Cspan2, television for serious readers. Good evening im john kenyon im the executive director of the city of Literature Organization prayed welcome tonight slip talks events featuring me the author joseph campbell. Botox is a series presented by the Iowa City Public Library featuring authors of books featuring political role of social engagement. In his new book, pulling failure in the u. S. President ial election, campbell looks at the long history of polling and its failures. You may recall dooey defeats truman or the shock of 2016 Election Results, campbell goes deeper. Looking at how polling methods in the way polls are covered in the media have affected our National Politics for decades. Campbell spent 20 years in journalism become a professor in Communication Studies program in America University in washington d. C. Hes also a writer historian, media critic and blogger. Hes authored seven books including getting it wrong, debunking the greatest myths in american journalism in 1995, the year of the future began. Tonight campbell talk but his brandnew book from the university of California Press, lost in a gallop. Following his opening remarks, i will moderate a q a section. Those watching live on crowd cast can submit a question by asking questions the bottom of the screen at any time during our program. Now, i would like to welcome w joseph campbell. Stomach thank you its great to be here. Its great to be back in iowa city if virtually. I spent productive. Time, about two and half years ago in iowa city going to the Gallup Organization papers in the university of iowas special collection. It was a very in revealing business. I must say i had a good time in iowa city. Wish i could be there in person rather than virtually. But this will have to do for no now. One of the surprises, mild surprises, about the gallup collection at the university was a small folder of cartoons about polls. These were cartoons were apparently collected by what of George Gallups sons. The dates of some these polling cartoons was after his death in 1984. I believe it was one of his sons to collect the cartoon. I have a few i would like to show. On the program tonight. Going to the next slide. This is a mildly amusing cartoon here in the wall street journal. Im not sure of the dates, but nonetheless it is kind of amusing. So is the next 12. A woman says, ive got to see doctor gallup ive changed my mind presumably a pole responde responder. And then the next cartoon is one that appears in my book, lawton gallup, early in the book on page nine disappears. The cartoonist, Richard Wright came up with this amusing characterization of rogue posters and when pollsters go bad they asked things like who cares what you think in response. With interactions. Such a small collection of cartoons but nonetheless amusing and kind of revealing. Tonights presentation will focus on a few cases, although not all. A few cases of polling failures in u. S. President ial election. And they will take up a few takeaways are reminders about election polls. I will offer a couple of suggestions what to keep in mind what to look for this fall. And then we will go to a q a. The presentation is drawn from i just published book, lawton the gallup, failure in u. S. President ial elections. In its book was brought out recently by the California Press publisher from work mls through four book projects. This talk tonight will consider a few cases i mentioned, specifically those from 1948, 1980, 2012, 2016. Pulling errors in those president ial elections. As are not a complete universe but it covers some of the betterknown ones. To the next slide, that one picks up dooey defeats truman election of 1948. As an epic pole failure. In which the polls got it completely and utterly wrong. George gallup and other pollsters forecast a certain victory for the republican candidate, thomas e dooey. And dooey ran what i call a glide path campaign. He very seldom evoked controversial points of view. He tried to smoothly run through the fall election, the campaign, and not upset anybody or take any controversial views or positions. Whereas harry truman ran a very aggressive campaign. And truman was recognized as being behind in the polls. In fact one pollster, Eleanor Roper one of the seniors and Public Opinion research and at 72 years ago today, 1948 that he was so certain that thomas ide dooey was going to win the election he was no longer going to be reporting polling. He would take polls but he just would not report the polls because he didnt think they added too much of the understanding of the race in 1948. That was emblematic. Ropers point of view is emblematic of the supreme they had income unchecked outcome of the election. Harry truman on by 4. 5 Percentage Points. After the election, one of the comments was the first time truman was the first candidate to lose in a gallop, but when in a wallet. Or. 5 is tory us, is fairly modest but pretty clear as well. And i think, although it is hard to measure the shock of 1948 was probably greater than that of just four years ago in 2016 when donald trump won unexpectedly. The shock will really ran deep. Emblematic of the shock was the front page of the chicago tribune, one of the early additions on the day after the election that declared dooey beat truman is one of the most memorable, Iconic Images of american politics. Harry truman was on his way back to washington from missouri and had to stop in st. Louis pretty had to stop at Union Station in st. Louis. They gave the front page of the chicago tribune. And he held it aloft of what is a very memorable photograph. After words journalists really criticize