Ve what changes are in store . Politicals to spread disinformation via social media, now a permanent part of the american election, i hope not. The Vice President for communications and strategy at the Brennan Center of justice, where she leads the and digital presence. Take it away. Lisa thank you very much. Is john on the call . Um, because he was going to be joining us to talk in particular about polling, but i do not think he is here yet, so we may have to wait for him. In any case, fortunately we have a panel here that can talk about anything. So i wanted to start out by noting that in the first 20 four 24 hours after the ballot counting started on Election Night, the media and polling classes had declared that the polling was the biggest loser. Playbook said the industry is a wreck. Lots of folks were talking about embarrassed to support trump voters that were skewing the results. Too many people are refusing to engage with posters because they do not trust the results. The president claimed on Election Night that the polls had suppressed turnout among his supporters, because they were predicting a democratic blowout. So the analytics guys say the average does not matter, posters pollsters say the average is better polls. So i am curious, which is it . What do you think was really going on out there . Karen i will take a stab at that. I work with cornell. This cycle we did jamie harrisons polling, some work for the congressional black caucus, and another one that i work on. We always have to be mindful that polls are a snapshot in time. I think we had the same problem this cycle as we have seen before, where we have seen some where we have to be careful how much weight we put into what polls are telling us. I did some polling work in the middle of the season, specifically looking for whether or not americans in battleground states thought that joe biden should select a black woman, or a white woman. We were intentional in the way that we worded those to try to get underneath unconscious biases. There is still an issue in how you ask the question can also skew the kinds of answers that you get. Again, i think the most important thing is to take it as a piece of information, a snapshot, but it also cannot be replaced by all the other information you would get on a campaign, in terms of what are you hearing when people are knocking on doors, what are you hearing when you are having phone conversations, what are you or looking to input are you getting or what are you seeing in terms of activity, like in early votes, who is turning out, where they turning out at and how is that matching up with the polling and what it is saying . And Something Else i will mention, i think that too much of our polling, we have to distinguish with what is done inside campaigns and of the sort of public polling, because those are done differently, and they are trying to find different answers. And and certainly think that one of the issues is they are very focused on the what, but not the why. When we do not understand the why, we cannot find the trends and understand what will change over time. One of the things that we did in in race, and a believer messaging. Stacey abrams, much of her success has been about valuebased messaging. Understanding the why helps to better understand the what and to get better numbers. Lisa anzo, welcome. We i glad to see you here. John, the question on the table was, you know, is the death of polling prematurely predicted . John i think that there are a couple things. Want to be a defensive poster, but it is not a good narrative. One is that, first acknowledge that polling is hard, so we are always working to innovate, given the low Response Rate to do better. He saw that we are doing better. Acknowledge that in 2016 2018. 20, different than of what isthe way perceived. It is really important. What i do or what your husband different. Is so much we spent a tremendous amount of doovation and money, what we to get noneducated white voters, rural voters, latinos, etc. Is incredibly laborintensive. It is very different than the dozens of poles that you see in the public, who are not using great procedures. They get dumped into these aggregators. I think it is a problem. You cannot see our polls because they are for our campaign, but we did not see that they margin differences. Everyone thing is that needs to google margin of error. The aggregators show that joe 1. 5 on average and he lost by 1. 5 in North Carolina. It is half of the margin of error. The spread could be seven points. I think that we learn certain these elections. Especially for democrat, talking about harrison in south carolina. I do avoid cooper. If we are not 50 , we will not win. Take a look at where jill biden was and look at his results. He got 48. 6. Michigan was at 51. There is a message to be learned. The margin is not as important. Democrats in battleground states get a pull. That becomes very important. Somebody in the questions asked something that is interesting. That itdence is there ensures Better Outcomes . Lets apply that to polling. Why does it matter . I am looking at the camera. Look, the money piece is a significant part. A lot of donors give because they want an outcome, but john just laid out why that is an improbable reality because there are so many factors. You put your finger on the most , which the most important question you should why. Get an answer to is Everything Else feeds off of the why. Problems, in my the work that i did as a , ididate and as the chairman ays pressed my posters pollsters to get at that question. With john. Of crappy lot polling. Those results, we are not going to see. Unfortunately for him, x, y, z company is doing something that they are paid to put out to sort of craft a narrative and lead the public into thinking a certain thing or a certain thing or certain way about their candidate or an issue. And then when the results come in, poor john is explaining why the numbers do not add up. And they are going, because of all of the crap put into the mix. I think the first thing that the pollsters have to do is delineate where the bad actors are. We know who the good actors are in the business. There has to be a greater exploration of that. I am engaged right now in