Eva Dale 0:00 From the heart of the Ohio State University on the oval, this is voices of excellence from the College of Arts and Sciences, with your host, David Staley. Voices focuses on the innovative work being done by faculty and staff in the College of Arts and Sciences at The Ohio State University. From departments as wide ranging as art, astronomy, chemistry and biochemistry, physics, emergent materials, mathematics and languages, among many others, the college always has something great happening. Join us to find out what's new now. David Staley 0:32 Joining me today in the studio is Thomas Wood, Assistant Professor in the Department of Political Science at The Ohio State University College of the Arts and Sciences. Welcome to Voices, Dr. Wood. Thomas Wood 0:43 Thanks for having me. David Staley 0:44 So, I noticed that you recently worked as Director of Experiments and Modeling for the Jeb Bush Presidential campaign. I have all sorts of questions about this, but first of all, what does the Director of Experiments and Modeling entail? Thomas Wood 1:00 Part of the effort of any modern political campaign United States is to take this voluminous quantity of voter specific data, and use that to make inferences about the way people make political decisions that likelihood that a person turns out to vote. Rather than treating the electorate as some monolithic set of blocks. The idea is that you'll treat individual voters, specifically as a function of what you anticipate their political attitudes will be. That's a specifically important task. During a presidential primary wherein you can't rely on partisanship to make political decisions. It's akin to having people choose between a list of types of vanilla ice cream. And so choosing from 17 kinds of vanilla ice cream, it turns out that very specific little components of mass attitudes tends to break people up. Whereas the decision between the political teams comes down to race and class, where a person lives, et cetera, et cetera. So in those primary races, experiments, and very fine grained machine learning techniques turn out to be useful. Turns out they are less useful in a general election, where we come back and we basically participate as two big political teams. David Staley 2:11 Why is that less effective at that level? Thomas Wood 2:13 So, we can think about a number of different ways. Just imagine your typical partisan friend, what piece of information would they be required to be furnished to abandon their political team in 2019 2020, it's like a person who traditionally votes Republican to not vote for the Republican presidential nominee, did not vote for the Democratic presidential nominee. It's basically an expression of your political identity. It's not like a decision that a person makes in face of some specific pieces of information, but decisions within our partisan primary. There, you're basically finding people with almost blank slates, what do they really know about these folks who are just introducing themselves to the national political environment? What do they really know about Elizabeth Warren, Cory Booker, Marianne Williamson, before they saw these folks, gabbing in each other on a debate stage, in the absence of relevant considerations, there's an opportunity for campaigns to be influential. David Staley 3:10 Well, and your title said experiments, Director of Experiments, you're performing experiments? Thomas Wood 3:14 Uh, some. David Staley 3:15 What do those look like? Thomas Wood 3:16 Those look like, you can experiment in the field, that happens rarely, wherein a campaign rather than assigning media exposure in the most strategically Trude possible way they randomly assign ads to be broadcast in a place visits to be conducted in a place to exploit the statistical average available through randomization is the reason for instance, we rely on randomization to figure out if a pharmaceutical affects some illness. We didn't do that many of those in the GOP campaign, if you recall, it was not the most politically famous conclusion to that race, a lot of high hopes, but they're not the voters unfortunately, had their say. So in a campaign, which is cut short, most of the experiments that have been done are being done online. And that is, you exploit the fact that the person you can find in say, an online labor marketplace willing to conduct and do surveys for, like 2030 cents a shot. If you condition on that person's demographic characteristics, that random person reacting online turns out to be a fairly reliable analogy for the typical voter watching TV in central Ohio on a Tuesday afternoon. And so for that reason, it's a very effective way for campaigns rather than just to cross their fingers that this particular messaging strategy will be effective. You can estimate with far greater precision before the messaging in some tech. David Staley 4:37 And so, you mentioned previously that there's an enormous amount of data that one has to sift through and this is just simply a function of modern campaigns: has data always played a role in political campaigns? Thomas Wood 4:51 So voter data, like the idea that campaigns would build relationships with a state party in some states and others It's the state government to say, here's the reliable list of folks whom we know turn up and vote in primaries and congressional elections in federal elections, those data have been reliably accessible to campaigns. Since 1970s, the real explosion has been in the consumer data that's been made available since, let's say, the late 1990s. And there it's I show students at OSU, a snapshot, I blank out the names, I think people are shocked at the level of precision that your pattern of consumer behavior provides to third parties as to the probability that you the kind of mortgages that you