Cass R. Sunstein
[Applause] Well, thank you for being here. Thank you for that very kind introduction. I’m seeing some new faces and some friends, and it’s an honour and a thrill to be here. So, the actual title of these remarks is ‘Why Societies go Whoosh.’ This is a technical term, w-h-o-o-s-h, and that question has three sub-questions. One is, why does social change happen? Why is it so hard to anticipate? And why does it seem to come out of nowhere? Those are the three questions on which I’m going to try to make progress. My hope is that the analysis is going to tell us something about Brexit, is that the word? About civil rights movements in general, about the rise of fascism in the 1930s, about contemporary populism, and about movements barely on the horizon, emphasising the unanticipated nature of many of the seismic changes that the world has experienced over, let’s say, the last 200 years. I’m going to fix on just one, and that is #MeToo, and use it as a case study, which can be generalised to lots of things, but that also has distinctive features. To vindicate the premise of my mystery, which is, “Why are these things so unanticipated?” Lenin was stunned by the success and speed of the Russian Revolution.
Now, let’s pause over that. Lenin, the Architect, if there was any, of the Russian Revolution, was stunned that it worked and that it worked so fast. Tocqueville reported that no-one foresaw the French Revolution. We’re talking about arguably, the greatest Political Sociologist ever, who said no-one understood that this would happen. More recently, the Iranian Revolution of 1979 was unanticipated. There we go, that boy had no idea what was going to happen, and the idea that the Iranian Revolution, which transformed a significant part of the world, was unforeseen, even by specialists, that’s a puzzle. More recently still, the Arab Spring was not expected by many of the best Analysts in the United States. I know that partly, ‘cause I was working in the US Government at the time, and I know that our great friends in the United Kingdom did not anticipate it either, and this is baffling, because the specialists in the United Kingdom and the United States, they’re really good at their job.
Now, these are cases of unexpected revolutionary activity. It’s also the case that social movements, large ones and small ones, often were not foreseen, even by those who spent a lot of time planning and plotting them. It’s observed often that movements tend to spread rapidly across countries, and from one part of one country to another. The reasons here are puzzling. Specialists often refer to two things: demonstration effects and contagion effects. The idea of a demonstration effect is comprehensible, suggests it can happen. It happened in Tunisia, maybe it can happen in Egypt, but it’s a little bit like a word rather than a framework. Demonstration effects, I think, are more comforting sounding for those in source of an explana – in search of an explanation than they’re really helpful. There’s another word that’s a little better, contagion, which refers to something that maybe goes viral, but here we have a metaphor, I think, rather than a framework. I’m going to try and give a framework, and it’s going to have a certain level of abstraction. It just has four moving parts, and I’m going to briefly tell you what they are right now, and then I’m going to try to explain them in a little more detail.
The first is preference falsification, and the idea here is that every person on the planet, with the exception of two, there are two in – who don’t, but everyone, aside from those two, actually, the truth is everyone, has inside their heads a belief or a desire that no-one except for that person knows. It may be a desire that involves injustice or wrongdoing, it may be a desire that involves morality or politics, but no-one except you knows about that. Self-silencing is often the reason that no-one but you knows about it. Sometimes, misdescription by speakers of what they feel and believe is the source. Not quiet, but falsehood. That’s my first moving part.
The second is diverse thresholds, and here the idea is that, for some of us, if we see something unjust or troublesome, we are there instantaneously. Our threshold for thought or action is really low. Others have very high thresholds. In fact, such high thresholds that unless basically, everyone they know is saying or doing the relevant thing, they won’t do the relevant thing. Most people are in-between. They’re not like zero threshold people, and they’re not like near infinite threshold people. They have to be roused. My third moving part, aside from thresholds and preference falsification, is interdependencies, which signals just the fact that most of what we say or do, at least with respect to morality or politics, is reactive to what other people are saying or doing. If one person is saying or doing something, we might think, “Crazy person.” If a thousand are, we might think, “Why haven’t I joined them yet?” If you put together those three: preference falsification, diverse thresholds and interdependencies, I think we’re on the road toward understanding the slide to my right and to my left, how that could’ve happened, but we have to add one more, the fourth moving part, and it’s group polarisation. Group polarisation means, and Brexit, I think, is a case study, when likeminded people get together, they often end up, after talking to one another, more confident, more unified and more extreme, and that can happen to two different sides in a sharply polari – on a sharply polarising issue. If you put together group polarisation, with preference falsification, diverse thresholds and social interdependencies, now we’re in business at explaining unpredictable social movements and revolutionary activity.
