Looking at Google -> US elections, I see that in many US states the percentage of reporting is >95%. This leads me to think that the last few percentages of votes in some states take much longer to count than the first ~95% of votes. Why?
asked Nov 5, 2020 at 8:46
-
4the good old 80/20 rule? joking aside, my guess is: lots of small counties and not so many huge counties.– FedericoCommented Nov 5, 2020 at 9:37
-
2Some states allow counting of ballots received after election day as long as they're postmarked before election day– waltzfordebsCommented Nov 5, 2020 at 16:30
-
Federico Pareto has spoken– Franck DernoncourtCommented Nov 5, 2020 at 20:49
-
2@waltzfordebs, actually, I believe that in all states , overseas ballots that have arrived 7-10 days post election day are counted.– BobECommented Nov 5, 2020 at 21:20
Add a comment
|
3 Answers
Federico said in a comment
lots of small counties and not so many huge counties
That's part of it. Another part is that mailed ballots take longer to process. They must be opened and then read, and the vote then recorded. Where both in-person ballots and mail-in ballots are counted by a scanner, the in-person ballots are counted as they are cast, and the total is available on each machine instantly when the polls close. All that remains is to transfer the count from the machine to whatever central tallying process is underway. Mail-in ballots have to be fed into the scanner.
Azor Ahai -- he him raises another interesting point in the comments, which is that problematic ballots are counted last. These will generally be more time consuming, so even if they are counted in parallel they won't be completed until later on, and they will in general represent a small proportion of the total count.
-
4Also, I would assume (but don't know for sure) that problematic and provisional ballots would be counted last (i.e. ones that can't be scanned automatically). I believe some states only count them if they matter. Commented Nov 5, 2020 at 16:30
-
1So counting the 80% totally good ballots is quick compared to fixing and curing the last 20 Commented Nov 5, 2020 at 16:31
-
1Then why has North Carolina been stuck at 94% for around 48 hrs now? Commented Nov 6, 2020 at 7:25
-
@DonQuiKong I don't know. Maybe you should ask a new question.– phoogCommented Nov 6, 2020 at 8:06
-
1@DonQuiKong, North Carolina (like Alaska) is a special case that probably needs its own question.– MarkCommented Nov 7, 2020 at 3:35
To (hopefully) clarify a comment for people who might not be aware of this: when you see reporting percentages, those are frequently not percentages of votes tallied; rather, they're percentages of voting precincts reporting. Precincts aren't uniformly sized, and precincts with larger groups of voters can take quite a bit longer to count than e.g. the Dixville Notch vote in New Hampshire, where the votes of all 5-10 eligible voters can be tallied in a matter of minutes after the polls open at midnight.
-
3I think this is often true, but not always. At least, I remember in elections past that news reports explicitly said "of precincts," while now at least some are saying "of the expected vote total" or something like that. It seems that the difference in approach is related to the very much larger number of mail voters.– phoogCommented Nov 5, 2020 at 21:07
-
3@phoog. Yes, this year, they are using the term of expected total because there is so much mail-in and early in-person voting. This answer is correct that "of precincts reporting" is how they usually do it on TV. I read this in a reliable article source before the election, but cannot remember where.– DamilaCommented Nov 5, 2020 at 21:33
-
@phoog That's a very good point — I've added a weasel word to my post to address this. Commented Nov 5, 2020 at 21:44
-
Actually, the same google page, supposedly based on AP, says precincts in German and just total in English. Commented Nov 6, 2020 at 7:26
-
I sort of think the shift from "% precincts reporting" to "% expected votes" may be more about giving a more consistent metric than specifically related to mail-ins (though it may've helped the initiative). From what I saw election night, %s seemed to be basically weightings of precincts by like population/registered voters/past rates than including any attempt to estimate (the very challenging) amount of mail-ins to come. For a while one day NC and GA reached 100% on many sites, when all precincts had values in. Then sites started to tweak it to estimates of mailins reportedly to be counted. Commented Nov 7, 2020 at 8:56
Because another Q was deemed to be a duplicate of this Q, I want to offer an answer to the question (from the closed Q): That "new" Q asks:
Why is the pace of vote counting so fast on the first day? Why does it slow down afterwards?
Consider the County where I live: On election day there were 212 precincts, each precinct has one or more voting machines. So that on election day there are actually 212+ machines that are counting in person votes simultanously.
Contrast that with the 3 counting machines (at the County Board) that are scanning all the mailed ballots.
Even allowing that the precinct machines do not process at the same speed as the more expensive high speed (at the county central), having more than 60 times as many machines running simultaneously, more ballots can be counted per hour.
Notice that I have not addressed the pre-canvass requirements. In my county there are (about) 450-500 people "checking-in and validating" each in-person voter at the 212 precincts prior to issuing a ballot. Contrast that with the (about) 25 people who are "checking-in and validating" the mailed in ballots prior to those ballots being counted.
