You may have seen or read accounts of the story that Joe Biden will appear to have lost in a landslide on Election Night, then will come back with mail-in ballots to win the White House. This little gem of a story is apparently meant to encourage Democrats not only that they will win even if it seems bad for them, but also to justify an all-out legal attack if and when they lose, just as happened in 2000 and 2016.
I personally find the notion silly, but it’s worth looking at the claim to see how it would really unfold if the Democrats tried this trick, to understand the group which made this projection, and to observe the assumptions which could undo the Democrats more than they realize.
First, the story of the Red Mirage. The website Axios interviewed employees of the Democrat-party scheme team Hawkfish. Axios calls Hawkfish a “thinktank” and an “analytics firm”, but that over does things quite a bit, as I will explain in a little while. Anyway, that article in Axios starts off by saying “it’s highly likely that President Trump will appear to have won — potentially in a landslide — on election night, even if he ultimately loses when all the votes are counted.”
Axios follows that statement with a brief explanation that the scenario is based on the idea that 80% of Democrats will vote by mail, while Republicans will vote the way they normally do. The Axios report was silent about how Independents will vote, but for this case I would have to assume Independents would also vote in their normal manner.
The idea is that Trump would gain a large lead early, but that the mail-in votes would eventually give Biden the win, and in Hawkfish’s opinion it would be a crushing 334-204 Electoral Win.
The Axios article and one from Bloomberg did make a number of statements which are relevant to discussing this scenario, as follows (important points in bold):
“Way more Democrats will vote by mail than Republicans, due to fears of the coronavirus, and it will take days if not weeks to tally these. This means Trump, thanks to Republicans doing almost all of their voting in person, could hold big electoral college and popular vote leads on election night.”
“Imagine America, with its polarization and misinformation, if the vote tally swings wildly toward Joe Biden and Trump loses days later as the mail ballots are counted.”
“That is what this group, Hawkfish, which is funded by Michael Bloomberg and also does work for the Democratic National Committee and pro-Biden Super PACs, is warning is a very real, if not foreordained, outcome.”
“Hawkfish CEO Josh Mendelsohn calls the scenario a “red mirage.”
“We are sounding an alarm and saying that this is a very real possibility, that the data is going to show on election night an incredible victory for Donald Trump,” he said.”
“When every legitimate vote is tallied and we get to that final day, which will be some day after Election Day, it will in fact show that what happened on election night was exactly that, a mirage,” Mendelsohn said. “It looked like Donald Trump was in the lead and he fundamentally was not when every ballot gets counted.”
An interesting detail in the Axios and other media reports about Hawkfish’s claim observes that the scenario giving Biden a 334-204 Electoral Vote win after Trump appears to win 408-130 on Election Night is just one of a number of models Hawkfish considered, is based on the idea that only 15% of mail-in votes will be counted by Election Night, and they do not reveal what the other models conclude, likely because those are less favorable to Biden, and possibly they are favorable to Trump.
Continuing, the Axios article also observes that the Hawkfish best-for-Biden model is
“Once 75% of mail ballots were counted, perhaps four days later, the lead could flip to Biden’s favor.”
“The methodology, described in detail below, was based in part on polling from FiveThirtyEight in August.”
“The ultimate results may well sit somewhere between these low-end and high-end scenarios and will also be impacted by who actually votes, and how voters’ views about their options change over the coming weeks.”
“Hawkfish is not just trying to educate the public about the possibility that Trump could prematurely declare victory, or try to delegitimize a Biden victory if it took days or weeks to determine.”
“The group is also trying to sensitize state and county elections officials, news and social media organizations, and the courts to the perils of premature results — and to the possibility of Trump and his team applying challenges and political pressure to reject a high share of mailed-in ballots counted after election day.”
“And the group is warning voters that rejection rates for mail ballots are higher than in-person voting.”
“To avoid having their votes thrown out, Hawkfish is advising voters to be extra careful about voting early enough and following all the instructions to the letter — or, potentially, putting on masks and gloves and going early either to safely vote in person or return the mail ballot in person.”
“Methodology: Hawkfish surveyed 17,263 registered voters in 50 states and DC, July 1-Aug. 16, 2020, to assess who people planned to vote for and whether they intend to vote by mail or in person at a polling place.”
“Responses were filtered for those described as definitely voting or likely to vote and weighted for state and national registered voter demographics.”
“The scenarios assumed votes at polls would be counted on election day itself (Nov. 3). A scenario taking a week to count mail ballots would translate to approximately 15% per day on average.”
“In another scenario, mail ballot counts took four days at 25% per day. For states that have had high vote-by-mail participation rates, Hawkfish assumed they would take two days.”
“election night vote counts will tend to be more Republican than the eventual final tally. That’s been the case for several election cycles, and there’s a good chance the effect will be magnified this year for two reasons: More voters are choosing to vote absentee, and because of President Donald Trump’s scare stories, more Democrats than Republicans will likely be mailing in their votes.”
