Is Harris's Rise Manufactured?
Unsurprisingly, people continue to not understand probability.
A lot has happened in politics over the past couple of months. We’ve witnessed a series of realizations among different groups (the White House inner core, the media, the general electorate, and political partisans of all stripes). Many are claiming, “See, this proves I was right all along!”
Now it seems some of these groups have changed to saying, “This doesn’t make sense. Therefore, this proves I was right all along!”
I want to zoom in on a particular group and their collective confusion. By no means do these people have a monopoly on cognitive dissonance or bafflement otherwise. In this case I’m talking about Republicans—in particular animated partisan Republicans.
The charge they are making runs somewhat along these lines:
Harris has always been deeply unpopular.1
She has never been a very successful political candidate.
She has been anointed as the Democratic Party’s presumptive nominee through an unusual process.2
The media is largely biased in her favor.
Since becoming the presumptive nominee, she hasn’t said or done much of anything.
Her rise in popularity and likelihood of winning is therefore illegitimate.
The facts are congruent with their assertions: she is unpopular and always has been, she’s been a largely unsuccessful political candidate when it came to actually getting votes, she is in a situation that bypassed the usual primary process, the media are generally in the bag for her, and she has been quite aloof up until just recently at the time of this writing.
They also can look to polling and prediction markets and seemingly get substantiation. This is where a little understanding of what they are looking at would help them a lot to see the error of their ways including why the argument above is bad reasoning—the conclusion does not follow from the premises.
Perhaps the most salient picture they can point to is this:
Low and behold she has seen a meteoric rise from the depths of nowhere. This has to be fraud!
Not so fast, my friends.
Let’s start with the fact that when you add Biden plus the others over this and the longer span of this election cycle, it is not too far off of where the single candidate, Harris, finds herself. The fact that there were lots of concerns about Biden leading into his disastrous debate performance gives us reason to believe his numbers were not representative of an “anyone-but-Trump” alternative.
To dig deeper, let’s get some more understanding from a couple of football analogies.
First is the backup-quarterback effect. People L-O-V-E the backup quarterback. He represents hope whereas the starting quarterback represents reality. If the starting quarterback is underperforming, people are quick to jump on the idea that the backup is actually the answer.
And if the backup has to come in, say in the case the starter gets injured, it is heads he wins, tails he doesn’t lose. If he performs well, then “see, I told you so!” If he performs poorly, then “well, he doesn’t get the snaps in practice . . . you can’t hold him to such a standard . . . he looked good considering . . .”
This explains some of the recent rise for Harris. If nothing else, Democrats and those sympathetic to a Trump alternative can take relief in that she is a candidate who can actually fully function outside the hours 10-4.
Second is the illusion of certainty. Consider a football game between relative equals—equals on paper at least. This could be Georgia versus Ohio State (two elite teams) or Texas Tech versus Boston College (two middling teams) or Kent State versus Nevada (two league-bottom teams).3 The point is they are relatively about the same and therefore competitive with each other.
Little factors will then have big, outsized effects on the probability of each winning. Further, in a binary probability4 one’s success must come at the other’s expense. Likewise, one’s failures must come at the other’s benefit. So Harris rising could simply be Trump fading—the likelihood has to go somewhere, and that is the most logical destination by default.
Back to football, imagine leading into the week of the game one team’s starting quarterback was rumored to be seriously injured right at the end of the prior game. Suppose further that the team’s backup QB had already sustained season ending injury in a prior game and the only remaining backup is a converted tight end. The betting markets might recognize this by having an opening line largely in one team’s favor even though that wasn’t what was to be expected a few weeks prior.
So Georgia opens up as a big favorite over Ohio State when OSU’s QB seems to be injured unable to play. Perhaps the line is Georgia by 7 points. Based on history, this implies about a 70% chance Georgia will win.
However, during the week it becomes clear that the injury rumor was false as clear evidence emerges that the OSU QB is fully participating in practice. As this becomes clearer and clearer, the betting line would start to move.5 At first it might become Georgia by 5 points. Then maybe 3.5 as it moves down to perhaps even. Suppose further that during the week the Georgia defense now has its own rumors of difficulty. It seems a key lineman has certainly sustained a hurt knee and likely will not play.
So Georgia is now fading just as OSU is ascendant. The line by Saturday morning, gameday, is OSU by 2.
If it helps the analogy, understand that the Ohio State quarterback has never been that highly regarded. He is effective and actually essential to OSU’s success, but he is at best simply an above average QB in the league overall.
Looking at hype around OSU and the fact that the media narrative is that “Georgia hasn’t played anybody” even though Georgia fans protest “our SEC schedule is not being respected!” doesn’t actually tell us much about why OSU might get right back in the mix to win the upcoming game.
Claims like “Georgia’s overrated” and “Ohio State is a fraud” are just fans yelling at fans. To understand the dynamics of who is likely to win, one needs to look at the fundamentals and understand the interaction between the two opponents.
Trump is a deeply and broadly unliked candidate. It would be very surprising to see him not in a tight race even if the other candidate is also a deeply and broadly unliked candidate.
It’s all relative.
They generally also get pretty opinionated on this point saying some version of “she sucks” or “neither I nor any smart person likes her”. These personal feelings are aside from the argument, and their inclusion pollutes the case they are trying to make.
I’m being very generous to the Republican’s talking point here as they scream “What about democracy!?!” completely ignoring that her ascendance while unusual is completely within the established rules.
While I want to say Trump vs Harris is most like Texas Tech versus Boston College since I think both of these candidates are buffoons who happen to be well known (keeping them from otherwise being from a lower division altogether), the sad truth is they are more like Georgia versus Ohio State since they are masterfully doing the politics including avoiding or taking loose policy stances.
Technically it is not binary since there are more than two choices. So mutually exclusive/jointly exhaustive is the most correct description.
For sake of the metaphor please ignore the fact that the opening line might have been withheld or that betting markets don’t usually move lines by as much as theory would suggest since they would then open themselves up to arbitrage risk.