Presidential Polling and Constitutional Law – What Happens Next - 12.13.2020

Larry Bernstein: Welcome to What Happens Next. My name is Larry Bernstein. What Happens Next offers listeners an in- depth analysis of the most pressing issues of the day. Our experts are given only six minutes to present. This is followed by a question and answer period for deeper engagement. I think you will find this discussion to be informative and provocative.

Larry Bernstein: The program is moderated to be politically neutral. Our speakers will give their opinions, and then we encourage you to make up your own mind.

Larry Bernstein: This week's topics include presidential polling, as well as constitutional law.

Larry Bernstein: Our lead off speaker today is Eric Kaufmann who is a political scientist at the University of London's Birkbeck College. Eric is the author of Whiteshift: Populism, Immigration, and the Future of White Majorities. Today, Eric will be discussing the shy Trump voter and how polling underestimated Trump's support.

Larry Bernstein: Our second speaker is Doug Massey from Princeton University where he is a professor of sociology. Doug will discuss the problems inherent in the non-response challenge to pollsters.

Larry Bernstein: What are the implications if only one in a hundred people respond to pollsters' phone calls? Are individuals who agree to discuss their political choices with a stranger on the phone representative of the population? Or is there something unusual about people who answer polls and surveys in this matter? If so, does that mean that polling by phone is fundamentally flawed?

Larry Bernstein: I chose the topic of presidential polling methodology because this is the second straight election that pollsters appeared to have materially missed the mark, which brings us to our final speaker in this segment, Joseph Campbell, who is a professor of Communications at American University.

Larry Bernstein: Joseph has written a book entitled Lost in a Gallup: Polling Failure in US Presidential Elections, which highlights that presidential polling has performed poorly not just since 2016, but since 1936. What is even worse is that the nature of the polling errors is different in each election cycle. This begs the question, why do we put stock in political polling at all?

Larry Bernstein: You may recall that what happens next to the discussion on polling before the election. And in that round table, there was a strong belief that the 2020 election, there would be a blue wave putting Biden

1 in the oval office Schumer as majority leader in the Senate and Pelosi as speaker with an expanded Democratic House majority. In addition, that we would know a lot early on election night. Well,

Larry Bernstein: I went to sleep in the wee hours thinking that Trump had won. And probably the biggest shocker, the Republicans almost flipped the house which was an outcome that no pollster had seriously contemplated.

Larry Bernstein: Clearly, there's much to learn about improving the quality and predictive value of political polling. My co-host, Rick Banks, will fully introduce our second segment on constitutional law later in the session.

Larry Bernstein: Just to whet your appetite, the speakers in this segment include Larry Kramer, who is a former dean of Stanford Law School, Mark Tushnet, who teaches Constitutional Law at Harvard and Ganesh Sitaraman who is a professor of Law at Vanderbilt.

Larry Bernstein: Okay. That is our lineup for today. With that, I'm going to hand the call off to our first speaker, Eric Kaufmann.

Eric Kaufmann: Hello. I want to address the polling miss, first of all, and why it took place. The first thing to note, of course, is the size of the miss this time. According to 538 on the popular vote, Biden was supposed to take it by eight and a half points. It turned out to be only about 4.4 points. That's a very large polling error.

Eric Kaufmann: More importantly, locally errors were as large as eight points in states like Ohio and Wisconsin, five or six points in Michigan, Florida. These are crucial states for those who are predicting the result.

Eric Kaufmann: But what I really want to focus on is the systematic direction of the error, not the error itself. It's a very difficult thing to predict an election to within one or two points. But the fact that the error went against Trump in both 2016 and 2020, I live in Britain and something similar happened with regard to the Brexit vote here that people assume that the Remain vote would take it.

Eric Kaufmann: Now, this is not the conspiracy that some on the right allege; however, it is also not the case that progressive cultural norms and narratives are also not culpable. And I want to make the point that in fact, they have a role to play in explaining some of this miss.

Eric Kaufmann: I want to now move on to the question of who the pollsters got wrong. And what really jumps out this time is the big skew with white college educated voters. White working class voters were predicted

2 wrong by the polls, but it was the white college educated voters that the ABC poll, for example, got off by 28 points, Pew between 18 and 23 points. These are massive misses.

Eric Kaufmann: And this is partly because this goes against the narrative that white college educated voters are smarter, too smart to vote for a buffoon like Trump, put off by his racism and his crudity. Could it be that this narrative was playing a role?

Eric Kaufmann: Second point is that the focus in a lot of the polls is very much on demographics and social factors, education, race, for example. But actually, most of the variation in support for Trump is within group, not between group.

Eric Kaufmann: So, for example, if you actually run a model of Trump voting, you'll find that education might predict 1% of variation. Attitudes to illegal immigration will predict 30% of the variation.

Eric Kaufmann: Anything that gets closer to those psychological attitudes, which some social psychologists term authoritarianism that is going to get you a lot closer to the answer.

Eric Kaufmann: So with this variation within social category that pollsters really need to start to try and control for and are missing. David Shore, for example, commented that in the General Social Survey, the number of people who say they don't trust other citizens is 20 points higher than a lot of the polls.

Eric Kaufmann: So the polls, even though they are correcting properly for education, are systematically missing these low trust voters and that correlates increasingly with voting for Trump. So you have a psychological and an attitudinal factor that's not picked up by the weighting.

Eric Kaufmann: The next point I want to make, however, is this question of social norms and political correctness which I believe plays an important role. A number of experiments have shown, for example in 2017, Lucian Gideon Conway showed that when you primed respondents with a passage about the importance of political correctness, support for Trump increased substantially.

Eric Kaufmann: Ashley Jardina in 2019 showed similarly that describing Trump's policies as racist tends to increase support for them amongst certain voters in the population.

Eric Kaufmann: It's interesting in this regard that a couple of Republican pollsters, Frank Luntz reports that, first of all, Republican voters were twice as likely as the Democrat voters to say they wouldn't share their views with their friends by 19 to nine points. Secondly, that many express resentment that pollsters would

3 misrepresent their views and that they didn't trust pollsters who often come with an academic or a corporate label.

Eric Kaufmann: And indeed, trust in universities has seen a substantial decline as has trusted media amongst Republican voters since 2015. From 31 to 14% in the media, trust in universities has dropped also by over 30 points between 2015 and 2017.

Eric Kaufmann: So we've got an increasing, what I would argue is a reactance problem, people reacting against what they see as a hostile or unfriendly pollster who doesn't represent them.

Eric Kaufmann: Now, we can argue about the merits of the case, but it's interesting that Josh Kraushaar cited a couple of Republican strategists who in interviewing suburban voters, heard the respondents complained about quote the excesses of so-called cancel culture, pointing to an environment where employees worried they can be fired or punished for heterodox views.

Eric Kaufmann: This has been backed up by a [inaudible 00:08:24] study, which recently found that 60% of Republicans with a master's or doctoral degrees said they worried about their careers if their views became known to their coworkers. In fact, 88% of Republicans say that political climate these days prevents me from saying things I believe because others may find them as offensive. Only 44% of Democrats do.

Eric Kaufmann: So we have a significant degree than of reticence amongst Republican voters, which I would argue resulting in both a certain degree of shyness and certainly reactance to being contacted by pollsters. And so pollsters are missing this group because of that reactance to political correctness. And we've seen declines in trust in important institutions associated with polling, such as the media and universities.

Eric Kaufmann: What then might be the solution to all this? And I think the solution beyond trying to roll back some of the excesses of political correctness and speech policing and norm violation policing is actually to think about how you might weight a survey by a question, such as a question on trust or a question perhaps on a key attitude that's relatively stable, perhaps a death penalty, something that can get away from simple crude social categories and allow pollsters to weight their surveys more accurately would, in my estimation, help to get around this problem of reactants and non-response bias that's increasingly plaguing polls and that cannot be corrected by simple demographic weighting. Thanks very much.

Larry Bernstein: Thank you, Eric. Our next speaker is Doug Massey. Doug is a professor of Sociology at Princeton. Doug, please go ahead.

Doug Massey:

4

I've been running surveys for most of my 40-year career. And one thing is clear is that surveys and polls are not doing what they used to do in American society. And this has changed for two reasons. First-

Larry Bernstein: Hey Doug, we lost you.

Doug Massey: Can you hear me now?

Larry Bernstein: Doug, please go ahead.

Doug Massey: Well, as I was saying, the reason that polling is not working as well as it used to is for two reasons because society has changed and second, because technology's changed.

Doug Massey: Polling really came of age in the 1950s when the United States was a homogenous middle-class society. It was overwhelmingly white, only 10% black, two-thirds of blacks lived in the South, 2.1% Latino and only less than 1% Asian and 92% of all Americans were native born. It was a period of historically low immigration. And the family composition was overwhelmingly nuclear families, either on the way to becoming a nuclear family or having been a nuclear family. 71% of women were not in the labor force, which meant that there was somebody home most of the time.

Doug Massey: It was a time of rise in median income and declining inequality with a relatively high level of generalized trust and a strong belief in government and faith in government. Most voters were moderate and there was a relatively high and consistent turnout.

Doug Massey: Today, we have a segmented society. It's only 60% non-Hispanic white, 15% foreign born, 18% Hispanic, 6% Asian, 13% black with a very unequal distribution of both income and wealth and a polarized polity that's fomented by social media, internet, cable TV, and the internet.

Doug Massey: This brings us to technology. The fifties was an analog world where communication was by telephone and one landline per household with a low frequency of calls. The media was basically three broadcast TV networks supplemented by a few local stations. Local radio and TV program were independently owned, not chain. Data processing was mainframe computing with mechanical adding machines.

Doug Massey: Now we've moved into a digital world, of course, where communication is cellular and internet with a high frequency of communication, dozens of specialized cable channels, syndicated talk radio, consolidation of TV and radio under a private ownership, chain ownership. Data processing is distributed computing with processing power changing every two years.

