Harris/Trump: Top Pollsters Talk a New PA Poll and More | Mark Halperin

[Music] welcome everybody thank you for joining on the eve of a holiday weekend very grateful to you uh I I'm confident though it's going to be worth your while uh I'm super excited about this I usually am about most of our sessions but especially excited about this to have a brand new poll out of Pennsylvania that's got some interesting questions and results that will run through you'll get the poll in my substack and on Twitter after we finish this session along with the video of the of the event you can share the full poll with anyone you want as long as you make sure credit uh to to Wick insights and to two-way and um and as always you can share the video our four pollsters are four of the best pollsters in America super glad to have David Burell of Wick insights who will go through his own the poll that he did in conjunction there are polling partners with two-way celinda Lake Democratic pollster who's worked with the Democratic party and and presidential candidates including the current one and eler who I worked with when I was at Bloomberg one of the best pollsters in America nonpartisan poster and Patrick gini a republican poster who's got a new piece hat in if you don't subscribe to a substack you should so we're going to hear from all of them as always on this platform this is not about fighting it's about learning peace love and understanding so they're going to have a very civil conversation and as we bring you all in if you want to be part of the conversation here on the two-way platform raise your hand when you come in you'll be equally civil uh and I hope that will extend on YouTube to the chat I know it doesn't always and that bothers some of you I can't control the chat but let's get started by bringing David Burell in right away to the Pennsylvania poll which looked at the presidential and some other questions of great interest so uh David come on in uh and um David's goingon to share his screen and um and uh and then run through the deck and it's going to start with the the methodology of the poll which is on the first slide so David go ahead and bring that in great go and share my screen let me know if you can see it um all right you seeing a slideshow deck yes yes sir there it is all right so the first thing David I want you talk about an address and it's on the first page of the deck which is how do you do your polling or live calls internet how do you do it sure great yeah so um first off thanks again for having me Mark really excited to go through this um so from a methodology standpoint taking a high level look at the poll um so as you can kind of see on the screen here it's, 1600 completes but the question is how do we collect these completes you know and and what methods do we use um so we uh pull random samples of likely voters and new voters and and and all sorts of voter cell phone numbers um and we text links to a survey to those cell phones and invite them to take a survey and that directs them to an online survey which some of youall probably receiv received before um so it's a method of random sampling probability sampling that we um employ um in order to hopefully get a good sample representative sample of Voters but we'll talk about that in a second and then very similarly we take a random samples of landlines and we actually Robo pull those landlines and 25% of the sample came from that type of um recruitment um so and and and just to be clear you people see the date on this this survey was done this week so it's very current but not done after the CNN interview so the extent you think the CNN interview had some dramatic impact in Pennsylvania you'll have to be disappointed but otherwise this is as timely a set of data as you could get today anywhere on planet Earth David great and yeah so for this be short presentation because we got a lot of great pollsters I want youall to get their insights we want to answer questions uh but we wanted to kind of go through three or four things one is take the look at the race what's the state of the race against um from the presidential ballot in Pennsylvania we want to look a little bit the hypotheticals of what if um Trump had chosen Haley what if uh if Shapiro was chosen for Harris's ticket and see if that would have impacted the race um as well um we're going to dive into the issues a bit we're going to take a look at whether you know try to tease out some insights on whether or not Kennedy's um dropping out of the race and endorsing Trump um and campaigning form is going to have an impact or you know where voters are on that and then the last thing is I'm going to talk briefly about a couple challenges that um are facing the polling industry right now um and so with that I'll just get started that's good with you mark yes sir please do all right wonderful all right so we start off just with the five-way ballot test um Donald Trump was up in the five-way ballot test by 0.9% actually so 1% rounded um not too much to go over here um this is you know we're seeing Pennsylvania's close in a lot of different polls out there um and we'll talk a little bit more about um what I think is driving um this specific Pole to look this way kind of closer to the end and we get other um posters uh feeding coming in on the board um second thing we wanted to look at is hold on get my screen to go um we split off the sample um to see what it looked like when you kind of changed the ballot up when you didn't have uh any vice president names on the ballot and it's just a head-to-head with Republican Donald Trump and Democrat Kamal Harris and you can see that was tied very similarly to the five-way um ballot test um when we looked at the hypothetical ballots that's where you started to see a little bit more movement um so and this is where we put uh the vice president's name in the question and in the answers um so they were selecting the presidential candidate with the vice president running mate and you know the one only one real that here that really jumps off the page is well there wasn't a big move with putting Haley on it and again this is not hypotheticals are very imperfect so you know you got to take this all with a grain of salt but there is Insight here um or there would have been an Insight if you know if Trump was performing Five Points better with Haley of course um but of course as you're kind of seeing here you have a very popular governor in Shapiro I think a lot of pennsylvanians are very disappointed that he was was not picked and you're seeing that he actually was outperforming you know significantly at least statistically significant for this poll um at her outside the margin of error um when we paired his name with Shapiro I'm sorry with Harris um the head-to-head against Trump and Vance um moving on from this of course we can always come back later I just want to get all the brains together as quickly as possible we looked at some issues oh sorry my apologies the next slide is actually the uh Senate race my apologies and you can see that um in this situation Harris is underperforming Casey by about three points here um he's got a two-point lead so in this poll right I know it's closer to four or five and some others but that's just where this one shook out um and is a very like Mark said a very uh fresh poll um so let's talk