himself having delegated the responsibility their legwork to the polls that they relied too heavily on pole poles. So, what went wrong in 1948 . A number of factors contributed to dooeys defeat of Harry Trumans upset victory. One of those factors was the fact that one of the thirdparty candidates, the Democratic Party split into three factions. It was the mainstream of dented democrats represented by harry truman. Then there was a progressive wing that Henry Wallace former Vice President under Franklin Roosevelt was a leader of. And then a third split in the Democratic Party was the states right party. And it was opposed to harry truman civil rights measures, and broke off and formed a separate party led by strong thurman of south carolina. During the election, the support for the Progressive Party dwindled dramatically. In the beneficiary lost support with harry truman and his campaign. That is one factor. Another explanation for what went wrong in 1948 was the pollsters stopped polling pretty close to the election by midtolate october. They were done with the polls. Announced in september is not going to take enough pulling resorts and a longer, he did conduct a poll late in october, but did not show much in the change so did not work for the pole at all. The pollsters figured not much was going to change and they did not continue polling right up until the end. This is a lesson by the way pollsters learn and relearn every so often in president ial elections. Another contributing factor probably was that Republican Voters were so confident the thomas e dooey was going to win, the hull in the press all said he was headed for victory, that many Republican Voters decided not to turn out, decided not to vote. That overconfidence translated into a deficit for thomas dooey. Those are some of the factors to explain the loss of doing 1948, the last of the pollsters as well. We move on to 1980 which was a another surprise outcome when news organizations entered polling realm in large numbers worn out doing their own polls or commission their own polls and polling by then, four years ago was more numerous than ever. In the polls indicated that president jimmy carter was locked into a very tight race with republican Ronald Reagan. In the polls were consistent in saying so. And yet, on election day, Ronald Reagan wins in a near landslide. In an outcome no pollster had anticipated. And after words, pollsters bickered and quarreled among themselves as to what went wron wrong. This spilled over and usually so into the public realm is the article from the Los Angeles Times suggest pollsters spat on why they aired so badly prayed what went wrong in 1980 . One of the factors was the fact that the only debate between the two major party candidates, reagan and carter took place very late in the campaign a week before the election. That seemed to have been a factor in tipping support to Ronald Reagan. The people could see he conducted himself well on the stage with carter. And he was not really as wowed in the centric as many people thought he was. That was reassuring. Probably contributed to reagans sizable victory. The other factors he pollster again did not pull up until the very end for they realize they ought to do that but for various reasons did not continue polling until the final weekend of the 1980 election. And also there is a Third Party Candidate John Anderson was running as an independent he was a republican running as an independent. For a while during the campaign the fall, it looked like he is to drain a lot of votes Ronald Reagan. This election day approached, anderson support dwindled and reagan was a beneficiary. Those are some of the factors that explain the unanticipated outcome, the near landslide that no pollster anticipated 40 years ago. You can take on the next line john . Stomach the next case is that of 2012 when the Gallup Organization was essentially alone and calling the election or estimating the election and mitch romneys favor. Throughout the campaign, gallup poll and kept signaling that mitt romney was ahead by four, five, six Percentage Points. And at the end of the campaign, gallup suggested was very tight race that romney was one point ahead. And in the end, president barack obama wins election at the real embarrassment for the Gallup Organization. It was also the year in which nay silver confirmed his status as an election or goal. In 2008 he had, through a poll based statistical model that he developed, estimated the outcome accurate in 49 out of 50 states. That was in 2008. In 2012, accurately forecast the outcome and he was recognized as a statistics guru, forecasting guru that help seal his reputation. Enter signaled the rise of Data Journalism as a way to also interpret polls and Public Opinion. maybe not as deep as 1948 put shock that night ran very deep because Hillary Clinton was widely expected to one the presidency and perhaps fairly easily, and what happened in 2016 is key polls in Battle Ground states, particularly the upper midwest, wisconsin, michigan, and pennsylvania, looked like they were going to give the outcome to Hillary Clinton, and had she won the three states she would have had enough electoral votes to win the election. Instead donald trump narrowly wins wisconsin, michigan and pennsylvania, and also takes Battle Ground states such as florida, north carolina, and ohio, and that combination of states swept him to the presidency. Electoral college victory. Hillary clinton clearly won the popular vote but trump won the Electoral College vote. What went wrong in 2016 . An argument still being discussed and arguments on the their side or many side being