5

Doug Massey: According to Moore's law, these developments have reinforced a strong segmentation and defragmentation of society and a fragmented media environment targeting specific demographics and political niches.

Doug Massey: All of these changes make it extremely difficult to do polling in today's world. It's difficult to reach the entire population on a telephone poll. Political and marketing come-ons often masquerade as surveys and make people suspicious. Technology permits screening of phone calls, keeping people from answering. Random digit dialing does not cover cell phones very well. People in households have multiple phones. Homes are often unoccupied. In fact, they're mostly unoccupied. There's a public overload with information and data. Of course, generalized trust has declined and faith in institutions has declined as pointed out in the first presentation.

Doug Massey: Now, we have a populist revolt based on anger, a lot of hatred and resentment. A rising diversity means you need larger samples to generate reliable results. It's difficult to predict turnout because non- respondents tend to be unreliable voters. A large fraction of the voters in 2020 turned out to be new or first-time voters and these people are very difficult to predict.

Doug Massey: So in the end, polls today have dropped normally to single digits pulling in, it used to be in the range of 60, 65%. Good polls would be 80 or 90%. Today, you're lucky to get 10% and they're down to single digits.

Doug Massey: If there's one thing that survey researchers know is that people who don't respond are not a representative sample of the entire population. People tend to not to respond for reasons. The reasons are, as Eric pointed out, not simply social and demographic, but they have to do with ideologies and feelings and emotions. These are very difficult to predict.

Doug Massey: So now, we're down to single digits and single digit polls are not doing the job they used to in any domain of life, political, consumer surveys or whatever we're looking at. Polling is under challenge because of profound changes in American society and the technology that governs it.

Larry Bernstein: Okay, Doug, thank you. Our next speaker is going to be Joseph Campbell. Joseph is professor of Communications at American University and he'll speak about failure in presidential elections. Can you go ahead, Joseph?

Joseph Campbell: Sure. And you can call me Joe, Larry.

Larry Bernstein:

6

Okay.

Joseph Campbell: In the run-up to last month's presidential election, the polls were mostly dismal. Perhaps, the most evocative characterization about polling performance this year was offered by Sean Trende. He's the senior elections analyst for RealClearPolitics, the RealClearPolitics online site.

Joseph Campbell: And a few days after the election, Trende took to Twitter to say, "The polls were a stinking pile of hot garbage, and there's really no two ways about it."

Joseph Campbell: A stinking pile of hot garbage. Now that's probably a bit extreme. It's amusing perhaps, but probably a bit extreme. In any case, the poll did, overall, suggest that a blue wave was in the making. What we got instead was something akin to a blue trickle. Instead of Joe Biden's leaving the Democrats to a smashing national victory, a blue wave, Republican candidates showed unexpected strength in races for the US Senate and the House of Representatives.

Joseph Campbell: Consider, for example, Republican Senator Susan Collins in Maine, she supposedly was a goner. All the polls in Maine were signaling that Collins's opponent, her Democratic opponent, was going to win and perhaps win easily. Collin's trailed in all the polls before the election. Yet, she won by nearly nine percentage points.

Joseph Campbell: The race for president this year was much closer than many polls had estimated. Take a few examples. The final Quinnipiac University Poll estimated Biden's lead at 11 percentage points. The final Economists/YouGov poll had Biden ahead by 10 points. So did the CNBC/Change Research poll, 10 points for Biden. The NBC Wall Street Journal poll also had Biden up by 10 points. And the final poll for CNN, which was completed during the week before the election had Biden up by 12 percentage points.

Joseph Campbell: Biden won the popular vote by four and a half percentage points. Now that's a healthy margin, but it's far, far from the blowout that many pollsters had anticipated.

Joseph Campbell: So the pre-election polls of 2020 now become an element in the unflattering lineup of prominent polling surprises and failures in US presidential elections. A mosaic that includes polling in the presidential elections of 1936, of 1948, 1952, 1980, 1996, 2000, 2004, 2012, 2016.

Joseph Campbell: Now, it's interesting to note that not all of those polling failures and surprises were quite the same just as no two presidential elections are precisely alike. Polling failures tend not to be the same either.

Joseph Campbell:

7

In a way, polling failure in US presidential elections can be thought of as akin to Tolstoy's unhappy families. They are unhappy in their own ways. Polling failures are distinct and history shows us that they are seldom duplicative. They surprise us in their own way.

Joseph Campbell: They are not at all like Dewey Defeats Truman, the polling fiasco of 1948 when all the polls, all the pundits, all the media outlets predicted Thomas E.Dewey would unseat President Harry Truman and perhaps do so by landslide margins. Didn't happen. Didn't happen. Truman conducted a vigorous cross country campaign and won the election by four and a half percentage points.

Joseph Campbell: This year, in the aftermath of yet another polling surprise, it is interesting to note that some pollsters and some polling analysts had sought refuge in the cliche that pre-election polls are not really predictions.

Joseph Campbell: They are but they say snapshots in time. Snapshots that capture a single moment and do not necessarily tell us what we will likely experience on election day. This point was made as a way to deflect criticism about poor polling performance. Polls are not predictions.

Joseph Campbell: And to that, I say, nonsense. It's a cop-out. It's a dodge to insist that pre-election polls are not predictions. To say so is a way of shielding errant polls from criticism.

Joseph Campbell: And here's why. Here's why this notion is a cop-out, a slight to shelter from criticism. There are, as we've seen in this year's election, many, many pre-election polls. Together, they offer many snapshots about voter tendencies during a presidential campaign. Those snapshots become more numerous as the election draws closer. These many snapshots presented by many polls together represent a highly detailed mosaic, a panoramic view of what the public is preparing to do when the time comes to vote.

Joseph Campbell: This information is useful. It's valuable in a democracy. Information about who is ahead and who is trailing has value in the robust marketplace of ideas so long as it is accurate.

Joseph Campbell: Besides, pollsters do measure their work against the final results and they have long done so. They don't do election polling with the intention or the expectation of being wrong, of being an error, of being misleading. They don't do polling to be wrong and in turn deeply embarrassed.

Joseph Campbell: Pollsters typically have devoted a lot of resources to their final pre-election polls so that stands a real good chance of aligning with the outcome at the ballot box.

Joseph Campbell:

8

Not only that. Pollsters have long tended to thump their chests when their pre-election polls have come close to the actual outcome. This goes back to the 1940s and George Gallup, who was one of the founders of modern survey research.

Joseph Campbell: Gallup took out prominent full-page ads in the journalism trade publication, editor and publisher to tout the relative accuracy, the relative accuracy of his pre-election polls in 1940 and 1944.

Joseph Campbell: More recently, the pollster Mark Penn indulged in some self-congratulation in an op-ed published last month in the Wall Street Journal, in which he mentioned the final Hill/Harris pre-election poll, which he helped to conduct.

Joseph Campbell: "That poll," Penn wrote, "got it about right. With a four point gap in favor of Mr Biden, we saw the momentum," he wrote, "we saw the momentum in that race seesaw in the final month and end up on Mr. Trump's side as he wrapped up his mass rallies."

Joseph Campbell: Penn also noted that prominent national polls, such as from the Wall Street Journal, NBC, CNN and Quinnipiac predicted 10 to 12 point margins for Biden, which he said would have produced a 40-state landslide. A result, he said, that on its face should have been dismissed as impossible. As I mentioned, Biden won by four and a half percentage points.

Joseph Campbell: I am by no means the first to maintain the pre-election polls should be treated as the predictive instruments that they are meant to be. A former president of the polling industry's leading organization, the American Association for Public Opinion Research, known by its acronym AAPOR, said as much 33 years ago. He was Phillip Meyer, a former journalist who developed the precision journalism technique that brought social science methodologies to the practice of daily journalism.

Joseph Campbell: In a journal article in 1987, an article titled Polls Are Predictions So Let's Stop Kidding Ourselves, Phil Meyer wrote, "The state of the pollsters' art has reached a point where there should be no need to whine about non-predictive snapshots in time."

Joseph Campbell: Meyer also pointed out a poll is an estimating device while on election is an exact measurement of the real world. By holding the device against reality, we can learn how well the device is working.

Joseph Campbell: As the 2020 election told us, that device is not working so well. Thank you very much.

Larry Bernstein:

9

All right, I'm going to open it up for questions and Joe, I'm going to start with you. You mentioned that we've had failure in presidential polling pretty consistently since 1936 and like a Tolstoy family, the errors have been different almost every single time.

Larry Bernstein: Can you highlight what has been the cause of those errors over time? Has it been that they've quizzed the wrong population? The populations weren't representative samples? Why were there exit poll failures? Can you just highlight some of the things that we've gotten wrong?

Joseph Campbell: Sure.

Larry Bernstein: And then the second point is that given the incredibly poor track record, why do we put stock in these polls at all?

Joseph Campbell: Well, one thing is that the polls are not always in error. Polls can get it right, have gotten it right. In the 2018 midterm election, a lot of the pollsters took a lot of credit for, gave themselves credit for getting it pretty right, for getting it pretty close. And that has led to a theory that Donald Trump being at the top of the Republican ticket somehow throws off the polls and the pollsters just really can't measure properly Trump and Trump's supporters and so forth.

Joseph Campbell: But the pollsters do bear some responsibility for these misses over the years. And indeed, they are quite different. In 1952, for example, that was the first campaign between Dwight Eisenhower and Adley Stevenson. And the pollsters were George Gallup, Archibald Crossley, Elmo Roper, principally the three national pollsters at the time. They had been burned in 1948 when they predicted that Thomas E. Dewey was going to unseat Harry Truman by fairly easily margins.