about some issues because I do think that there's some key things that are interesting here um one of them is uh obviously you're going to see stopping inflation uh as the number one um issue for people um legal immigration but preserving the Integrity of the American Democratic process is also really high up there and that was also high up in 2022 it really dipped down after that um you know uh know saving democracy and and you know and um threats to democracy as I think how we were all asking it back in 2022 um but it went down and it's now creeping back up and and what you're going to see in a couple different slides is it's actually creeping up not because the of the left but the Democrats starting to index it higher than they were before it's more because the Republicans are um and The Independents are so you're kind of getting a combination of that doesn't just mean Jan 6 to people um like it kind of did two years ago um taking a look at um which you so we list a number of desirable outcomes that I think most people believe desirable outcomes and we ask which candidate Trump or Harris would uh be more effective they think would be more Ive in achieving those outcomes um and as Mark said last night I'm sure a couple of these are making um Trump's team pull his hair out uh because I do think that it's real that Cala on some of the issues has been able to close the gap and I think that that matters and the question is will it sustain um as Mark was saying yesterday that you know Trump's team thinks they can through advertising and just more exposure they can fix this but it's it was it's a real shift um so and again to highlight what I was saying before preserving the Integrity of the American America's Democratic processes is right here and and it's concern for both I mean you can kind of see both sides are are are concerned about this okay um view alignment um so this is you know kind of standard um which one's views on the issues do you think aligns most with yours and you're kind of seeing um where that landed not too many surprises here again we can come back to any of these anytime so what about RFK um you know I I I think that I have a lot of thoughts about the polling surrounding RFK um dating back to his high points and his low points and what was real and what was not real um but what you're seeing here is you know from the visualization from the cross tabs where we say um you know if you have a fav or unfavorable opinion of if you have a favorable opinion of RFK what you're saying here is you um are supporting Donald Trump if you don't if you have an unfavorable opinion of RFK supporting Kamal Harris so I don't I don't know how this is ultimately going to impact the ballots right but if you're this is a pretty good slide to show if you had a favorable opinion of um RFK who his people are going over to Trump's side more than they are to Harris aside um and I'm not saying this isn't saying if you were going to vote for him before he dropped out it's just saying if you had a fa a opinion of him um so moving on one of my fun slides is so this is a bit about methodology so Mark may want to talk about this a little bit more but y'all I'm sure we've all grown how inaccurate polling can be sometimes and how it changes all over the place and on this slide uh that we're recognizing have been really recognizing the past few years um when I was talking to you about the methodology we were talking about using you know probability sampling um a lot of the polls you see out there now um not Mi any names it's not a bad thing but uh they're using panels um which is people who are opted in and paid to take surveys so it's it's a different type of um it's a certainly a very different type of uh uh sampling technique um that has challenges because it's not random sampling right and uh so and there's ways to improve it it's great for ad testing but there are issues with it so our methods are supposed to be random samples um and what you're seeing here which was not the case um years ago and and and also I think that these numbers would have changed four months ago which we'll talk about in a minute is when we're naturally doing these random sampling the the Democratic um Republican independent party identification right this is party ID not necessar registration um came in kind of how we think it should have um if you go look at a number of the polls out there right now um this is about where they're having their party as well um but at the same time so that's com in looking good um education is coming in terrible just naturally so this is halfway through the poll we start looking and saying hey where do we need to start adjusting our sampling techniques to hit our quotas we say oh wow we have way too many postgrads and a few too many fouryear degrees coming into the survey um so why this matters is is there's issues with random sampling that did not exist in my opinion years ago and I have my thoughts about why that is and we may get to that later um and and it's driving some issues with the polling and an example of that would be that um and look two of the three polls I looked at in Pennsylvania prior to running this one didn't even add po postgrad degree they add they had four years or more so you might not even be able to catch that so even though you've asked education you actually have way too many postgrad degrees in your survey and you know if we allow this to continue to happen just to let you know the impact of it is the independence um would have been breaking towards more towards kamalin you would have been winning in this poll so it does make a difference and this is just one example of issues that I'm seeing across the board and Survey sampling that are driving big changes in polls and I think that naturally if you had have gone right after the Biden debate um I don't think the postgrad would be this High I think because they're much more likely to be conser liberal and they're dejected and not taking surveys so there's some new techniques we're having to uh address or or utilize in order to try to get representative sampling um and address some of the issues with the polling going all over the place as it has been over the last you know few years really probably since 2016 maybe even before that so that's that's it for the presentation hope it was David thank you that great David thank you unshare your screen we only have selinda for for about 15 more minutes Sinda so grateful to you because we all know how busy all of you are but love to get your reaction to to what David just presented and including if you would weighing in on this question of of how a pollster keeps from oversampling and looking for likely voters folks who with post uh Advanced degrees and what that means so David really interesting Poland and great to be seeing you uh it is a little bit more conservative than most of the polls that we're seeing but I think you know Pennsylvania is really tight and as other Battleground states have been moving in Harris's Direction this has been one that is a little more resistance in part because people don't realize but you would that Pennsylvania is the oldest of the Battleground States think it's Arizona or something but it's not it's Pennsylvania um so that that obvious that a age distribution uh works against teroris because she's she has such a dramatic age