[inaudible conversations] good afternoon, everyone, good afternoon. My neighboring is michael, resident fellow of American Enterprise institute and its my pleasure today to welcome to aei profess or Cass Sunstein. He was the administrator of the Office Information of regulatory affairs, commonly known as oir from obama administration. Hes written a dozen books and herest hes here to talk about his latest books which is called republic. I think that a reader would not have pronounced the word that way 20 years ago but its an important and interesting book so please welcome Cass Sunstein to aei. Okay. Well, its a thrill to be out here and you all do amazing work and ive learned so much from whats produced here on legal and regulatory matters and i really thank you on behalf of zillions of people. I was blessed or cursed for the last several years to live in the waldorf towers, which is pretty heavy stuff from a boy from massachusetts and i lived there because my wife embassador of united nations. That was fine except one of them had become quite a good friend after several months in the sense that i knew him and talked to him every day and it was a little awkward that he was calling me by a name other than my own and i said to him after a few months, you know, its Cass Sunstein, you can call me cass or mr. Sunstein, he looked at me and said thats unbelievable. Thats amazing. You look exactly like mr. Power. And that was intriguing to me in the sense that he was not a rational person, he was updating basically his beliefs based on the new information and he believed that it was more likely that there were two people who looked exactly the same Walking Around his building than that the embassadors husband had a different name from her. And given his prior beliefs that was not irrational. It turned out to be wrong but it was not irrational. Okay. Here is the United States of america in many respects today where people are asserting their correctness and the falsehoods emanating from the other side. This has implications not only for our capacity to handle problems that cut across peoples ideology but also affects how people think about one another so much so that theres a phenomenon partyism that isnt as ugly as racism which in some respects is larger, the number of people who would be unhappy if their child married someone of the political opposing party is higher than people who would be unhappy if their child married somebody of a different race. Amazed by this finding i a few years tried to find out whether people would be more unhappy if their child married someone of a Different Political Party than if their child married someone of the same sex. We are not there yet, but we are getting there which suggests that the intensity of partyism is growing rapidly and as of 2017 its immensely higher than it was a couple of days a couple of years ago, probably a couple of days ago. A couple of decades ago when people didnt care if their child married someone of a Different Political Party and this is showing up in other measures of animus so people discriminate an employment decision against people whose Political Party is different from their own. Okay, here are some issues where partyism is causing problems. Whatever you think should be done about infrastructure chances are good, something should be done about infrastructure and the intensity of disagreement between republican and democrats is actually relatively modest on that issue and yet Nothing Happened yet, aumf is talk for use of military force, to get an authorization for the use of military force in areas where the level of consensus across democrats and republicans is pretty high is really difficult. With respect to immigration and judges, we could have much more agreement than we are observing the Nuclear Option taken by the democrats first and now the republicans is testimony to the difficulty of moving beyond Party Affiliation as a way of producing generally agreeable results. In washington, my job in washington i notice in some meetings with some people of a Political Party that i will not name a very sharp disagreement between their privately expressed moderation and their publicly expressed immoderation and i should just say i observed that among between both Political Parties with a kind of poynant plea if it would not cost us our jobs. Social media on capacity to retain their jobs if they have moderation. That is whats very visible to them. Okay, im going tell you about three empirical studies which are not involving the internet precisely but replicating, i think, an experimental setting what is happening every hour of every day on twitter and facebook. The first experiment comes from colorado and the idea was to get people from bolder, which is left of center together to discuss Climate Change, affirmative action and samesex unions and to get people from Colorado Springs which is right of center to do exactly the same thing and we did a reality check to make sure that our Colorado Springs participants were right of center and boulder people were right of center. What we did amazing they were. What we did record anonymously and discuss after they talk today one another. What interested me as one of the three authors of the project, was only one question. How would their views shift in their anonymous predeliberation statements to their anonymous postdeliberation statements, what would those discussions do to how they thought about the three issues when they were recording their judgments in a way that no one would ever see except the except experimenters. People got more unified before they started to talk, im kind of worried about Climate Change, maybe we should have an international agreement, not after the discussion. There were people in boulder who thought affirmative action has some problems. I think its a form of race discrimination. There are some diversity in boulder