Robert Wiblin: Hi listeners, this is the 80,000 Hours Podcast, where each week we have an unusually in-depth conversation about one of the world’s most pressing problems and how you can use your career to solve it. I’m Rob Wiblin, Director of Research at 80,000 Hours.
Today, I’m speaking with Professor Cass Sunstein.
Cass is a Professor at Harvard Law School who focuses on constitutional law, administrative law, and behavioral economics. He was put in charge of the White House Office of Information and Regulatory Affairs during the Obama administration’s first term. Studies of legal publications between 2009 and 2013 found Sunstein to be the most frequently cited American legal scholar by a wide margin.
Many listeners will be familiar with his book, co-authored with Richard Thaler, called Nudge: Improving Decisions about Health, Wealth, and Happiness, which laid out the case for libertarian paternalism.
But folks may not know that he was also a pioneering author on animal welfare law, writing Standing for Animals in 1999, The Rights of Animals: A Very Short Primer in 2002, and co-authoring Animal Rights: Current Debates and New Directions with philosopher Martha Nussbaum in 2004.
He has also been a promoter of cost-benefit analysis in government and the law, writing The Cost-Benefit State in 1996, and recently publishing The Cost-Benefit Revolution in 2018.
Finally, he has also taken an interest in existential risks, in 2007 releasing Worst-Case Scenarios, an attempt to explain how society should avoid both under and over-reacting to the most severe threats we face.
Today though, we’re speaking about his new book How Change Happens, a study of how social cascades can change our culture overnight, and how this can be used for good and ill.
The book ties together a number of important concepts that psychologists have been developing over the last few decades, and the first 50 pages seem like important reading for any activist or politician trying to build support for their ideas.
Professor Sunstein is an incredibly busy guy, so I was only able to get an hour with him. To avoid wasting time having him repeat the basic thesis of his book, which he has done plenty of times already, Keiran put together an abridged version of a lecture Cass gave at the Harvard Law School Library in April which outlines all his key points. Thanks to Harvard College for allowing us to use that presentation.
It lasts 38 minutes and we’ve put it at the start, which means we can dive straight in to questioning how accurate his model is, and what implications it has for people who want to have a positive social impact.
We’ll get to that in just a second, but first, I wanted to let you know that we recently released our 2018 annual review, which outlines what we ourselves at 80,000 Hours worked on last year, and how much impact we think we had.
We’ll link to that in the blog post and show notes for the episode. We still have a bit to go to close all our funding needs for this and next year, so if you also feel passionate about 80,000 Hours’ mission of solving the world’s most pressing problems, and would like to support our work, the show notes include a link where you can find out how to donate.
Alright, without further ado here’s 38 minutes of Cass explaining How Change Happens, followed by our conversation about the book.
Cass Sunstein: Okay, so the title I wanted to give for this book was Why Societies Go Whoosh, W-H-O-O-S-H, but the publisher said, “No, that would be too obscure and undignified.” This is also 25 years in the making. A long, long time ago, I did a first draft of what is now chapter three, which is about how law changes social norms, but I got stuck for the 1990s and the 2000s; I got a little distracted. This basically emerged recently and I’ll tell you a bit of research that helped orient its emergence. In Saudi Arabia, there’s a custom that wives don’t work unless husbands say it’s okay. That’s the custom. The vast majority, the overwhelming majority of young Saudi men think it’s fine if their wives work.
Cass Sunstein: It’s also the case that Saudi men overwhelmingly believe that other Saudi men think it’s unacceptable if their wives work. The private view of young Saudi men is it’s fine. Maybe it’s good. The perception of young Saudi man is that people like them think it’s not good and not fine. There’s an opportunity there. In the relevant research, the experimenters told young Saudi man that, “Actually, guys like you think it’s fine if wives are working outside of the home.” Guess what happened? As a result, the number of Saudi women applying for jobs in the relevant group increased dramatically four months later.
Cass Sunstein: Okay, I’m going to be using that as a clue to the answer to the question about How Change Happens. What I want to give some particular attention to is why change often seems to come out of nowhere. The Civil Rights movement of the 1960s, completely unanticipated. The fact that Brexit worked, not completely unanticipated and it hasn’t quite worked yet, but the fact that it succeeded was widely not expected. The rise of feminism in various parts of the world as different moments in history came as a shock to many people who embraced feminism. The rise of fascism in the 1930s, including the success of Hitler and Mussolini, who knew that could happen?
Cass Sunstein: We’re in the midst of an assortment of social movements. Some of them we can glimpse; some are barely on the horizon and one thing that’s highly likely is that the best prognosticators today are going to get it wrong. I’m going to start with some pretty simple remarks about why the unexpected nature of social success, including transformative success, is not as baffling as it appears to be. I’m going to apply this to #MeToo on the ground that it’s of interest in itself and #MeToo tells us something more general about social change in numerous fields.
Cass Sunstein: Here’s a way to get clear on the mystery that motivates the obsession I’ve had. Tocqueville, probably the greatest sociologist ever, at least one of the candidates if they gave an all-time greatest sociologist award: Tocqueville would be on the shortlist. He reported that no one foresaw the French Revolution. Nobody. Lenin was stunned by the speed and success of the Russian Revolution. Lenin. If anyone was an architect, it was Lenin. He had no idea. The Iranian revolution of 1979 was unanticipated, including by the participants. The idea that the country could turn round as Iran did, that was not expected. The Arab Spring was unanticipated by the best analysts in the United Kingdom and the United States.
Cass Sunstein: I saw this up close because I was in the White House at the time. When things happened in the Arab Spring, the leading experts were amazed. They didn’t see it coming. Okay. It’s common in social science circles to refer to two things: demonstration effects and contagion effects. But, that might be like trying to explain the success of opium by its dormitive qualities. Opium induces asleep. Opium is said by some pranksters to induce sleep because it has dormitive properties. That’s not an explanation. That’s a restatement. To say that the success of social movements is a product of demonstration and contagion effects is like an explanation by noun, rather than an explanation of a phenomenon. It doesn’t tell you anything.
Cass Sunstein: Okay, I’m going to try to make progress by referring to three things: preference falsification, diverse thresholds and interdependencies. Elaboration will come, but for starters, everyone on the planet, including everyone in this room, has some desires and beliefs, and experiences in our heads that we have told either no one or only our closest family and friends. The fact is that we will silence ourselves about some of the things we want. It may involve Harvard. It may involve Massachusetts. It may involve the United Nations. It may involve the United States. We shut up, or worse, we misstate what we actually think. That’s preference falsification. The second point involving diverse thresholds is that for some people, and you know them, don’t you, injustice, and they’re there. I had a friend in the Middle East a number of years ago, where we witnessed, he and I, a father beating up a child. It was probably his child, and it was on the street, and he was just punching him. My friend, who was Irish and had a temper, and about five foot seven, he just ran up to that guy and said, “Stop hitting that child.” Now that was a low threshold.
Cass Sunstein: He didn’t need anyone to support him before he did that. He was there, I followed him and supported him. He had a low threshold. I had a slightly higher threshold. Some people, with respect to injustice, have very high thresholds. With respect to the third part, interdependencies, most of us are really reactive to what other people say and do. If one person is doing something, embracing, let’s say, a green New Deal, or calling for animal rights, we might think, “Crazy person,” but if 1000 are embracing an idea or a movement, we might think, “Why haven’t I joined them already?” If you put together a preference falsification, diverse thresholds and interdependencies, the difficulty of anticipating social change, and even large-scale social transformations becomes much less puzzling.
Cass Sunstein: Now, what I’m going to do… I’ve told you the text, and like any self-respecting law professor, the text is much shorter than the footnotes, and here comes the footnotes. Ready? They’re not going to be in small font. Footnote one: with respect to preference falsification, people might say they like an existing status quo when they really don’t or they might change the subject when the status quo is raised, or they might hear a little voice in their head which they turn off. Here’s some words from the best book I’ve ever read on Nazism. It’s the best book because it’s not only revealing, it’s also cheerful. You can read it without crying. It’s written by a journalist who went back to Germany in the 1950s and spoke to former Nazis and found that to his at least mild surprise, he liked everyone.
Cass Sunstein: They were all good people. One of them said this when asked about opposition: the former Nazi named Karl said, “Opposition? How would anybody know? How would anybody know what somebody else opposes or doesn’t oppose? That a man says he opposes or doesn’t oppose depends on the circumstances, where and when, and to whom, and just how he says it and even then, you must still guess why he says what he says.” Now, that’s offhand remarks by someone who wasn’t a social theorist, but who lived under Nazism. Its profound. He’s suggesting that the existence or absence of opposition is contingent on what’s permissible, what social norms are. To that extent, Karl is referring to the fact that in Germany, as in every society, people live in a state of pluralistic ignorance, which means we don’t know what is in other people’s heads. People might seem content with the status quo, or miserable about it, when in fact what they’re thinking inside, if you could see a thought bubble, would be very different. If they’re silent, it’s very hard to know what they’re thinking.