Okay, now I’m going to just elaborate these ideas, put a little more skin on the bones, and then shift to #MeToo. Okay, preference falsification means that we live in a world of pluralistic ignorance, in which we are ignorant of what other people actually think. Under regimes that are oppressive, in one or another respect, preference falsification is everywhere. Because of oppression, it’s really hard to know what other people think, and this helps explain the unexpected nature of the Iranian Revolution. Even in societies which aren’t oppressive, in the sense of authoritarian, human beings often conceal or don’t reveal what they actually want. They might say that the status quo is fine, when they don’t think that, or they might say nothing about the status quo, which means their friends and neighbours might have absolutely no idea what they actually think. Of all the books written on Nazism, I have a favourite, and the reason it’s my favourite is it’s the most human and, bizarrely, the most cheerful. It’s cheerful because it’s written by a Journalist, who went to Germany in the 1950s, to talk to former Nazis, to figure out what they were doing and thinking, and what he found, to his amazement, was he liked them. He didn’t like Nazism, but he liked the former Nazis, they became his friends.
Here’s something one of his friends said when asked, “What did ordinary people think under Hitler? Wasn’t there opposition?” And his friend said this, “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 upon the circumstances, where and when, and to whom, and just how he says it, and then you must still guess why he says what he says.” That’s a profound statement by someone who had lived under Hitler and had no academic anything about the phenomenon of preference falsification. Opposition? How would anybody know? For those who want to protect – predict a social movement, the problem is that social norms can draw a wedge between what people express publicly and what they believe privately. The law matters if freedom of speech is violated and if dissent is punished, but even if there’s no restriction on freedom of speech, if people will be shunned or disliked or compromised in some way, they might not merely silence themselves, they might say that they’re happy with the status quo when they aren’t at all.
I gave you some words from a citizen of Germany under Nazism. Here are some much more recent words, basically the day before yesterday, from a Computer Programmer 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 he might start whispering. He won’t speak loudly. That is too scary. After all that time, even outside Syria, you feel that someone is listening, someone is recording.”
Now, let’s turn to the matter of diverse threshold. Different people will require different levels of social support before they will join a movement or say what they think. Some people are rebels by nature. Every one of us knows one, yes? Some of you look like you might be them. Let’s call them rebels, without being pejorative, zeros, not because the number connotes low, but because they will act even if they have zero social support. They might well be isolated. No-one may join them, in which they may look nuts, or at least radical. Other people, and this is more common, require at least a little social support. They won’t move, unless someone else does, but if someone else does, they’ll be prepared to rebel too. Let’s call them the ones, they need one person.
Others require more than a little, and you know what we’re going to call them? This is fancy math here, the twos. The twos don’t – won’t do a thing, unless they see the zeros and the ones, but if they do, they’re going to move too. The twos, of course, are followed by threes, and tens, and hundreds, and thousands, all the way up to the infinites, defined as people who aren’t going to oppose the status quo, no matter what. Now, here’s the kicker. Outside of science fiction, it’s not possible to see people’s thresholds. We may not ourselves know whether we’re zeros, ones, twos or threes or fours, on particular issues. We might be surprised to find we’re twos when we thought we were tens, or that we’re hundreds when we thought we were threes. We might turn out to be really surprised.
The American Revolution, I’m here to tell you, was defined by an interaction between thresholds of various kinds and preference falsification, people who were concerned about independence, shut themselves up until they got the requisite level of social support. And then, whoosh. Thomas Payne put it this way, “Our style and manner of thinking have undergone a revolution, more extraordinary than political revolution of a country. Our style and manner of thinking have undergone a revolution more extraordinary than an earth-shattering political revolution. We see with other ears, we hear with other ea” – we see with – we don’t see with other ears, that would be miraculous. “We see with other eyes, we hear with other ears, and we think with other thoughts than those we formerly used.”