springs. After they talked different the diversity was crushed. In their private anonymous statements, the second thick that happened was they got more confident. The people in Colorado Springs, some on samesex union werent sure, they didnt like them so much, after they talked they were sure. The third thing that happened and most disturbing and illuminating is they got most extreme. The people in Colorado Springs were to the right, the people involved in boulder were to the left here but they were like here, this range after our little experiment they were like here. They were operating in different political universes in their privateanonymous views. Now, what i just described is a very artificial experiment, thats xa social media constructs, a capacity of experimental groups. Why did that happen . Identify actually seen the tapes of the discussion. Here is what you would see in realtime. First in Colorado Springs, the place where they started kind of skeptical about a Climate Change treaty and in the end very, very skeptical, the number of arguments in their discussion that supported the International Treaty were few. Thats not amazing, thats a statist call inve itability, the number of arguments that oppose an international were numerous, also inevitability. And in boulder exactly the same thing happened in the mirror. That is on affirmative action, some people didnt like it very much. Most people liked it plenty. The arguments that supported it crushed the arguments that undermined it and people, oh, gosh, those are the arguments are supportive of that, that produced confidence, unity, extremism. The second thing is slightly more subtle and you can see this in realtime and i think each of us can see this in our daily lives whether it involves what products to buy, what people to like, what political view to hold on an issue and here is the phone onlyna. Most people who dont have a ton of information, tend toward the middle. They become tentative because of humility. They become confident and more extreme. Confidence is increased by crobbation and you can see that in both cities and corroboration makes for more intensity of commitment to the view and that helps for the finding, the third involves reputation, people dont want a group of people who tend to say, affirmative action is great, they dont want to look like racists or idiots in front of one another and they end up saying like affirmative action plenty even if they privately didnt ten minutes before and in their private anonymous statement of view it would be very awkward to say what they said publicly, they dont actually believe, so their private view lines up with their public view. Here is the second of the three studies and this does not involve professors or think tank people trying to construct something. This involves the world and benefits the study from a s srendipidous and the competition is reagan, bush appointees, obama, clinton appointees, obama, obama, reagan appointees and bush, bush, clinton appointees, meaning three are appointees, three appointees, two d and two rs and thats it. Thats all that can be possible. Okay. Here is the headline finding. While the Political Party of the appointing president is a pretty good predictor of how a judge is going to vote in a contested case, not fantastic but pretty good. The Political Party of the two other judges, of the president who appoints two other judges in the panel is at least as good and often a better predictor of how the judge is going to vote. Got it . You want to know how judge x is going to vote, dont ask whether obama or trump appointed judge x, do ask whether obama or trump appointed the judge with whom x is sitting. Here is the most dramatic finding, while in the aggregate data theres a 13 percentage point difference in the likelihood of a liberal vote between r and a d appointee, roughly 13 Percentage Points which is concerning but not catastrophic from the point of view of the rule of law. The likelihood of a liberal vote on a ddd panel is frequently 30 to 40 higher than the likelihood of a liberal vote from a judge on an rrr panel. That is to say a d judge on a dd panel shows phenomenally liberal voting patterns and r judge showing conservative voting patterns where phenomenal shows how they vote in aggregate. What makes the startling finding in my finding that the d in ddr panel, if youre with me, the ds have the votes, if it involves whether the epa greenhouse regulation is lawful, they can do what they want, they have the vote. Why is it that the dds show more moderate patterns than the dd voting panels. I think its about information at least in part. What are they hearing . A d on a dd panel, thats boulder and what makes statistics not expected is that what we are observing is legal professionals, judges answering not a political question but a legal question and even so the selection to random draw into Something Like an information cocoon produces more systemically extreme results. Now the data i told you understates the magnitude of the phenomena. We just collected votes, by we me and a team of law lawsuit who is are not in mental institutions now im very relieved because counting many thousands of votes, thats not fun. They did it nonetheless. What i didnt ask them to do because its much harder is to explore the relative extremism of the opinion rather than just the up or town vote. So we are asking did a woman when in a sex discrimination case, did epa regulation get upheld or not, thats what we are asking. We are not asking whats the reasoning, theres every reason to think on a ddd panel, whoa and so to an r and r panel. What was the professor technical for omg . Okay. [laughter] here is one other story of updating, you heard the waldorf story and the other story comes from star wars and i did a little book on star wars so i researched star wars and there was a debate between forgive me, did you think