Cass Sunstein: The law matters if people lack freedom of speech and if dissent is punished, but social norms are often the villain of the piece, if there’s a villain, in the sense that they mean that people might be ostracized or shunned, or punished in some way if they say what they really think. I gave you some words from the 1950s. Here are some words from basically yesterday, from Syria. When you meet somebody coming out of Syria for the first time, you start to hear the same sentences: that everything is okay in Syria. Syria is a great country; the economy is doing great. It’ll take him like six months, up to one year, to become a normal human being, to say what he thinks, what he feels. Then they might start whispering. They won’t speak loudly. That is too scary. After all that time, even outside Syria, you feel that someone is listening, someone is recording. Okay, the Syrian computer programmer, in that case, is the same as the former Nazi, is the same as the young Saudi men who were saying something even to their spouses, which is different from what was actually inside their minds.
Cass Sunstein: Second moving part is diverse threshold. Some people require no support at all before they will say what they think or join a movement. They might be courageous, foolhardy or just deeply committed. We can call them, and this isn’t pejorative, the ‘zeros’ in the sense that they need nothing to join a movement of one or another kind. It could be white nationalism. It could be Nazism. It could be a liberation movement. If no one joins them, they’re going to be marginalized. They’ll look foolhardy, extreme, or possibly nuts. That’s the technical term. MIT Press didn’t let me put that in the book. Other people are going to require some social support, like me, in the Middle Eastern country; I supported my friend, but I needed him to go first. People like this won’t move unless someone else does. If someone else does, they’ll prefer to join, too. Call them the ‘ones.’ Others require more than a little; they need two people, so they are the ‘twos.’
Cass Sunstein: The ‘twos’ are followed by people who, not shockingly, have numbers assigned to them all the way up to ‘hundreds’ and ‘thousands’, including, eventually, the ‘infinites’ defined as people who, for one or another reason, won’t challenge the status quo no matter what. Okay, here’s the kicker. It’s extremely difficult to observe people’s internal preferences in light of preference falsification. It’s even harder to get at people’s thresholds and we ourselves probably don’t know what our thresholds are. In the Iranian Revolution, people who participated in the revolt were amazed that they did. Some of them turned out to be ‘fours’ and they had no idea. Others turned out to be ‘seventies’ and they might have thought that they were ‘infinites’.
Cass Sunstein: Consider the astonished words, now we’re actually talking about a large part of the dynamics of the American Revolution, John Adams wrote with amazement, “Idolatry to monarchs and servility to aristocratic pride was never so totally eradicated from so many minds in so short a time.” I have a friendly amendment to Adams; I don’t think he has it quite right. It’s not as if there was idolatry and servility eradicated. It’s that people who were silent about their resentment and distress, weren’t silent anymore. So, it seemed like a ratification although it was instead a kind of unleashing, Thomas Paine put it, I think, more precisely. He said, “Our style and manner of thinking have undergone a revolution more extraordinary than the political revolution of a country.” Now, for Paine to say that in the midst of the American Revolution, that’s drama. “More extraordinary is what’s happened to our manner of thinking. We see with other eyes, we hear with other ears, and think with other thoughts than those we formerly used”. I think Paine is also speaking of preference falsification undone. His language doesn’t quite get at that. Any reading of the American Revolution shows that’s what’s happening. Preference falsification and diverse thresholds are turning in a direction of movement because the ‘zeros’ and the ‘ones’, and the ‘twos’ started to act.
Cass Sunstein: Okay, last moving part is interdependencies. Everything depends on who is seen to have what… done what and exactly when. Diverse thresholds are one thing; whether people are going to move depends on whether the zeros go first and are seen to have told that father to stop hitting his kid, whether the ones are seen to have joined a movement, let’s say, for #MeToo; and then the twos and the threes, and the fours. If that’s what happens, we’re going to see a movement and it’s going to succeed. Everything depends on the distribution of action and the thresholds. If there are no zeros or if no one sees any of them, no rebellion’s going to occur. If there are a few ones, the status quo is going to be safe. If most people are tens, or hundreds or thousands, the same is true even if there are plenty of twos and threes, and fours. Okay, those are my three moving parts.
Cass Sunstein: I have to add one fourth component, which helps explain why social inflammation sometimes occurs. That is rooted in what I learned from a failed academic project from a few years ago. The failed academic project was a group decision making. In our work, we found how outraged individuals are about corporate wrongdoing and in trying to study jury behavior, we put individuals into computer-generated juries. We had a bunch of jurors; we created statistical juries and we took the median judgment of the statistical juries as the likely predictor of what the jury would do. Critics of our paper, only seven people read the papers so there weren’t a lot of critics, but of the seven, four pointed out that maybe the jury wouldn’t end up where the average or median juror was; that there was something artificial about our study. We followed it with a very large mock jury study involving actual deliberating juries, hoping to prove that we were right, that the median was the best predictor. We were wrong. People in deliberating groups, if they’re mad, end up much madder than the median individual. People in deliberating groups, if they’re feeling lenient, are more lenient than the median individual.
Cass Sunstein: Groups end up going to a more extreme point in line with their pre-deliberation tendencies. Having found that with respect to juries making punitive judgments, punitive damage judgments about corporate misbehavior, we followed it up with a study of political beliefs. The most currently salient example is climate change. If you get a group of people a bit concerned about climate change, and think there should be an international agreement, after they talk to each other, they are terrified about climate change and think we should sign an international agreement right now. If you have a group of people who aren’t worried about climate change, and thinks maybe it’s a hoax, even if the distribution of views is varied and some people think it’s probably a problem, international agreements may be a good idea, after they talk to each other, they think the whole thing is ridiculous. It’s made up by the environmentalist elite. Forget about it. This is real people in Colorado, actually, who went whoosh to the right and whoosh to the left, depending on their pre-deliberation tendency. That can fortify the three moving parts that I’ve emphasized.
Cass Sunstein: Okay, here’s the attempted resolution of the mystery of unpredictable movement. First, it’s really hard to know what people’s preferences are. They can’t be observed. Second, people’s thresholds are even harder to ascertain. Even if we know people are really upset about something, to know what their thresholds for action is, is really tough. Even if we could solve those two problems, it’s very hard to know in advance the nature of social interactions. Now, I’m going to give the weak version of the unpredictability claim, and then a somewhat stronger version. The weak version I’m going to give with some confidence. The stronger version is tentative. The weak version is that it’s just an empirical challenge that we’re not close to being able to surmount. The stronger version is conceptually impossible and humanity’s never going to be able to solve this.
Cass Sunstein: Here’s a gesture toward defending the stronger version. With Google searches, we can find out something about people’s private preferences. If people are Googling a lot, let’s say MIT Press Books published in April 2019, we have some reason to believe that even they were embarrassed to say they’re interested in MIT Press Books in 2019. That’s not the coolest thing that they’re actually interested in. I’m trying to give you an example that’s not as salacious as the examples that principally provide the data for figuring out what people really care about. There’s a book on this. It’s not as ridiculous and a little more off color than the MIT Press Book.
Cass Sunstein: Okay, so even if we could find from Google searches something about what people really care about involving politics or products, we wouldn’t know what their thresholds are. Even if we could know both of those, we wouldn’t know who interacts with whom, when. That’s very hard, I think, impossible to know because it depends on accidents, often. We might know that probabilistically, some movement is more likely than it would otherwise seem, if we could observe people’s private preferences, but we won’t know whether it’s going to happen or not because we won’t know who’s going to interact with whom and when. No one has that kind of presence.
Cass Sunstein: There’s probably a Black Mirror episode in production right now which is intended as a response to my more ambitious claim where it’s predictable, but I don’t believe in them. Okay, here’s an upshot. It’s often tempting at hindsight to say that some movement or reform was consistent with history’s arc, or was the product of some cultural disposition. It’s more often true that it’s a product of some small, random or serendipitous factor of who did what, when, of who heard what, when, or whether a butterfly flapped its wings at the right moment. History is only run once so it’s very hard to prove this. History doesn’t allow for randomized controlled trials. But when we think that practice or status quo or regime A fell, we often think it was bound to fall. It really wasn’t. It happened to fall. The same is true if it doesn’t: it happened not to fall.
Cass Sunstein: There’s an amazing Spanish Netflix show called something, “If I Hadn’t Met Her”, or “If I Hadn’t Met You”, which is profoundly about exactly this. Counterfactual histories are not observed outside of Netflix and science fiction, but they illuminate the power of serendipitous or seemingly random factors. Okay, having said this about social movements, it must be added that if we have clarity about preference falsification, about diverse thresholds and social interactions, we have some good clues about how to start or stop a social movement.
