US presidential election: Why it “feels” like 2016 once again

It is ear­ly days for want­i­ng to pre­dict the out­come of the 2020 US pres­i­den­tial elec­tion based on finan­cial mar­ket behav­ior, as the mar­kets still have a lot of time to swing around before elec­tion day, which they con­tin­ue to do. In 2016, I could derive the elec­tion out­come from finan­cial mar­ket posi­tion­ing when the stock mar­ket closed at 4pm on November 8, so still pri­or to any exit polls hav­ing been pub­lished. This year, the mar­kets are so far out of their range of nor­mal behav­ior (one need only men­tion the extreme and pre­vi­ous­ly unseen dis­crep­an­cy between the rel­a­tive S&P 500 price lev­el and the cor­re­spond­ing volatil­i­ty index VIX) that at this point I am not sure a rea­son­able pre­dic­tion will be pos­si­ble at all. We shall see.

Anyway, pulling togeth­er the clues that I have picked up so far, it sim­ply “feels” like 2016 once again.

To a logi­cian, the reas­sur­ing thing to observe in recent days has been that, when dri­ven to an extreme, illog­ic will be chal­lenged even regard­less of ideo-log­i­cal affil­i­a­tion. Two exam­ples that I, like pre­sum­ably many peo­ple, found remarkable:

One may also be inclined to see par­al­lels between pub­licly label­ing those with oppos­ing views as “deplorable” and pub­licly hold­ing that “vot­ers do not deserve to know” about (at least) one of the Democratic candidate’s actu­al intentions.

Then, of course, those fore­cast­ers that made pre­dic­tions cor­rect­ly reflect­ing the actu­al essen­tial out­come in 2016, i.e. a Republican pres­i­dent in the White House, pre­dict that the same will be the case after this elec­tion as well.1

Enthusiasm lev­els, as indi­cat­ed by ral­ly crowd sizes and FB like num­ber devel­op­ment, seems to clear­ly favor, once again, the Republican candidate.

Finally, the “miss­ing piece”: In 2016, a mem­ber of the Wall Street Journal edi­to­r­i­al board pub­lished an arti­cle a week or so before the elec­tion warn­ing vot­ers that by “pulling the lever” for Hillary Clinton, they would assume “respon­si­bil­i­ty for all the bla­tant cor­rup­tion that ensues” (or so, cit­ed from mem­o­ry). Now, the entire WSJ edi­to­r­i­al board asks ques­tions with regard to Joe Biden and China-relat­ed corruption.

So why might – and “might” it is for now – most poll­sters get vot­ers’ actu­al inten­tions so wrong once again? If it is true, as Trafalgar Group alleges, that only “one in five” or so Republican lean­ing vot­ers is even will­ing to par­tic­i­pate in a tra­di­tion­al, in-per­son poll, because the dom­i­nant media have cre­at­ed an atmos­phere where it is “offi­cial­ly” con­sid­ered unde­sir­able to sup­port the cur­rent US pres­i­dent, then even if one leaves aside the oft-alleged over­sam­pling of Democrats, one would have to “fear” that those vot­ers reg­is­tered or iden­ti­fy­ing as “Republican” that do respond will be rather those that tend to be crit­i­cal of the cur­rent pres­i­dent, fur­ther skew­ing the result toward Democrats. Under these cir­cum­stances, resort­ing to auto­mat­ed ways of polling and, espe­cial­ly, includ­ing the “neigh­bors” ques­tion (“Who do you think most of your neigh­bors will vote for”? in addi­tion to “Who do you intend to vote for?”) that leaves respon­dents room to express their views with­out direct­ly declar­ing them­selves, as the Trafalgar Group does, seems like a log­i­cal thing to do.

Interestingly, yesterday’s IBD/TIPP poll result details includ­ed this: “50% think most of their neigh­bors will vote for Trump, and just 34% think their neigh­bors will most­ly back Biden.” This dif­fer­ence has jumped from 12 to 16 points with the lat­est poll. But, then again, one might won­der whether the dif­fer­ence has by now pos­si­bly become too large to still be indica­tive in a sim­i­lar way it may have been in 2016 when the “neigh­bors ques­tion” was far less wide­ly discussed.

Sometimes, in hind­sight, it can feel as if it may actu­al­ly have been the “anec­dotes”, rather than the sta­tis­tics, that bore the deci­sive sig­nif­i­cance. If one sees a tweet such as this one from a rap­per of col­or with 12.5 mil­lion fol­low­ers, it may, in the end, have “said” more than a thou­sand polls.

_____
  1. https://news.stonybrook.edu/facultystaff/maverick-modeller-helmut-norpoth-predicts-another-win-for-trump/; The Pollster Who Nailed 2016 Says the Polls Are Wrong, Trump Will Win; the IBD/TIPP poll, also cor­rect in 2016, now seems to have start­ed to come around to this assess­ment, too: https://www.investors.com/news/trump-vs-biden-poll-race-tightens-like-2016-ibd-tipp-2020-presidential-poll/[]
Share on facebook
Share on twitter
Share on whatsapp
Share on pinterest
Share on vk
Share on linkedin
Share on xing
Share on google
Share on email
Share on print
Logic is all about relating individual pieces of information in a consistent way. “Consistent” means non-contradictory. An apparent tendency of two pieces of information to appear inconsistent calls for express differentiating clarification, with the logically required depth of differentiating discussion depending on the extent to which both pieces of information appear similar.
We had expressed ourselves confident to be able to predict the outcome of the 2020 US presidential election with a high degree of confidence based on financial market behavior. What we figured out – with the highest degree of confidence – was that we were maximally uncertain about the outcome (keeping us in silent consternation). Which in hindsight would appear exactly as the correct assessment to have been made.
An Egyptian man takes a hostage at knifepoint in the Milan cathedral, threatening to slit his victim’s throat. A once reputable Italian paper tells its readers that police “convinced him to lay down his weapon and release the hostage”, when in fact, for everyone to see on video, police had to forcefully overwhelm him. The full and unredacted video, in turn, explodes on social media.

Leave a comment / join the discussion

Logic is all about relating individual pieces of information in a consistent way. “Consistent” means non-contradictory. An apparent tendency of two pieces of information to appear inconsistent calls for express differentiating clarification, with the logically required depth of differentiating discussion depending on the extent to which both pieces of information appear similar.
We had expressed ourselves confident to be able to predict the outcome of the 2020 US presidential election with a high degree of confidence based on financial market behavior. What we figured out – with the highest degree of confidence – was that we were maximally uncertain about the outcome (keeping us in silent consternation). Which in hindsight would appear exactly as the correct assessment to have been made.
An Egyptian man takes a hostage at knifepoint in the Milan cathedral, threatening to slit his victim’s throat. A once reputable Italian paper tells its readers that police “convinced him to lay down his weapon and release the hostage”, when in fact, for everyone to see on video, police had to forcefully overwhelm him. The full and unredacted video, in turn, explodes on social media.