relationship uh and you saw that somewhat in your polls there's also a big gender gap in uh in Pennsylvania and your data shows that as well and I think this is uh emblematic of all of the close States we've got to win women by more than we lose men that's really the formula for success and there are two parts to that it's and an is an expert on this but it's real whacka you know you got to keep men Sullen but not mutinous and you got to win women enthusiastically um and then the turnout of African-American votes is always up for grabs in Pennsylvania um so uh that's all of issue this education piece is a very big deal and education is a stronger predictor right now than um a lot of the variables that we're used to looking at and the education Gap is bigger than some of the other gaps we're used to looking at like uh the gender gap the marriage Gap the age Gap so this education piece is very big most of the private posters have some way of dealing with this it people have become aware of it and I would say all the private polls uh are either setting quotas stratifying their sample although the census information on education's not very good and I'd be really interested to know what methodology you're using but one way or the other um and most of the private polls have postgrad for exactly the reason you're suggest because you really get a bounce with the postgrads um so that I think the private pollsters are quite aware of it and we are constantly critiquing a lot of the public polls and saying it's too educated it's too educated it's too educated that's why we're looking better so that is a very good thing for you to lift up and I think it's a very good criteria and people have no idea how non-college educated America is I mean two-thirds of young people are not going to have four-year degrees and people think every young person is on a college campus not really um and in fact if you look at the white male new young people Trump is winning them um and RFK probably helped him with that particular audience so um it Hur him with a lot of other audiences but um so I think the college thing that you're lifting up may be the most important thing for people uh on this Zoom to remember and to take away and it's really good that you lifted it up I want to add one thing and get y'all's feedback on It ultimately as well is and we're we're trying to you know as y'all know in the space it's rapid fire when you have clients paying for polls and those are sometimes the best polls right because you got it it's not just the public ones we put out there right um we're also seeing issues because it illustrates the point that you know that and if you had true like random problem sampling like that would not happen like it it there's something driving it and we're also seeing I don't think it's the only thing I think it's the one we've kind of learned from from 16 and 20 and we're catching at least privately but we're seeing stuff if we could spend more time on it also on uh geography um you know saying you got every you got the right mix of urban Suburban um rural small towns well where are they um and we're seeing it's like the coming it's not distributed right and these areas are very different so we're looking at ways to bring more data in to say how can we like create better you know non-standard Urban urbanist Geographic um segments to look at things that and say what yeah we have the right people in urban areas but they're all in this type of urban area or something like that so we're seeing that too it's just it's a sampling problem that's what the key takeaway I want people to understand like you said there there's issues I think I think most of the private posters stratified by geography we certainly do and we and so that protects a little bit we don't get all of our Urban sample in si City Center or in subur we we stratifi pretty aggressively uh the other thing that we do is we and this is something that that caught a lot of pollsters not an and us actually but a lot of the other pollsters in 2016 was um that we had different response rates and now and we at the time were replacing not Trump a lot of people were just replacing someone that they couldn't get with someone else and that meant they often were retracing a trump voter with a Biden voter we even at in 2016 were modeling what kind of voter we thought you would be and we replac a trump voter with a model Trump voter um and now that's pretty common practice and there may be other techniques that people are using too so response rate matters a lot as well yeah Patrick I'd love for you to talk about Pennsylvania about David's poll and about methodology in whatever order or ratio you'd like sure well I agree with the assessment that this is the one this is the state it's both it's both the most it is both the most important state in this election and the one that um you know I think it is ultimately going to prove the hardest of the Rust Belt states for Harris to crack and that's for a few reason one of which Sinda mentioned which is the age uh the age Factor um I think to some extent Joe Biden was actually did have a little bit of a HomeTown effect in Pennsylvania where you actually split the state in down the middle in half in the eastern part of the state um close to where he grew up sort of influenced by the Philly media Market he did three or four points better relative to 2016 than in the western part of the state where Democrats have been bleeding support um for election cycle after election cycle um you know I think the fact that Joe Biden is Northeastern Catholic probably helped him right that was probably a factor that helped him not something that's discussed and not something that really Harris can replicate um so I think losing um this matters those kinds of factors obviously matter less and less but they still matter in a very very very tight election um and so um I want to also point to to to the issues we've been discussing um there's a post this week by Mark Harris uh Republican strategist in Pennsylvania that um talks about this exact same issue that he is seeing in his polls of the state where um what you're seeing is a distinct drop off not of white workingclass voters generally but rural white working class voters so even when you do all the stratification by education when you do the quota uh you know by party either by party ID education geography what have you um if you're not taking into account the interaction of geography and demography right um what you're gonna what's going to end up happening is those rural white workingclass voters who are probably you know three to one Trump at this point um are going to get replaced in the survey um by maybe a white ring class voter in Northeast Philly that's for Trump but not quite 7525 is more like a 6040 voter and as a result you might get results like the New York Times poll which New York Times poll I think they are an excellent poll however uh you know in the last in the last survey they had at Pennsylvania um it showed Trump ahead by what among white non colge voters by I think it was 10 or 15 points which just doesn't really wash as a as but all polls are going to have those you know and you know I don't think we should be cross tab diving or or or or or picking apart polls too much because all polls are going to have some disparity like that but just to