made but nonetheless appears that polls in these states either ends their polling too early or failed to weight their percentages, failed to weight their results in a statistical adjustments that pollsters inevitably mack. Failed to statistically adjust for college ed noncollege educated voters who went to trump fairly heavily. That is one interpretation of some of these erratic polls in key midwestern States States ao pretty clear that trump picked up undecidedded voters in large numbers toward the end of the race and he had more undecided swinging to him than Hillary Clinton had swinging to her. So that combination of factors was enough to probably tip the Electoral College to donald trump. And it is a scenario that in people suggest could happen again, could happen again in 2016. My research into polling failure in president ial elections suggest no two Polling Centers are the same either so not likely well have a car been copy of 2016 this year. But well see in eight or nine weeks. What do thats case tell us . What are takeaways and reminders . Obviously pretty rare for a president ial election not to be characterized by some sort of polling dispute. Polling controversies are common place in president ial elections and he welcome expect to see them this year. The types of polling failure, the variety is not the same. We have seen just in this brief presentation four different types of polling failures. Theres the epic failure of 1948. Theres the landslide that pollsters dud not foresee in 1980. Theres the venerable pollster who get is wrong, the Gallup Organization in 1980. I mean, 2012. And also, the another type of polling failure are polls in key states that upset the national outcome. As what happened in 2016. Wisconsin, michigan and pennsylvania. Its also interesting that polling to the end of the campaign through the last weekend before the voting, before election day, is a lesson that not all pollsters have learn or always put into effect. We see this happen time and again. Saw it happen in 1948 and a few key polls in 2016. And another takeaway is that polling failures often correlate to journalistic failures, and in the sense that journalists often take their lead from polls, from a preelection poll. Polls are central to oh how journalists understand and interpret president ial campaigns. It is essential that polls are stolen how they set and fix and pursue the campaign narrative. So when polls mess up, journalism can falter, too. So journalistic failure is often equated to polling failure. Its something that we dont always keep in mind. Finally, what might we see in 2020 . What are we likely to see this jeer in polls will be more numerous than ever. Seems like its hard to escape the polling deluge and were only in september. Eight weeks away. And even now poll results seem to be all over the place. Real clear politics, an enveriable and very evenhanded political aggregation site just today posted several polls that show joe biden is ahead of donald trump by a range of 2 Percentage Points to 12 Percentage Points. So, polls seem to be all over the place, and some are suggesting in a close race, some are suggesting something less than close. So well be seeing probably polls with some erratic swings to them. One reason for that is that polls are being done by a variety of methodology anymore. Phones, cellphone, robocalling, internet panels, even social media platforms are being tasked for insight into Public Opinion in president ial elects. There is no single Gold Standard for polling any longer. That Gold Standard used to be random digit dialing telephone calls with a live operator but the Response Rates to those calls, to those random digital dialing telephone calls, Response Rates are dropping have dropped into the Single Digits so makes is very difficult and very expensive for pollsters to use the method and try to get a good crosssample of opinion. So people in the polling business are looking intensively for the next Gold Standard if you will, the next approach to polling that is going to be reasonably accurate and reliable and not terribly expensive to do. Thats a difficult combination and pollsters have been looking at this for a number of years the first internet based polling was done in 1999, so we are still in this period of a good deal of experimentation and churn among pollsters trying to find the next best standard. Its also important to keep inman that polls are not always wrong. They are done by people who are professionals and have a strong stake for the most part in their outcome and being accurate and reliable, and offering the public a good idea, a reasonably accurate clue what is going on in the race. But polls have been wrong often enough as we have seen in this very small sample tonight, polls haveline wrong often enough and have a checkered record so its advisible and its note a bad idea to treat them warily to be a little bit skeptical about polls and the polling numbers. And polls like the one i just referred to showing a race between biden and trump ranging from 2 Percentage Points to 12 personal opinions is evidence to treat women if the skepticism and wariness specialty is far out. The closer we get to election the better and more accurate the polls can be expected to be and we dont what theyll be but we can expect surprise developments in the final days and weeks of the campaign. We have seen this happen often enough and happens with maybe not predictability or regularity but happens onenough we can and eight some Major Development to happen in the latter part of the campaign. We saw this in 2016 when james comey released a letter saying the fbi