have, the kind of credit cards that you hold, how often you engage in overseas travel, they append to that all the state records and local records on the kind of heating system your house has. So we can make incredibly fine grained decisions and predictions on exactly the socioeconomic status of a particular person. I'll give you an example. This is an example of people enjoy the companies that sell frozen steaks through the mail, I don't participate in a mock, I'm not making any endorsements. People, the stakeholders tell me this surprisingly good for the fact that they were frozen sent to you in the mail those companies, it's a more successful lucrative business model to sell the data on who is subscribing to the steak in the mail, then to literally sell the steak just because it's such a reliable marker of a person's high socioeconomic status. And they're selling those data to people attempting to sell insurance and automobiles. And one of the very small marketplaces for those data are political campaigns. David Staley 6:30 So, what changed in the 1990s that led to this sort of explosion of consumer data? Thomas Wood 6:35 The easy digitization of these data, like how cheap they became to store and disseminate. And so, it used to simply be the case that all these Visa card would be sitting on reams and reams and reams of consumer data, but was it really worth their while to package it up and sell it to other people, hard to digitize, it's hard to distribute. There's not that many people who have the statistical computational expertise to exploit it, the massive reduction in cost of this technology, like the massive explosion in capabilities, and you can just work on a desktop computer, and run very complicated models to predict political attitudes across a very big site like Texas. And so for all those reasons, it's not something that you require people who have decades and decades of technical skills. Now a person who walks out of like an elite college like Ohio State with a degree in statistics can sit down and afternoon extract value from data. David Staley 7:28 So, the fact that you were named Director of Experiments and Eodeling, this is not unusual for campaigns, presumably, all campaigns have someone in that kind of role? Thomas Wood 7:36 Yeah. The conventional wisdom is that Democrats are probably a little more sophisticated at this, that the Democrats are probably better equipped at reaching out to the academy, but you know, it's a competitive process, and so Republicans have made massive strides in recent years. David Staley 7:50 How did you end up in this role? Thomas Wood 7:52 I had worked on the Romney campaign four years previous, and I had... David Staley 7:56 In the same sort of role? Thomas Wood 7:57 Yeah, in a more junior capacity. And I had, that was when I was in graduate school, I had enjoyed the experience remained in contact, I continue to consult for some campaigns. It's a pretty small world, while presidential politics appears to be, you know, like the world stage, the number of people actually involved in is actually pretty small. It's not quite as lucrative as people imagine. It was almost like a hobby that people have, and so I own some goodwill, and I was glad to participate again. David Staley 8:19 What can you tell us about your experience on the Jeb Bush campaign? Thomas Wood 8:22 Obviously didn't politically go the way people had expected. David Staley 8:26 No one blamed you, I assume. Thomas Wood 8:28 Well, maybe, not to my face. Interesting opportunity to observe the rise of Trump from inside the party, where people who study campaigns have probably been fairly circumspect about Trump's capacity as a general election candidate, the way he understood the media environment and the incentives that the news media companies were providing to Republican candidates really demonstrated a very sophisticated appreciation for the strategic environment. So, gosh, people are sort of concerned about the possibility that Trump can be a reliable indicator of what politics is going to be like in this country going forward. I would suggest if you're anxious to make sure that doesn't happen, people need to start thinking about the way that parties conduct primaries, and less about the way general elections are conducted, because that's basically where we saw a guy like Trump managed to walk into American public life having no previous political experience. David Staley 9:24 So, your principal area of research when you're not engaged in campaign work, but your area of research is in what you call campaign effects. I'm curious, what do you mean by campaign effects? Thomas Wood 9:35 There's a broad consensus among of journalists to write about campaigns and journalists to write about politics, that campaigns can reliably influence political behavior. I think it surprises people that political scientists, by and large, are far more circumspect about that possibility. I was just lecturing on this yesterday. In fact, you do about as well predicting the outcome of a presidential campaign looking at 18 until the economic data as you do predicting the outcome of the race from the polling data, it's conducted the weekend before the election is held. And so it's interesting to sort of think about if that is a reliable empirical case, if elections are so predictable, why the polls move so much? And why do we invest all this time and campaigning? What's the answer to that? Well, we don't know. Maybe we can regard it as a vitally important democratic symbol. And we all enjoy participating and it's an occasional festival for us to affirm our shared commitment to democratic process. Maybe presidential campaigns just very, very weird compared to other typical election campaigns and Ohioans, I see you ever just Voted sticker, I can affirm