highlight in both 2016 and 2020 the polling error was far worse in these Rust Belt states than it was in the Sun Belt states um it was far worse in a state I mean it was the worst it wasn't it was pretty bad in Pennsylvania the worst state worst offender both both years was Wisconsin um and that's because and I think you know you look at where the Trump vote is in these states and it happens to be a very rural vote and in many cases a vote that uh does not look like a traditional Republican voter um in many cases these are registered Democrats and still registered Democrats in the file uh and you've seen Republicans make pretty strong registration gains in Pennsylvania because as you have these older rural Democrat s coming off the file or switching parties but if you have the right let's say party registration mix in your poll you're going to probably get it's going to be a more democratic sample because you're going to get partisan Democrats and partisan Republicans when you have a pretty significant slice of that electorate that is not a very partisan Democrat that is that that just hasn't updated their voter registration yet great thank you Patrick Ann is going to share her screen and and run through some thoughts she has uh and then we're going to I'm going to ask you a couple questions that I'm interested in we're going to open up to everybody else so and go ahead and share and and take it away well thank you Mark and and thank you for sending the data for me to do a little bit of a dive uh as a pollster and I'm sure that the other three would agree we like to look at a lot of numbers all in one piece of paper to see if we can spot Trends and um the Dynamics internally to the poll so David I very much appreciate you're pulling here and I thought some things jumped off the the page a little bit so you tested a number of horse races and that's what I have across the top here the overall numbers as you showed Was a onepoint Race for Trump when it was five way and just a head-to-head it was a tie when you they mentioned the current vice presidential candidate it gave haris a onepoint edge if they swapped out Haley on the Republican side that went even a little more strongly for Harris that's a point it's about the same but Shapiro is the The Game Changer in a poll in in Pennsylvania so here are the overall numbers like selinda I'm interested in the gender gap because I think the the difference between what women win with how Democrats win with women and how Republicans whenn with men is very telling about the whole size of the race so among men you see consistently that Trump has the advantage and sometimes it's as high as 14 points um when they mentioned Haley it drops down to just Five Points McCormick and the senate race is looking to win handily but you see this big number swap to Harris leading if Josh Shapiro were her running mate so a complete swap from where the overall numbers are with men to a switch to the other side with women as celinda said you have to win more women than you lose men if you're a Democrat and you don't really see that happening it's pretty consistently Harris all the way across the biggest number is in the two-way race um Casey winning by about the same margin with women and then I'm always interested in who might be swing voters so I pulled out the people who self-identify as moderates and the margins are huge Harris wins by 33 points in the five way and as much as 43 points if it were Shapiro there but they're pretty consistently very high numbers so I don't know um the role that moderates are going to play in this they're 32% of this sample so it's not a small uh slice of the electorate here but that's a big number you don't see the same Thing by pulling out people who identify neither as Republican nor as Democrat but as Independents and that's a smaller slice of the sample but you see it's red so meaning the the advantage goes to Trump across all of these races it's a tie if they mention the current vice president name I'm not sure what that signifies but it is a ninepoint advantage to Harris with in the hypothetical where Governor Shapiro is her candidate it um again it goes back to Red in the senate race so that Independents look to be more on the Republican tilt moderates and I i' I'd be curious David if you have some thoughts about this and say are are exceptionally strong on the um Democratic side well yes my first fault is I hope we didn't code moderates wrong on that uh on the cross tabs we brought the data in last night um but um so no that that's surprising to me um and and again I do know that if if we again did not control for certain things such as education if that changes like you said it is such a strong predictor right now you know if a more post educate postgrad group turns out to vote then those Independents start to flip over to um Harish pretty quickly just because there's such a strong correlation between um an independent's education level and who they're going to vote for in this election um but no I I I'm surprised to see this and I need to I'm going to have my team I'm texting them now to check and make sure that we had that unless y'all y'all have seen this something that high in the other polls I'm wondering if something was wrong with the cross tab to be honest on the moderates okay let's let's wait and see if you get a real-time response and if you could unshare I want to ask one question I'm gonna not ask both my questions but before we lose Linda I want to ask about Reproductive Rights and go ahead and put up the this the the tweet that I did and and this is something you've seen Donald Trump in the last week yesterday on two issues previously TR social saying he'd be so wonderful for women in reproductive Freedom he clearly is feeling the pressure of this issue and and the potential it has to cost him the race not overstaying the case so go ahead and put up that tweet if you would and and selinda I just want your reaction to this this issue which is when you see it you'll see a number of Republicans have won quite handily in races in important States who were pro-life staunchly pro-life Doug duy in Arizona Brian K in in Georgia Glen yunan Des santis these people are are more pro-life than Donald Trump and yet they've been able to win in competitive States particularly the two that are Battlegrounds here so slenda I'd love before we lose you to just comment on that question of why is this seemingly more of a threat to Donald Trump than it was for Brian Kemp in Georgia than it was for governor Ducey in Arizona and others besides the four listed here hold on got go ahead ah thank you uh one of the things David to reassure you there are a lot of moderate damps in PA and so uh maybe your cross tabs wrong I actually think your cross tab is probably right okay good I didn't look at that when I I shout it out to high are good so first of all I would say some of those Democrat some of those Republicans were not thought to be uh anti-choice uh just like Donald Trump was not thought to be anti-choice in 2016 I mean we test it over and over again and people said said like he's I we said I was moderating groups in Michigan he's going to defund PL Parenthood and people laughed at me and said he's probably paid for three abortions that plan Parenthood don't be ridiculous um so I think there is a post-dbs world and a pre-dbs