Joseph Campbell: In '52, they were very, very, very cautious. And they figured that the race was close, that there was a substantial undecided vote that could swing either way and that if it went to Adley Stevenson, that Stevenson had a chance of winning that election.

Joseph Campbell: So the pollsters were really, really very cautious and their caution blinded them to what really was going on. And that ended up in a 39-state landslide for Eisenhower. It was a landslide that no pollster saw coming.

Joseph Campbell: We saw something akin to that in 1980. It wasn't quite the same, but it was akin to that. When Ronald Reagan and Jimmy Carter were in a three-way race but John Anderson for the presidency.

Joseph Campbell:

10

Carter and Reagan looked to be very, very close as the election wound up. Most polls, even after the only debate of the campaign between Carter and Reagan which took place a week before the election, even afterwards, the pollsters were largely saying that this race was too close to call. In the end, Reagan won in a near landslide, a near landslide that no pollster had anticipated.

Joseph Campbell: Now, sometimes the polls are thrown off by and pollsters are thrown off by late breaking developments in the campaign. In 2000, for example, most of the pollsters figured that George Bush, the younger, was ahead narrowly ahead of Al Gore in the popular votes.

Joseph Campbell: What happened about three or four days before the election was the disclosure by a television station in Maine that George Bush, George W. Bush, had had a drunken driving arrest on his record, arrest and conviction many years before near the family compound at Kennebunkport, Maine.

Joseph Campbell: Bush had never disclosed that. He never talked publicly about that. And the reporter who broke the story had a main television station had the effect of upsetting the whole race and Bush probably lost the popular vote. He did lose the popular vote in 2000 and the probable factor in that loss was the late disclosure of his drunken driving record. It was enough probably to swing the popular vote to Al Gore.

Joseph Campbell: So there are many factors. There are many reasons why polls can get it wrong. Sometimes, this late breaking phenomenon is a major factor and there's just no way that pollsters can control for that.

Joseph Campbell: James Comey's announcement late in the 2016 campaign, in late October 2016 may have thrown enough votes, perhaps undecided voters, into Trump's camp and deprived Hillary Clinton of enough votes in key states to deprive her of the electoral college victory.

Joseph Campbell: She won the popular vote, but lost the electoral college. Principally because three key states in the upper midwest, Wisconsin, Michigan, Pennsylvania, surprisingly went to Donald Trump by narrow margins. The margins were so close that it could have been related to this late October surprise that James Comey sprung on everybody, by announcing that he was going to and the FBI were reopening an investigation into Hillary Clinton's private email that she had used during her time as secretary of state.

Joseph Campbell: That announcement late in the campaign really shook things up. So there are many factors that can screw up polls. Polls are fragile. Polls are not perfect instruments by any means. It does take an awful lot of money to do polls correctly these days. And that's probably another factor. So I hope, Larry, that that addresses some of the questions-

PART 1 OF 4 ENDS [00:31:04]

11

Joe Campbell: That's probably another factor. So I hope that addresses some of the questions or some of the points that you're asking about.

Larry: Yeah. So, let me bring in Doug for a second. Doug, you were mentioning about all these other surveys that we have, nonpolitical, and I'm thinking more, some longitudinal studies, particularly. What can we do to try to get a representative survey if so few people are responding? That seems to be the clear challenge that as you said, is we've collapsed from an 80% response function to something like 10 in certain political pollings. It may be as low as 1% and that the key to these statistical approaches is that we have a random sample, but if we don't, what can we do? What does that mean for us?

Doug: It's very difficult to adjust ex post facto. You have to know why people didn't respond in the first place and then do a counter waiting scheme to be able to correct for that. As Eric pointed out, initially, the factors that are causing people to not respond or to not tell the truth are not easily measurable. They're not things like demographic categories or socioeconomic categories. They're much more complex views on hot button issues like immigration or political correctness. Those are what get people motivated to respond or not to respond or to respond truly or not. So, it's really a conundrum and has been plaguing survey research for quite a while. Longitudinal surveys, you mentioned, those are surveys that survey the same people over and over again over time. Those are very expensive, but they generally have much higher response rates because of the survey researchers cultivate the response group and keeps them in contact with them, frequent updates with them.

Doug: That's a different kind of survey, but those are extremely expensive to run, costing much, much, much more than the kind of polling that was done in the 1960s and 1970s. Polls haven't always been that bad. In the sixties and seventies, the polls were pretty good. Even in the 1980 election, although national polls stopped polling the week before the Reagan Carter election, Pat Caddell, the pollster for Jimmy Carter. Was doing tracking polls on a nationwide basis. He could see Carter's support collapsing in the final week of the campaign. He actually took Carter aside before the campaign, before the election was over, before the election had been called and said, "You're going to lose this election, I think." Carter was devastated, but put his polling was correct at that time. So, the track record isn't quite as bad as I think it's often made out to be in the past, but it's gotten bad now because we're down to single digits and digit response rates. We don't always know why people are not responding and so we can't cut our weight properly.

Joe Campbell: Larry, this is [crosstalk 00:34:36].

Larry: Go ahead.

Joe Campbell:

12

Let me just step in for a second and note that Caddell's polling was not public. This was private polling that he shared only with the Carter campaign and his Republican counterpart, [Wirthlin 00:34:48], also was tracking a similar advantage for Ronald Reagan. In fact, he did not find that the polling at the end of the campaign suddenly went one way or the other, that Reagan had a substantial lead at least during the last week of the campaign. So, the public polls though, were quite clearly showing that this was a very, very close race. It was quite a surprise when Reagan won that 39 state landslide or near landslide.

Larry: I want to go back to longitudinal studies just for a second. So, the USC survey did a longitudinal study in preparation for the Trump, Hillary Clinton election in 2016. It's result was a kind of a Trump plus one at the end. The result was that people said, "Oh, that's incredible. This was the only poll that showed Trump doing so well," and they were fully transparent. They had an African-American male who was a Trump voter, and that caused a lot of controversy. There was an article in New York Times about whether or not how much weight should be given to that individual. This time in the 2020 race, that longitudinal study showed again, Biden winning by more than 10. So it fell to this, it had the same problems as the other polls. So, I wonder if that means if you guys suspect whether or not longitudinal polling, even though more expensive, may not create much more value because its problems associated with a not representative sample. Doug, maybe you can start with that.

Doug: Well, I think longitudinal surveys is not the kind of fast, quick mechanism that you can use for political polling on an ongoing basis. It's more for fundamental research survey work and no, it's not a panacea to solve all these problems. It's much more basic than that.

Eric Kauffman: Yeah. I mean, it's worth saying too, that another approach, which I think the GSS and other surveys undertake is to more intensively recontact people who don't respond and put money into that. So, that is another route to improving accuracy perhaps, but again, very expensive.

Larry: Hey, Eric, question for you, Eric. You mentioned maybe layering on top of what questions to ask during the polling to maybe help you reference what kind of audience representation you have. You mentioned asking questions about trust, asking questions about immigration, but what other things can they do? Can they tease out maybe religiosity views on race, views on childcare, or any tax policy for the wealthy? What sort of questions do you think would be adding value to properly incorporate these from other studies?

Eric Kauffman: Well, I think you want something that's relatively stable that isn't something that can be manipulated, let's say, by a political campaign, like support for the wall, something that's very clearly a politically manufactured constructed issue. So, something like can people generally be trusted, a question on that, a question on support for death penalty, views on childcare or even one of the big five personality dimensions is called openness to experience. I think these would all be useful if we could benchmark them ideally in the exit polls and then begin to start to form a view as to what share of the electorate, what the electorate looked like in terms of things like trust, things like openness, then we could start

13 doing what we do with validated voter rolls in the census and to form a picture of what the demographics of the electorate are.

Eric Kauffman: I think it's going to be important to try and get at some of these psychological dimensions in order to weight properly. Otherwise, you'll just consistently, particularly when politics now is increasingly revolving around this so-called open closed rather than left right dimension, which is more about openness to experience or perhaps people who want faster change, slower change. That is linked also to this trust dimension even more, so you're getting a systematic error in the same direction. I mean, that's the other point is if the polling errors for Trump votes and also for Brexit here have gone in the same direction each time. That's, I think, what's also important to point out. These aren't just random errors. These are errors, which I think we can, because they're moving in the same direction we can start to try and correct. I think these sort of psychological items might be one way of doing that.

Larry: Yeah. I think what's particularly embarrassing is websites like Fivethirtyeight saying, "Biden has a 90% chance of victory," or Hillary Clinton has a 95% chance of victory." When they assume a confidence interval where they conclude that if Biden's up by 10, then they have a two standard deviation basically has them Biden winning, when in fact maybe the distribution is not that random, nor do they have the competence interval right. It's one thing to say, "Look, it looks like Biden's in the lead," but to say that he has 95% chance of victory, that doesn't do anybody any justice.

Eric Kauffman: Right. Yeah. I mean, of course you have to know what your target population is to calculate that confidence interval right. So, it's a chicken and egg a little bit. You have to be able to have some idea of what the electorate should look like. I think not just in these demographic terms, but also to some degree in psychological terms if you really want to get this better, really, and correct for this kind of error that seems to be happening in the same direction, time and time again.

Joe Campbell: I was just going to say that in defense Fivethirtyeight.com, in 2016, Silver was projecting Clinton at a 70% chance of victory. There were some polling forecast models that had others that had Clinton up ahead as high as 98 or 99%, but that did not include the Fivethirtyeight. This time, Silver was saying that there's a fine line between what could be a Biden landslide in a very narrow election. So, he was kind of hedging his bets a little bit if you will, but nonetheless, that's what turned out. It was not a landslide, but it was not really a nail biter either in terms of the overall popular vote in the country.

Larry: Well, I mean, I think that currently in the United States with California representing such a positive vote getter for the Democratic party, if it's 70, 30 Democrat in California and California represents around 10% of the country, the Democrats are going to win the popular vote by four and then tie elsewhere in the nation. It's those other States that it matters so much and to have gotten the battleground states off by such an enormous amount, I think, is where the real problem was, not so much in the popular vote. I don't know what to make of that, what you think of that.