had renewed or reopened an investigation interest Hillary Clintons private email server and that slowed her momentum at the end of october. And may have been enough, some analysts suggest may have been enough to tip enough votes in key states to donald trump so he could win a Electoral College victory. We also saw in the year 2000. George w. Bush, he was ahead narrowly against al gore in most of the polls in the popular vote in 2000. But just a few days before the election, news was reported in maine that george bush had had a dui, a driving under the influence arrest and conviction years before, near ken kennedy near the family residence and he had not reported that dui and the news shook his campaign and may have been enough in a tight race to tip voters from bush to gore so that al gore won the popular vote and george bush narrowly won the Electoral College of a 37day ordeal about the outcome in florida where the election hung. So its likely that we will have some sort of late in the game election surprise. So those are few suggestions and takeways about what to expect and id be happy to field questions. Thank you very much for that. That was a great overview of the book and if folks have questions i would encourage them to again click the ask a question tab at the bottom. They can ask a question. I did want to get started while people are submit something questions and submitting some questions. One reason we were excited to have you come here is the connection that George Gallup has to the university of iowa and iowa city here. George gallup earned a bachelors degree in political science, masters and doctorate in college from the university of iowa, was the editor of the daily iowan student newspaper. A lot of connections and i know as you mentioned you did some research at the special collections at the university of iowa libraries for the book. I was wondering if you were able to clean thing about gallups personality or insights how the man actually work . From the collection of his papers at the university, one of them how he early in his career was a real advocate, a real evangelist for accuracy in election polling and would take great pains to really describe polling as his polling as having been very accurate in president ial races. Were talking about 1940, 1944, even 1952 and would even when races didnt really align with his polling outcomes he was inclined to say, yeah, we had a very accurate election. So he was perhaps more important than any of the other early pollsters. There werent many. They were polling rivals and contemporaries to George Gallup but gallup was inclined to try to find a way to emphasize the accuracy and reliability of the polling and polling data. One reason he did that was to make the case that if election polls were accurate or reasonably so, then we can expect all other polling to be pretty accurate as well. He called election polls the acid test of Public Opinion research. He didnt really think election polling was all that valuable but he really put a lot of emphasis and a lot of faith in the outcomes of elections and used them to emphasize the fact that other parted of the Opinion Research and polling field could be considered to be accurate. This is one takeaway from the research i did at the university there. Another takeaway was how gallup he has reputation for being this avuncular guy and a very pleasant person who was calling everybody, my friend, and i think theres truth to that but also had another side to him, he was very, very inclined to criticize critics and to take them on publicly and to make them to really challenge their arguments and their criticism of his work and Public Opinion research as well. So he was an evangelist for polling and Opinion Research, and he was very aggressive about doing that, and much of the public story about George Gallup doesnt get into that, doesnt get into the aggressive side of George Gallup and that is one takeaway i found in the research there. Good. One followup about gallup and then were starting to get questions. I was struck in reading the book, you mentioned that he was on the for of Time Magazine in may of 1948, and i was just trying to think of if there would be a pollster today whose name anyone knew other than perhaps nate silver and thinking it was obviously a very different time when a pollster could be that prominent the public sphere and do you think that pollsters maybe hide behind their companies now because sometimes those mispolling experiences could be somewhat personally embarrassing . Interesting question. I think thats theres something to that. Pelting does hot have the same singular identity as it had 7080 years ago when George Gallup was starting out and he did become as he was the evangelist for early Public Opinion research and although he called his company in early days something other than the Gallup Organization, the American Institute for Public Opinion research, but nonetheless he became identified with Public Opinion polling at an early stage, and one of the reasons that happened was that he was the one who was producing polls on a regular basis, and had many prominent newspapers lined up as clients, as subscribers the washington post, the boston globe were early describers of the polling information andhe is not polling week in and week out on election issues. He is polling on policy topics. Most of the gallupoling was on Public Policy issues and only a fraction really was preelection polls. So he became more readily and more quickly and more prominently than his rivals, the guy who was identified with polling polling and pollsters, and in 1948 youre right, Time Magazine