world uh but people didn't uh we pulled against junan on abortion and people did not believe he was anti-choice um or that he would act on it and you know he ran a very good two-tiered campaign where he was right-winger by night in moderate by day um and and that two-tier he got away with it it really worked um so that's Point number one point number two is people don't want the states to act on this they want a national action one way or the other I mean the pro-choice voters want National guarantee by two to one voters want National over State action the pro-choice voters want National guarantee the pro-life wers want National ban so the presidential and after the Supreme Court decision the presidential position is a lot more relevant people think oh I don't think my Governor is going to particularly focus on this I don't think my Governor is going to act on this for example with Camp people think I don't think uh Camp's really focused on the abortion issue one way or the other he's not moving on it um so I think that is part of the reason and then um a much more aggressive campaign has been run against Trump on the abortion issue and it's been a two-year campaign and and um it it really moved members it moved from 35% to 76% who thought he would sign a ban and that's when he must have seen the same numbers because he came out and said I won't sign a ban and had Vance say I won't sign a ban um so he's trying to have it both ways and they have been trying very very hard to rightsize themselves because he thinks his party went too far and you know like on the IVF issue the dirty a little secret IVF is very expensive 46% of the people who have IVF treatments are Republicans because Democrats can't afford ibf um Democrats often have insurance that doesn't cover IVF so this I think Trump knows his peers he knows uh that that the party has gone too far and he wants desperately to get this issue uh he wants desperately to have it both ways he wants this issue to be less of a focus and and those Haley numbers makes sense to me because they're pro-choice Republican college educated independent moderate women of the Haley vote those are the people that have gone to uh Harris so I'm not surprised when you put Haley up it's like yeah I'm going with both my girls yeah slenda before you go if it can if you need to go you'll just tell me if Trump continues these gambits free ibf saying he's not going to vote for the Florida Bill uh ballot measure even though they kind of pulled it back saying saying on true social he'll be great for Reproductive Rights is what how would you rate the chances of success that he can take the edge off this issue and not be seen as anti-woman and anti-abortion be really interested in an's opinion too I think it's tough because we still got the Supreme Court I mean the law is the law and unless he's going to uh reinstate a national guarantee for abortion and one of the things that was great about him saying I'm going to leave it up to the states the day after after that uh Arizona made their decision and they didn't send it back to 1960 they send it back to 1864 so proving the state some states are bat crazy and that two-thirds of the vote that doesn't want to leave it up to the states doesn't want to leave it up to the states and are impacted by people movement like the Illinois voter is very impacted by Missouri women they've set up whole clinics along the border um so I think voters no matter what he's doing voters think it's your Supreme Court one of the other things that I think is more Salient Democrats never voted and Independence never voted the Supreme Court the right voted the Supreme Court not true anymore after dos and one of the things when you say to people you have to tell them how many judges there are because they don't know how many judges are in the Supreme Court but when you tell people that Donald Trump could appoint three justices of the nine and that could take him to a 7 to2 margin people are like whoa that will not be good idea and so I think that Supreme Court narrative can help reinforce the abortion narrative slenda thank you if you do need to go we're grateful to you for joining okay so thank you for joining I appreciate it uh Patrick and then and then an and then David brief comments if you would on the abortion issue and then we're goingon to open it up to to the group but Patrick what do you think of of Trump's chances of diminishing the impact of this one issue which he's obviously quite concerned about yeah I mean I think I would uh I think both candidates are really trying to to whitewash in some degree like the vulnerabilities that they have on the issue starting with you know obviously Harris moving um pretty strongly in the direction I mean changing many of her positions but particularly uh embracing uh going much further than Biden did in her rhetoric on illegal immigration and border security uh and really running an ad out of the Bill Clinton 1996 playbooks talking about um you know how she was a border state prosecutor notice they don't mentioned California uh and um and I I think likewise uh Trump is trying to do the same thing on on the abortion issue right um and um in in many ways returning to um I think that 2016 view of him right where um you know he was not um you know he had to Plate the religious right with uh you know the idea that I'm going to uh n Nate Supreme Court Justices off this list uh provided to me by the Federalist Society because I don't really care very much about the issue the issues I care about are trade and immigration and I'm just going to Outsource this whole piece of it so he does that and it comes back to fight him right I mean it comes back to say and it comes back to you know a political situation that he doesn't want and he never wanted but um it was never um you know his priv it as president to really kind of steer things steer Court appointments right in a way that other presidents um would um and so um you know to some degree he's returning to I think his initial political instincts from New York um when he was vocally pro-choice and you know I don't know what his actual position and you how his actual position has changed over time um but I do think he has a nose for where voters in the center of the electorate are on this yeah an love for your brief comments on abortion then I want to go to the community questions yeah I think it's important to remember that in 2015 2016 U Donald Trump made his bed with the evangelicals and that is the Cornerstone of his support and if you think that that there aren't more Trump voters to be recruited out there because of how how polarized things are you have to worry about that rock that even angelical Rock uh trickling away because they were very much in it for the abortion question they were from focus groups I did around the state they're all about saving babies and if that's not the the worry anymore and they listen to what Trump is saying this time to go I don't know why I'm voting at all right I think that's more more of the danger then that they will switch parties got it thank you all right debate goe I think the debate's going to matter and I think one thing it does is is probably limits one of Trumps gives him an easy answer and and and something that people probably don't know I mean not everyone's following what he's saying about this