Cass Sunstein: The Chinese government is actually quite alert to some of this, such that the recent data with which, to its credit, the Chinese government has cooperated in the provision of, Gary King (Harvard political scientist) finds that the Chinese government on social media is not censoring disagreement and dissent. Not a lot. If you want to say negative things, that’s fine, as a first approximation. But if someone says, “We have a protest movement that’s meeting Thursday on this street and there are a lot of people coming,” that’s not welcome and that may find itself taken down. That’s clever in terms of the prevention of social disruption, that if there’s advertisement about a large movement at a time certain, then group polarization may kick in, then preference falsification may undone and then we may have the social interaction that can get things moving.
Cass Sunstein: Okay, this is a bare-bones picture. Let’s introduce just a few supplemental points. Here’s some words from a woman basically, the day before yesterday, not old, from North Korea. “It never occurred to me that I could or would want to do anything about it. It was just how things are.” That’s an arresting comment. The most arresting word is ‘want.’ “It never occurred to me that I could or would want to do anything about it.” Now, what makes that word important is that this isn’t like the Syria tale or the former Nazi. This is a story about preferences that have adapted to the status quo. Her wants were affected by where she found herself.
Cass Sunstein: In some cases, people’s preferences are adaptive to background conditions, which means they do not have a voice in their head saying, “This isn’t good.” There’s very puzzling data from subjective well-being studies, which finds basically that women’s happiness relative to men has been going down since the 1970s and absolutely. In the period in which women’s equality has been on the rise, women’s subjective well-being has been decreasing. That’s a paradox. One explanation from the path for the paradox is that the well-being data from conditions of, let’s say, more serious inequality is a product of adaptive preferences. If you live under circumstances of inequality, you may not be rebelling because the inequality seems part of life’s furniture. Okay, if we’re dealing with preferences that are adaptive or partially adaptive, then it’s inadequate to talk about preference falsification.
Cass Sunstein: That idea is too simple. Okay, second point. The word ‘preference falsification’ is under-descriptive. Often, what we’re talking about is people’s experiences and values, not their preferences. Sure, they’re concealing or falsifying what they prefer. But they’re also concealing their deepest moral beliefs, and most searing of all, what actually happened to them. They’re either silent about it or they’re lying about it. I’m acutely aware that the incidence of sexual abuse in the United States of children and adults is such that the number of people in this room who’ve experienced it or who’ve experienced it in their family is a lot higher than zero. That’s a case study in preference falsification being inadequate. We’re talking about silence about experience or about moral convictions. That’s worse. That’s fake news.
Cass Sunstein: We also have to note that if the goal of those who are in favor of the status quo, and it may be a great status quo or a bad one, is to maintain itself, there are a lot of options. You can allow dissent and disagreement until it becomes too visible. It might be an outlet. You might make concessions hoping to retain the existing arrangement by doing that. You might bring out guns. You might try persuasion or you might actually take advantage of what we know about the four mechanisms where the fourth, the supplemental one, is group polarization, to try to stop things before it gets out of hand.
Cass Sunstein: Okay, I want to concretize this by talking about #MeToo. My hope is that whatever social movement you’re interested in, or excited by or alarmed by, this will involve mechanisms that are completely adaptable to it. If they’re not completely adaptable to what you’re interested in or alarmed by, tell me about it because that will suggest the incompleteness of the account. Okay, with respect to sexual harassment and sexual assault, preference falsification, needless to say, has been pervasive. Victims have often said all was or is well, when not, or they’ve silenced themselves. As noted, that’s not enough because what many women and many, but fewer men did not reveal, what they kept private, was a set of experiences alongside evaluative judgments about those experiences. Experience falsification is really the engine here.
Cass Sunstein: Second, different women had and have different thresholds for disclosing their experiences and their judgments. I’m using women here deliberately, acutely aware that many men, including many men in this room, I hope not many, but at least some men in this room have experienced something like what I’m describing; but because the disproportionate number are women, it might efface something not to say that. Okay, some women are ones. Others are twos. Others are tens and others are hundreds, are infinites. The infinites are especially interesting. They might have some loyalty to the perpetrator. They might not want their lives disrupted.
Cass Sunstein: They might cherish their privacy. Some might have no clarity on what their thresholds are and they… and we will learn only ex post. Here are some words from Beverly Young Nelson, who accused Republican Senate candidate Roy Moore of having assaulted her in 1977. She’s very precise on what made her speak out. “I thought I was his only victim. I would probably have taken what he did to me to my grave had it not been for the courage of four other women who were willing to speak out about their experiences. Their courage inspired me to overcome my fear.”
Cass Sunstein: Third, social interactions are crucial to #MeToo. #MeToo has benefited from the visibility of those who spoke out and the multiple interactions made possible by social media. Taylor Swift was a contributor here with her lawsuit against unwanted touching, but Alyssa Milano was the largest instigator. Within 24 hours of her initial tweet, and I find this almost impossible to believe, but it’s so: within 24 hours of her initial tweet asking for #MeToo be said for those who had such experiences, 45% of all use… Facebook users in the United States had friends in their network who posted #MeToo. Within 24 hours – not a lot less than half. Once the ones and the twos spoke out, the threes and the fours were safer, or emboldened. Okay, this account is also very bare-bones and I think we want to emphasize just a couple of points for present purposes. One of the not very recent, but relatively recent electoral findings in behavioral sciences points to the power of descriptive social norms. I’ll give you an example. Do you all know about Twitter? It’s one of the new social media platforms. Have you heard of Twitter or not yet? Do you think I’ve lost my mind?
Cass Sunstein: Okay, so on Twitter, there’s a guy I follow who’s a behavioral finance guy, who wrote a book called, I think, ‘Behavioral Finance’. It’s a good book, but as its title suggests, it’s not challenging Stephen King for number one on the bestseller list. He tweets very frequently, “My book is doing better than expectations. Thank you for the support,” which is, I think, literally true and smart. He’s suggesting that his book is selling well and it’s probably doing better than expectations, which were zero. He’s not going to lie and say he’s doing great, but he say… says that. Now, what he’s doing is invoking a descriptive social norm, to the effect that people are buying his book. We know that if people think that people are recycling, or doctors think that they are not giving antibiotic prescriptions as much as they thought, or that people are paying their taxes in numbers that are really high, we can increase the volume of people who engage in the relevant behavior. That is, invoking the descriptive norm often creates a self-fulfilling prophecy, for better or for worse. Okay, that’s an old finding.
Cass Sunstein: The new finding, which is not widely known, but I think is as electrifying is, is if you don’t have a great existing social norm, point to an emerging social norm. Say, people are increasingly doing X, or X within your realm, your group, is the current trend. The reference to the dynamic social norm often fuels behavior and #MeToo has benefited from both of those. Okay, we have to note that with respect to #MeToo, we’re not speaking only of revelation of preferences and experiences, but also about the transformation of preferences and beliefs, and values.
Cass Sunstein: I’ll tell you something a little bit personal here and this isn’t very characteristic, but I find what we’re now describing, both relevant to politics and law, and powerfully relevant to personal life. Here’s my example. A long time ago, when feminism was just starting in legal circles, I was interested in it and started writing a bit about it, editing a book about it. I told my mother, who’s no longer alive, that… she asked, “What are you working on?” and I told her. She thought that was not a good idea, she said, “You’re a constitutional law guy. Don’t you work on the Administrative Procedure Act?” She was a very supportive mother and said, “Why are you working on this? That’s crazy stuff.” I said, “Well, there’s some really good work being done.” She said, “Stick with the Administrative Procedure Act.” Finally, I told her what Catherine MacKinnon and Martha Minow, and others were working on, and went at some length about my reading and she said with three words that she’d never said before and she never said after; she paused and she said with indescribable emotion, she said, “God bless you.”
Cass Sunstein: What I thought was in those three words was an assortment of personal experiences she herself had had, that she never told anybody about, and that in… with those words she was signaling; which is to say that #MeToo, as in many social movements, is not just about the revelation of preferences, beliefs, and values. It’s also about their transformation; most obviously, in this case, on the part of perpetrators, but equally relevantly on the part of victims. Any social movement doesn’t just unleash pre existing values; it casts a new light on existing experiences. It produces fresh ones. Part of the point of civil rights movements of multiple different kinds, right and left: one of their greatest achievements, and I hope, if there’s anything you remember from these remarks, it’s this, is to transform a sense of embarrassment and shame into a sense of dignity.
Cass Sunstein: Okay, final words; recall in this light, the testimony of a computer programmer recently from Syria. When you first meet somebody coming out of Syria for the first time, you start to hear the same sentences, that everything is okay. It’ll take like six months, up to one year, for them to become a normal human being again, to say what they think, what they feel. Then they might start whispering. They won’t speak loudly, but eventually they might. Thanks.
Robert Wiblin: Thanks again to ‘The President and Fellows of Harvard College’ for allowing us to use that copyrighted presentation. We now go to my conversation with Professor Sunstein. I knew I only had an hour with him and had a very long list of questions to ask, so we dove right in.