Joe Campbell:

14

This is [Joe Campbell 00:00:42:52]. Yeah. I think quite right, that California is going to give the Democratic candidate an overwhelming majority. I mean, this year, it was almost 30 percentage points for Biden, and there's just no way that the Republican party is going to overcome that kind of popular vote deficit elsewhere in the country. There was some discussion that Texas, of all places, might go Democrat. It did not, but that's the only place where they could begin to make up that kind of deficit. With New York trending Democratic as well, the expectation is that that Democrats are likely to win the popular vote in presidential elections for years to come.

Larry: Just to follow up with something Eric said about using exit polls to help us properly gauge going forward our analytics of our population. I guess Eric, what I want to push back against and Joe, I'm going to lean on you for some support on this, is that exit polls themselves are also deeply flawed. In the 2004 election with Kerry, you highlight in your book, Joe, that the exit polls were completely wrong. So, I worry that using the exit polls as a fixed help in this process may be leading us into another rabbit hole.

Eric Kauffman: Yeah. Yeah. I mean, I think there are different instruments we can use. I mean, there are also surveys that have intensively, like the GSS that intensively trine and reduce non-response by recontact. That's another way of validating. But I do think the exit polls have some advantages, not only the much higher response rate, but the fact that they don't, perhaps because they are associated with an official process, wouldn't engender the same reactants that you might get from a pollster that's associated with either a university or the media. But of course, I agree. I think we need to use a number of different methods, but I think we need to get beyond the census categories, just the demographics, because so much of the variation is not to do with the demographics. So, I think we got to use whatever instruments we have that are less contaminated. I still think the exit polls are useful in that respect.

Joe Campbell: This is Joe Campbell. There is a lot of concern about exit polls overall though. It was not just in 2004 when, as you say, Larry, John Kerry was seen as headed to victory based on exit polls that year. In fact, the exit polls prompted George Bush to brood at the White House for a couple of hours or so thinking that he had lost just like his father had lost after a single term and exit polls signaled that night, that Kerry's victory. In fact, one of his top aides believe that he had one and he referred to Kerry as Mr. President. I think we've seen exit polls go, if not completely off the rails as they did in 2004, but to be misleading often enough to be very, very wary and some journalists who cover election polling and do so fairly well, like Nate Cohn of the New York Times, as well as Nate Silver at fivethirthyeight.com. They really have deep reservations about exit polls and what they tell us.

Joe Campbell: There is so much that we rely on exit polls, not just for who's winning, who's losing, but why people voted the way they did. If exit polls are suspect, if they're in error, if they're off, then all that valuable information about how the election turned out the way it did is questionable. That's really very serious.

Larry: Let me give you an example. [crosstalk 00:46:57]. Sorry, let me give you an example from the...

Eric Kauffman:

15

Sorry, go ahead.

Larry: No, go ahead.

Eric Kauffman: Well, no, I was just going to say, I mean, I think it would be important to look at a range of exit polls. I mean, and certainly the ones in Britain recently have been very good. It just depends, I guess, on the year, but it's one thing to say how correctly do they predict the final outcome, but I just think also just for looking at some of these breaks by other variables and using those perhaps to weight future surveys, I think they're useful from that perspective. It's not to say they're going to get it exactly right either, but I think they're probably better than polls based on a several percentage point response.

Larry: Let me throw out a couple of things that came out in the exit polls in this presidential election. One was that Hispanics voted much more for Trump than expected. Second is the black males voted more for Trump than expected, and that white males voted less for Trump than expected. How much value should be placed on those conclusions based upon these exit polls, if any? Anybody?

Doug: Well, an exit poll's only as good as the precincts that you got in your sampling and how correctly they're weighted to reflect the population. That's a very tricky exercise to get done correctly. So, there's a lot of unreliability in exit polling, depending on the weighting scheme and where you're doing your exit polling.

Joe Campbell: This is Joe Campbell. I agree that exit polls can be very dicey in that regard, as well as the fact that early voting is taking place in greater numbers than ever before, which complicates it. It doesn't make impossible exit polls, but it really does complicate the taking of exit polls. So, some exit pollsters do go to early voting locations and conduct exit polls as much as they have done on election day at various precincts in the country. But it's still a bit of a roll of the dice. As I said earlier, relying on for so much information about why the election, went as it did, and if those exit polls are in error, if they're off, then our understanding of the election is off. So, the points that you just made a few moments ago, Larry, they could be right, they could be wrong. We really don't know for sure, unless the exit polls are on target, and they're [crosstalk 00:49:33].

Eric Kauffman: It's Eric [Kauffman 00:49:36] here. I just want to say that with regard to Hispanics and Asian shift, I mean, we also can look at aggregate level data and particularly from heavily Hispanic areas, such as the Rio Grande Valley and Miami Dade and Imperial County and various other parts of the country. You can certainly see in many of these locations, actually all over the country, a shift amongst these groups towards Trump. So, I do think that tallies pretty well with what we see in the exit polls. I think if you look at the swing, that's also come out in the aggregate data. I think those two points really would suggest that something did happen in this department. I'm happy to talk about why I think that's occurred, but I'm going to let others have a chance.

16

Larry: There was a survey by Trafalgar in which they ask, not who you're going to vote for, but who is your best friend going to vote for, hopefully taking away some of the, to tease out your concerns about bias. These are techniques that some surveys have used to detect racism. They.make you choose two or three out of four and then vary the number of questions to look at a larger population to tease out some answer. I'm just wondering what you guys think about structure of polling designed to ascertain the nature of themselves, or do you think that, for example, finding out if they asked who your friend or your daughter or your father, how they're going to vote, whether that would reflect information that would be potentially more valuable? Doug, do you want to start with that one?

Doug: Well, it doesn't get around the non-response problem, which is fundamental. I don't think that those techniques in the survey research world, asking about friends and who your friends are to vote for, have been validated very well. The one that was described where you can vary the questions so that it's impossible to know which one the person's answering, those work pretty well, but they're complicated to administer, and you can't really connect an individual to a particular response. You can only develop an estimate in the aggregate. So, I'm not hopeful about that avenue.

Eric Kauffman: Yeah, I would, Eric Kauffman here, I'd add I think I saw a paper where someone tried that list experiment with Trump voting and did not pick up concealment. So, it would seem like the non-response problem is the big one. This, on its own, probably wouldn't address that.

Larry: Eric...

Joe Campbell: This is Joe Campbell, Trafalger has gotten some cachet as the pollster that predicted the 2016 election most closely in key states. I think that some of that has been lost in the 2020 election. I don't think that their election was all that great this year. They didn't do national polling, but they did some key state polling.

Larry: I think they did pretty well. [crosstalk 00:53:07] I think they called North Carolina and Michigan was going to be way close. They said Wisconsin was going to be really close.

Joe Campbell: They were off in Michigan, they were off in Michigan. I think that they had Trump ahead by five in Georgia where he lost by less than a percentage point. So, I'd say it's a mixed bag for Trafalgar, but they certainly weren't the pollster of the year in 2020 as they were in 2016.

Eric Kauffman: But I think, I mean, one thing they did do in addition to asking about how your friends voted is they have a running panel where they're asking about a lot of non-political questions. So, maybe they are able to reduce suspicion that way and then just throw in the odd political question and get away with it more

17 perhaps. So, maybe it's that rather than the question about your friends is why they're getting it a little closer.

Joe Campbell: This is Joe Campbell. I think it's really intriguing how the polling industry at large is trying various techniques to try to get around this non-response issue. I think that the gold standard of random digit dial live operator telephone polling is recognized as being on its way out, if it's not already out, as the gold standard and just what is going to replace it is uncertain. There is an awful lot of experimentation that's going on. I think what Larry was mentioning related to Trafalgar is an example of that kind of experimentation. The field is rich in experimentation. They haven't yet, pollsters haven't yet landed on the new gold standard, if there is indeed a new gold to be found. The internet is really one of the more alluring aspects of contemporary polling research, but no one's really figured that out yet, either

Larry: A question for Eric, we've obviously as Americans, the audience is 95 or 98% American. So, we focus a lot on this US presidential race, but Brexit was a huge deal in the UK, an enormous percentage of the population voted for it. The results were shocking to the pollsters and I call it the intelligence of the UK. What lessons about polling, what lessons about understanding the UK did those pollsters miss for that incredibly important vote?

Eric Kauffman: Well, I think first of all, they didn't get it too wrong. That's one thing on the Brexit, but they did get it wrong and they got it wrong in the same direction as the Trump polls, but it wasn't out by the same margin that the polls are here. But I think it's the same general issue about a shy voting, by shy, we're talking about people who, for whatever social reasons or whether they're reacting against people they don't trust or not answering, are systematically answering less. I think that's kind of the lesson really from all of these misses that are in this direction of being against populace. I think that there has to be some awareness that has to be corrected for somehow and that you can't just keep on business as usual. So, again, I would urge trying to take into account some of these non-demographics, stable non- demographic attitudes, or psychological dispositions, if you can, to wait at least to some degree and see if you can do a bit better.

Larry: I would like to end this segment on a note of optimism, in this case, related to improvements in survey and polling methodologies. What do you think, what do we need to do to get this thing better as we go forward? Eric, do you want to start?

Eric Kauffman: Yeah, sure. So, I think really that we have to recognize now that trust and resentment of elites, which includes media and universities and so on, is now systematically correlated with voting in western countries and that pure demographics are not going to capture that - just controlling, waiting for education is not going to be enough to correct for that. You may get non-college voters, but they may be the high trust non-college voters. You really have to somehow figure out how to get low trust and particularly amongst college voters, low trust college voters. We know that highly educated voters, they're not all voting against populace. They're, in fact, very politicized and more polarized. So, we need to somehow find a way, as I mentioned, to try and weigh by some of these non-demographics in order

18 to improve the accuracy of polling, particularly in a climate when electoral politics revolves increasingly around these cultural open-close dimensions.