I think Payne didn’t quite get it right, the Revolution did not involve different ears and different eyes, it involved the revelation of things that had been suppressed, until you got the requisite level of social support. Interdependence points to the fact that the behaviour of the ones, the twos and threes depends on who sees whom doing what, and exactly when. To see how the conditions might be ripe for this slide to work, and this is eerily descriptive of a book on the unanticipated Iranian Revolution. Here are the conditions are ripe. The zeros go first, they move. Then the ones see them, they move too, then the twos, then the threes, and so forth. Under those assumptions, a rebellion’s going to occur, given two things: the right distributions of thresholds, the right kind of visibility, and the right kind of interactions. It’s crucial to see the conditions have to be just right. If there are no zeroes, or if no-one sees any zeroes, we’re not going to get that. Nothing’s going to happen. If there are a few ones, the status quo is likely to be safe. If most people are hundreds or thousands, no worries, even if there are some ones, twos, threes and fours.
Okay, preference falsification, thresholds, interdependencies, and here’s my fourth and last piece. I did a study a few years ago of climate change believers, basically worried about climate change. I got them to talk to each other, for a short period, and then to record, anonymously, their views about climate change. They record their views anonymously before they talk, then they record their views anonymously after they’ve talked. Here’s what happened. At first, the group was, on average, concerned about climate change. There was diversity of opinion. After they talked, everyone in the group was terrified about climate change, and they were unanimous. The little group went whoosh. as a result of basically 45 minutes of conversation. At the same time, I created a little group of people not concerned, actually several, not concerned about climate change. Not unconcerned, but concerned – not very concerned. There was diversity in that group. Some thought an international agreement was a – probably a good idea, some weren’t sure, but, basically, on average, they weren’t worried about climate change. As a result of discussions with relatively likeminded people, that group became wholly unconcerned about climate change. It ended up thinking the idea of an international agreement to control greenhouse gas emissions was terrible, and they were unanimous about that in their private expressions of view, guaranteed anonymity.
What we saw, in both groups, was polarisation to a more extreme point, in line with their antecedent tendency, and with respect to issues that I bet you’re thinking about. Exactly that has been witnessed recently in, let’s say, this country, where people associating with likeminded others go to a more extreme, unified, competent, confident position, and so do others. I certainly observe this in my country. The group polarisation phenomenon created artificially in the climate change experiment is created naturally, every minute of every day online via Twitter, which are machineries for associations among likeminded people.
Okay, if we put these factors together, the mystery with which I began, I think, starts to dissolve. We can now see why large movements, including revolutions, may be impossible to predict. We can see why Tocqueville’s claim, that no-one anticipated the French Revolution, is not baffling. It’s a natural product of the best account we can offer of why social movements succeed. First, we don’t know what people’s preferences are. By hypothesis, they can’t be observed. They’re in people’s heads. Second, we can’t know what people’s thresholds are. They too are unobservable. Which people are twos, threes and fours, in the Iranian case, the historical studies show that many participants in the revolution had no idea that they were threes or fours, but once things started going, they sure found out. Third, and this is the most important point of the three, we can’t anticipate social interactions. We can’t say who will say, “We can’t know in advance who will say and do what, and exactly when.” Even if we knew people’s preferences, and could know their thresholds, we can’t know in advance who’s going to interact with whom, the ones and twos and threes and fours. This suggests that even if new technologies make it increasingly possible to identify private preferences, and they do, by Google searches, you can learn what’s people’s in heads – that’s in people’s heads more than ever before, that’s knowable, but even if you can do that you aren’t going to be able to predict social movements. You’ll know something important, if people secretly hate some practice or status quo, the likelihood of a successful movement is higher, but that’s not sufficient. We need to know people’s thresholds, and that’s really hard. Even if we could overcome that obstacle, we need to know the interactions. Who’s seeing whom when? No-one has that kind of prescience. The answers to those questions, unanswerable questions, are going to be determinant of outcomes.
Okay, I’m going to make a modest claim here, with some confidence, and then a more ambitious claim, with tentatively. The modest claim is, just empirically, it’s too hard to find this stuff out, and for the foreseeable future, we’re not going to be able to, even with the transparency that we get through Big Data and online behaviour. The more ambitious claim, ventured tentatively, is, in principle, we can’t ever learn these things, that the nature of social interactions is such that it cannot be predicted in advance, even if we have all the data in the world. Okay, if this is right, we can also see that successful movements and reforms are often explained as inevitable products of history’s arc, or of cultural dispositions, but it’s better to see them as products often, usually, of serendipitous random or small factors, of who did what when, of who heard what when, of whether some kind of butterfly flapped its wings at the right moment.
