

So, using the magic of ~ A L G E B R A ~, we can ask: what would the turnout *among registered voters* look like given different shares of the vote accounted for by EDR?

In short: if 10% of people were EDR—a low share—then you'd need 80% of registered voters to turnout, which is high but not insane. If 15%, you'd need 76%; if 20% (which is high, but remember that EDR scales with turnout) just 71%, which is Kim's # here: https://twitter.com/KimStrassel/status/1324093625890791424

If people have objections to my math, I could be told so! But I think in general, dividing turnout by RV is misleading, and the apparent disparity can plausibly be explained by WI's high rates of EDR.

They aren't obviously incompatible—you could have both lots of pre-mailed ballots from RVs *and* lots of same-days. But my point is mostly that there is a plausible explanation besides conspicuous voter fraud for this disparity.

I don't know enough about the breakdown of ED vs. mail-in votes in WI to say

https://twitter.com/AndrewCQuinn/status/1324363476291194880