Trump still leads in latest NYT Siena poll

Lead cut from two points to one point, but still a momentum blocker as other pollsters started to show post-debate movement for Harris

I am always skeptical of any sort of post event poll that is supposed to be showing a bump or bounce for a candidate. They usually invest a whole lot of time to questions regarding whatever event just took place, they lead people into a particular narrative and then later in the survey they ask about a vote.

So reading the questions and crosstabs of some of these more "traditional" pollsters, there is an awful lot of questions that could be seen as "leading". I understand asking a question or two about the debate such as did you watch, who do you believe won. But when you start getting into the specifics (dogs being eaten in Springfield for instance) now you are moving past the normal polling and getting into an attempt to form or even change an opinion.

As the polling has shown, the first set of polls that came out after the debate showed almost no movement either way. But those are generally a type of tracking poll, or a poll that asks about the demographic of the person (gender, party, ideology, college, etc...) and then asks about plans to vote. But those types of polls do not bring up current events. The questions are the same from week to week or day to day. They do not change with the circumstance. This is one of the consistent things necessary to be considered a tracking poll.

However, the polls that people like Nate Silver and the 538 crew see as gold standard are the more elaborate phone sampling which probably takes 10-15 minutes per interview, where there may be several (even dozens) of questions asked prior to asking about the election and the vote. These are designed to actually highlight and bring awareness to current events to see how those current events might be shaping votes. I would say when looking through the various polls that came out in the past week or so that the more questions were being asked about the debate, the more likely it was that you would see a bump for Harris.

Now, this might be a curiosity issue for some. They really do want to know if an event (convention, debate, assassination attempt) changed the outlook. But shouldn't that just be left up to the people to have already decided if the current event changed their vote? If you have to talk to them about something before asking the questions about voting, then you are not actually gathering an honest sample of the electorate. You are simply gathering a sample of people who first had to answer questions and think about a current event "before" reaching their conclusion. Anyone who changed their mind (or made up their mind) somewhere in the middle of the survey actually is skewing what is actually happening in the real world, as in the real world, 99.9999% of the public were not answering a bunch of questions about current events before deciding who to vote for.

Now in theory, that would suggest that some of these pollster that showed an increase in Harris support will show her losing some of that bounce as the next round of polling moves in and a new set of events are brought up. Of course, if it is actually a long-term increase in support, then it is not just a bump or bounce and the gravity of the underlying election will not bring it back down. Time will tell if there was any real shift in the electorate based on the debate. But based on the extremely small amount of voters who remain undecided, I would expect very little if any real long term effect from that one event.