to the listeners that your host does vote, maybe like those local campaigns are qualitatively different than a presidential campaign. But the extent that we can find evidence that the individual strategic choices made by the campaign's can reliably affect electoral outcomes that have been very thin. Eva Dale 11:04 Did you know that 23 programs in the Ohio State University College of Arts and Sciences are nationally ranked as top 25 programs, with more than 10 of them in the top 10? That's why we say the College of Arts and Sciences is the intellectual and academic core of the Ohio State University. Learn more about the College at artsandsciences.osu.edu. David Staley 11:30 So, you've recently written an article and I'm going to read the title of the article, and I'm interested in your findings here. So, it's called "Taking Fact Checks Literally, But Not Seriously: The Effects of Journalistic Fact Checking on Factual Beliefs and Candidate Favourability". What were the findings? Thomas Wood 11:47 This sort of is the capstone of a couple of years work that I've been with a colleague of George Washington, we've been working on the role of fact checks factual interventions, especially on these fraught political questions, these areas where we expected Americans to be very biased by ideological concerns and less motivated by accuracy ambitions. The particular set of findings that we've basically managed to move a lot of people to our side is that factual interventions reliably improve Americans factual understanding their political accuracy, even when such a move comes at the cost of their ideological commitments, their preferences for a candidate, you show Republicans Donald Trump making misleading statements about for instance, the level of violent crime, this was a very important topic of, of misinformation by President Trump and by candidate Trump, you tried to get people motivated? No one's talking about it at the highest level of violent crime and 30 years, 40 years, etc, etc. That's an interesting area to provide corrections on because it's both very politically salient. And it's so wildly inaccurate, as I'm sure most people who know something about the way the FBI studies this question. were, in fact that decades, low levels of homicide, low levels of rape, low levels of basically every violent crime, it's fallen, so far, in fact that sociologists and criminologists it's challenging even our accounts like we wouldn't have expected the forest fires it had. So the idea that you would show a Republican Donald Trump making those misleading statements, and then they would proceed to the correction, the conventional wisdom before Ethan porter at George Washington and I started doing research was that not that factual corrections would be ignored, but the factual corrections would compound ignorance. There's this kind of what people call the backfire effect, it's sort of blossomed into this little cottage industry of research and Facebook used to cite the backfire effect for why they wouldn't administer corrections on their platform, et cetera, et cetera, et news media organizations have spoken out, like what is the role of journalism given that we know that people have this pathological response to factual intervention that cetera, et cetera, turns out the backfire effect is not a reliable statistical finding. And the paper that we're talking about at the moment was co authored between Ethan me and the two political scientists whose paper started this research. A decade ago, we collaborated in a friendly but adversarial attempt to replicate this finding. And that literally, but seriously indicates that we can reliably improve factual accuracy, but it has no downstream implications to people's other political attitudes. That is, it won't go on to a fact, the extent to which you evaluate the speaker whom you've just seen a correction for, it won't change your willingness to vote for a corrected personal to oppose a corrected person. David Staley 14:25 That would seem to have implications for the period in which we're living, which many have said is a time of fake news. Thomas Wood 14:32 Yes, and "post truth" is the other one. David Staley 14:33 "Post truth"? Thomas Wood 14:34 Yeah. So, the democratic implications of this might be grave, but I want your listeners to understand that there's almost no evidence at all consistent with the most negative evaluation of Americans levels of political knowledge. We have administered surveys for decades and decades measuring Americans chronic scientific knowledge, for instance, the American public has never been more scientifically informed there may exist right now. We measure understanding of very fraught political questions like climate change, the proportion of Americans who were saved to the scientific consensus on anthropogenic climate change has never been higher. We ask divisive questions threatening our religious beliefs. So the state is the only wealthy country where religious attendance is at a very, very high level. There's never been a high proportion of Americans who are saved to the premise of Darwin and evolution is a source of variation in living things. The way they sort of understand just how weak the claim is for a post truth environment is to ask the people who make that claim, when would you regard was the truth environment? If we're living in a truth environment, when would you regard as the high point, is the Victorian ideal for Americans familiarity factual stuff, and you're looking at me as if you can pry the answer? David Staley 15:46 I don't know if I have a ready answer, the 50s, I don't know, the 50s, 60s? Thomas Wood 15:50 So, it's interesting, so we're basically tripled the proportion of Americans who come and do these wonderful things on college campuses. Would i strike you as