you heard something about star wars, this seems surprising or terrible, one little star wars story, debate with Great American screen writer and george lucas on killing the main characters on returning of the jedi and lucas says, luke isnt going die, well, kill laya, laya has to die. Its not nice, you dont go around killing people and then kasdan says he will kill yoda. Kasdan gets in his soul, he gets real and gives a speech about art, im saying that the movie has more emotional wage if someone you lost along the way. Hes talking about culture and art and lucas says very quickly, one sentence, i dont like that and i dont believe that. Now, notice the beautiful architecture of the sentence, not liking proceeds and probably helps account for not believing. The waldorf story, my friend, didnt particularly disliked the idea that i had a different name from my wife, he didnt believe it, it seemed that couldnt be true. Here is about the emotional commitment proceeding the belief. Okay. Here is the last study. How goodlookinglook dog do you think you are on a scale 1 to 10. I have good news for you, thats you. [laughter] now, what do you think having heard my news for you. Lets do a second experiment, ask the same question, how good looking do you think you are. I have some good news for you. Thats ewe. [laughter] okay. Here is what the data suggests that people are asymmetrical updaters and some credible outsiders say they they are 8, im a 7 or 8. Ive learned. If they say im a six and outsider says theyre a four, they say, no, thats not true, im a six. People believe the good news, not the bad news. If this is true for many personal things with respect to the risk of having diabetes, insomnia, enfir tellty enfir that sounds like dr. Seuss, i apologize for that. People will rely more with good news than bad news, bad news theyll say thats noise and good news theyll say, okay. If you zap it, then the good news badnews effect disappears. Okay, so what i was interested in and am interested in is how this works for political information. So here is what we did. We got initial study which i will tell you about, this has been replicated with more people. We got basically 300plus americans recruited them and sorted them in three groups based on answers to questions about the environment. Strong Climate Change believers, we Climate Change believers and moderate Climate Change believers, creative names. The data we got is not surprising in terms of their antidepressanted anticipated warming. The weak 3. 6. The only thing thats interesting there that the weak believers are not at 1 or 0, they are 3. 6. Probably roughly with americas bottom is. Here is what we did and how we try today figure out how people ree react today information. In the good news condition, youre not going to have a mouse in your house, youre better looking than you think, we told people, actually scientists have come in with terrific news likely increases 1 to 5degrees. Stuff like that people are seeing on wall street or facebook all of the time. We gave the other half bad news, scientists have some tough information for you. Its worst than we thought. 7 to 11degrees. The weak believers in climate are like the people with respect to appearance and getting the good news, their estimate fell by a full 1 degree from 3. 6 to 2. 6 and given that the baseline is low, 3. 6, that falling is very dramatic, they really updated getting the good news, getting the bad news, they were unmoved at all. Zero impact. Now, in terms of social science thats not going to get any prize because it fits with everything youve heard before and just applies it to politics. It is, i think, politically, explosive suggesting, you know, think Affordable Care act, immigration, terrorism, the effect of the minimum wage, theres a bottom who will treat good news as very informative, gun control and bad news as who paid them our top showed exactly the opposite pattern. These are the Climate Change believers who arent panicking but the most worried top third of america, they were far more moved by the bad news than the good news. Getting the bad news their average estimate jumped by 2degrees, getting the good news fell by less than half of that which suggests that systemically people who are really scared of Climate Change will be jumping in terms of their level of fear when they get new scientific information suggesting things are bad and theyll be dropping just a little bit. When they get good news suggesting the problem isnt as large, okay, what youve just heard, i think, is basically a really simplified version of what happens on the social media every day in realtime where people are getting both versions of our interventions and where a bunch of people are reacting as asymmetrically. Whats the inspiration of the study, we dont know, one possibility is, i dont like that and i dont believe that. Motivated reasoning. The waldrof tale. Given my prior conviction its going to get really hot. I hear science suggesting its not. There must be identical twins Walking Around. That cant be true. Exxon must have paid that. Theres others that say thats Environmental Group climate, i dont believe that where the good news is more credible to me and thats not a motivated reasoning story, the story given in your prior convictions what youve leashed from the new information. Okay. Almost done. Facebook, 2016, our success is on getting the people the stories that matter to them the most. This is direct quote. You could look through thousands of stories every day and choose the ten that were most important to you, what would they be, the answer should be your news feed, it is subjective personal and unique and defines the spirit of what we hope to achieve. Really . What they are speaking is architecture of control in which algorithms or individual choices are creating ddd and rr panelses. Jane jacobs, the hero of my little book wrote a gr