Larry: Doug, what about you?

Doug: I think, if we could start asking the big five personality dimensions on polls regularly to get a handle on how that distributes for the entire population, that might provide an alternative way of weighting, if we can determine the degree to which these different personality characteristics are predictive of different voting patterns. So, I think we have to start collecting data on new kinds of concepts and putting them in polls on a more regular basis.

Larry: Thank you. Joe?

Joe Campbell: Larry, you said an optimistic note, is that what you wanted?

Larry: Yeah. You want to make a pessimistic note, it's fine. Go ahead.

Joe Campbell: No, no, no, no. Optimistic note refers to a point I made a little earlier about the experimentation, the churn in the field that despite the reverses and uncertainties and embarrassments in polling, there is still an awful lot of experimentation that's going on in the field. That, they haven't really found the new gold standard for polling, but there's an awful lot of experimentation going on. I think that's very favorable and perhaps, in the long run, will lead to some improvement in election polling. So, that's an optimistic note, I think, to not lose sight of.

Larry: All right. With that, that ends this segment. I'm going to turn the call over to Rick Banks to introduce the constitutional law segment next. Rick, take the floor.

Rick Banks: Okay. Thank you, Larry. This is Rick Banks from Stanford Law School. This segment will focus on constitutional law and, in particular, the Supreme Court. I'd like to first invite our speakers from the first segment to please join the fray in the Q&A section of this discussion. We are focused today on what might seem a simple question, is the Supreme Court broken? If the Supreme Court is broken, who broke it? How did it get that way? If it is broken again, then how to fix it? What should we do going forward? What would be both feasible and desirable or perhaps desirable, but not yet feasible? What changes need to occur so that we have a Supreme Court that functions more as it should?

Rick Banks:

19

We're joined with three outstanding legal academics. I'm honored deeply to have them together. Leading off will be the current president of the William and Flora Hewlett Foundation. Also, my former dean of Stanford Law School, Larry Kramer. Larry is a graduate of Brown and University of Chicago. He is well-traveled as a law professor. Before coming to Stanford, he taught at Harvard, University of Chicago, NYU, the University of Michigan, and Columbia. He was one of the very shortlist of faculty that we long sought to recruit to Stanford. The entire faculty in fact, could not have been more delighted when he agreed to join the faculty as the dean.

Rick Banks: He is the author of the landmark book, very directly related to our topic today, The People Themselves, Popular Constitutionalism and Judicial Review. When Larry left Stanford a few years back to assume his philanthropic role at the Hewlett Foundation, I joked with him sometimes that he was assuming the enviable, yet challenging task of spending, given the size of the Hewlett Foundation endowment, literally $1 million every day. Well, Larry took to that task enthusiastically. I am delighted that on this day, he has chosen to take a break from that task to join us in this important conversation. Thank you so much, Larry, for joining us. Our second...

PART 2 OF 4 ENDS [01:02:04]

Rick: Thank you so much, Larry, for joining us. Our second speaker today is Mark Tushnet. A long time member of the faculty at Georgetown University Law Center, Mark is now a member of the faculty at Harvard Law School. During his long and accomplished career, he has become one of the most cited law professors in all of the legal academy. I first encountered Mark Tushnet speaking many years ago, decades ago. In fact, it was before I even went to law school. His book, The NAACP's Legal Strategy Against Segregated Education was actually published the year I graduated college. That was one of the many books I read as a young man trying to determine my path that prompted me to apply to law school, hoping one day to follow in the legal path that he blazed. Most recently, he has authored the book Taking Back the Constitution: Activist Judges and the Next Age of American Law. This book directly pertains to our discussion today. It is a continuation and elaboration and an enlargement of ideas that he began discussing decades ago in his 1999 book, Taking the Constitution Away from the Courts. We are delighted to have Mark Tushnet join us today.

Rick: And finally, the third member of the panel, no less esteemed but somewhat junior, is Ganesh Sitaraman. Ganesh is a professor of law at Vanderbilt University, and he is also the director of the program in law and government. He writes widely across many fields, constitutional law, foreign relations, administrative law. For more than a decade ago, he has been an ongoing advisor to Elizabeth Warren in her many political roles. Ganesh is an extraordinarily productive law professor, one of the people that makes me feel that I'm definitely not doing enough. You see, he publishes, in recent years, almost a book a year. Between 2017 and 2019, he in fact did publish three books. In 2017, there was The Crisis of the Middle Class Constitution: Why Economic Inequality Threatens our Republic. And he followed that up in 2019 with two books, The Public Option: How to Expand Freedom, Increase Opportunity, and Promote Equality, and most recently, The Great Democracy: How to Fix Politics, Unrig the Economy, and Unite America. We're delighted to have Ganesh join us. So, our first speaker will be Larry Kramer, and then Mark Tushnet, and then Ganesh Sitaraman. Larry, I turn it over to you.

20

Larry Kramer: Thank you, Rick, and also for that very kind introduction. So, we assume today that if the Court has the power of judicial review, that is the power to say a law is unconstitutional and so unenforceable in the case before the Court, that the Court's interpretation of the Constitution must also be final and binding on everyone else. No matter how controversial or debatable, everyone else in government and society is supposed to accept and get in line behind the decision of these nine judges in all other instances. We call this stronger notion judicial supremacy. The courts are supreme over the other branches of government in interpreting the Constitution. It's actually a pretty weird idea, if you think about it. We say we have a system of democratic self-government, meaning the people have final say over the meaning of their laws, except for the really important one, control stops at our most important law, which is controlled by nine judges appointed for life.

Larry Kramer: Having fought a revolution whose whole purpose was to replace monarchical power with popular control, it would have been bizarre for the founding generation to create what amounts to a mini monarchy over their most important law. Yet, today, people nod when told that the most important reason to vote for this or that presidential candidate is that he or she will choose Supreme Court justices. Really? Doesn't that signal something very awry in our democracy? In fact, judicial supremacy has been generally accepted only in the past 50 years or so. For most of us history, the rule was not that courts had final say over the Constitution.

Larry Kramer: Think about it this way. When designing a democratic constitution, you need to balance two considerations, accountability and independence. On the one hand, you need government to be accountable to the people. That's the very definition of representative democracy. On the other hand, you need government to be independent enough to make hard decisions and act on them. The US Constitution does this balancing by creating multiple branches that have shared and overlapping authority. Each branch balances independence and accountability differently. Different electorates choose them. They have longer or shorter terms, and so on, so that they respond in different ways to political events and can check and balance each other. When they disagree and act differently, the solution comes through whoever the public ultimately supports, which is what makes this system democratic. Political leadership means acting and then persuading the democratic people that your position is right. When it came to courts, the framers went for strong independence by giving federal judges lifetime appointments and salary protection. That's because they were thinking about the role judges play deciding individual cases, which was the experience that they had. They weren't imagining judges playing an important policy role, because almost no one yet understood the possible role of judicial review, which is why it wasn't explicitly included in the Constitution.

Larry Kramer: When such a world did emerge, the Court's independence needed to be balanced with accountability. That was done by using tools the Constitution clearly makes available, tools like Congress's power to determine the Court's size and makeup, to fix its budget, to decide the scope of its jurisdiction, and to determine its rules of procedure. And these tools were used throughout American history whenever the Court pushed a political agenda, by presidents like Thomas Jefferson, Jackson, Lincoln, and both Roosevelts, not to mention the reconstruction Congress, so hardly a rogues' gallery here. The idea was that the Court could exercise judicial review and its decisions will be final in the case before the Court,

21 but not final as precedent for the other branches, which could act differently if they believed that the Court was wrong. If the Court went along, as it usually did, the issue was settled. If not, the political branches might break down, as also happened a lot.

Larry Kramer: But if the issue mattered enough, the political branches and the Court stuck to its guns, the political branches could push back by adding judges, shrinking jurisdiction, and so on. If, that is, there was political support for it. Attacking the Court is always controversial and takes a lot of political will, but it did happen on occasion, and the mere possibility that it could happen produced a nice balance that kept the Court's power in check. In this way, disputes over constitutional meaning were decided through the same system of checks and balances as other questions. Different branches of government with different understandings pushing for their views, with the issue ultimately decided by whose views were most persuasive to the nation as a whole.

Larry Kramer: So, what changed? How did judicial supremacy become the norm? It's a political story. The idea of judicial supremacy actually emerged in the 1790s, championed by conservative federalists who were fearful of democracy as represented by the French Revolution. It was repudiated in the election of 1800, which brought the progressive Jeffersonians to power, a point made clear in the 1803 Marbury decision, which was actually a retreat by John Marshall to the progressive Jefferson's position of judicial review without supremacy. But the idea never went away completely, and for the next 150 years, conservatives, however defined by the politics of the moment, supported aggressive claims of judicial supremacy, while progressives, however understood at the time, rebuffed them. Mostly the Court avoided the conflict by restraining itself. In the few instances where it pushed the issue, the other branches pushed back, and the Court backed down. Those are the instances I mentioned before with Jefferson, Jackson, Lincoln, and FDR, and so on.

Larry Kramer: Then came the Warren Court, which was something we'd never seen before, an activist Court that was also progressive. And the left flipped, embracing the opportunities judicial intervention suddenly afforded to advance their agenda. Conservatives, for their part, continued to support judicial supremacy, objecting to what they saw as the Warren Court's kind of Footloose method of interpreting the Constitution, but without denying that the Court's word is final. And the debate shifted from who has final say over the Constitution, the Court, everyone now agreed, to how that say should be exercised. This is a point, I should say, first made by Mark Tushnet. There was a little debate over constitutional interpretation before judicial supremacy became the norm. Once the Court became final, however, how it interprets became a lot more important, hence today's debate over originalism versus a so-called living Constitution, which really only started in the 1980s.