issue right now and so I think you know one of his uh liabilities and one of his vulnerabilities in the debate it's going to probably help him just he's got he's got answers that people going be surprised about and want to hear I mean like uh selinda said you know from whatever 30 something to 70 something that uh percent said he would do National ban and so um but I don't know if he'll win people back with just advertising but that might be kind of a help for him I think yeah okay great Ash come on in if you want to get a question and please raise your hand got as many as you you can in a remaining time Ash tell folks who don't know where you are who you're supporting and any question you want polling methodology specific public data whatever you'd like yeah so uh Pittsburgh suburbs uh on the Harris side of the fence love the experts um I'll say a quick few comments and go back on mute so my kids don't ruin the the chat here but just a few things in a margin of er race I look at a few things so enthusiasm head thehe head favorability Gap and then what's happened on the voter registration right so I look at some of the latest scallop numbers Harris is up to 78% I think in enthusiasm from Democrats to 62% uh for Trump big so just for the group like that number is the highest since bomber 08 I'm gonna Ash I'm just want to ask ask a question because I do want to get folks in so if you got a question sorry um my question was what do you guys make of all the the the voter registration information that came out today that showed young black women up 175% increase black women up almost 100% what do you guys make of those numbers that came out comparing this ago thanks Ash where was that data that was um was on CBS news today it's from Target smart it's just using voter registration data okay Patrick do you know about that yeah no I've seen that I've seen that I've also seen really just consistently throughout the year Republicans have actually increased especially like in a state like Pennsylvania but have increased their share of vote registration in across States part of this is again uh it's not necessarily A reflect it's a trailing indicator it's not it's a a change in attitude so all year uh in really leading up to this Republicans have been I think doing a little bit better on the voter registration front but that's the caveat the asterisk on that is that's when Biden was in the race and you had a demoralized democratic base as long as Joe Biden was in the race and so um I think there's probably a fair amount of pent up energy on the Democratic side in the last few weeks in a state like Georgia in particular um you've had probably Democrats playing some catchup now thing in a state like Georgia we don't know if they're Democrats or Republicans right when they register in a lot of States we don't know if they're Democrats or Republicans in Pennsylvania we do but in Georgia we don't until they pick a primary you know until they vote in the primary we don't know but um looking at specifically these specific demographics where Biden was really lagging behind um but Harris is has been able to make up some ground now that's said I I I think like you know I think in the polling we've done the polling we've released we're see Democratic enthusiasm Spike but it's spiking to the level that Trump enthusiasm has been all year um so it's not that that that there's necessarily one side has the advantage both sides are very energized and I think that's going to mean higher turnout um you know by at least 10 million votes compared to when Biden was in the race and getting to 2020 levels or not necessarily I think that maybe maybe we will get close to 2020 levels in absolute numbers obviously population growth since 2020 I don't know that we'll get to 2020 but um but we'll see okay an or David do you want to weigh in on that I'm going to go look up the CBS News repord okay keep thanks for the tip all right Stephen go ahead no I didn't have much to add in addition to what Patrick said okay Stephen come on in tell folks who don't know where you are who you're supporting and and your question about polling I'm in um uh San Francisco Bay Area uh County and uh I'm voting for Trump and my question is really I'm I'm a numbers guy so when I see you know typically that no single women tend to go gravitate toward Democrats and then if you look at who's going to school more and more wouldn't that just normally you know if more women are going to school wouldn't that just pump the education numbers higher Democrat and I mean so does it mean the same thing as it used to be because you've got people that are going to like General Studies they're getting degrees in things that maybe don't require degree they don't know what they want and you know they're still being bartenders but now they were used to be non-educated now they're going to be registered as educated it's it's the same that demographic is changing I I'll just chime in here Stephen that that the the and I don't have it in current data but not only do single women Trend toward Democrats so do single men so there's as much of a marriage Gap whether you're married or not is a predictor pretty good predictor of which way you're going to vote David no I I agree I think that I mean if you ask me right now you know if you take a look at a lot of different voting groups and Will trump do better or worse than he did in um in 2020 compared to 2020 this year I'd say single women especi I mean probably almost in all age brackets I think could probably do a little worse on um whereas married women I don't think he'll do worse I think he'll do about what he did which wasn't good um um but I do I will say that the single men as as as which just mentioned are left leaning as well especially the younger ones but they're um they're trending back to the direction of trump compared to the 2020 numbers at least in our in our research okay Stephen thank you for the question appreciate it uh Michelle tell people who don't know where you're from who you're supporting and your polling question yeah uh I'm from Pennsylvania I'm voting for Harris I'd never vote for Trump uh I was just wondering if there's a cross tab about uh educated not school educated but High information voters versus low information voters and what that tells you you mean specifically in David PA or just in general well since we're mostly talking about David's poll yes yeah um no we did not ask uh question that would give us an indication of you know how much they're how closely they're following the race and that self-reported uh data is not always that reliable as well um so but it definitely can give you insights um but we didn't have one on this one I'd be curious if an or Patrick has anything to add on that theyve seen in other polls it's not something we've I think it's an interesting question it's not something that we've we've tracking too much because it's kind of hard to measure Ann and Patrick is that is that something you'd aspired to know and if so is there a way to get at it that's not self-reported that would be meaningful and coste effective I would say no there's not a way to get that other than What's self-reported um there this this election cycle is so different in so many things that are breaking major news stories it's it's almost hard to ignore what's going on with the campaign