Now, this claim I’m making about the serendipitous nature of social change is hard to vindicate, and the reason it’s hard to vindicate is history has only run once, so we don’t have, aside from counterfactual history too, so we can’t see, but the claim is, it’s often tempting to think that a practice or a regime or a status quo was bound to fall. That temptation should be resisted. Often, it’s better to say it happened to fall. The same is true if it doesn’t fall. It happened not to fall.
Okay, I want to make two points to complicate this very spare four-part claim about social change. The first part, which is essential, but very different, is signalled by these words from a woman in North Korea, also recent, basically the day before yesterday. “It never occurred to me that I could or would want to do anything about it,” meaning the situation in North Korea. “It was just how things are.” Now, that’s a staggering two sentences. “It never occurred to me that I could or would want to do anything about it, it was just how things are.” She is not describing preference falsification. It’s not the story of Syria, the Computer Programmer. She’s saying that her desires themselves were adaptive to how things are. What she wanted was an artefact of how things are. That’s not about preference falsification, it’s about desires being formed by the status quo, and that’s a phenomenon where people – what people want is often a product of their sense of what’s possible.
That is an extreme case. A less extreme case involves partially adaptive preferences, in which people are often aware that something is wrong or bad, but the voice in their head is quieted so much that it’s not really preference falsification. The second point is the word ‘preferences’ is under-descriptive, or perhaps misleading. We might do better to talk about people’s experiences, that’s what they’re falsifying. Okay, let’s turn, if we would, to Gwyneth Paltrow and Harvey Weinstein, and talk about #MeToo, and what I’m going to do is bring the pieces in contact with something very concrete. With respect to sexual harassment and sexual assault, preference falsification has run rampant. Victims have silenced themselves, including, I’m very confident, many people in this room, more women than men, but some of both, who have not said to many people, or to any people, what actually happened. That form of preference falsification is probably best understood, in some cases as something even worse, which is experience falsification, when a misstatement of what actually happened, or silence about what actually happened, is the driver of a status quo, and also, which makes the status quo highly vulnerable, because once people start to speak, then whoosh.
Second, different women have and had different thresholds for speaking out. Here are the words of Beverly Young Nelson, who accused a Republican Senate candidate, named Roy Moore, of assaulting her in 1977. “I thought I was his only victim. I probably would have taken what he did to my grave, had it not been for the four other women, who were willing to speak out about their experience. Their courage has inspired me to overcome my fear.” She was a five. Maybe she was a four, temporarily she was a five. Social interactions, the third point, are essential to social movements, and in particular, to #MeToo, which has benefitted, as have many social movements, from the great visibility of those who spoke out, and the interactions made possible by social media.
After the Actress Alyssa Milano tweeted, “Say #MeToo if you’ve had an experience,” within 24 hours, 45% of all US Facebook users had friends in their networks who had posted #MeToo. 45% of the many millions did it. Once the ones and twos spoke out, the threes and fours felt safer or more emboldened. Group polarisation is certainly a mover of the success of #MeToo, in which likeminded women have spoken with one another and become more confident, more – and more unified as a result. The important qualification here is that #MeToo must be seen as a case study in what happens to adaptive or partially adaptive references, and I’ll tell you a personal story that just occurred to me today, in the context of preparing these remarks. Hadn’t thought about this for about 20 years.
When feminism was first arising, I was keenly interested in reading about it and writing a bit about it, and I talked to my mother about it, and she was very sceptical. She said, “Why, Cass, why are you working on this? You’re a constitutional law person. You know, what are you talking about? You know, women have lives, men have lives, what’s this about?” and then I started describing to her what feminists, in the United Kingdom and the United States, had been saying, and she paused, and it was very uncharacteristic of my mother to pause. She said, with more feeling than I will be able to muster as I recapitulate what she said, three words that she never said to me before and never after. She said, “God bless you,” and in those words were, I could feel, a set of experiences, which she had, to which she had never given voice, even to herself, until basically that moment. #MeToo has a lot to do with a mechanism of change by which preferences unadapt, or by which partially adaptive preferences are unadapted as social norms shift.
