CNN poll good news for Trump

Tied with registered voters and a one point Harris lead with "likely voters" in what has become a consistent (but odd) pattern.

For reference, the CNN/SSRS poll taken at the end of Sept in 2020 had Biden up 16 points. Now perhaps CNN has learned their lesson and is no longer creating total messes of polls, but this is an interesting result for CNN.

One of the things I have noticed in this election cycle that is quite different from previous election cycles is the concept that "likely voter" samples have shown better results for Harris and Democrats than the "registered voter" samples. This is a complete turn of events from years gone by and I am really not coming up with the plausible explanation as to why there has been a change. Certainly Trump has made inroads with the blue collar worker and many of your the college educated and wealthy voters have gone from Republican to Democrats over the past few cycles. But those are reliable voters in any election.

What has always been a problem for Democrats and turn out is not in the working class or suburban class voter, but with the inconsistency of the younger and the extremely poor voter (who both still seem to vote Democrat). I am not sure how this has changed in our current election situation from how it has worked in the past. It is hard for me to believe that the inconsistency of younger and the very poor (think project and urban minority voters) is no longer a problem, or that otherwise reliable blue collar voters are going to balk at voting in 2024. But obviously something significant has changed and the obvious reasoning is lost on my right now.

Generally speaking, the GOP has held a 1-2 point advantage when you use the likely voter sample versus the registered voter sample. As the registered voter sample grows (due to all of the automatic registrations during drivers license and such), one would believe that there would be a larger number of registered voters who will not "likely" vote in the election. To think that these automatically registered voters are telling pollsters that they would vote for Trump, but might not vote seems like a hard sell to me. But what other real differences are there from years past?

So what we are seeing in 2024 is basically about a two or three point difference between what we would have previously expected a "likely voter" vs "registered voter" sample and what we are seeing in 2024. A poll like the CNN poll where Registered voters are tied, would have generally seen Trump ahead with likely voters by one or two points. Now it shows Harris ahead by one. I would love to see the inner workings of how the pollsters are defining the likely voter and screening out the less likely register voter? Are they using the same reasoning and questions as in the past, or have they adjusted in order to provide more favorable results to the favored candidate?

Just another thing to think about.