in may of 1948 put him on the cover and called him the babe ruth of Opinion Research, and its interesting because six monthses of after the consider story he suffered perhaps the most embarrassing setback in election research, the Dewey Defeats Truman election. That upset with the one that to the poll temperatures did not see pollsters did not see coming at all. An interesting character. His no one has done a biography of George Gallup. Not something i had necessarily an interest in or the time to do, but i suspect it remains to be done. Maybe thats your next project. So we have a question from one of the folks in our audience. Curious about the value of polling relative to your colleague, allen lickman and his prediction success. Yeah. Dont know much but legmans 120 13 keys to the president ial election outcome he claims he can predict how the popular vote will turn by a variety of what he calls keys and they include Foreign Policy issues, economic issues, and whether theres a Third Party Candidate or whether theres a an internal candidate in the president s party. Those are i think four of the several keys he uses, and he claims Great Success with those, and i have never spent enough time to investigate. Others have and have great doubts about his 13 keys but i dont know independently whether how effective they are. He claims great accuracy. Its interesting, too he makes the prediction fairly early in the fall campaign, not like on the last weekend before the election. Perhaps early has september and at that time like as potentially premature because we have seen over the years how polling numbers and the dynamics of president ial campaigns can shift dramatically in the last several days or last several weeks. So, if lickman is making predictions early in the fall might not be in time to catch the changes and the shifts in campaign dynamics. One question asking but the impact of proliferation and wide adoption of social media and what that has done for polling and also looking at facebook in some news accounts as appearing to take sides politically in some of their decisions related to advertising and fact checking. How do all of those things have had an impact on polling . Its hard to measure the impact that social media. Its pretty clear there must be some kind of impact. All kinds of people on twitter, facebook, instagram, et cetera. And its intriguing that pollsters have tried to figure out a way in which they could tap the social media posts and try to figure out whether social media is a leading indicator of changes in Public Opinion, and researchers have come at that from a number of different directions. Its theres nothing yet that is clear or persuasive that social media, particularly twitter, is a the predictor in changes and shifts of Public Opinion. Theres a fair amount of suggestion or suspicion it ought to be an early indicator of changes in the direction of Public Opinion, but so far pollsters have not figured out a way to tap that in any kind of reliable way. 2001 reason is not everybody is on twitter. Tens to be a hangout for journalists. And so there may be a twitter bubble that some journalist are into that they follow people who have likeminded and just getting reinforced all the time. So its not necessarily the most representative platform but an intriguing option, and pollsters have been looking at this for a number of years trying to figure out there is a way to top social media as a precursor in terms of change in Public Opinion. Another change to politics that someone has a question about is how well have pollsters adapted to the near yearround election cycle, particularly the shift to digital. Seems particularly here in iowa, seems like as soon as we elect one president , everybody starts to come calling. I dont know if that will happen again in 2021 but its certainly getting longer and longer every cycle. How has that been affected how that hat affected polling. Interesting question. In iowa you have a great are familiarity with that short cycle or repeating cycle than most of the country. It puts a burden on pollsters in the since we have to keep pollig anding and resume our polling and candidates for the democratic nomination . In 2020 . 20 candidates or so. Something like that, yeah. So trying to get a sense of polling numbers in such a large field is really a daunting task and can be very expensive, and that is one over the drawbacks of preelection polling. It can be very expensive and pollsters are trying to figure out ways to make it less expensive but the former Gold Standard i mentioned earlier, the random digit dial live operator telephone polling, very expensive. Very expensive to do that, and so i dont know. The proliferation of pollsters is a constant. Every year we have more and more pollsters, nate silver at 538. Com his predictions and polling analysis site, ranks more than 400 pollsters and gives them grades from a to f in terms of reliability in track record. And that is awful lot of pollsters. Its not like 400 pollsters are out there week in and week out generating numbers. That were supposed to digest but its suggestive of this huge number of polls that are being out there and is a said earlier, today, real clear politics posted ten different polls with numbers all over the places. So, its tough for consumers, day in and day out, week in and week out to make sense of the numbers. Early in the election cycle in 20192020, there were candidates who seemed to be really likely to do well in the polls, the early polls had them up, and it didnt last. Kamala harris is an early beneficiary of polling numbers and drop off and Pete Buttigieg was