Originally Posted by James Lacey
Polls are not just raw data. Pollsters change the weight of the data based on educated guesses. They know the poll is not a representative sample, so they massage the data based on educated guesses about why it's not representative. These educated guesses can be wildly wrong. For example, the poll's respondents will have a particular demographic distribution (age, sex, race, education, etc.). Still, the pollster knows the poll's distribution does not match the electorate's distribution, so they weigh the data by using other data to make the poll represent what they think the electorate looks like.
So, the poll may have 1,000 respondents, 600 of whom are women and 400 of whom are men, but they know from 2020 data that women only represented 55% of the vote, so they will weight the poll data to reflect that. In other words, it reduces the influence of women voters in the poll. However, they underestimate female voting participation in 2024 if they are wrong about female turnout. This is the simplest example I could come up with, but they also adjust based on party affiliation, likelihood to vote, region, etc. I guess the point I'm trying to make is that given all the assumptions that go into polling, it's impressive they are as accurate as they are sometimes. On the other hand, I largely ignore national polls. They are pretty worthless in a contest like this. In a contest like this, the policy wonks and people with real election experience who can interpret and predict what's happening in the campaigns and the electorate are often more valuable than the pollsters.
Even Nate Silver says Kamala will probably win if she keeps the Blue Wall (MI, PA, and WI). Harris is not doing badly in those states, especially given Trump's recent foo-bar at his NY MAGA rally. Other policy wonks and experts I read and keep up with say that all the polls tell us is that the race is too close to call using polls. There are many other factors to consider, such as lots of women voting, lots of young people voting, lots of first-time voters voting, etc. We all know most of those votes will be blue votes.
Meanwhile, Trump is hemorrhaging his core base of Republican males. So, he needs new voters, but where will they come from? Nowhere, from what I can see. Other factors are that Trump's campaign is undisciplined and out of gas, and Trump is more incoherent and belligerent than ever. Of course, his base eats it up, but he can't win with his base. All he's doing is alienating more voters and driving them away. There are lots of other reasons why things on the ground are looking good for Kamala, too many to go into completely, but one that I've been getting some enjoyment from is all the lawsuits the MAGA minions have been losing when trying to suppress the vote under the guise of "election integrity." Polls are essential, but they don't tell the whole story.
I could not agree more. Great writeup, James.