That latter point suggests something extremely important that is related to Brexit, which is that social moments – movements are often not about the revelation of expressed preferences, experiences, beliefs and values. They may be about the transformation of preferences and beliefs and values, not taking something out of a had that had been silenced, but putting something new in there. Any social movement, for good or less, helps to change preferences, values and beliefs. It does not merely elicit pre-existing judgements, it produces fresh ones, and we need the French Revolution as our star performer on this front, and this is a sentence I hope I can leave you with, though it’s not quite my final. Part of the point of the French Revolution and part of the point of #MeToo is to convert a sense of embarrassment and shame into a sense of human dignity.
I conclude with the statement from the Computer Programmer from Syria, which I hope you will hear now in a somewhat different light. “When you meet somebody coming out of Syria for the first time, you start to hear the same sentences, that everything is okay, that Syria is a great country. It’ll take, like, six months, up to one year, to become a normal human being, for people to say what they think, what they feel. Then they might start whispering. They won’t speak loudly, but, eventually, they might.” Thanks [applause].
Thank you so much for your patience and indulgence, and we now have two and a half hours for comments and questions. Sir.
Member
Thank you for your presentation. Cheers. Thank you for the presentation, it was very interesting. I’m just wondering, these four knowledge problems that you identify, why would they not be, sort of – for your nudge, why would they not represent a similar problem? How can a regulator tell, ex ante, with a nudge, whether they’re being paternalistic or simply nudging with these tables?
Cass R. Sunstein
Okay, so, with a nudge, you probably will have to do some work to know if it’s going to have an effect, but if I told all of you that you have a book by default, the likelihood that you would keep it would be higher than if I told you that you had to go upstairs and get the book. In other words, if we just gave you the book by default – I’m sorry to say this isn’t going to happen – we gave you the book by default in the next five minutes and said, “If you don’t want it, you can leave it on the chair,” would you do that? Or, “You have to go upstairs to get it, but it’s going to be free.” In the first, more people are going to end up with the book. Default – the default approach matters. So, we often know that the default is going – we often know that a nudge is going to have an effect.
If people are given a warning about the use of a product, on average, most of the time, their use of the product is going to be different than if they’re not given a warning. So, a default rule and a warning typically work. Whether a nudge is paternalistic isn’t an empirical question, it’s a normative question about, “What’s our definition of paternalism?” and that would not be vulnerable to empirical falsification. So, if you said – if you give people information about the nutritional content of food, that’s not paternalistic, that’s just a claim about what paternalism is, and it wouldn’t be subject to empirical refutation. Though, you could have an understanding of paternalism, in accordance with which giving people information is paternalistic.
Member
Can I ask – oh, sorry, I’m looking for someone else. Just using your example, how can we tell the difference between whether – so, in your example of putting the book on the chair, how do we know whether we’re putting the book on the chair or we’re asking you to go upstairs if we can’t preference falsify, we have these diverse thresholds, etc.? As in, we effectively don’t have that empirical…
Cass R. Sunstein
Okay, so, the great Philosopher John Rawls had a footnote which he never published, which said – which says, “We post a signpost. No deep thinking here, things are bad enough already.” So, that’s my answer. The preference falsification, social interaction, diverse thresholds, group polarisation issue confound the prediction of the success of the French Revolution, may – are relevant to the extreme difficulty of predicting who’s going to be President in 2020, make it not surprising that Brexit and Donald Trump were both surprising. We shouldn’t be surprised by surprises in light of this, but the effect of a default rule is independent of all this. So, if default rule sounds like social science speak, I’ll give you a homely example. Homely not in the sense of ugly, but in the sense of mundane.
If the default setting on a printer is double sided, people are going to use less print – paper than if the default setting on a printer is single sided, even if, in both cases, they can decide on the spot, “This time, I want to go one or the other way.” So, the default rule affects outcomes. We just know that, the data is overwhelming, and this stuff is related to a different kind of problem. So, this is just to say around the corner is the behavioural insights team of the UK, which is fantastic. They’ve contributed enormously to data on the effect of nudges, and we have predictable impacts, we know a lot. With respect to large-scale social movements, the likely outcome is much more challenging, because we just don’t know.