another and his numbers dropped off. Early polls early in the cycle theyre not very revealing, theyre not prophecies by any means. You quoteed nate silver in your book who was saying that trumps victory in 2016 was, quote, the most Shocking Police Cal Development of my lifetime and i just wonder you say its difficult for the layperson to parse all of these different polls. Heres someone whose entire profession is to parse those polls and to give them weight and credence and he was surprise by the result. So what hope there is for the rest of us to figure that out . There are any tips you would give. You mention people named to not ignore them but look at them with some skepticism as well. Thats right. To treat them warily and recognize polls are not always wrong but have been wrong often enough. And 2016 is a classic case. And its a nuanced criticism as well because many leading pollsters point to the fact that in 2016 the polls in aggregate called the election fairly closely for Hillary Clinton. The popular vote. The popular vote. May have been one percentage point or 12 percentage opinions off from the outcome in aggregate and thats a and some pollsters opinion to that, thats an impressive historically good, accurate result. It was delivered, though, because Hillary Clinton wiped out trump in california. She just won california by a popular vote margin that was overwhelming, exceeding anything that president obama did in and it was enough to wipe out trumps advantage, popular vote advantage in at the rest of the country and the resulted in her winning the popular vote by a margin that was pretty close to what the polls in aggregate were saying. So, its a knew nuanced story but polls and polling failure and where the poll dozen wrong. Theres not one sinkle template that explains all polling failure. Theres a variety of those and we saw one of those in 2016. Its interesting, think about that, because we have seen polls now talking about biden being ahead of trump or the polls tyingenning or so maybe shows snowing Different Things especially on a national. He whats the value of National Polls in president ial elections when we still have the Electoral College, the state races are where it comes down to. Is there a need beyond the it layings of being able to perhaps predict the future . Those National Polls . There is. I think a National Poll will give you a general sense of where things might stand. And given the conflicting numbers or the somewhat conflicting numbers i referred to a couple of times tonight think dont necessarily give us the motor accurate of snapshots. Could be a fairly blurry snapshot. But overall, National Polls, if the margin is large enough, it is going to signal that there is a strong unlikelyihood that the person, the candidate who loses the popular vote will win the Electoral College. It has to be a fairly close race for that to happen, for that scenario to take place and thats what happened in 2016. Trump loses the National Popular vote by, what, 2. 1 Percentage Points but win the Electoral College fairly easily. If trump had been defeated my clinton by 7 or 8 Percentage Points nally probably no way he would have won enough states to win the Electoral College. Theres an indicator built interest the National Polls but at this stage even eight weeks out were still kind of tentative in what these polls are telling us. And taken together they may signal biden buys is ahead but we have been numbers shift dramatically enough, late in the campaigns, that not always but often that we should treat these numbers witha degree of wareieness, a degree of skepticism. There is would chat on social media in the last couple of weeks talking about if biden would happen to win the popular vote by x percentage, this is the actual realistic possibility hell actually win the Electoral College and that the personal went up the possibility of winning the Electoral College went up. Theres a relationship there and the exact numbers may vary with elections because no two elections are ever quite the same, but nonetheless, a substantial National Vote victory, popular vote victory, usually translates into an Electoral College victory as well. We have an interesting question here, kind of gets also that 2016 election and how the numbers were so different from what people expected and the questioner asks, do you give any credence so the thought that pollsters got it right nor 2016 put the Election Results were manipulated . Dont give much credit to that at all. To do that would be a recall mass of conspiracy. Keeping in mind that elections, national electionness the United States are state level elections typically. Theyre not run bill the federal government. Theyre run by individual states and localities, so to have that kind of manipulation to throw the election off would have taken an awful lot of work and i dent see any evidence for that at all. This suspicion by the way was raised in the aftermath of the 2014 no 2004 for election front john kerry and george w. Bush when bush won reelection. It was very close race in ohio and there were irregularities in ohio that lead many people to argue there had to have been systemic vote stealing in that state. There was to my way of thinking never any Persuasive Evidence that pointed to that conclusion but the suspicions were there and one reason for the suspicion was that the exit polling in 2004, the day of the election, people telling pollsters what they how they voted and then they collect all the data and aggregate it. The exit polls in 2004 were signaling pretty clearly that john kerry was going to win the election maybe by three or four Percentage