Robert Wiblin: I know you’re a very busy guy. Thanks for coming on the podcast, Cass.
Cass Sunstein: A great pleasure and honor.
Robert Wiblin: Yeah. Listeners will have just heard you giving a presentation explaining the core content of the book, How Change Happens.
In my mind you bring up four key concepts: preference falsification, differing thresholds for action, and group polarisation once people start organising themselves. You then add these together and suggest that they, along with the randomness of who talks to who when, makes social change fundamentally unpredictable.
Let’s dive into discussing how accurate all that is and, I guess, what implications it has for all of us who would like to make a better world. The book has been out for three or four months now. What do you think people have most misunderstood or underappreciated about it?
Cass Sunstein: Well, I’ve actually been really gratified that people have focused on the idea that there’s an interaction between social norms and the unleashing of what’s actually inside people’s heads. The principal theme is that we all have in our minds, goals and commitments, and aspirations and experiences, to which we don’t give voice. Sometimes that’s good because our commitments might be awful. We know it and so we shut up about them, but sometimes it’s not so good when our commitments are honorable or justified. There’s something in our community that makes us not say anything. People have really picked up on that. I think it resonates with both the right and the left. I confess, I haven’t seen misunderstandings, so pleased about that or maybe just, this is motivated reasoning so that I’ve blinded myself to the misunderstandings.
Robert Wiblin: Why did it take you decades to put together everything you needed to finish the book?
Cass Sunstein: I started with what’s now chapter three, the ‘Expressive Function of Law’, and I thought there was a core of something there, but I couldn’t figure out how to do it. It took, I think, subsequent work on ‘Group Judgments’. That’s now chapter two, and how like-minded people in the… end up going to extremes and how to think about that. Sometimes that’s part of democracy that’s best and sometimes it’s… creates horror. The exposure, probably, to the work of Timur Kuran, who writes on preference falsification and he became a coauthor in the late 1990s, that… post-dated, and probably my immersion in behavioral science, which started after actually, thinking about the expressive function of law, was what ultimately made the book possible.
Robert Wiblin: Have you heard any good arguments against the vision of how society changes that you’ve presented in the book? Have you gotten any pushback that’s changed your opinions?
Cass Sunstein: Some people have said something that’s in the book, but maybe not highlighted adequately, which is that often the consequence of a norm change is that you will create new values and commitments that didn’t exist before. I’ve heard from many interlocutors that the emphasis on unleashing what’s inside people’s heads isn’t exactly wrong, but it’s excessive. If you think of, let’s say, the movement for animal rights or the movement to ban smoking in public places, or the movement for preventing discrimination on the basis sexual orientation; a lot of that isn’t about liberating people to say what they think, but about changing values.
Cass Sunstein: That interacts well with another concern that people have raised, which is that people often adapt their preferences and their values to a sense of what’s possible. Jon Elster, the great Norwegian political theorist, has spurred work on the adaptation of preferences to what’s around and people have said that that deserves a great deal of emphasis. It is in the book, but it could have been foregrounded a bit more.
Robert Wiblin: Yeah, that’s one of the questions I had about the book. How much do you think social change is driven by these falsified preferences? I guess, especially in a country like the United States where, although you can be socially sanctioned for saying things that are unacceptable, you’re unlikely to be thrown in jail or beaten up or killed or anything like that, which, I think, is where Timur was originally doing his work on preference falsification. It was about revolutions and dictatorships.
Cass Sunstein: Yes. I think a lot, even in the United States and the United Kingdom, and France, and Canada, countries that do have either formal freedom or formal freedom plus cultures of, let’s say, welcoming dissent, I think still a lot. I’ll give a few examples; the movement for same-sex marriage, clearly, it’s the case that the closet was a devastation for the movement for that. As the closet started to open up, then something started to move. It’s much bigger than that, though, that heterosexuals who were fine with same-sex marriage or who were fine with equality or even wanted it, they were closeted, too. The existence of a shifting norm unleashed people like my mother, heterosexual, but my mother talked as if she was homophobic until at a certain point, she said, “That’s ridiculous.” She always had a voice in her head saying, “That’s ridiculous.” That happened to many millions of people in free countries. I think John Stuart Mill was on this point, that we often underestimate the extent to which conformity pressures are squelching behavior even in quite free nations.
Cass Sunstein: Mill, his lived experience with an illicit love affair with Harriet Taylor, which ended up creating terrible disruption in his relations with his family and his friends. I think he spoke for that and what Mill encountered when he violated a norm in his own community. All of us, I think, are in a sense, afraid of and I mean… that’s a strong statement, but I mean it that strongly; not like we’re terrified or cowering. In some sense, all of us are scared of what happened to Mill.
Robert Wiblin: Maybe the line between these two can be a little bit blurred because it seems in the case of gay marriage, probably lots of people had never really thought about it. They didn’t have some hidden preference in favor of gay marriage, but it was the case that because people who were gay were not speaking up about their own experiences, that they were failing to be persuaded where they could have been persuaded quite easily, if this… if the conversation was happening much more in the open.
Cass Sunstein: Yeah, so one big part of the movement for gay marriage is the unleashing through the softening of social norms of gays and lesbians to say, “I love this person. I want to marry him. You’re not allowing me?” The other thing that happened was that many heterosexuals who went around laughing in a homophobic way or thinking, “Same-sex marriage? That’s ridiculous,” but while in their head, they didn’t really think that. They shifted. As you say, you’re quite right that a number of people didn’t think about it much or thought of the ban on same-sex marriage as just part of life’s furniture and as the possibility opened up, they started considering it. What I want to focus on, because I think it’s particularly intriguing for extremely rapid social change, is when a dam breaks and many social movements, large and small, are exercises in dam breaking, where a social norm operated as a tax on behavior, so that you couldn’t do it without facing something like a fine. Then the tax diminishes, or what was once a tax becomes a subsidy, so you’re better off if you say what you think.
Robert Wiblin: Yeah, I guess I wonder what lessons there are here for activists, especially ones who are willing to campaign on a wide range of issues, whichever ones they think are more promising? Do you think it’s easier to get social change when people have hidden preferences that agree with you or rather than cases where you have to actually go and persuade them because they sincerely do disagree?
Cass Sunstein: The first is much easier. The fact that President Trump got elected and that Brexit came out the way it did, is a tribute to the existence of pro Trump and pro Brexit sentiment that had been stopped by social norms. Once the norms started to shift, then something extremely surprising in both cases could happen. That’s the easier one. Activists do have some lessons, which are; first, if you can draw attention to a social norm that people haven’t realized is in place; that is, people are with you. You don’t know it. Say what you think; that can be extremely effective. Many activists are, at some level, aware of that. The other thing, which is very recent data, not even sure it made it into the book, is that if people see a norm not as in place, but as emerging, then they are inspired to join, partly because then they feel freed, but partly for, to your point, they might not know what they think, but they want to be on the right side of history. For activists to say, “Increasing numbers of people are,” is smart.
Robert Wiblin: In light of that, one point you make in the book is that it can be incredibly hard to figure out what people’s hidden preferences are, and what are their… what are people’s thresholds for actions, and sometimes even they don’t know it. It seems like people campaigning for social change, both progressive and conservative, maybe they should be spending half of their resources just engaging in very extensive public surveys to try to figure out where people already agree with them because those campaigns are just likely to be so much more successful than ones where they banging their head against the wall because people actually just sincerely disagree. Yeah. Do you think it’s maybe something political parties or campaign groups should invest serious resources in finding out?
Cass Sunstein: Well, I think to tell people how to allocate money is hazardous business, but certainly, it’s a wise approach, both to target existing views, which people haven’t felt free to signal and there’s research that can help you find out what those are. Google knows a lot about that. There’s a nice book called ‘Everybody Lies’, which are things that Google knows. Some of its really surprising. It’s an avenue into uncovering where private preferences are and to have as a strategy that to disclose a fact… I’ll give you a little study from Saudi Arabia, which is that Saudi men, by custom, have authority over whether their wives work outside the home. Most young Saudi men actually think it’s fine that their wives work outside of the home, but they think, most young Saudi men think that most young Saudi men think it’s not fine. They think they’re isolated in their openness to wives in working outside of the home.
Cass Sunstein: In the experiment, once Saudi men were informed that most Saudi men actually think like them, then the number of Saudi women applying to join the workforce grew dramatically four months later. That was a research study, not a feminist program. There’s a clue there about programs of all sorts.
Robert Wiblin: Yeah, it’s very interesting. It seems like you could just almost conduct scattershot surveys of… try to find lots of cases where maybe people’s private views are somewhat different than their public views, and where they agree with you, and then just publicize the results every time you find out that, in fact, lots of people already agree with you. Yeah.