because of such startling news development so my guess would be there are fewer who know next to nothing and we've seen that as a as a decent slice of the electorate in past Cycles I think that's going to be smaller this time around it Patrick I I don't really have anything to add I think that uh everything I've seen it's it's fairly even but um the challenge with polling is what we we really really do want to know about those people who aren't paying very close attention to the election because even if you know they're evenly split right that's where the undecided vote is going to be people who are paying close attention generally know where they stand politically and so I'm really interested in that other half of the equation yep all right Michelle thank you come back regularly enjoy having everyone particularly those of you who live in the states that are picking our president Jason come on in tell folks who don't know where you're from who you're supporting and ask question I am from the San Francisco Bay in the city in county of San Francisco itself um I had been supporting the couch when it was Trump the Biden and now I'm Loosely in the kamla Harris Camp my question is in terms of pulling motivation to work uh rather to vote it seems like this is definitely going to be a turnout election rather than a convincing election and so for me I'm I maybe a little bit less concerned about the inches in the middle of the candidates and where the UN Independents are going to fall and maybe more you know what amount of excitement there is to vote for Trump to vote for K haris to vote at all and so how do you actually you know discern someone's likelihood to vote given that everyone wants to think oh of course I'm gonna go and vote and I'm engaged and I'll get out there when in reality we know that you know that's simply not the case yeah Jason let me just say from my point of view I'm not as expert as others I think every election is a base election and a persuasion election I think it's a mistake to ever say this election is just one or the other particularly if it's close it's going to be both um we talked a little bit about likely voters uh and and and Patrick David an I'd love for you guys to say when people look at polls and as we increasingly move away from registered to likely how do you know if a poll has got a good screen does it matter that much it's very difficult to to even find a definition of what they what the poster has decided is going to count as a likely voter we have a very simplistic approach and if you look at it on paper you would say this can't possibly work uh but we have a single question that ask people How likely are you to vote will you definitely vote probably vote might or might not vote or probably not vote so the to to the extent that it has worked very well the beauty is that there's a positive social uh positive social answer in probably voting which is a step off of definitely and we don't take you if you say probably we only take the definites now is it true that there are some people who tell us they will definitely vote don't that's true are there some people who say they will probably vote who do vote that's true so far and all we can tell is how accurate the polls have been it works out in terms of the cross-pollination there can can I add maybe a third category because I think there's going to be record amounts of early voting so is there like another category you would add of have already voted yes when and and that's they're just now starting to to allow people to apply for an absentee ballot absentee voting depending on what state you're in is going to happen in October uh some maybe some states late September once that is started in a state where we're polling we say or have you already voted that that's the best kind of likely voter you've got because they're vot's already in the camp all right thank you Jason appreciate it come back off and I'm glad you're off the couch appreciate it all right unmuting mandarava if I pronounce that correctly we're close to it yes yes all right tell everybody who where you are who you're supporting and you're pulling question please I'm in Los Angeles and uh and what what were the other questions you wanted me to answer who you're supporting Trump trump okay and what's on your mind about polling are there and please answer honestly now that you would do otherwise oh crap hello embarrassing oh God okay um are there pollsters with bias that are intentionally putting out inaccurate information yeah you guys get that question all the time too as do I um so who wants to have at it it it there's lots of Suspicion on the left and the right if there's a poll that they don't like that the poll was done crookedly in order to influence the outcome in your experience guys does that happen well I think it but one thing I know happens is when internally you get a good poll whether you're trying to make it good or not as the pollster of the campaign it gets put out there right so the the bad polls don't get put out there so um and so we you know when we're doing internal polling and sometimes that's what you're seeing they'll do press releases on it and so forth it's David D David I want I want to interrupt you right there's pollsters who are legit who who just have a bad poll because even if you're doing a perfect methodology one and 20 or so polls are bad but are there people whose polls get covered in the news who are malicious actors who who cook the books does that exist in your experience or not because people on the left and the right both believe that it does you don't need to name names but does that exist you ever see polls covered in the news either on the left or the right or in the in the regular media and you say I I know for a fact that that person is not legit so I would suspect that maybe I don't know anyone who does that I can't obviously this is a world we you know Patrick and and Anna we work and I can't say yeah this this is the person who always does that so I can't say that I know anyone who does but you know um it would not surprise me um if if there are you know some groups out there um that are doing it you know the thing I'll just I'll chime in with is that for me and I'm unaligned by party and I don't work for candidates and all of that but I say to my clients of all ilk the best news I a poll is they were going to give you is the truth because if you don't know what's what's the most accurate read of the electorate how is it that you're going to respond so I don't know that they're doing a service to their candidate or to their their U supporters to think that the race is better than it is or and and and and vice versa I don't think it does a service I'm not going to name names but I I can tell you that there are some of the most famous pollsters in America some of the most famous I'm not saying you three or anybody like you but people who a lot of people on here have heard of I've Been Told throughout my career that person more I'm talking about more than one person that person doesn't do real polls that person does polls for their clients to make their clients feel good or bad depending on how they want their clients to feel or that person puts a public polls to influence it so I've been told that about very prominent pollsters I don't know that it's true but I certainly know that