Points but wasnt until late in the night on Election Night when the when more data came in from individual precincts that signaled that president george n the election. So exit polls were off markedly in 2004. Another example of how polling can go awry, and it really led to the suspicion that the election had been stolen in ohio. Again theres to my way of think nothing persuasive argument or evidence that was presented on that, and that John Kerry Campaign was not really interested in pursuing such allegations. Its emblem matic how poll emalmostattic how polling errors can lead to suspicions and claims and argument about conspiracies to steal the election. To folks watching you can submit questions. Another interesting one. Are there certain groups that pollsters find difficult to poll but who might play a significant roll in the outcome of an election and how do the pollsters then account for the opinions of those folks when theyre trying to put together theyre data . Thats a good question. It speaks to what is sometimes called the shy trump phenomenon. Pollsters respondents to poll dont tell the pollsters who they really support for perhaps social desirability reasons and may say im undecideds or im going to vote for the democrat, im not going to vote for trump, but in reality they do plan to volt vote for trump. A number of studies that looked at the shy trump phenomenon and have come away with really inconclusive findings. Its not real solid, not real concrete, that this is a phenomenon that happened. Nonetheless, its pretty clear that trump outperformed his polls in a enough key states in 2016 to win the election, and not just . Wisconsin, michigan, and pennsylvania but in florida, and north carolina, and ohio and other places. He is outperforming his polling numbers. And that suggests there might be something out there akin to a shy trump phenomenon, which is supporterses are not saying that they are going to be voting for him. That is a tough thing to find out. If poll respondents are not telling pollsters, yes, this who is im really going to vote for or claiming to be undecideds when theyre really not, what can a pollster do . Its a hard thing to interpret and pollsters dont try to do that. They accept the responses they get. They also apply what are called likely voter screens to the response. They did questions designed to determine whether this respondent is really likely to vote or not in the upcoming election, and sometimes these likely voter screens have several questions, many questions, sometimes theyre fewer questions but all intended to weed out those people who say, yeah, im going to vote but have no intention of doing so. And that is where polling becomes sometimes more art than science. More art than science in the sense that how do you know for sure that youre weaving out the right people . You dont. Put you have a really good sense that these folks are probably not going to vote so we wont count them. If the likely voter screen is too tight your week out people who will vote and thats something you dont want to do. On the other hand if the likely voter screen is too loose and your letting people w. H. O. Who are say theyre going to vote but dont vote. This is a conundrum that polls at the have debt with for. Decades and never figured out how to do it and its been said, perhaps accurately, that there are as many likely voter screens as there are pollsters out there. And it could be the case because different pollsters have different wives trying to figure out who that electorate is going to be and its important to get close to that because if you dont, you run the risk of throwing off your results. If you figure this out you can change professions is what youre saying. Thats akin to a line that George Gallup used wednesday one of his critics back in 1940s, i think. If you figure this out, it will save me a lot of headaches. Youll be the guy who finely figures out how to determine who is going to vote and who is not. That is a revealing comment in the sense that how pollsters have grappled with this conundrum for so many years and still havent solved it, and perhaps theres no way of really 100 solving it. Thats a good segway to a question i wanted to ask talking about your former profession as a journalist and your writing about pollsters. Obviously poll at thes and journalists have symbiotic relationship and they feud a lot. You write journalist take a lead for polls and tend to be unforgiving when opolls surprise them in a president ial election. Is that fair if theyre surprised its means the polls got it wrong or is there Something Else in there . Well, there is a lot of factors in there, and part of it is a tradition among journalists, prominent journalists to really have suspicions put polls and whether theyre going to be accurately interpreting and analyzing and presenting Public Opinion as it is. Poll bashing used to be a from intent feature of american journalists. We had dan rather, cbs news, even edward r. Measureow, the broadcasting journalism legend hat suspects about polls and suspicions about polls and some of them, i arianna huffington, severe poll basher and hated polls. Rescue cohen couraged people to throw off their results and make them look bad when results come in. We dont see so much of that anymore, of overt hostility, overt poll bashing but the low level suspicions about polls i think do remain. One reason we dont see much poll bashing knee. Is the fact that nate silver, Data Journalism is on the rise and he is kind of muted, i think, the inclination that many journalistes have to bash polls. Another reason is that the yroko and the gem