To predict the success of a movie or a song is sometimes described as more art than science. I don’t like that, it’s neither art nor science, it’s – science doesn’t allow us, because we need to know who is going to be buying the new songs by a Taylor Swift wannabe, and whom do they know? And that’s really hard to know in advance. I think, in principle, it’s impossible to know in advance. I’m not sure of that. That’s a different kind of problem than what’s going to happen if every time you buy sunscreen in London, there are big bold letters that say, “Use it only under conditions X, Y and Z, and this kind of sunscreen isn’t going to protect you from skin cancer.” That will have an effect. Yeah.
Member
I need a microphone, I think. Thank you. Three questions. First of all, do you have any observations about Psephologists, Pollsters and advertising, and, secondly, do you have any policy recommendations or views about social media?
Cass R. Sunstein
I have nothing to say about Pollsters, but I wish them well. Social media, with respect to these remarks, there’s something straightforward, isn’t there, which is that social media often creates echo chambers, which are vehicles for polarisation. So, if people are sorting themselves, either into an information cocoon, they may end up unified, confident and extreme in a way that sharply splits them from other members of the human race, who’ve sorted themselves into a different information cocoon, and that we observe. So, social media, and Facebook, in particular, up until, I’m not sure what year, with the Facebook case, 2016, was very enthusiastic about the notion you can design your own information diet, and Facebook had a post which said, basically, that’s what we’re providing you with our newsfeed. Facebook’s been much more circumspect about that since, on the ground that if people can choose their own information diet, we might get, let’s call it, unwarranted extremism, where it’s because people are just talking to likeminded others, and that is a challenge for self-government, created by something which, in most respects, is wonderful, which is people’s ability to talk with each other.
Member
And what about fake news in perspective of social media?
Cass R. Sunstein
Yeah, fake news is a problem of exactly this kind, where maybe if we can get, say, just a little – few more words on fake news, if you go outside and encounter someone you kind of know, or even someone you don’t kind of know, and ask, “Do you know what the weather report is for tomorrow?” you don’t expect them to lie to you. So, you think – or if you say, “How do I get from here to some place near?” you don’t expect them to lead you to something – a dangerous area. And so, when we hear from other people, we are credulous in a way that assumes knowledge and good faith, on the part of people who talk to us, and that’s good, that we’re credulous in that way, and it’s, on average, useful, but it creates an opportunity for terrible things, where we can be told by someone, who’s the equivalent of the person we kind of know or don’t know on the street, that up is down, or that two plus two equals five, and we think, “Oh.” Not because we’re, you know, not thinking straight, but because we’re generalising from ordinary experience to a context in which it’s not like that. And I confess that, you know, my main subject is law, and when I read media on law, I have a reality check. I know the primary materials, but there are other areas that I know very weakly, where I read the media materials and I agree with – I accept them, where, maybe, there’s an effort to skew or to press self-interest for something, and it’s very hard to get yourself in the mindset of thinking, “Oh, there may be a political skew, or self-interest, or something, on the part of a member of the human species whom I have no particular reason to distrust,” and the seriousness of the fake news problem, maybe we can locate it there, where all of us are treating other speakers as if they’re, you know, in good faith, basically, on average, but in some cases, it’s not so, and Russia certainly has exploited this with agility, in the electoral context.
Member
Thank you.
Cass R. Sunstein
Yeah.
Member
Hi. You’ve explained really eloquently the – how these movements magnify, become highly visible, and using the #MeToo as an exemplar, feminism and the F word have become sayable again, they’re hugely popular, they’re high visible. Ivanka Trump’s a feminist, Theresa May’s a feminist, Sadiq Khan’s a feminist, right? How do you then move from this hyper-visibility to changing the defaults, so that it doesn’t stop at the visibility, but it actually engenders true social change?