“There’s nothing wrong with slow counts of absentee ballots. It’s more important to get the count right than to do it fast.”
“Absentee votes take longer to tabulate because they’re safe. It takes time, for example, to check signatures and otherwise ensure that a ballot is legitimate.”
“The reason for heavier use of mail voting in 2020 is simple: It’s the pandemic.”
“Democrats tend to gain in counting after Election Day because of who votes when (for instance, older voters tend to vote as soon as they can while younger voters tend to wait until the last minute). This pattern has held in states where Republicans dominate as well as in California where Democrats hold most offices, so it’s about when people vote, not who is counting the ballots.”
“Trump will cry fraud regardless of evidence or results. We know this because he’s falsely cried fraud before, even in 2016 when he won.”
“The conclusion we should draw from all of this is that the count will take a while in many states — especially those where ballots need only be postmarked, not delivered, by Election Day. And we should expect the count to move toward Democrats after the initial tally. The news media should prepare everyone for these facts in advance, and then cover the results with them in mind. Just as good reporters have always cautioned readers about early returns and disparate voting patterns in various states, this year they’ll need to be alert to different absentee practices.”
“Many states count their absentee ballots rapidly; the California pattern of taking weeks to get it done is relatively rare. It’s possible that normally quick states will collapse this year given unprecedented use of mail voting, but most have been preparing for that possibility and are unlikely to be completely overwhelmed. As Politico’s Steven Shepard points out, large media organizations will also have exit polls, and while those are hardly perfect, the people who run them have had many cycles to get used to early and absentee voting.”
“The basic principle here is very simple: All votes cast legitimately should be counted. In 2000, we didn’t know who the next president would be for a few weeks, and it was still fine. The good news is that even with the various slow counts, if the presidential election isn’t close then we’ll probably know the winner promptly in any case. If the election is close, then waiting is pretty normal. And when Trump claims fraud, just ignore him.”
So there it is. I will next walk through what I think will happen if the Democrats really are foolish enough to dump 80% of their votes into mail-in votes.
So first, let’s talk about the number of voters. There were stories in 2016 about 200 million registered voters, which was a bit high compared to 158 million actual registered voters and 138 million actual voters in 2016. I mention that because estimates of registered voters for 2020 are likely to also be high. My best guess is that we’ll continue to see around 61.8% of the voter-age population actually vote this year, which would be 157.7 million voters. If Party Affiliation is consistent with 2016, that gives us 52.6 million Republican voters, 57.3 million Democrat voters, and 47.8 million Independent voters. Let’s also assume that 90% of Republicans vote for Trump and 5% for Biden, 90% of Democrats vote for Biden and 5% for Trump, and to be even-handed, let’s say the Independents split 45-45 for each. I’m not going to play games about the Electoral College, but will focus on the aggregate popular vote, since any victory there for Trump would mean winning the Electoral College.
So we start with Election Day numbers. Effectively 10% of Republican and Independent votes will be delayed by Absentee and Mail-in votes, while 85% of Democrats will be delayed (or denied) by using Mail-in votes.
Election Night: Trump 62,394,750 votes, Biden 29,461,500 (65,843,750 votes left to work)
E Day + 1: The remaining Republican and Independent votes are counted, plus 15% of the Democrat votes are addressed. However, errors in submission will remove about 20% of the mail-in votes submitted.
So, E+1: Trump adds 1,324,454 votes, Biden adds 6,031,118 votes
Trump now has 63,719,204 votes to Biden’s 35,492,618 votes
E Day + 2: Another 15% of the Democrat votes are addressed. Again, errors in submission remove 20% of the mail-in votes submitted.
So, E+2: Trump adds another 242,783 votes, Biden adds 4,370,102 votes
Trump now has 63,961,987 votes to Biden’s 39,862,720
E Day + 3: Another 15% of the Democrat votes are addressed. Again, errors in submission remove 20% of the mail-in votes submitted.
So E+3: Trump adds another 242,783 votes, Biden adds 4,370,102 votes
Trump now has 64,204,770 votes to Biden’s 44,232,822
E Day + 4: Another 15% of the Democrat votes are addressed. Again, errors in submission remove 20% of the mail-in votes submitted.
So E+4: Trump adds another 242,783 votes, Biden adds 4,370,102 votes
Trump now has 64,447,553 votes to Biden’s 48,602,924
E Day + 5: Another 15% of the Democrat votes are addressed. Again, errors in submission remove 20% of the mail-in votes submitted.
So E+5: Trump adds another 242,783 votes, Biden adds 4,370,102 votes
Trump now has 64,690,336 votes to Biden’s 52,973,026
Last Day, E Day +6: The last 10% of the Democrat votes are addressed. Again, errors in submission remove 20% of the mail-in votes submitted.