strange that the population, which was so much less educated, was much, much more exposed to misinformation in religious environments that etc. Would it be weird if that population turned out to be more, you know, factually informed. And of course, that's what we observe that the American population is very, very reliably, factually informed. But my research would suggest that you're looking in the wrong place. If you imagine that the levels of misinformation were a factor in Trump's election. There's very little evidence consistent with that possibility. And so it's basically that we became anxious about these two things at once a racial populist achieved, elected office, the United States. And at the same time, we became convinced that the American public was becoming like unmoored by reality and allergic to factual interventions. And so one of those things definitely happened, but the other thing just might feel right, but there's very little evidence for it. David Staley 16:50 Is this the subject of your recent book, "False Alarm"? Thomas Wood 16:53 That is right, that encapsulates, gosh, the 15 to 30 experiments with that, we do a lot of these. In the last four or five years pursuing this, we've had some success getting articles published, and this book has now been published. David Staley 17:03 And your focus is on the United States, obviously. How does the United States, I guess, compare to other democracies? Thomas Wood 17:12 This is tough. So, if you want to compare it to other democracies, we've got some good research from the other wealthy democracies in Western Europe and Oceania. But the place where funders are very motivated is looking at these sort of like more fragile democracies, where there's the concern that political agents are being very effective in ginning up group animosity in concealing the extent to which a ruling class is engaged in expropriation of public goods, et cetera, et cetera. And there, we have to sort of shrug our shoulders, there's very little of this research that's been done in a non western setting, to the extent that we can be responsible for improving that a little bit we are collaborating with, with a British factchecking nonprofit at the moment, we're conducting experiments in South Africa, Nigeria, Argentina, and pairing them with a British sample, I would anticipate that voters in less reliably democratic regimes behave pretty comparably to the way that Americans behaved. But this is why we gather the data and do the science hopefully, yeah. David Staley 18:14 So, I know another area of research that you work in, and these are your words, you work in the unorthodox dimensions of American political attitudes, and I wonder what that means is that include things like conspiracy theories, those sorts of things? Thomas Wood 18:27 That's precisely what it means. David Staley 18:28 Oh, okay, so what's that research look like? Thomas Wood 18:29 People tend to get the shackles up when you say conspiracy theories, and so David Staley 18:32 Maybe so. Thomas Wood 18:33 Yeah. They'll say, well, I'm not into conspiracy theories, I'm into truth theory. So it's like, oh, gosh, well, you know, the stories about the JFK assassination of fluoride in the water supply, because it's not recording conspiracy theories. This is encapsulated by a set of work that I started to underway with my PhD supervisor at the University of Chicago. He and I recently co authored a book, we became interested in conspiracy theories, basically, during the opening stages of the Obama Presidential campaigns and eventual presidency, this very widespread belief that Barack Obama was concealing a foreign identity and his religion wasn't as he proposed, and we were among the first people to start putting this on national surveys. We had survey providers tell us, you can't ask that question. That's so weird. The average American survey respondent, no one's gonna have heard of the story. We were basically vindicated when 20 to 30% of the American public seems to entertain the possibility that there was this shadow government going on that shadowy forces control the American political system that the global financial crisis had been engineered by the Federal Reserve. The war in Iraq, had been concocted by shadowy interest in the American government to seize the oil supply, etc, etc. These are remarkably widespread accounts. I guess the two big things I'd like an audience like you guys who are just like crapping on the cause, hearing this Australian accent talk about conspiracy theories, I'd say two things. One is that folks who believe in these stories and folks who have a more sort of conventional factually informed understanding of politics have basically the same media diet. It's not the case that folks are hanging out in their grandmother's basements and spending a lot often on obscure chalkboards. People who watch sitcoms and the football on their Sunday afternoons are the kinds of people who also entertain these on their face implausible accounts. And then the final thing is that it's not a recent occurrence. The best way to study attitudes before the emergence of public opinion research in the 19 century is to look at the folks who are writing letters to the American newspapers, and American newspaper letters pages in the 19th century, were replete with accounts of Masonic plots, and attempts by the Russian royal family to interfere in American politics, etc, etc. It's a more fundamental question in American attitudes. Sorry, that answer was... went long, I apologize. David Staley 20:50 Not at all. Thomas Wood 20:50 You got talking about these things I'm interested in, I run like a toy across the floor. David Staley 20:54 And your interest is infectious, obviously. You've co authored a book called "Enchanted America: How Intuition and Reason