Cass Sunstein: Yes, completely. President Trump does use this, completely bracketing whether we like or love or don’t like President Trump, but he’s very intuitive about this, saying that most people think X, and he’s often not wrong. Most might be a little too strong, but many and that is a little bit social proof.
Robert Wiblin: The rate at which you seem to write books is pretty phenomenal because just two months after publishing ‘How Change Happens’, just last month you’ve published a book called ‘Conformity’, which goes into even more detail about this issue of group polarization. I’ll stick up a link to an extract from that book, which I found very thought-provoking. How confident are you that the evidence for group polarization will stand up in the replication crisis? Are there any circumstances where groups moderate rather than extremize? It seems like they always be extremizing or the world would just be absolutely wacky.
Cass Sunstein: Okay, so this is… these are great questions. Actually, the ‘Conformity’… my ‘Conformity’ book, I based on lectures that I’d never published, that were published in 2003. I had a fondness for them and I’d been working on them over the years, and I relatively recently thought maybe I should put them together as a book. ‘Conformity’ took about 16 years and the other took 25. I don’t feel… I feel like the tortoise, not the hare. That’s that. On the replication crisis, I have maybe an idiosyncratic view, which is what happens in Syria these days, or what’s happened certainly over the last period in Syria. That’s a crisis. Climate change is arguably a crisis. I think it’s fair to deem it a crisis. If there’s an earthquake, there’s a crisis. The idea of a replication crisis strikes me as, it has become the term, but it strikes me as… what’s the…?
Robert Wiblin: Hyperbole.
Cass Sunstein: To say it’s hyperbole is an understatement. It’s true that it’s hyperbole, but it’s… I mean, only academics, I think, would think that the failure of some of their research to replicate as a crisis rather than learning. Now, the fact that we’ve gotten more disciplined about ensuring that a finding isn’t a one-off or a finding isn’t an artifact of a particular, let’s say, fraudulent or inadequate design, that’s very important. That is progress, so that’s great. I was very obsessive about the group polarization material. It’s been replicated countless times. We don’t have just one group polarization study. I’ve been involved in numerous ones myself and it’s always happened, but I’m a very modest contributor to the empirical research. It’s a very robust finding.
Cass Sunstein: Now, the finding means that like minded people talking with one another typically end up in a more extreme position in line with their antecedent tendencies. That is, in typical form, what will happen within deliberating groups, but I tried to be clear in the book ‘Conformity’ and it probably made its way in ‘How Change Happens’ also, about the circumstances when it won’t happen. Jim Fishkin, an excellent political theorist at Stanford, has done a deliberative opinion poll where he does not observe… group polarization and one reason is, he supplies people with information that isn’t self-generated by the group; information for and against various positions. If you have some epistemic intervention, then that will undermine the mechanisms that lead to group polarization; so that’s one. If you have a group of people that consists of, let’s say, two sets of subgroups which are equally opposed, then they aren’t going to polarize. They’re just going to entrench themselves typically.
Cass Sunstein: Now, if they have some softening capacity, then something surprising can happen. Typically, two, four people who think Israel is the greatest country and never done anything wrong; four people think Israel’s the worst country and they’ve done everything wrong. They’re not going to move each other; they might end up liking or hating each other, but they’re not going to polarize in the way described. A cool finding is that for, what are called, #Eureka problems’, where the answer, once announced, is clear to all, we observe consensus on the right answer, not group polarization. It’s why a large group of people can solve a crossword puzzle where someone will say, “Four letters. What’s the word for going south, down?” That’s how they feel it and they say, “That’s it. That works.” For that, you won’t observe polarization. Also, some things can happen in individual groups, where you could have a charismatic or extremely informed participant who thinks that the dominant tendency is wrong.
Cass Sunstein: Then the group could… I haven’t done any experiment where this actually happened, though I’ve done a lot, but in the world, it happens where a charismatic informed person can be like someone who says, “Down, four letters going south,” and moves the entire group in his or her preferred direction.
Robert Wiblin: Yeah, I know a group that has explicitly aimed to use group polarization to radicalize its own members and make them more committed to their work, and I guess they hope that these radicals will then form a vanguard that will go on and convince the rest of society of their views. I got to say, this makes me feel pretty squeamish because it’s deliberately making your members less reasonable and have less of a balanced overall opinion. It does seem like it could be justified in some cases, if that’s the only way that you could get an outcome that was desirable. Do you have any feelings about that approach to social change?
Cass Sunstein: I share your view. Everything you said I agree with. Suppose you believe, let’s say, that climate change is a serious problem and we should do something about it. The preferred way to move in the direction of that goal is to persuade others that it’s true; not to have, let’s say, a climate change camp where people are deliberately heating themselves up, so to speak. Now, if it’s climate change camp, where people are exchanging information so that they know what they’re talking about, it’s designed for learning and moral commitment rather than for group dynamics; that’s better. On the other hand, you can imagine circumstances where, let’s say, a polarization entrepreneur is trying to combat something horrible and doesn’t mind the fact that the social dynamics in the group are going to create an extremely committed and an extreme set of people. That’s not out of bounds, but it would be a little bit of a, I think-
Robert Wiblin: You really got to hope you’re right.
Cass Sunstein: … yeah, an emergency situation.
Robert Wiblin: When you get groups together, sometimes they must extremize just because they become more informed, if they share information and it turns out that an extreme position is the correct one. It’s not always the case that moderating is the right thing to do. Do you have any thoughts on norms that groups can follow that can help them tell the difference between warranted and unwarranted extremization?
Cass Sunstein: Okay, I think you made a very important point and let’s press a little on the factual predicate. You’re right. It’s a double-edged sword. First question is why do groups end up going in a more extreme position in line with their pre-deliberation tendency? Why, if you have a group of people who think that occupational safety is the number one issue and workers are dying in extremely high rates, why do they end up being more extreme in that position after they talk with one another? The reason, the principal reason, I think, the data, I’d be more careful in… you know on reflection, than saying it’s the principal reason, it’s not less than any other reason, is probably the more accurate way to put it; is just information exchange.
Cass Sunstein: Within a group that’s worried about occupational safety, by definition, the number of arguments that support the concern will be more than the number of arguments that undermine the concern; and so if people are listening to one another, they will hear more arguments in favor of the position and that’s going to lead to polarization. Now, that has nothing to do with anything invidious; it doesn’t have to do with bounded rationality. It just has to do with information exchange within a group.
Cass Sunstein: Okay, but here’s the kicker in terms of rationality; that people in groups are not inclined to discount for their group’s composition. There’s insufficient thinking that, “Okay, I’m learning from these people,” but these people aren’t the only people in the world and maybe the information flow within the group is unrepresentative of the information that’s there. So even if information is the driver of polarization, it can lead people in not very good directions. If that happens, what you want, I think, is a group where people can participate in enclaves of like minded types, that’s part of freedom of association. But you want to make sure also, that social architecture, let’s say, makes it easy or likely that participants within the enclave will be exposed to other stuff, too. And if they are exposed to other stuff, it’s not like to hold it up as if it’s a piece of dirt and say, “Look how ridiculous it is.” This may sound a little bit abstract, but you can think of a newspaper or the BBC, or Facebook as being able to create enclaves or to broaden, let’s say, the enclave exposure, the other stuff.
Robert Wiblin: I saw an interesting paper recently suggesting that people are very worried about, yeah, like groupthink in online communities, but it actually seems like people’s real life groups are even more enclaves of political views. It’s like you think about your housemates, your family or your colleagues, they’re almost even more grossly selected than the friends you encounter on Twitter, or Facebook, which is an interesting spin on things.
Cass Sunstein: I’d be careful about the data on that. The data goes both ways. And it may be that within your neighborhood, the amount of epistemic diversity let’s say is lower than the amount of epistemic diversity on your Facebook page. That might be, I don’t think we actually have a clear answer that. But that might not mean that the epistemic diversity on let’s say, Facebook, which is greater by hypothesis, we’re not agreeing, but we’re hypothesizing, that that is helping. Because it might be your neighbors, you talk to them about the weather and your families, and activities there, not about politics or causes.
Cass Sunstein: But it might be in this, it might be on Twitter, and on Facebook it’s the newsfeed which is potentially a polarization machine.
Robert Wiblin: I guess one thing is you might like the people who you meet in real life to a greater degree and so take the views more seriously. Where it’s a lot easier to have contempt for people on the internet.
Cass Sunstein: Yeah, that is so.
Robert Wiblin: Yeah, I wondered as I was reading the section on group polarization, whether it’s possible that that part of it could be an illusion, because basically, people want to fit into a group, and so they start like misleading, they start lying about what their views are. You’re basically getting preference falsification within these experiments, because people are just trying to go along to get along.
Cass Sunstein: I have data on this.
Robert Wiblin: Oh great.