Jimmy Breslin has pass away and dan rather has retired so the demons have left the scene but there is a suspicion and an easy inclination to finger point among journaliists to pollsters and say you have get it wrong. After the 2016 election a lot of pollsters pushed back on the argue. That journalist were making that the polls got it wrong. Theyre say, no, look at the positive vote. We came very close, historically great percentages. And although i think thats a bit of an exaggeration. Nonetheless emblemat quick of the tension between pollsters and journalists. Another factor the decline of poll bashing among journalist this fact that many prominent news organizations have been doing their own polls or commissioning their own polls for many years, the New York Times and cbs news, their polling goes back to mid1970s. Cnn is another polling organization, wops washington post, wall street journal so they commission their on polls and prominent news organizations have a stake in polling which wasnt necessarily the case wendall run and rope when gallup and roper were getting going in the 1940s. You had a great quote from George Gallup where he says with the same certainty we know we can be right most of the time we know well be wrong some of the time. Has to be that way. You live by the law of probabilities. Will there ever be a point at which be public agrees with him, that thats something we have to live with or do you think always be that pushpull between pollsters and the public and journalist in there somewhere as well. I wonder if the public deep down recognizes that, that polling isnt a pursuit that will have some degree of error built into it. Its not the solid 100 reliable operation that many people think it is. Maybe dean down people believe, yeah, if youre going to tap Public Opinion its going to be difficult and maybe theres going to be some measure of error and always some measure of error in polls anyway, and its often described as the margin of sampling error in the news articles in the fine prison. Its plus or mine minus three Percentage Points and theres always going to be some measure of magic error in any Public Opinion poll. So i wonder if perhaps the public deep down kind of realize this is a fallible operation, its not infallible, and that maybe we should recognize that. Maybe im being too optimistic in saying so but i think anybody who takes even the cursory look at polling will realize that theres many variables, many factors that can asset can set a poll off and make it wrong and thats important to keep in mind as well. Thats a bit of Public Education that wouldnt be bad to embrace. Pollsters over the years, including George Gallup, said we have to do a better josh educating be pock but what we do, and they said it often enough to make one suspect they really didnt do a very good job of educating the public about polls, and you can even you hear such comments even these days in the second decade of the third decade of the 21st century. We say poll at thes saying we have to do a bert education how we do our work and probabilities and margin of sampling arrow and its one of errors and its one of those recuring themselves that pops up in the history of Public Opinion research. As much as people complain but polls as soon as their candidate is favored in the polls they grab hem and embrace them. Feels like well always have the pushpull. I think youre right and its a natural response for people to say, yeah, bush Trump Supporters would say, yeah, only two points off. Look the rasmussen poll. Or even a better poll in my view is the Emerson College poll in boston and the latest Emerson College poll in the president ial race the end of august came tout and said its a twopoint race as well. So Trump Supporters can point to those and say its a tighter race than you anticipate. Biden supporters point to other polls and say its a near landslide in favor of the democrats and can point to evidence that suggests that. I can the welcome expect to see we can expect to see some narrowing of the polls toward the approach of election day, and some boom say this is a phenomenon called herding in which pollsters intentionally they put their thumb on the scale and bring their results in line with others, and so you see a narrowing of the results as election day approaches. Herding is suspected but difficult to prove and whether its a reality or not is something that pollsters have debated for many years but there is a narrowing that does often happen. In 2016 we saw a bit of a narrowing although polls had Hillary Clinton ahead by five or six points and we had others, a handful, couple maybe that showed trump ahead in the popular vote by a couple of points. So, even at the end of the 2016 election we had some narrowing and then a fair amount of discrepancy between the high end clinton polls and the high end trump polls in terms of the advantages. Those who fine polling fascinating are right in the thick of it right now. So thank you for writing this book and sharing your information tonight. We are at the top of the hour so i think well go ahead and bring our evening to a close. Id like to thank you for coming and sharing with us here in iowa city virtually. Wed love to get you here personally one of these days. Thanks to the Iowa City Public Library, our partners the lit talk series and the university of California Press for helping to make this possible and thanks to everyone who turned out today for our event. You can watch the website of the iowa city unesco city lib tour at www. City of literature. Org and i hope that everyone has a bless isnt evening. Thank you very much,. Thank you, john