So E+6: Trump adds another 231,222 votes, Biden adds 4,162,002 votes
Trump finishes with 64,921,558 votes (53%) to Biden’s 57,135,028 (47%)
What this means is that this scenario does not really play out the way Hawkfish claims it will. I think they know that, that mail-in ballots will fail to be validated the way they expect, that voter turnout will be different than they expect, from numbers to party affiliation participation, to support from different days. That raises the question of what Hawkfish hopes to accomplish.
That brings up Hawkfish’s short history. Hawkfish was founded in 2019 using money from Michael Bloomberg, whose short-lived Presidential campaign may fairly be described as a train wreck. This matters, because Hawkfish was supposed to be a data ace for Bloomberg, but they failed to prepare him for an effective campaign. After Bloomberg dropped out and Biden became the Democrats’ guy, Bloomberg offered the services of Hawkfish to Biden, who was reportedly not impressed. That suggests that Hawkfish needed something big to convince Biden that they mattered as an asset, and the scenario where a complex crisis condition could mean that data analysis might swing the outcome. That the assumptions in the scenario blow the claims apart, suggests that in this case Mr. Biden might be wise to be skeptical of Hawkfish.
I now move on to the significant statements I bolded earlier, with some comments:
“Way more Democrats will vote by mail than Republicans, due to fears of the coronavirus” – The virus is not really much of a motive to vote by mail. It’s not likely that most Democrats will change how they vote this year.
“it will take days if not weeks to tally these” – Tries to set up reason to delay certification. This is crucial, as judges are very reluctant to change an outcome of an election once it is certified. There’s no reason to think valid ballots which are received or postmarked by Election Day cannot be counted inside of 7-10 days at the most.
“Hawkfish, which is funded by Michael Bloomberg” – Bloomberg desperately wants to be a player, and Hawkfish is his Hail Mary.
“every legitimate vote” – you will hear that phrase a lot. It will be used the same way we hear about ‘mostly peaceful protests’, in order to continue protesting the outcome.
“The methodology was based in part on polling from FiveThirtyEight in August.” – This matters because 538 was and is left-leaning, and they do not compensate for poll bias, even when empirically established.
“rejection rates for mail ballots are higher than in-person voting.” – For good reasons. Mail-in ballots lack the identification procedures used in Absentee voting, and do not compare at all to in-person voting verifications.
“Methodology: Hawkfish surveyed 17,263 registered voters in 50 states and DC” – this matters because while Hawkfish refers to likely voters, they did not in truth base their analysis on that smaller respondent pool, and so this calls their claims into broader skepticism.
“election night vote counts will tend to be more Republican than the eventual final tally.” – This is not necessarily true, but Hawkfish says this to help build their case that delay in counting is important to determining the outcome.
“Absentee votes take longer to tabulate because they’re safe.” – No, they take longer because proper verification of absentee votes is more time-intensive than in-person. Note the tacit attempt to link absentee voting with mail-in voting, which is not at all the same thing.
“it’s about when people vote, not who is counting the ballots.” – dishonest attempt to discount the potential of election fraud, which has been demonstrated to be more likely at the point where votes are counted.
“Trump’s falsely cried fraud before, even in 2016 when he won.” – this is a lie from Axios and the Left. Trump refused to say he would ‘accept the results and concede’ when asked more than a month before the election. Trump reasonably said he would see what happened, and act according to specific conditions. It should be noted that it was Clinton who refused to accept the outcome, who claimed on no evidence that she had won the critical states, and who recently told Biden not to concede “under any circumstances”.
“the count will take a while in many states — especially those where ballots need only be postmarked, not delivered, by Election Day.” – another effort to lay foundation for ignoring statutory deadlines for declaring a winner.
“The news media should prepare everyone” – a subtle reminder that the media does not report, it advocates.
“large media organizations will also have exit polls, and while those are hardly perfect, the people who run them have had many cycles to get used to early and absentee voting.” – media exit polls have been wrong fairly often, including 2016 exit polls which led Clinton to believe she had won. Exit polls this year do not have any automatic credibility.
“In 2000, we didn’t know who the next president would be for a few weeks, and it was still fine.” – False. The 2000 election is a textbook example of what not to do. First, the media bungled calling Florida and some other states, Gore’s conduct after the election and during the recounts and court cases may fairly be described as rude and surly, crowds became angry and bitter, and the court decisions were reported with bias by the media, implying dishonesty and partisan motives by the court, damaging public confidence in government for quite a while. It’s not something we would be OK with repeating this election.
So, with all that said, what is the main takeaway?
I think Joe Biden is in much, much worse shape than we have been told. Trump has rebuilt support in critical Red states, and has taken crucial leads in swing states, especially with Independents. Worse for Democrats, there is little excitement for Biden, leaning to the real possibility that Democrats will stay home in unprecedented numbers.
If that happens, the only course left to the Democrats will be to accept defeat, but do so in such a way that blames the Trump campaign for somehow cheating Biden out of a respectable outcome. So the Left is cooking up excuses to claim that millions of votes for Biden will be denied, simply because the mail-in voting is working out just the way they always knew it would.