Larry Kramer: As a practical matter, judicial supremacy was made operational by de-legitimating the devices that had been historically used by the political branches to push back. That is to say, de-legitimating ideas like Court packing, jurisdiction stripping, and so on, and fostering an ideology that the Court is supposed to have final say over the meaning of the Constitution. Not surprisingly, the role of the judiciary grew exponentially, an expansion that has continued uninterrupted, whether the Court has been liberal, like the Warren Court, moderate like the Burger Court, or conservative like the Rehnquist and Roberts

22

Courts. Given the mess we have today with the Court expected to settle so many plainly debatable issues and the huge controversy this creates around every appointment and the bad incentives it creates to appoint children to the courts so that they will last there forever, it seems pretty clear this has been a mistake and we should restore the legitimacy the devices whose constitutional authority is clear, and that's served us well in the past, restoring a constitutional balance that works considerably better than the one we have today. Thanks.

Rick: Larry, that's fabulous. Mark Tushnet, you're up next.

Mark Tushnet: Thank you. I'm going to pick up on some of the themes that Larry's introduced. Progressives and conservatives are dissatisfied always with judicial activism. The problem is that they see activism occurring in different cases. Now, the obvious solution if you're dissatisfied with judicial activism is to opt for judicial restraint, but the problem is that, as we've experienced historically, you're in favor of judicial restraint when the other side controls the Court. You're in favor of judicial activism when your side controls the Court. And as we've also experienced, as control of the Court shifts from progressive to conservative, you'll come up with the conservatives, who've been arguing against the Court, arguing for judicial restraint, now will come up with an account of constitutional review that will explain activism. And so, conservatives no longer say, "Well, we're in favor of judicial restraint across the board. What we're in favor of is judicial interpretation of the Constitution in accordance with the correct theory of interpretation. They should be activist when the Constitution says they should be and restrained when it says they shouldn't be." And that's their account of the theory of originalism, as Larry suggested. Progressives have the sort of flip side of that, in the theory of living constitutionalism.

Mark Tushnet: Okay. So, now we're in this problem, which is nobody likes judicial activism, except when they control the Court, and everybody sort of knows that their current situation vis-a-vis the Court, whether in control or out of control, is not a permanent condition built into the constitutional scheme of things. And so, everybody sort of knows there's a problem. Conservatives don't face it right now, but they sort of know that they could face it in the future. Liberals didn't know they had the problem until the Rehnquist and Roberts Court, but they're afraid of doing things that would prevent them from regaining the upper hand, if they got control of the Court.

Mark Tushnet: So, what can you do about this? Well, the one thing that we know we can't do is solve the problem by appointing justices, who somehow, you hope will be committed to a general principle of judicial restraint. First of all, we can't do that because the process is highly politicized. And second, you can't get them, the nominees, to credibly promise to act with restraint. So, just my favorite current example is that in Justice Barrett's confirmation opening statement, she said, well, look, the role of the courts is not to displace policy judgements by the elected branches, and I'm not going to do that. But all the work is done by defining some things as policy choices, rather than constitutionally impermissible choices. So, she said she was going to be refrained, but we know she's not, because she's also an originalist.

Mark Tushnet:

23

So, how can you do that? How can you get around the problem? Well, there are structural solutions that have been proposed. The one that has gathered the most sort of support is term limits for the justices. Approval polling suggests on the order of 70 to 80% support for term limits. One problem with term limits is that the easiest way to do it is through a constitutional amendment, but amendments are really difficult. There's a theory that you could do it through statutory changes. I think that theory's okay, but the statutory changes couldn't take effect immediately, and that means it's difficult for politicians to get excited about imposing term limits.

Mark Tushnet: Court expansion is an interim measure. People talk about expanding the Court, but then there'll be retaliation, and the Court will end up at 57 varieties of justices. There are fancier solutions that don't have much political purchase yet. I actually favor Court expansion, but as a temporary measure, I have floated the idea of you send a message to the justices by cutting the number of law clerks they have from four to two. They've cut their workload in half. Why not cut their staff in half? But ultimately, in order to achieve a world where everybody is satisfied with the level of judicial activism, you have to change the underlying political culture. You have to address the issue that Larry flagged of judicial supremacy, and I think Larry's book on popular constitutionalism is the vehicle for that kind of cultural change. Thank you.

Rick: Okay. Thank you very much. That was wonderful. Ganesh Sitaraman?

Ganesh Sitaraman: Thanks, Rick. So, I want to ask the question, why are we talking about Court reform today? And then talk about what some of the proposals are in a little bit more detail. I think there's two reasons really why we're talking about Court reform today. The first reason is that in the last few years, we've seen extraordinary constitutional and political hardball take place over the Supreme Court. As you all know, Merrick Garland was nominated by President Obama, blocked from even getting a hearing, and kept open a seat on the Supreme Court that was President Obama's to fill by right. Then when the same exact scenario happened with the death of Justice Ginsburg this past year, Republicans went ahead and confirmed Amy Coney Barrett days before the election. This exercise of hard ball of power politics that really had no consistency to it was one thing that caused a lot of consternation among Democrats in general about the Supreme Court and the way that the Republicans were treating it.

Ganesh Sitaraman: The second reason why we're talking about Court reform, which overlaps with the first, is that we have a system right now in which our country is deeply polarized and in which our legal communities are deeply polarized. And now, in which the Supreme Court's polarization overlaps with partisanship. This was not true historically. But right now, the president who appointed each of the justices lines up with the ideology of the justices. Historically, the New Deal when FDR's proposals were getting struck down. There were Democrats in the majority doing that. Brown versus Board of Education was by a Republican. Roe versus Wade was written by a Republican. A dissenter in that case was a Democrat. We had a kind of scrambled system for a while, and that is not true anymore. And so, a great fear is that the Court is going to block actions by liberals, and that half the country or more will strongly oppose Court decisions, including on topics where there is extremely strong majority support. So, if the Court is facing an increasing legitimacy crisis in terms of its decisions and its role in our constitutional and political

24 system, there's a concern, not just for the Court, but for what that means for respect for the rule of law and our constitutional system more generally.

Ganesh Sitaraman: So, what to do? There've been a number of proposals. Mark mentioned a few of them. I'll say a little bit more in detail about them. First is to expand the size of the court. This has gotten a lot of attention. One positive thing about it is it is the most constitutional of any of the proposals that can be done just by statute. The case for it is that what proponents say is that you need a Supreme Court that will not just align with Democrats' proposals. It's mostly Democrats proposing this, but you really need a Supreme Court that will approve of structural reforms related to democracy and the rule of law. And that if we have a Supreme Court that is okay with things like voter suppression, campaign finance, is against things like campaign finance regulation, you will have this slow and continued disintegration of our Republican system of government. The cons that people propose with this approach is that it will lead to a tit for tat system in which when the other side is running the country, they will expand the court, and this will just spin out of control over time.

Ganesh Sitaraman: I am not so persuaded by this approach, and the reason why is if you game it out, there's two real conditions that you're looking at. The first is a hundred percent chance of a conservative Court for as far as the eye can see, or some percentage chance of a democratic Court for a while, and then maybe a Republican Court after that, only if Republicans can win the House, the Senate, the presidency, and are able to pass another Court expansion bill. Of course, part of the argument, though, from proponents is that Republicans have a harder time winning elections if elections are fair, and that the reason why they are able to win is because of voter suppression, non-majoritarian institutions, and the like, but reforms to those things that would be upheld by a Democratic Court packed Supreme Court would force Republicans to have to change their views, which may mean that you have a stable equilibrium over time.

Ganesh Sitaraman: Okay. So, that's one. Let's look at another proposal, term limits, which Mark also mentioned. The basic idea here is nine justices serve for 18 years with staggered terms. A new justice picked every two years. Each presidential term means there'll be two picks. Each Senate gets one justice to confirm. Current justices reign on Court for life, so there'd be more than nine for a bit. New justices would retire to the lower courts after the 18 year terms. Now, advocates say this would mean more predictability in vacancies. Appointments wouldn't turn on things like the deaths or judges deciding when to quit. It would also prevent multi-decade justices and gerontocracy. It would help the trend of ever younger nominees, and it's not explicitly partisan. So, those are all the positives, but everything has trade offs and I think this proposal has some problems. First, it would likely mean that the Court becomes an issue in presidential and midterm elections every single year forever. A new justice every two years means campaigns will be about the Supreme Court every year. If part of the problem is that our Court has become too political and has become a part of our electoral and political process, that seems not ideal.

Ganesh Sitaraman: Second, justices themselves might become more political because of post judicial options. They could cash in, become a lobbyist or a talking head, or run for Senate, governor, or the presidency. We've had justices throughout our history who've run for office or who resigned to run for office, and there's a

25 question of whether opinions will increasingly be designed for future campaigns. This is why I'm also unsure that the 18 year plan will stop the push for younger justices. Why not still appoint someone who's 40, who can retire at 58 with a national profile and run for Senate, governor, or president, or be appointed Attorney General, for example?

Ganesh Sitaraman: Finally, if you're a Democrat who wants reform to immediately do things like safeguard voting rights, reproductive rights, worker rights, climate legislation, et cetera, the 18 year term plan doesn't really do anything for you. In the context of if that were to happen in the spring, if Democrats were to keep winning elections and Stephen Breyer doesn't retire, the Court would be 6-4 in 2021, 6-5 in 2023, not tied until 2025, and the first chance Democrats would have to have a majority on the Court would be in 2027. So, that's not a great proposal if what you're interested in is attempting to safeguard democracy in the short run.