Cass Sunstein: One of my projects involved citizens both to right of center and left of center in which their views were collected anonymously, pre-deliberation, then the group’s went to a verdict publicly on climate change, issues involving racial discrimination and same sex unions. And then they were asked to record their views anonymously after. Their anonymous views were really close to the views they expressed in groups, not the same, but really close, which supports your view. There is a degree of preference falsification, but in this study it’s small, and the shift in views from what happened in the private, anonymous view, before they talked to like minded others, and how they recorded their views anonymously after they talked to like minded others. That was what I was most interested in, and its really dramatic. Their anonymous view shifted, as a result of talking, speaking with others.
Cass Sunstein: That fits with some of the original conformity experiments, where you have people defying the evidence of their own senses, in order to agree with the group. And a certain number of them did say, “I knew what I was seeing, and they were crazy, but I didn’t want to look like an idiot in front of everyone.” But others said, “I thought this is what I was seeing, but since everyone else thought otherwise, I went along with them. I guess I was wrong.”
Robert Wiblin: Yeah, I’m sympathetic with those people because it seems like just in normal life going along to get along, is actually potentially a good strategy for often achieving your goals, because you just don’t want to be making trouble and enemies all the time.
Cass Sunstein: Yes.
Robert Wiblin: Yeah.
Cass Sunstein: The first account of conformity and also group polarization, which feeds into the change theme, is that people are learning from others, and they actually do change their views. The other is that they are not learning, but they are trying not to be unpopular or be excluded. And so they’re really fighting themselves. I’ll tell you a personal example, which I’ve actually never talked about publicly so I’m pleased to do that here, because it is closely connected to your point. I worked in the White House for four years, and the position I had, I had to be confirmed by the US Senate. I almost didn’t get through because of my work on animal welfare and such. It was truly a nightmare with death threats and ridicule. Ridicule is okay, death threats, not a whole lot of fun.
Cass Sunstein: Since that I have been much quieter about animal welfare, animal rights. I’m going to start to get less quiet and I haven’t been completely quiet. But I didn’t not notice the extent of, let’s say, conformity pressures that were surrounding me.
Robert Wiblin: Yeah, I’m sorry to hear that. I think I would probably button up if I started getting death threats as well. This leads into my next question, because I was going to say, if we’re to believe the book, it seems like rebels and contrarians and people who are willing to accept death threats can have a very outsized impact on society’s directions. Typically, their ideas wouldn’t take off, but every so often, they just set off a massive cascade that goes far beyond anything that they could have directly accomplished. Do you think we should pay more attention to idea rebels and contrarians and potentially try to become them if we’re willing to bear the personal cost involved?
Cass Sunstein: Well, certainly, to pay more attention to them is really important, though, to pay more attention than to for what is a natural follow up question. To pay more attention to them to think about how change happens is crucial. If you think about the rise of the feminist movement, the rise of the civil rights movement, there are people who were and are by nature, rebels, they don’t need social support, they’re relatively indifferent to the informational signals they’re getting from the majority. They think the majority is full of nonsense, and they’re relatively indifferent to staying in the good graces of people. They say what they think. And so to pay attention to them, even if their names never end up in the history books as essential is… that’s really important.
Cass Sunstein: The category of rebels is so large, and it includes people who are literally crazy, or people who just have anger management problems, or people who are fastening on something that is a cause that has no merit. I wouldn’t want to glorify them independent of what it is they’re doing. People who kill people are often fighting for a cause, which they sincerely are committed to. There’s a paper that has been cited many times but hardly anyone’s ever seen it, because it’s never been published. It’s called ‘Misfits and Social Progress’ by a superb behavioral economist called David Hirschleifer. And he has written the paper, but he hasn’t published it. I’m trying to get him to publish it. But it’s right on your point about understanding the extent to which people have very low thresholds for acting, and are impervious more or less to social pressures, are important for social progress.
Robert Wiblin: Do you think that the kind of people that you know and potentially listeners in the audience are too cautious about being outspoken, because they’re worried about how it might limit future government careers or other careers? Or do you think they’re not cautious enough? Do have any sense?
Cass Sunstein: Well, I would give a reckless answer, which is, yes, they’re too cautious. I think the median human being errs toward undue caution. The median human being doesn’t err in the direction of civility and grace. There’s almost in my view, no way of erring if you’re being civil and gracious. That’s close to bedrock. But most human beings have moral convictions that deserve airing, or experiences they’ve had that are shared, and they shouldn’t happen to other people and to ratchet it up in a way that’s respectful and civil, is usually a good thing.
Robert Wiblin: Do you worry that if we had too many ideas entrepreneurs or too many rebels that society could become a bit of a jumble? They’d be like too much conflict and too rapid of a social change. That maybe we need a lot of people to just be boring to hold everything together?
Cass Sunstein: Well, so it’s a good question, and we need to know what society we’re living in. If we’re living in a society which is fundamentally good, then to have respect for the informational signals given by others, and to have people concerned about what other people think about them is really important. If you live in a society that’s a horror show of injustice and oppression, then the norm should be one of struggling against it. If you’re on a continuum, my country I think is close to the first, closer to the first, not perfect by any means, but not a horror show, probably can do a little more.
Cass Sunstein: Now I’m in favor of respecting the law, so civil disobedience I think would be a last resort, that is an instrument for chaos. We don’t want lawbreakers, courtesy and civility are good. But it’s often good for self as well as for society for people to say what they think. Now we want to be careful about that because it’s all pervaded by substance. If people think that people of certain skin colors are inferior, it’s not so great if they’re broadcasting that, or if they think that people of certain religious convictions should be in jail or deported. That’s not very good to have as a welcome statement.
Cass Sunstein: In the abstract, norms in favor of self-silencing might be really important. They make stability and peace possible.
Robert Wiblin: You mentioned that the Chinese Communist Party allows people to air grievances it just doesn’t allow them to coordinate to promote a revolution. And part of the reason they do that, is so that they can keep tabs on what people actually think, so that they don’t become unaware of massive discontent in the population. I wonder, a lot of people in countries like the US or the UK, they put their activist energies into trying to discourage people from expressing bad views that they don’t want to get a foothold and don’t want to spread. But I wonder whether this does create this risk that then you have a lot of people who disagree with them who are biting their tongues, and then it can all explode all at once, and maybe you want to moderate that a little bit so that you can have a more accurate sense of what people sincerely think.
Cass Sunstein: Let me offer if I may my limited understanding of what’s happening in China, based on my limited understanding of the empirical research, but I think this is what’s happening in China. Social media dissent, as you say, is generally permitted. I’m not aware that the reason for permitting dissent is to keep tabs on what people are saying, that might be true, but I don’t know that to be true. The idea is, if there’s dissent that’s okay. The Chinese government’s response appears to be to add to the system a lot of different voices, that is voices of celebration of how things are going, not to censor. But if someone says on social media, “We’re having a demonstration in a specific place at a specific time, please come.” that may well be taken down.
Cass Sunstein: And that in terms of protecting the stability of the system, that’s clever, and it fits with a lot of the things we’ve been discussing. A concern is that you can create a cascade or a polarization machine if you have like-minded people ending up in a certain place at a certain time. You might give a green light or license to activity that had formerly been pressed against by social norms. That seems to be a clever strategy. I think, to your point that if a society let’s say that’s at risk is imposing, self-consciously through the government or through private institutions, political correctness, let’s call it writ large, not just left leaning institutions of higher learning their political correctness.
Cass Sunstein: It may be that the people will feel that they’ve been treated disrespectfully, they might feel their dignity has been assaulted. And it’s possible that there’ll be hell to pay. I think a lot depends on what types they are and the firmness with which they hold their convictions. This could be modeled in very precisely economic terms. As a first approximation, if you can get away with producing whether you’re a government or a private employer, if you can get away with inculcating a norm of a certain kind, that’s not a silly project, even if some people are going to abide by the norm, while they’re gritting their teeth.
Robert Wiblin: What’s your best guess for a currently hidden position within the United States that a social entrepreneur could take advantage of? Do you have any guesses for what the next revolution in ideas might be?
Cass Sunstein: It’s in the nature of the beast, that it’s going to be really hard and that I think is interesting. The reason it’s going to be really hard is that first, you have a hard time knowing what’s inside people’s heads. Google knows probably more than any institution ever has, but it doesn’t know a whole lot really compared to what’s there in people’s heads. And even if what’s in people’s heads, you don’t know what their thresholds are for acting or speaking out. They might have in their heads I care about pesticides a lot, but they might have a lot of things to do, so to do something about pesticides would take maybe a lot of social support, or a feeling of crisis. Even if you knew what’s in people’s heads, and even if you knew what their thresholds were, social interactions are essential to getting something off the ground.
Cass Sunstein: You need the right people talking to the right people at the right time, and that’s, I think in the nature of the thing, not possible to anticipate. There’s a terrific new paper Duncan Watts is one of the coauthors of the difficulty of predicting in real time what’s going to happen in history. His study is actually of a much more tractable question, which is whether people in the national security area who are reporting to the United States State Department about historic events in real time, do they actually know what was a historic event as by decades later, what became historic? The answer is, actually they didn’t. They weren’t completely random, but basically they were way off the mark. And Arthur Danto, the philosopher of history actually tried to explain this in 1965, that if you’re an ideal observer in real time seeing something, and I’m far from an ideal observer, in response to your question.