some political professionals make that accusation against famous pollsters of course it's true of course of course it's true all right you say Patrick you're you're totally silent do you want to say anything about this or well no I mean I I it's much much the same as David I I mean look I do I do I know for a fact that there's anybody who is doing anything such as I mean just just say like either making up data or saying they're conducting a home when they don't um I think that is unlikely um with that said there was a pretty big controversy right in 2022 when um you had certain averages uh certain polling averages that only admitted certain pollsters and many of those pollsters um you know we're putting out data saying there's going to be a massive Red Wave and then it turns out that there is no no red wave and like an said Republicans um you know are disappointed um now with that said I could could I see how uh maybe a poll in the last few days of an election uh um uh if um you know comes in a certain way or comes in um yeah I could I I could see that I sometimes get polls to come back and I personally say I think this is just too Republican or I think this is too Democratic and I'll tell my clients that but um but certainly um there are uh you know there is a certainly a motivation on in in this um political media ecosystem in this partisan media ecosystem to only um you know even if people aren't conducting fraudulent polls and people aren't making up interviews to only amplify the things that confirm a pre-existing bias right right that definitely does that definitely happens thank you for the question Kent come on in tell folks who don't know who you support where you are and ask your polling question thank you Mark for calling on me I am in Atlanta uh I support Trump in this election um my question uh is given the Trump uh under uh over performance uh to polls particularly in the blue wall in 2016 and 20 how are the pollsters adjusting for that um have they really tried to model exactly what the uh electorate actually said in 2020 and 2016 uh in the 2024 election y probably the biggest question question I get maybe most of all and for some the biggest question out there um who wants to have a shot at that vital question well I mean first of all I don't think that we know whether or not a pulling error like that's going to happen again pulling error is kind of random although in the newsletter Mark you alluded to I did uncover a chart from Pew that looks at polling error over time over every election cycle since the the start of polling and it does turn out that um I said I've said it's random but but it does turn out that does at least in the postwar period um that Democrats do tend to be under overestimated a little bit more by polls but but it's not a by a great amount um look I I think that um polls are only one source of information um you need to be looking at election results you need to be looking at um you know how has the state trended previously uh what did the polling say about these States versus in 2020 versus what it said uh what the actual results were previously and a couple weeks ago I was seeing in the polls uh you know showing uh KLA Harris doing a lot better in the rust Bel States then uh I think she should have based on um you know what the results actually look like in 2020 there seem to be an idiosyncratic shift her way U across the Battlegrounds and in the rust Bel States but you know it turns out a week later um that has subsided and I think it's subsided um based on the fact that um you know we just have more polling but also um things are changing pretty rapidly when polls are taken in one state it's very looks very different than when polls are taken in another state um so I would say if you get a poll out there right that says Wisconsin is going to vote you know five or six points to the left of Pennsylvania I'm gonna call BS on that like I don't think that's going to happen I think that those two states are going to vote pretty much in line with each other now whether that's for Harris or for Trump I don't know yet but things like that are things like strange indicators in the data are things that I'm gonna I'm gonna be looking for David or and you want to comment on that I have so one thing that you'll recognize and we poster scale some as an industry get to take a little bit of false credit for it is it looks like the polling is not that bad and we keep saying that after the election that's because our polling get better as we get closer to the election day to be honest and that's why you look at the polls and what people remember is you know 16 when uh Clinton was up you know eight points and multiple key swings and swing that she lost right and then then like or somewhere like Pennsylvania I can't remember what the average was and same thing in 20 and um even though as celinda said earlier we do we do stratification and we're trying to control for the things but there's issues that are we're missing a l i mean we miss a lot can get missed and we have we don't have the magic of answer for it if we did we would that wouldn't have happened in 20 again and it would happen in 16 so but I do think as you get close to the election and why do people take polls anyways like people are most civically engaged across the board when it gets really close to election so some of those voters that even Patrick was uh talking about earlier who are you know in the rural areas that maybe we're missing right now in those um segments that we think we have them in now they're starting to answer polls a little bit higher rate so you know all the pollsters we get a lot better as we get close to election because I mean a lot do people do because U the incentive to take polls kind of equal across all these people are very C civically engaged because the election around the corner like most voters aren't on calls like this as an I think an said they're getting information now like there's not a lot of there are low information voters but like people are you know getting exposed to 10 times more than you know they were even five years ago so um that's so you know we'll see if it's happening again this year we'll see if you know for some reason once again it's it's you know overpredicting um the Democratic candidate because we're in August and September that seems to what happens but I think it ends up leveling out as we get closer to the election yeah Y and the Trump campaign's counting on that all right five o'clock so we're going to stop thanks to uh to an and David and Patrick and celinda all of you ask questions you'll get an if you're subscriber to my substack you'll get an email from me shortly with David's full poll the slides he showed plus some additional data from the survey a link to Patrick substack which you should subscribe to and a link to the YouTube video of this effort today you can share all of that with anybody you want again if you're a journalist please give credit to David Burell insights as well as to twoway and we'll be back uh if there's breaking news over the weekend otherwise 9:00 a.m. Monday even though it's a holiday for the morning meeting with Sean Tim and me and again grateful to Ann Patrick David and selinda have a good evening good holiday weekend and we'll see you no later than Monday morning 9:00 am eastern time thank you [Music]