conversation in maryland with hard trump folks about stuff that is not factually true. And i do not know what or how and they are like siding the stuff, and i am like, that is not real. No, that was wrong. You have these dual realities at play and reparable Companies Like john and campaigns that are trying to navigate in this space are having a hard time doing it. This goes back to the question of why. If you can fashion polling around that, it will be harder to muck up the answers because it will reflect more of the reality than what we have seen in 18 or 20. I think going to that point, if i may get to that question in a different way, again, i appreciate that because i agree the why matters. But you brought up something important, the poll matters. It is from the Sugar Company and it says that sugar is good for you, you know, we have to be as consumers of information, we also have to be mindful of where the information is coming from. And johns point was well taken that internal polls, what youre are seeing internally, that is used that is important to democracy because it is used by campaigns and the candidate to get a feel for where the voters are and what the climate of the environment that you are running in is like. For some candidates, black women have a hard time raising money in part because polling it is used to show viability and it is not the only thing that should be used. We are all talking about Stacey Abrams, i worked for her early on in her gubernatorial race, when nobody was giving her money because they said the polling showed it was impossible. It cannot be done. No way georgia will ever turn blue. God love her, she had not just polling, but other data to show why the why that she believed there was a case to be made. So part of it is the source and the intention of the polling. When you are talking about, if we invest in a 10 year plan, we can turn georgia blue, that is different from saying that we are trying to sell a narrative that says a certain narrative that we want to be able to go out and sell, and using polling as a messaging tool rather than an instrument to try to understand, again, what is happening. Not to say we will not do that message testing, but to understand what resonates with people we have to understand that its important to democracy, but it is used in different ways and we have to be aware of that as well when we are looking at the results and trying to understand what they are telling us. You made me think of this, karen, that every state is not the same. Some states are really difficult to poll. Florida is one of them. The Latino Community is not homogeneous. You have to use bilingual phone banks. Higher percentage of cell phone than any other ethnic. You probably need to do text and web to get younger africanamericans, for example. Florida is like five different states and a couple different countries, in terms of how you approach it. For some reason, michigan never really, what we would consider a good batch of state public pools. Do not know why, but that is the case. You could say the same thing about nevada. Some states are just for again, i could give you theories, but they are easier to poll. Georgia was basically dead even and it turned out to be a dead even race. But there were some dynamics going on that i believe was a hangover from Stacey Abrams, where africanamericans wanted to settle the score about a race they feel like was taken away from them. And you do not have the volatility with whites. If you can get to your 31 , you probably will not go further, right . Where in michigan you will see the movement away with noncollege whites. Every state is different and has its own quirks on how to approach it. And some states have a bigger universe of, if you want to say swings, than other states. Lisa you have played the senior role in many races over the years, working with the Republican Party, did you see anything in the selection that surprised you, you turn you did not expect a turn you did not expect . Thank you for having me. It its always great to have a discussion after the election, something that starts the conversation for the next battle, but in a good way it brings combatants together to express ideas that are not in 140 characters. So, thank you. I have talked to a lot of republicans and we thankfully we thought that public polling was off, so we were counting on having a plan for a close election. Um, people the challenge when i talk to republicans is there is frustration. Campaign so that is really your compass. And i believe every thing john is saying to be accurate because you won, and winners get to say that, but there is a lot of of lot frustration with republicans because the republican results of 2012 were disappointing, and we blamed the polling. Told they did not have enough cell phones. So, in 2014 we fix that. When the pulling was off again, wheres battlegrounds, we were told we did not have the right mix. In 2016, we did not work on education. So, i think that has been a huge frustration for the republicans. And i know that there is a lot of folks saying, no, we had it right, but if you look at the decisions on spending, the Republican Party has to reconcile that we have been outspent, we have been outspent ever since barack obama got into the campaign. That was a big moment for american politics, we saw a big explosion. I think we will be north of 15 billion spent on campaigns this year. If republicans want to continue to govern as the majority party, they need to think about two things, one is so there are no blowouts, and if youre going to get out funded, you have to have better data and better looks into where the electorate will be. It is easy to say now, as a republican we had good numbers internally, but then you look at last minute spending decisions after labor day and columbus day. Maybe reflect on places where we could have spent more efficiently. It doesnt mean that we should throw all of our pollsters out, and they are all bad. What it means is we need to crack this. It is frustrating on the strategy side when you feel like you cannot trust the numbers. I agree with the sentence from sentiment from michael that polling has become supercheap with digital polling, it is superfast. You can get 1000 