Cass Sunstein: That even if you had an ideal observer, he or she wouldn’t know the sets of interactions and subsequent events that have to happen in order for something to become real. Now, having given all those disclaimers, I’ll give you the one example, I think, where it’s not completely reckless to anticipate something possibly happening and that is animal welfare. The reason is that it has some features in common with gay marriage and with feminism, where there are many people who are concerned about animal welfare. But they don’t say so because of social sanctions, or they don’t say so because they’re just focused on many other things too.
Cass Sunstein: What we have is a potential there, given the existence of preference falsification, or openness, we have the potential for a lot of people doing something, and economic developments are making it more feasible. I just had my first Impossible Burger a few days ago.
Robert Wiblin: They’re really good.
Cass Sunstein: And they’re really good. To distinguish the impossible burger from the best hamburger you’ve ever had, I think is hard. And I am not an investor in the company. As the availability of substitute becomes easier, the moral conviction may come to the fore.
Robert Wiblin: Yeah, I was actually going to suggest that as possibly one that we’ve seen evidence that is not going to take off, because objections to the treatment of animals have been around for a long time, and many people agree, but people are eating more meat than ever before, and it hasn’t really had the virality or it hasn’t created a social cascade over the last 40 years. I was going to suggest, maybe this is evidence that this idea just isn’t really ripe for social cascade, and so possibly people should move on to other issues or attempt other approaches to try to solve it. Because there just aren’t enough people at every level of willingness to speak up that can create a full blown avalanche.
Cass Sunstein: Well, a famous American baseball player once said, predictions are hard, especially about the future, and so I’m cautious here. But John Stuart Mill wrote this objection of women a long, long time ago, and the feminist movement didn’t take off in his direction for decades and decades and decades and decades and decades. The idea that slavery was a bad thing was around at the time the US Constitution was ratified. In fact, it was around big time. It wasn’t until the Civil War, that it was abolished. So the fact that the animal rights movement or animal welfare movement has had, let’s say, incomplete success over the last 40 years. I wouldn’t take that as a powerful evidence that it’s not one of these areas. I’m not even sure it’s evidence at all, given the alacrity with which movements move once they start.
Robert Wiblin: Yeah, there is some evidence that there is suppressed opinions that are ripe to change. If you do public opinion surveys, you find that the vast majority of people say that they think it’s wrong, to cause farm animals to suffer, and then when you present them with the inconsistency, with the fact that they eat meat, they resolve it by saying that they think that farm animals don’t suffer, but that actually the conditions are good, which is obviously a farcical response. So I suppose it suggest if you could like really… You’re suggesting if you could make it cheaper to be vegetarian or vegan, and you could like make it very difficult for people to resolve that inconsistency by just claiming that farming is actually humane, then maybe you could get like a quite a rapid change in behavior.
Cass Sunstein: Right. The Behavioral Insights Team in the United Kingdom, with whom I’m privileged to work occasionally has easy, attractive, social and timely EAST as a mantra for social change. If you make something easy, the likelihood that people will do it increases often much more dramatically than one would ever expect. So we have in the United States, 15 million children enjoying free nutritious school meals, these are poor kids who are entitled to the meals, because we have a policy that automatically enrolls them. They don’t have to be in, they can opt out, but it’s really easy, they’re just in. And so to make Impossible Burgers or vegetarian alternatives easy would have a big impact. Social refers to the social norm, the S in EAST, A means attractive. To have something that is not seeming fussy and elitist, and high minded, but instead attractive.
Cass Sunstein: There’s something I learned very recently after the book that’s closely connected with one of the themes of the book, which is Amazon sells certain packages in what’s called frustration free packaging. And I know this because I get electric shavers and electric razors to take them out of normal packaging is a nightmare. So I leapt at the idea of frustration free packaging, and it’s a wonderful godsend. And then I looked up frustration free packaging, what’s it about? It’s about sustainability. It’s an environmental thing, it’s all recyclable, it’s less solid waste, no plastic, and it looks as if the real motivation is environmental, but they don’t sell it that way. They sell it as frustration free packaging. That’s the A, attractive, and T means timely. Just have it so that people see the relevant thing when it matters.
Robert Wiblin: But one way that I thought moral circle expansions like animal welfare, or worrying about ways that we’re harming future generations, that maybe different from previous revolutions is that in the past you could have gay people speak up in favor of gay marriage, because it affects them personally. But we’re now at the stage where we’re talking about animals and future generations — groups that can’t speak up for themselves and can’t advocate for their own welfare. Do you think that’s going to be a barrier potentially, to getting these moral revolutions to take off in the same way as past ones have?
Cass Sunstein: Yes. I mean, with climate change when I worked in the US government, we did things including a clean power plan to reduce greenhouse gas emissions from power plants, a fuel economy standard for automobiles, one for trucks, energy efficiency standards for appliances, and the effect in reducing greenhouse gas emissions of all these things was massive. One of the motivations for these greenhouse gas reduction initiatives was protection of future generations. While the current administration has repealed some of these things, it hasn’t by any means come close to repealing all of them. And so the future generation idea is something to which current people are willing to attend, and empathy is real, with the number of white people who marched for civil rights really high.
Cass Sunstein: The number of men who have worked on behalf of sex equality is also really high. The Supreme Court of the United States, actually very controversially to many, ruled that sexual harassment is a form of sex discrimination, and it was either all or almost all men who joined that ruling. I think of the nine there were eight, and a moral motivation. It was a legal judgment, but it had a moral component to it.
Robert Wiblin: I was thinking about the social cascade model, it seems like in order to have an idea explode and go to saturation where almost everyone agrees with it, you need to have people at every stage of willingness to speak up. I guess you have these like number system, that we’ve got like zeros speak up anytime, no matter what. Then you got ones who speak up, if the zeros speak up, and twos who speak up if the ones speak up and so on. It seems like if you have a break in that chain, where it’s like we get to five, but then there just aren’t many sixes, sevens, eights, and nines. But there’s a whole lot of people who would speak up, but then they’re at level 10. You can get stuck, you get to this hump, where there’s a bunch of people who agree, but then you just can’t get any further without I guess some external shock to push the 10s to start speaking out. Do you think that’s a right way to conceptualize this?
Cass Sunstein: Completely. It’s of course a very stylized version of reality, and it reduces the complexity of human behavior to a few numbers. But like many stylized, it’s informative, You need for many social movements to start, you need people who will do it even if they have no social support. Rosa Parks was one. I think Peter Singer, the philosopher who works on animal rights is another. Martin Luther King’s probably a zero, meaning he’d go even if he didn’t have much social support. And then there are people who will join them, they’ll be their first lieutenant, but they need to see the zero first. And then there are others who need to see the first lieutenants and the zeros. This is how movements often develop. But you’re completely right that if there are no twos or no visible twos, then the threes are just going to go about their business, and you can get stuck if you can’t find fives and sixes.
Cass Sunstein: Often a social movement will happen, even a very dramatic one that will change a country, but it wasn’t inevitable. It seems inevitable in retrospect, it wasn’t it needed the fives and sixes to be visible and present. And things that we never see or music that never gets popular, movies that fail. It’s a product of the fives and sixes don’t appear. They either aren’t there, no one sees them.
Robert Wiblin: Yeah, I guess it means that everyone is essential, because it’s like, as soon as you break the chain, then there’s potential for the movement to stagnate. I guess, maybe I guess people should look out for like, what things are there… kind of the next person who can speak out on this issue to push it along, and you can seek out lots of cases where you’re the next person who’s willing to push it forward.
Cass Sunstein: Yes, fortunately, we’re talking about a temporal cue with people with numbers assigned to them. Societies are sufficiently large, that the number of fives and sixes may be 400,000. And if some of them are visible and willing to move, because they see the people before them, then something will happen. If you look at what happened in the Arab Spring, by the way, it has many features in common with what we’re discussing. Where the US government and the UK government did not anticipate the Arab Spring. It’s a case study in cascade effects, both within countries and across countries where what had to be done was there were rebels, the famous incident in Tunisia, which was very important. There are rebels who are seen and then there are ‘maybe’ rebels, and then there are ‘maybe maybe’ rebels, and then there are the ‘maybe maybe maybe’ rebels. That’s basically what happened in many countries a few years ago.
Robert Wiblin: Is there anywhere in your model for outspoken activists turning people off their ideas? It does seem like you can have the wrong people who pioneer a movement and give it a bad name. And they can make it even harder for other people to speak out, because then they have to affiliate with these other people whose just seem crazy or generally unappealing.
