Using SCIENCE to prove what we all know in our gut about current “polling” of the 2020 race for POTUS
Let’s start with the known results from the 2016 POTUS election…
- Trump (45.9%) 62,890k
- Clinton (48%) 65,845k
- Others (6.1%) 8,066k
- Electorate = 136,891k
Current (as of today) polling from RCP (not going to bother to link it because it will change probably within the hour)…
- Biden 50.8% (+8.1%)
- Trump 42.7%
- Others/Undecided 6.5%
So, to try to keep this relatively simple, let’s explore 2 scenarios…
Scenario 1: Trump matches his 2016 raw vote number. Basically, we’re saying that we’re buying-into the polling that says Trump is gaining votes among ethnic minorities but is suffering a simultaneous “White Flight” among, mostly, upper-middle class White Wine Moms. Almost certainly true on the former, dubious at best on the latter, but we can assume if both are true then the raw number total is essentially a wash.
Where does that leave us after Election 2020…
- Biden (50.8%) 74,927k
- Trump (42.7%) 62,980k
- Others/Undecided (6.5%) 9,587k
- Electorate = 147,494k
What does this mean? Well first, it means that Biden – who can’t draw flies to his “rallies” and attracts precisely ZERO enthusiasm anywhere, is somehow, miraculously, going to pull-in >5M more votes that Obama did in ’08 and nearly 10M more than Clinton in ’16. Meanwhile the electorate is going to balloon by more than 10M voters overall. (See the Addendum, below, for some context on this projection.)
Remember, that Obama in ’08 was a “Historic First!” type of candidate who attracted a great deal of enthusiasm to his campaign. He also had the benefit of running against perhaps the weakest GOPe candidate in decades (Juan McVain) AND in a year when, thanks to War Fatigue and the general ineptitude of Bush 43, the GOPe was rated with the electorate just slightly higher than herpes.
And yet, the polls (under this scenario) are suggesting that Joe Biden – Joe. Freaking. Biden. – is going to pull-in more than 5-million more votes than Obama did in ’08. And note, this is before we parse-out the “Undecided” voters from the roughly 6.5% of the electorate not currently going to Trump or Biden – meaning that Biden stands to gain perhaps has many as 2M additional votes from that cohort, above and beyond his already historic projected haul.
That seems… unlikely. To the extreme. I don’t care how much vote fraud we’re talking about here.
So let’s look at a different scenario. Here, we’ll consider that the electorate has not grown since 2016 and is fixed at just about 137M voters. Granted, this is also a somewhat dubious assumption, but I believe more likely than assuming it will grow by over 10M over the last 4-years, particularly considering that there is no “Historic First!!” on the ballot this year.
Where does that leave us on 04NOV2020…
- Biden (50.8%) 69,541k
- Trump (42.7%) 58,452k
- Others/Undecided (6.5%) 8,898k
Obviously there will be no “Undecided” voters on election day, so let’s say that the Others/Undecided break roughly 25/25/50 between Trump, Biden, and Others. That presents us with the following final result…
- Biden (52.4%) 71,766k
- Trump (44.3%) 60,676k
- Others (3.3%) 4,449k
This result is not quite as incredible as the first scenario, but is still pretty ludicrous. To believe this, you have to assume that Biden is going to pull-in more than 2M more votes than did Obama in ’08 and almost 6M more than did Clinton in ’16. Further, you have to believe that Trump is going to hemorrhage more than 2M voters over the last 4-years, net, despite what is widely-accepted as significantly growing support among ethnic minorities (even a modest net 10-point swing with Blacks and 5-point swing Hispanics results in a roughly 4-point swing – +2 Trump, -2 Biden – in the overall electorate, which would have been more than enough to swing the (irrelevant) Popular Vote to Trump in 2016).
So, to summarize, to believe current polling you have to believe…
- That Joe Biden, who again attracts no enthusiasm from any demographic and who’s “rallies” attract crowds that would be small for candidates for State Representative in New Hampshire, is going to bring-in 10M more voters than we saw in 2016 who are going to go overwhelmingly for him – that as a result he’s going to draw bigger raw vote totals than either Clinton in ’16 or even Obama in ’08, AND
- That Donald Trump, despite significant growth in support from ethnic minorities over 2016, is going to bleed overall support by a net of several million votes
Individually, these are a stretch. Taken together they are many bridges too far to be credible. As such, it is simply not possible that current polling is an accurate reflection of the electorate in 2020.
Or if it is, we are well and truly lost as a country.
Addendum: for reference, here was the total vote and turnout (as a percent of voting-age population) for the previous 21st century POTUS elections
- 2000 (Bush 43/Gore): 105,594k (51.2%)
- 2004 (Bush 43/Kerry): 122,349k (56.7%)
- 2008 (Obama/McVain): 131,407k (58.2%)
- 2012 (Obama/Rmoney): 129,253k (54.9%)
- 2016 (Trump/Felonia): 136,891k (55.3%)
It’s estimated that there are roughly 152.7 million registered voters in the USA in 2020. To assume a vote of 147.5 million in 2020 is to assume nearly 96.6% turnout of registered voters (over 56.7% of the estimated 260 million voting age population). This sort of turnout is what pollsters are counting-on for their models to have any validity at all (to say nothing of how they’re skewing the electorate).
FWIW, projecting a static electorate from 2016 results in 89.6% of registered voter turnout, 52.7% of the voting-age population.