Cass R. Sunstein
That’s a great question. So, we probably want to get very concrete, I guess, so if the problem is the lack of representation of women in high positions, then there are things that can be done, that can help, such as self-consciously ensuring that the pool of people considered includes women, and that often can significantly increase the number of women who end up chosen. Here’s a very cool demonstration of that. If decision makers are asked to evaluate candidates, kind of, one-off, under certain conditions, men will do better, on the part of, let’s say, unconsciously biased Judges. But if they’re asked, “Compare them. Compare them, which would you choose between them?” then the women do much better, because the judges see, “Oh, I would choose the man and then the woman – and over the woman, unless I’m looking at them both at the same time.” But looking at the same time, gender becomes less salient, I’m looking at credentials, and the women are better. So, that joint evaluation practice has been found to reduce discrimination, but pool expansion is a way to do it.
If we’re talking about sexual harassment, then to have inculcation of a norm by people who are in positions of authority, men or women, can do the job. If we’re talking about sexual violence, then a) that, and b) criminal enforcement. The idea that – I mean, in some circles, the idea that sexual violence would be considered acceptable is – that’s almost unthinkable, to me, that sexual violence would be considered acceptable, though it is, in many places of the world, including places in the world that are less exotic than we’d like to think. One of the evident effects of #MeToo is to make sexual violence not only a crime, but unthinkable, in the sense of so abhorrent that it’s off limits, and there are areas where that’s happening, either because of observation of what the victims of sexual violence have experienced, or because of authority from people, women or men in positions of power, who say that’s not tolerable. There’s a zero-tolerance policy. So, here’s a way to put it that – okay, so, President Obama, inside Government, and he said it outside of Government, so I’m not violating any confidences, has a great phrase. “Better is good,” and he uses that in response to concerns that some problems are so large that the steps we take to address them seem modest. His answer is, “Better is good,” and on, you know, countless dimensions, #MeToo and its predecessors have produced better. Is better good enough? No, but it’s good.
I’m conscious that we have four hands, and they are all, unless we have something very complicated going on, male. Sir, yeah.
Member
Thanks very much. I have a question about the third point, interconnectedness. Are we more interconnected now than ever before? And I think the Arab Spring and the recent social change has been cited as a result of new technologies. If we are more interconnected, then does that make social change more likely, or do new technologies actually make networks shallower and less robust, because revolution and change is something that happens on the internet, and we don’t have skin in the game?
Cass R. Sunstein
Okay, it’s a great question. The interdependency point is that whether people will move, depends on whether they see other people moving, and, other things being equal, the increased visibility of others will increase the likelihood that people will move. So, we should see, other things being equal, more social movements brewing as a result of social media and the internet than we otherwise would. That should be. Now it may be that there are countervailing factors, so that, just as you see others moving, you see large numbers of people challenging the move, which could counteract.
There’s a little bit of data here, which bears directly on your great question, and it comes from China. There’s a Political Scientist named Gary King, who studied the behaviour of the Chinese Government under – with effect to the social media – with effect to social media, with considerable co-operation, interestingly from the Chinese Government, and his basic finding is, if people on Chinese social media, citizens of China, say negative things about the Government, it’s not censored. If you say the Government is doing bad things, the leaders are making mistakes, even if it has the degree of commitment and energy in it, evidently, it’s allowed. And the response might be to add material to the stream that is very positive, so people see, and this is the countervailing, see positive effects, but it’s, kind of, there’s a degree of allowance.
By contrast, if people on social media say, “We’re having a meeting at this specific place on Thursday night. We’ve got 400 people coming, please join us,” that will be taken down, and that’s smart, in terms of an understanding of how social movements take off, because the meeting and the reference to large numbers of people already attending, loosens the norms that falsify preferences, suggest the – that the thresholds are being met, and try to get out the threes and fours and fives, and are, kind of, real virtual evidence of social interactions, suggesting we’re going to go whoosh a little bit. And so, social media monitoring, on the part of private or public sectors, can be enlisted to cease this stuff. And it’s intriguing to think how a good Government and, you know – and you might think that, in any many respects, the Chinese Government is a good Government, or you might not, or any good Government, should think about these things in the context of protest movements that challenge it.
One bit of data that’s highly relevant to your question, and quite recent, is that we know if you want to produce more sustainable behaviour, more environmentally friendly behaviour, if you draw attention not to people already doing the thing, which might be a lie, and shouldn’t lie, but to the emerging norm in favour of doing the thing, you know, where fill the thing – I’m being very articulate with that word – with environmentally good behaviour, one or another kind, you say, “Increasingly, people are doing this.” That can be effective in creating a self-fulfilling prophecy, and it’s connected with overcoming the threshold problem and the social interactions problem. Sir.