Divide Our Politics". Is this a further example of these kinds of unorthodox dimensions of American politics? Thomas Wood 21:09 Sure. As scholars of public opinion, we're trying to always look to media appetites, but if it wasn't media appetites, the answer we came up with was consistent with the anthropological evidence on this when the Trobriand Islanders when they would go fishing with inside of the beach, no ceremonies, no animal sacrifice, just get in the boats go out and fish when the island has had to fish outside of the at all. And out of view of land, massive religious festival, massive cultural practice, resolve this source of anxiety through magical thinking, like use magical interventions to provide safety because how can pre scientific people know what the weather is going to be like whether we encounter shock from away and if you think about it, if you're a baseball fan, you wonder what it is that ballplayers are so responsive to magical ritual. Why? You see before the battle walks into the box, they have like this, like very clearly rehearsed ritual of what part of the uniform they adjust which ear they tog who may look to in the dugout, why these necklaces that promise to channel a person's CI in the most useful possible way to improve a person's physical performance, because they're in a high anxiety environment. And they have no other material way to resolve that anxiety. For many Americans, that's basically the situation that politics presents, politics is a is an amorphous cloud of things, the more they try to understand it, the more complicated it becomes. And from that amorphous cloud emerges sources of anxiety, foreign manufacturers threaten your job, foreign terrorists, threaten your physical safety hear in your little Ohio town, etc, etc. People have used attitudes that are akin to those magical processes, those cultural processes to resolve those sources of anxiety. David Staley 22:57 How does that get expressed in politics? Thomas Wood 22:59 Well, that gets expressed in a lot of ways. So the fact that education that most poorly educated members of the public are the most reliable to indicate the adherence to these conspiracy theories, the fact that people who feel social anxiety born from like society, so people who have the most negative impressions about the probability that America have a recession, that America will be engaged in a war, that there'll be a terror attack, United States in the coming period of time, are also those who are most conspiratorial is observational evidence consistent with the possibility that these conspiracy theories are serving a sort of a process to achieve emotional environmental homeostasis that is returned with a position of emotional equilibrium. And so it's been a productive area of research, if only because the intuition that people are hanging out online and adopting the things they observe is such a seductive intuition. And I want to assure people that there's very weak evidence consistent with a possibility. David Staley 23:56 I'm curious to know how you ended up as a political scientist, as you ended up someone interested in these issues. What was your journey to this stage? Thomas Wood 24:03 I grew up in Australia, as your audience has probably figured out, I spent a period of time at the University of California as an undergraduate, and Americans sort of take for granted how fantastic research universities are in this country. The idea that we can have a college like Ohio State where huge sums of money are invested in the study of human cognition, political institutions, social institutions, etc, etc. You basically have to be a country as wealthy the United States to justify that expenditure. So even countries were like Australia's have perfectly lovely place to live and you'd walk around that Ohioans would feel very at home and as a streetscape. It's very evocative of an American streetscape. huge investments in like biomedical science or natural sciences, but the social sciences a very, very much competing over the scraps there. So we are spoiled in this country. The land grant colleges have built elite institutions of social research. It's the place to be and so also I made the misfor of marrying an American and now have little American children, so my hooks are slotted into me. David Staley 25:04 Tell us what's next for your research. Thomas Wood 25:07 So I'm interested in the way that Americans conceive of democratic values. There's this increasing conventional wisdom that we're seeing the withering away of liberal values in the face of polarization that is, Americans aren't willing to engage in like democratic tolerance and democratic restraint, because it's so important for them to have victories over their opponents. My suspicion is that that's basically wrong. And I'm guessing that public universities this huge expansion of higher education that we've invested in the last 30 years have been pretty successful, inculcating nice democratic values. David Staley 25:41 Do you see yourself engaged in any more political campaigns in the future? Thomas Wood 25:45 Probably not the things I've said on this podcast. It scratches an itch. It's a fun place to be. It's a high stress environment. Most of the people listening to this podcast, many is Central Ohio had the opportunity to door knock every two to three years. It's a fun opportunity to be involved and so probably yes, but hopefully not leaving lovely Central Ohio again. David Staley 26:07 Thomas Wood. Thank you. Thomas Wood 26:09 Thanks for having me. Eva Dale 26:10 Voices from the Arts and Sciences is produced and recorded at The Ohio State University College of Arts and Sciences Technology Services studio. Sound engineering by Paul Kotheimer. Produced by Doug Dangler. I'm Eva Dale. Transcribed by https://otter.ai