Cass Sunstein: That’s very important. It’s not in the book, but probably should have been. I’m trying to think of examples but it’s clearly the case, that if you have someone who is unattractive in any one of a number of ways, or is a hypocrite or something, that they can put the whole movement in disrepute. I think one reason by the way that I’m struggling to think of examples is that we focus a lot on movements that succeed, the anti-smoking movement, the feminist movement, to a large extent certainly the civil rights movement, Brexit, Trump. The movement for buckling your seat belt. These are all within memory, successful movements.
Cass Sunstein: The movements that fail are elusive to the mind. But there’s a counterfactual in which they didn’t fail, and science fiction writers are very good at developing them. It takes imagination, but the fact that it’s a counterfactual history, a movement that never got off the ground may well be a product of a, let’s say, the opposite of charismatic leader, or a really strategically foolish leader. And then we don’t know about the movement.
Robert Wiblin: Yeah, one case that I’ve heard about is I think the animal rights movement in the UK. The well was really poisoned by a few people who engaged in terrorist actions against animal labs with animal testing. I don’t think they actually killed anyone, but nonetheless, it really discouraged I think moderates from getting involved because they kind of had a violent reputation. And it kind of… It leveled out for a while.
Cass Sunstein: That’s great. I mean, it’s not great, but it’s a great example.
Robert Wiblin: Do you have any sense of how you can go out and find out about like movements that almost succeeded or crashed? It seems like we need much more of a database of these and much more of an understanding of them?
Cass Sunstein: Yeah, well, there’s some historical work, and this is something I’ve scratched the surface on only, like of constitutions that failed, which is a way in, because constitutions are often statements of basic principle by people who have views that never got any anywhere. To look at constitutions that were produced, that’s one avenue. To think very inventively about movements that never succeeded, they could be religions, and there are variants of Christianity. This is all extremely interesting, and there’s a very good book that uses informational cascades to think of the… to explore the network theory, to explore the triumph of Christianity. Some of the work on Gnosticism and the Apocrypha is very interesting in this regard.
Robert Wiblin: Yeah, I like the idea of doing constitutions or I guess yeah, theology, because those ones are nicely written down, or they’re nicely legible in a way that some other like social movements that are just about people’s opinions and groups are less written down. You put it in the book that if a group consists of just people, then it tends to be less polarization than if it consists of Republicans or defenders of the second amendment or pro choice activists. I guess given the importance of reasonableness and just having true beliefs for the effective altruism community, do you think we should just tend to identify as people who are interested in effective altruism rather than call ourselves effective altruists, which is a bit of a cringe term anyway?
Cass Sunstein: Okay, so the fact that solidarity can diminish dissent and lead to more polarization, more unity, more confidence, more extremism, that’s not necessarily a bad thing. If you have a movement let’s say for human decency, and people become extremely unified and committed to the idea of decency, and I’m using that word as a placeholder for a value that is unambiguously good. The fact that its citizens for human decency, rather than people is instrumentally good. So it’s Hume, you can’t get an ‘ought’ from an ‘is’. The fact that a particular identification or label can increase the likelihood and intensity of polarization, that might be a good thing.
Robert Wiblin: Interesting. Yeah, I guess I feel like with effective altruism, yeah, not falling into the trap of just having grip solidarity, where we all agree with one another is especially important, because the whole idea is that we’d be willing to change what we’re doing and what we think and in response to new evidence, and so it should be like an underlying set of principles that can adapt to new circumstances rather than a campaign that runs its course.
Cass Sunstein: I’ll give you an example that I’ve thought a little bit about, I’m a fan of effective altruism, but not an expert. I have been to one of the meetings and I admire what’s being done there. There’s an online journal out of Australia called Quillette, which is often described as part of the internet dark web. And it basically is, as I read it, I’m not a ‘read everything in it’ person. But as I read it, it’s trying to be empirical about questions where often some orthodoxy is in the not unreasonable view of the authors, defeating empiricism. It might be asking questions about gender difference in a way that borrows from evolutionary psychology, or it might be writing about the firing of some person on a university campus in a way that reflected unwillingness to listen, or allow a presence of someone who had, let’s say, a right of center view or right of center interest.
Cass Sunstein: It’s a free thought entity, and I’ve been admiring it. But I’ve seen recently that the people in Quillette themselves are very self-conscious that they might become their own echo chamber. And they don’t want to do that. If you are by design let’s say a anti-political correctness group, there’s a risk that you’ll become your enemy. To their credit, they are alert to that, and they don’t want to be that.
Robert Wiblin: I know you’ve got to run in just a second, but I’ve got two quick questions on your career and what you’ve learned from it. I guess, if you were starting your career all over again. Imagine that you have everything you know now but you’re 18, in your old self’s body finishing high school in 2019. What do you think your plan for having a social impact ought to be?
Cass Sunstein: Not to have a plan for having a social impact, to be your plan for having a social impact. I think if someone says “I’m going to have a social impact”, that’s actually self-defeating, it’s a little like saying, “I’m going to try to get to sleep.” That’s not a good way to try to get to sleep. To care about something. Myself, in terms of my career, I’ve not focused on having a social impact. If I’ve been lucky enough to, it’s been a byproduct of pursuing interests.
Cass Sunstein: My wife who was the US ambassador to the United Nations, and really has had a social impact, her advice for having social impact, I think I would embrace which is, “Know something about something.” And I think I would tell my younger self, “You’re going to be more lucky than you could possibly have imagined, and forget that I said that now go and live your life.” I’d say, “Do what you love.”
Robert Wiblin: Is the reason for that, that it’s more motivating, and that you’ll likely throw yourself into it in a bigger way? Or perhaps that do people have good instincts about what interests them and does that line up with impact?
Cass Sunstein: Well I think if you try to have a social impact, you don’t have anything yet that you want to have a social impact with. If you care, let’s say about highway safety, then if you want to have a social impact with respect to highway safety, the first thing is to learn everything there is to know about highway safety. I guess the second thing is to say something that’s relevant and helpful to people about highway safety or to do something with respect to highway safety.
Robert Wiblin: Do you think once you’re an expert and in the middle of your career, that thinking about what steps can I take to have a social impact now to fix this problem, is this more sensible than when you’re just a kid?
Cass Sunstein: I may have an idiosyncratic answer to this, because I’ve never thought, “How do I have a social impact?” Not once has that occurred to me. I have thought, I guess, “How do I make my writing clear.” And when I was offered a position in the government, I was aware that if I didn’t screw up, I could have an impact. And that was not uninteresting, that was thrilling. If someone asked me to help with something that has a social impact, I’m completely honored. But to think, “How can I have a social impact?” It feels a little too abstract, right? If it’s “How can I help reduce highway deaths?” Or “What can I do about the opioid crisis?” That’s… The opioid crisis by the way, unlike the replication crisis, that’s a crisis. People are dying. I think I would focus on the issue that concerns you. And to have a social impact, Hitler had a massive social impact and he was… [inaudible 01:37:21].
Robert Wiblin: I should have said positive, maybe. Yeah. How have your years in government affected your academic work, if at all?
Cass Sunstein: Thank you for asking. I think by anchoring me in practical things more. And in a way that’s not ideal. So I think academics often do best if they let their minds go to all sorts of places. I had a good friend in government who told me after two years, this is a very good academic and a very good public servant who said, “Get out.” And he said, “Government ruined me. I can’t think academically anymore, and it’s going to ruin you, get out.” I stayed a little while longer than he wanted, but I never forgot his words.
Robert Wiblin: That’s interesting. Do you think that the mindset’s just too different?
Cass Sunstein: Yes, so there are many academics… This is I think a pretty interesting subject. Many academics I saw go into the Obama government, who were terrible at it, because they were really good at coming up with creative ideas, but the idea either couldn’t or shouldn’t be implemented, and they weren’t good at doing the solid work of turning a good policy into reality. I tried early on to have my ears big and my mouth small so that I would just learn from people who knew how to do government. The people in government often whom I greatly admired, they wouldn’t be good academics. They’re not writers, and they’re not ‘idea’ people. They’re amazing at figuring out how to pull levers to do something which actually is helpful. It’s a really different skill set.
Robert Wiblin: Well, I really appreciate you finding a full hour. Unfortunately, we’re out of time, but you’re such a prolific author, there’s probably a good chance we’ll have another book to discuss with you maybe in a couple years’ time.
Cass Sunstein: Thank you. I really enjoyed it. Fabulous questions.
Robert Wiblin: Yeah. My guest today has been Cass Sunstein. Thanks for coming on the 80,000 Hours Podcast, Cass.
Cass Sunstein: Thank you.
Robert Wiblin: The blog post for this episode will link to the papers and books that came up. The show notes will also take you to our annual review and information on how you can support 80,000 Hours if you think our work is having a lot of impact.
The 80,000 Hours Podcast is produced by Keiran Harris.
Thanks for joining, talk to you in a week or two.