How to avoid both exuberance and despondency this election season

The race is still tied

In my last post, I discussed how the degree of elation on the left was well out of line with the data at hand. True, Trump and the Republican Party had been caught flat-footed by the ruthlessness with which the Democratic Party jettisoned Biden, but Harris was leading by margins smaller than we had seen not just in 2020 but also in 2016! Inasmuch as so many Democratic partisans were already popping the champagne, their sense of a victory already won was premature.

The US left has a sense of itself as less emotional and more data-driven, but two episodes this year would quickly dispel that notion for any impartial observer.

1. The unwillingness to accept that Biden was a poor reelection candidate for 2024.

   1. Biden had a “stutter” right up until the moment he embarrassed himself and his party on national TV.

   2. When Robert Hur described Biden as “a sympathetic, well-meaning, elderly man with a poor memory,” he was immediately dismissed as a partisan hack. He was, of course, nothing of the sort.

   3. The 2024 State of the Union performance was graded on a very, very generous curve.

   4. Poll after poll going back to 2023 were dismissed when they showed that:

      1. Biden had unfavorability ratings higher than Trump

      2. At least 70% of the public did not want him to run for reelection

      3. Biden was running behind Trump in most battleground states.

2. The irrational exuberance and the sense that the race was essentially over when Harris replaced Biden. There was nary a murmur of recognition that she had weaknesses as a candidate. Instead, the woman who only weeks before had been seen as such a weak candidate that she was one of the main reasons it was supposedly a mistake for Biden to decline to seek reelection was now seen as the second coming of Obama.

None of this was based on any look at the evidence. The boring reality is this: the race has not changed much. In ditching Biden, Democrats went from being headed for certain defeat to having a fighting chance. That is all. The news cycle hasn’t changed this, except perhaps insofar as it temporarily made some excited left-leaners more eager to answer pollsters.

It is the job of the media to sell news. A boring race does not generate as much revenue as an exciting one, with one candidate being unexpectedly up, or the underdog catching up, only to be left behind again. There is nothing inherently wrong with wanting to watch a horse race. The only problem is that it makes people alternatively elated, complacent, and despondent, even when nothing fundamentally changes.

If you’re anything like me, you will prefer to have a way to ground yourself to an analytical framework, so that you might limit the influence of your day-to-day emotions on your understanding of what is going on with this race.

The first place to start is that polling, while being the single best tool we have, is fairly inaccurate.

Presential polls since 1998 have been off by an average of 4.3%.

This isn’t has helpful as it seems, since we have no way of knowing which candidate will be overestimated in any given year. Still, you have to start somewhere, so let’s take that 4% and call it our margin of error. When the gap between the candidates is less than 4%, the race is a toss-up. Harris has never led or been behind by more than this in polling averages.

Another thing to keep in mind is that the smaller the gap between the candidates, the less predictive polls are.

A lead of less than 3% is still pretty much a coin toss. But a lead of up to 6% only results in victory about 7 in 10 times. That is still not very accurate. In order to be more confident in polling results, you need to do 2 things:

1. Focus primarily on polls of battleground states. The election is not fought nor won at the national level.

2. Adjust for the margin of error.

The latest NYT poll shows Harris behind, which is causing a lot of heartburn on the left. But that is not a good way to interpret things. Let’s stick to our analytical framework.

Here’s what NYT is showing about the 7 states where the election will be decided:

What’s the proper way of interpreting this? The race is a toss-up in every single battleground state. It’s a toss-up this week as it was last week and the week before and the week before. Nothing has changed. Anything that works by sampling could give you similar results.

I don’t put much stock in the concept of momentum because almost all of it can usually be explained by sampling. Now, if you see a swing of 5+ points from one week to the next, that’s probably real. But a +1 that turns in to a +2 that then becomes a -1? Meh. Nothing anyone should get excited about. Right now, the map looks exactly the way it has looked since Harris became the nominee.

Whether you lean left or right, there is no more reason to celebrate or panic than there was before. At this rate, unless something dramatic happens in the next 8 weeks, we’re only going to be certain about the election results after the votes are counted.