Ganesh Sitaraman: A couple of other proposals that I have proposed with my colleague, Dan Epps, the first we call the balanced bench, and you may have heard of it because Pete Buttigieg adopted it during the presidential campaign last year. And the idea here is expand the size of the Court to 15, require that there be five from a Democratic approved list, five from a Republican approved list, and those 10 would pick other five from the Courts of Appeals to serve for a year at a time. If there's no agreement, the Court won't be able to sit for a year. And the purpose of this is to reduce the power and importance of any individual nomination, so there aren't death matches over and over again. And it's also to reduce the power of individual justices, so we're not seeing arguments always pitched to the idiosyncratic views of a Justice Kennedy, for example. This approach also recognizes that partisanship matters, but it forces justices to find compromise on personnel and indirectly as a result on outcomes, so it uses partisanship to try to disarm partisanship.

Ganesh Sitaraman: A final approach that I'll talk about, we call it the lottery or panel approach, and this was floated by Bernie Sanders during the campaign last year. The basic idea is that every judge on the federal Courts of Appeals would be appointed as an associate justice. And so, the 180 or so Supreme Court would hear cases in panels of nine, randomly picked. This is similar to how things work in the Court of Appeals right now. And they would go to DC for two weeks for oral arguments, and then they'd write their decisions from home. To address problems of panel composition, there'd be a super majority requirement to overturn federal statutes. No more than five on any panel could be appointed by a president of a single party.

Ganesh Sitaraman: Now, the upside to this is appointments would be much less of a big deal. Justices would basically just be Court of Appeals judges for their lives, and occasionally, they would do a few SCOTUS cases. There'd be no generation serving justices, no rockstar justices, no celebrities, no random vacancies, no salient confirmation fights. Yes, Court of Appeals nominations would be more contentious, but if you haven't already noticed, they already are. Republicans under McConnell have understood the importance of the appeals courts. And if anything, this approach might force Democrats to pay more attention. The final positive is that this would be a system that would operate under a veil of ignorance, where justices wouldn't be able to pick their own cases. A different panel would hear them. And as a result of this

26 panel approach, you'd also get massive amounts of geographic and professional diversity with justices coming from the whole country, not just from Harvard and Yale and not just from New York and DC.

Ganesh Sitaraman: So, the bottom line is what proposal you're interested in partly depends on the problems you're trying to solve. If you're really worried about age and gerontocracy, you might like 18 year terms. If you think Democrats should have their policy preferences adopted immediately, you might like Court expansion. If you're concerned about partisan balance, you might like the balanced bench, and if you are worried about celebrity justices and want more diversity on the court, you might like the lottery system. But design should follow from what your goals are, and that's the discussion we need to have.

Rick: Wow. This is a sobering conversation. It seems that you three all agree that we are at a point where there's widespread skepticism of the political neutrality of the Supreme Court, that that sense of the politicization of the Court leads to skepticism toward law, and then if we have skepticism toward the rule of law, that leaps to perhaps the undermining of democracy. So, this is not the time for optimism yet in the conversation. But one preliminary question is, if this is the case that the stakes are so high, and if the situation we're in now is so unprecedented, why would we not expect Supreme Court justices to themselves police themselves as an institution, so that they can maintain the legitimacy on which their roles depend?

Mark Tushnet: So, this is Mark. One reason is that most of them, I think it's fair to say, were appointed primarily because of their views about the Constitution rather than their views about the Court's role in the political system. They understand they can promote their views, only if the Court is sustained as an institution, but there's a trade-off. And one question for each one of them is, well, how much will I draw back on my, at this point it's mostly conservative views, in order to preserve the Court in the long run, meaning my ability in the long run to carry out my conservative views? In the book that you mentioned at the outset, I call this the 80% solution. So, you rule with your conservative values 80% of the time, but 20% of the time you temper them so that people will say, oh, the Court's not merely conservative. It does have a view about preserving its institutional role. The problem is that individual judges can think that, but aggregating them and getting them to think that in the same case is a tricky matter, and you can't guarantee that. And so, overall, you might find individual judges making this trade off, and you might actually find each judge making the trade-off, but in different cases, with the effect that it looks like they're just out to advance conservative goals.

Larry Kramer: And let me just add, Rick, the key is the extent to which the justices will do that is exactly correlated to the extent to which they think they're at risk institutionally. So the purpose, the whole purpose of selling the ideology of judicial supremacy is to increase their power. It's basically a pushback that says you may disagree with what we said, but that's not your job. If you want to change it, you have to amend the Constitution or wait for some of us to die or retire and appoint other people like us, because we de- legitimated every other form of pushing back on them. So, what Mark says is exactly right. You see it in John Roberts. He knows the Court's going to have his name and that it's going to weaken the Court as an institution if every important case is 5-4 on ideological lines, so he's found ways to kind of temper that appearance, although he has less and less space for it.

27

Larry Kramer: But the solution is to create some credible sense that the Court actually can't just do whatever the heck it wants, that the nine of them weren't put there. I was once in an event where one of the Supreme Court justice sort of boasted that the more people hated his decision out there, the more the American public thought he was wrong, the more he felt good about it. And that's a kind of crazy attitude, if you think about the role of the Constitution.

Ganesh Sitaraman: This is Ganesh. I would just add to amplify one little clause in what Larry just said, which is it is harder for them to retain themselves as a body, when there is a 6-3 balance versus a 5-4 balance, because persuading two people is more difficult than one to restrain themselves. So, all of Mark's points I think are right, but are made more difficult the more imbalanced to the Court gets.

Rick: There is one other point to throw out, which is historically a lot of the justices, it's not all of them but a lot of them, and on any given Court, a number of them would have come out of politics themselves. And so, they would have a really keen sense of how it worked, and of what it means to take responsibility for a decision, as opposed to just making one. And as the Court got more and more important, our ability to put anybody on it who'd ever actually done anything like that has gone away. So, now you have a Court that has nine basically legal technicians. That's all they've ever done. That's all they've ever been. They're good at that. Of course, the big, important cases, the ones we care about, the reason we care about the Court are ones where the law kind of runs out before you get to that decision. And they don't have any of that sort of wisdom that you might get from being in the fray and then being pulled out of it. And they have ideology, and that's what we see operating over and over again with rare exceptions.

PART 3 OF 4 ENDS [01:33:04]

Larry: And with rare exceptions.

Rick: And how is it as a society, we have collectively gone along with this? It seems that collectively, we have so much to lose from seating Democratic power to a group of unelected people. How is it that we've taken such a wrong turn?

Larry: Mark, do you want to go first on that? We all have views on that, I'm sure. Mark's written on it.

Mark: Yeah, well, I think my take on it is that it's a result of the, call it, general elite domination of the political system in both the Democratic and Republican parties. With the decline of, for example, labor organizations as a central factor in the democratic coalition, some sense the transformation, even, of the Civil Rights Movement into a national movement, rather than a grass roots level activity. And there are undoubtedly deeper explanations as well. I'm not a specialist in this kind of thing, but I gesture in the direction of globalization as a transformation of the economy that makes democratic control over

28 economic life within a single nation much more difficult and so on. But now that's well out of what my area of expertise is. And I just can say, well, that's sort of sounds right to me, but there's no reason for you to think that I'm got any special insight into that.

Larry: Yeah, I would only add I think a big part of it is there is this, I find it somewhat bizarre, but deep romantic attachment on the part of the left to the Warren Court. So even though for the 150 years before the Warren Court, the court was a dependably conservative institution. And even though the 50 years since the court has used the credibility earned by the Warren Court to basically undo everything that it did, an occasional bone gets tossed, but for the most part, the left remains so attached to this romantic notion that the court is there to protect individual liberties, even if it doesn't do it most of the time. And so they don't necessarily like where the conservative court is going, but they will protect its ability to take us there to the death and the right is of course excited with what it's been able to do.

Mark: I just want to qualify that a bit, just as there's a meme on the left of, But Gorsuch, that is as a criticism of, well, we'll put up Trump because he gave us Gorsuch and Kavanaugh and Barrett. I don't think, there's a problem for the popular critique, popular constitutionalist critique of judicial supremacy, which is posed by Roe and Obergefell. Now these are not, I don't think, bones thrown to the progressives, these are significant accomplishments of the progressive agenda. Now it would have been better to accomplish them through political means, but I don't think they can be dismissed as insignificant.

Larry: Well, all I would say is Roe, to me, is part of the Warren Court, it's tail end, it's true the Chief Justice himself was no longer there, but the court doesn't turn exactly on a dime. And I agree with you on a Obergfell Phil, it's a pretty significant win, but does it weigh against the trashing of voting rights? Just this huge long list. And again, for me, this is not an issue of left, right. The way I think about it actually is the court just needs to be a bit more closely connected to where the actual country is and have a little bit less say, but to the extent that they were trying to understand how the court has had its power grow that way, what I'm saying, I think is that one way or another, the right has been delighted with what it's doing with rare exceptions and the left has been willing to continue to defend it because despite what it's doing, although with rare exceptions.

Rick: Okay. So we're going to turn to solutions, but before we get to the big solutions that you all have discussed, let’s consider a middle solution. Some people advocate that the problems with the Supreme Court could be fixed at least to some extent through tweaks or maybe substantial changes to the confirmation process itself. Is that naive to think that there are changes we can make in the confirmation process that would help to diminish the political politicization of the court overall?

Mark: I can't imagine any. The confirmation process exists as it does because politicians have created it in that way, and they've had incentives to do that, which you have to change their incentives in order to get them to change the confirmation process. There are individual things that might have some effect. This is a fantasy at the moment, but it would be very interesting to see what the confirmation process would look like if president Biden were to nominate a, let's say, sitting Senator to a vacancy on the Supreme

29

Court. Would it look the same as it does now? But I don't think system systematically there's a way of changing the confirmation process without some prior changes in political, as I've called them, incentives.