Member
Cass, do you want to take a brief look at three to finish, and then…?
Cass R. Sunstein
Okay, yes, I do want to keep – take a group of three. Oh, that was a rhetorical question.
Member
I found your concepts around social support and polarisation very interesting, and it, sort of, gets into some of the dynamics, perhaps, of what’s going on with Brexit. You did mention at the very beginning that you were going to refer to populism, and it’s intrigued me that right across Europe, in a whole range of very different countries: Greece, Spain, Italy, Hungary, Sweden, everywhere, Britain with Brexit, America with Trump, why is it all happening more or less the same time? Why have we got a wave of it?
Member
Hi, [inaudible – 52:54] on behavioural economics. A simple question, why it took so long for behavioural economics subject to be known by everyone? I know it’s been around in science for a long time, but it’s seemed recently, after your book, it became recognised. I studied in 2003 and n- one knew it.
Ed Dowding
Hello, Ed Dowding from Represent. How would you improve or modernise the political process to reduce the friction, improve the legibility of the state, so that the will of the people would be more evident, and we don’t need this kind of revolution? Or is this kind of revolution necessary to create durable change?
Cass R. Sunstein
And these are great questions and I’ll start with behavioural economics. So, the reason behavioural economics, on some accounts, slowly became influential or used is the extreme power, and by that I mean both analytic power and institutional power, of rational choice economics, which was so dominant for so many years, and for that to fall, or at least be intruded on, takes both time and effort. And that’s okay, because the rational choice economics, a lot of Nobel Prizes, extremely well deserved, and amazing people, and much of it is true, it’s like any intellectual movement, that it has to prove itself, and here there was a particular superhero. Gary Becker is his name. I knew him. He was a University of Chicago Professor, who was very sceptical of anything that deviated from rational choice economics, really empirical and openminded in that sense. It just took a while to prove itself, and even now, you know, what the domain is, and so forth, people are trying to get right. I think that’s it.
Third question is a very powerful question, which is to make the political process more legible, so it reflects the will of the people. That’s very important. Having worked in the Government, I think as important is how can Government work to help people have good lives? And that might mean, you know, hands off, it might mean markets. It might mean requiring that automo – all automobiles have cameras, so you can see behind. It might mean doing things to combat obesity, which can take years off people’s lives. So, I would want to think, and here we’re in the land of Bentham, Jeremy Bentham, who’s both a Democrat in the sense that you are, and also, a focus on what are the human consequences of a Government – what Government is doing. And in a time of contestation, legibility is really important. I think it has, I’m not sure how to put this, an older sibling, which is called human welfare, and that I would put, you know, in the biggest font you can have, and bold letters, human welfare. So, much of politics is expressive. It’s about how can we express our values best? And I worry that expressive politics neglects the human consequences in favour of, like, a punch or a hug, and legibility is often a good partner of human welfare, but it needs that partner. Legibility can feed into, sometimes, terrible noise. So, it’s important, but okay, there’s that.
On populism, that is also a fantastic question, and I think two answers are possible, and I’m not sure which one is right. One answer is the idea that there’s been a contagion effect, or something like that. It’s an optical illusion. I don’t think this is correct, but it might be correct, where what’s happening in Greece and Spain and Italy and the United States and Britain, they’re all different things. There’s some term, populism, that we use to unify them, but that’s, in a way, an artefact of language. There’s something diff – very different happening in them. I think that’s probably wrong. It’s probably more as if there’s a set of concerns and fears. A lot of them have to do with displacement and humiliation, I think, on the part of ordinary people, who are in all of those countries, who have felt emboldened, either by what’s happened in other countries, or what’s started to happen there. So, whether you love it or not, it’s not so different from everything else we’ve been discussing. It’s not so different from #MeToo, and for those who are enthusiastic about the populist movement, so-called, to see this as an opportunity to make for a change that’s overdue, the understanding of the mechanism’s helpful, and those who are very worried about them, to see the assault on dignity or self-respect, that the populist movements are responsive to, is a good first step [applause].