personal samples for less than 50,000 and generate interest at your website. And people start looking at the averages say that this is the truth, this is what is happening. And it is frustrating for professional polling because, you are right, you can engage sentiment, but when people say head to head this is where they were on a media poll, i do not think anybody serious really takes that into account. But the problem is you have seen the weaponization of polling. And that has become a problem that the classic case was the race for governor where there was polling all summer and right before the election came, polling said it was going to be a close race, and are candidate was underfunded and lost. But there is a sense among republicans that with the polling continuing to show a bias against republicans, that it is unreliable, but also designed to suppress donor enthusiasm first, and then voter enthusiasm second. And the real frustration is, even the republican posters have heard that is where they see this where it says the candidate is up or down 10, they get a poll that says their candidate is only down two, so they turn the dial to get closer to herd. That is incredibly frustrating. Againste willing to go the herd. This as anay isictment, but it frustrating when you see republican polling for firm that are consistently with the herd, getting consistent results. Data, it ise bad extremely frustrating. I have spoken to are extremely frustrated and do not know what to do. They do not know how to combat it. Results. Ve good at a previous panel, there was discussion about how the message is not working. It is hard to do that when the polling is not credible. It is hard to, as a strategist, to drive a candidate in a certain direction if they do not believe what you put in front of them. So i laid out all those frustrations, but i will close on a positive note, which is when i saw joe biden in georgia, i thought, they must be crazy, why does he have biden in georgia . I knew it was going to be a close race, but i congratulate him on keeping georgia tighter than i thought, because i thought that that would go in our favor. And they got it right. You are throwing me one bone. Thanks, rob. It is good to be king for a minute. But you went to a point i think is interesting for this other piece of our discussion, which is looking at social media. If polling is click bait, as rob said, we do not have a journalist here, but we have people that have spent time on tv talking about it. We have patrick here, who spends a lot of time inking thinking about this. What should the media be doing differently when it comes to reporting on polls . Patrick as a person that studies Public Opinion for a living, certainly looking at all the public polls and academic polls that will be coming out. I think there are good and bad things to draw from this election. Postmortem study to figure out what went wrong with polls. The problems they found were peopleey were not enough with a College Education and there were a lot of late deciding voters. I do not think either of those of high we had a lot with matching samples to estimates. The exit polls that i have looked at did not show any evidence of voters moving towards trump. I do not know if you saw anything different in your private poll. What that suggests is there was a problem, and it makes sense isause the pandemic disrupting how people vote and disrupting how they decide to the. Even in what we think are going , the the highest quality best indicators of independent polling research, it is about is answering the phone and who will pick up the phone. Educatedcollege professionals. Other thing is, this is an incredibly high turnout election. On the one hand, there were good butrts to solve the problem 2020 in the polling presented some tough challenges that evidently did not overcome. Polling results with actual results. Ohio, iowa and michigan. So far off that you have to wonder how it happened and how it seemed to happen in a pattern that we saw. It is probably too simplistic of an answer that you did not voters. Gh and he wouldback not get any different. You have to do better than that. I do not know. You never know how they do their process, but it was about getting the right type of College Educated voter. Too many people in the service industry. Factory. O work in a toin, we do a lot of work make sure that we are getting the right makes. The second is, we found that when you are getting verl interviews, they are to educated. That is a problem. Intervieware given an and the educational level is too high. . Ow many the best you can get is doing cell 30 of phones. , 60 of cell crosstab and at a look at those results, which are want republican you your clients doing multiit is difficult to figure out what that makes is. Looking at it, pick the senate race. The difference between cell phones, landline, text to lead and online, you want that. Probably one of the few they hadposters good results for biden. Did you see evidence of latebreaking move for trump . That is the other thing i was going to point out. Where there states from the second to last poll, where we saw movement . Yes. Been continuing to be movement down. Florida is the perfect example. There are a couple things that i think that we need to take on. I said this and somebody thought it was brilliant, but i do not think it is brilliant at all. You are asking one of the things i have found is the reaction people have to the answer, that language does not resonate with me. Places,found in some sometimes there was a reaction to the language. I would just say whatever. Where yes, in places they intended to vote. They were not sure how it works. Say as our country becomes more diverse and our electorate is more diverse, we need to be we need to bring in a lot of different thoughts about what questions we are asking. Questions about a white woman versus a black woman. They are not going to answer that question that way. There are things that you have to do to give the person little bit of cultural distance because you do not want to answer that question. Not not racist, but will necessarily get the answer. Math betterw the than i do. I know messaging and understanding the why. Of the things that we found out