Ganesh: I think this goes back to where we started, which is if the problems with the court that we have identified are a function of broader sociological trends, like polarization, partisan alignments, and so on, and then as Mark mentioned, the narrower political incentives that specific members of the Senate who are themselves influenced by the broader social trends, then the small technical fixes are highly unlikely, I think, to address the broad social trends. You either have to harness those features themselves, or you have to have much bigger structural reforms.

Rick: Okay. Well let's talk about the structure reforms. So putting aside the question of feasibility at the particular moment, which structural fixes for the Supreme Court do you each think would be most desirable assuming it was political will? We can take it in order? Let's go Larry, Mark, Ganesh.

Larry: Sure, I'll go with the variation of something, both Mark and Ganesh talked about, which was a proposal put forth by Roger Crampton and Paul Carrington back in 2005, I think. And they would have a new justice appointed once every two years with the nine most junior ones sitting at any given time, the other ones stay on the court, they can sit on lower court case, they can do all sorts of things. So I think it's pretty clearly constitutional. I do think it solves most of the problems. So it actually is implemented right away as soon as you begin appointing new justices, the nine most junior ones sit. I think once there are regular ones every year, two things happen, which is the likelihood of the court being way out of whack with where the rest of the political system diminishes, if doesn't disappear, and the consequences for any single appointment also actually go down because there's going to be another one in two years.

Larry: It actually mimics most of the rest of the world. When they set up their constitutional courts, they gave them the same power, but they structured the judiciary to prevent what happens here. So they have limited terms that are staggered. They have requirements for super majority, I would do that. Go back to the filibuster actually for getting an appointment through and so on. And they have not had the kinds of problems we have systematically in this country. So I think there are good reasons to think that the political dynamic is itself partly, at least as concerns the court, a product of this structure and if we could create a more sensible structure, we could fix the politics, at least as regards to the court to some extent.

Mark: So I actually signed on to the Crampton Carrington proposal when it was initially circulated. I do think there are design detail issues that need to be worked out. It's actually, it's not clearly, put it more affirmatively, I think it's really hard to say that once this is implemented, and you get a new justice appointed, the most senior justice currently sitting, who I guess is Justice Thomas, no longer sits on the Supreme Court as part of the panels of the court. I think it's really hard to justify that. So you still have the problem of, the actual outline of it takes three or four appointments before the composition of the

30 court actually changes under current circumstances. I also think, as Ganesh suggested, there probably should be, as there are in other places, limitations on post service work that the former justices could do. But those are design details, I think the idea of term limits is the easiest thing to do. And it is, I want to stress, the most popular.

Mark: I quoted this survey of 70 to 80%. There are these draft things from the US Congress, from the constitution center with a progressive constitution and a conservative constitution. And both of them have 18 year term limits written into their amended constitution. So there seems to be, of the things that could be done, not feasibility, but the most support for term limits.

Ganesh: So I'm partial to the two proposals that Dan Epps and I have proposed, which is why we proposed them. I think the reasons why that I'll just shake out here is I'm pretty skeptical of the 18 year term limits plan because I'm not convinced that it will solve the problems, I think that there's a good chance that it will make the political problems worse with the Supreme Court. And that it doesn't actually hit at the problems that we are trying to solve today as compared maybe to the problems that folks were thinking about when the proposals were first outlined. So I tend to be a little bit more skeptical of the 18 year term for that reason. I think that court expansion is, as I said in my remarks, if your interest is in the structural democracy questions, and you're very concerned about a court causing further problems in those areas, like on voting rights for example, or you just have policy preferences that align with Democrats and you would like a court to do that, I think that is probably the best approach for you to adopt.

Ganesh: But I think the reason why we proposed the two that we did is that we were really trying to find ways to lower the temperature of appointments, but in a way that understands that you cannot get rid of the deeper polarization problems that we face right now. And so that what we have to do is try to harness those dynamics in order to disarm them in the case of the balanced bench or a water down this importance so significantly that it is very, very difficult to try to gain the system through the courts, which is what the lottery approach tries to do.

Larry Bernstein: This is Larry Bernstein, I have a question. What we haven't talked about is something more radical, maybe ending judicial supremacy in its entirety. I guess my two-part question is, if we did get rid of judicial supremacy, how would you do it? And is that constitutional? The constitution doesn't seem to mention judicial supremacy at all, that was just an internal judicial decision from Marbury. But why not just get judicial supremacy? Why not have Congress pass a law that says that acts that are considered unconstitutional could be reversed with two thirds of the vote of the House and the Senate and confirmed by the president and that would be that? Have you considered radically reducing judicial power as a means of dealing with this problem?

Mark: So, this is Mark, I'll jump in on this. I think there there's a good case to be made for developing some sort of what's called another nation's override power, a power in Congress to override decisions of the court, interpreting the constitution where Congress disagrees with the interpretation. There are many

31 serious design issues associated with an override power. Just for myself, for example, I would want it to apply only to decisions invalidating national legislation, not decisions invalidating state and local laws. But again, those are design questions. I think that it would be interesting to develop the argument that you could implement an override power simply by statutes in the face of an argument that I think, content of which is obvious, that is that although judicial supremacy isn't written into the constitution, it has becomes such a settled component of our constitutional practice, that it has constitutional stature of a sort that can be displaced only by an amendment. But an override power is an interesting, I think, Judge Bork actually proposed it at one point, the Canadians have it. It's an interesting idea.

Larry: This is Larry, let me jump in. So again, we want to distinguish judicial review from judicial supremacy, right? So judicial review, I don't think any of us are talking about getting rid of, which is the idea that if a case comes in front of the court and a law is presented to the court to enforce, and they think the law is unconstitutional, they should not enforce it in the case. The question is what sway should that decision have in other contexts in other cases? And you can do that formally with something like an override. As I say, judicial supremacy as an accepted doctrine across the political board is a really pretty recent idea. And so we managed, it's a part of a political dynamic. So, my view is if you structure the court right, and that's one of the things that a staggered term with regular turnover can do, and you have the capacity for political pushback by the tools that are available in the constitution, you have effectively taken care of that problem, you don't need it.

Larry: It's a political dynamic, not so much as, think of an equilibrium point and the court will always go as far as it thinks it can. Right now it's gone so far because the ideology of supremacy has basically persuaded pretty much everybody else that whatever the court says you're just stuck with, and that's not the case. So I think creating any sort of formal device would actually seem to me, that would be really hard. I think that might need a constitutional amendment at this point. In fact, I'm sure it would. Whereas just restoring devices that were already there in the constitution, the capacity to control the makeup of the court, its size, its jurisdiction, its budget and so on, are tools that worked throughout American history and would work reasonably well, I think would work well enough.

Larry: We have de-legitimated them, that has been what the selling of the ideology of supremacy literally is. And that's what changed. So I would just change that back. Mark, I am curious about something you said earlier, which is, I'm not quite sure why nothing in the constitution says that if you've been appointed to a court, you have to be on that court. So I don't see any problem with adding justices and saying to, say, Clarence Thomas, you now no longer sit and decide the panel cases in the Supreme Court, you can sit on lower courts, you can play all the other roles Georgia judges do, but you don't have a right to sit in and decide Supreme Court cases in particular. And a lot of the other proposals-

Mark: So, this gets into the weeds, but my view has two parts. One is the simpler version is that the commission that Clarence Thomas has says that he's a Justice of the Supreme Court. And the commissions under the Carrington Crampton proposal or these other things I would say, justices of the Supreme Court for 18 years and then blah, blah, blah. And so there'd be a difference in the formal commission. The second thing is there is an argument based on how two aspects of the appointments

32 clauses work, that people can be appointed only as justices of the Supreme Court, that they can't have this twofold appointment. I don't think it's a great argument, but it is around.

Larry: And I'm the first one, I think, the holding of Stuart v. Laird was that you can actually change those commissions.

Ganesh: I would just add there's a simple fix to this, which we use in our lottery approach, which is you just appoint every Court of Appeals judge also to be an Associate Justice and then you redefine what the job of an Associate Justice is, which is, the job of an Associate Justice is to hear Court of Appeals cases and some set of Associate Justices will occasionally get to hear final appeals at the Supreme Court level.

Larry: I think that would work, but it wouldn't solve Mark's transition point. He's making a transition point about the judges who were appointed before. Changing the commissions of the ones who were already appointed.

Ganesh: Right. So I think, I agree that there's a question there and none of these things are constitutionally bullet proof, but I think there's a constitutionally plausible argument that you could make that you could even say the commission that he has as Associate Justice of the Supreme Court remains. It is just that the Supreme Court now is an institution that has 180 people and it hears cases in a very different way than the current court does. We've just changed the institution, your commission remains as it is. Your role has just changed because of the contextual factors.

Ganesh: I just want to add a couple of things to the question and the conversation for those who may not know. One is that to the point of Congress overturning court decisions, and Mark mentioned in Canada and in other countries, there is an ability of Congress, the legislature to do that. That is possible also in the United States, in statutory cases and it would not require a constitutional amendment for Congress to overturn court decisions in statutory cases, simply because Congress writes statutes and can rewrite them as they will. It's just that over a number of decades, Congress has not taken the opportunity to override Supreme Court decisions in statutory cases as frequently as it once did. And there are proposals including one that I have to create an institutionalized mechanism for Congress to do that kind of review on a fast track basis. That would be one more narrow way to try to rebalance some of the power shifts that have happened.

Ganesh: Another proposal that has been talked about quite frequently, on Twitter and in other places recently is the idea of jurisdiction stripping, which is basically saying that a court is not allowed to hear a certain case, does not have jurisdiction over a set of issues. And there's different proposals for how this might work, whether it's stripping the Supreme Court of jurisdiction to hear cases or stripping all federal courts to hear cases. But for those who are interested in this, and they usually pitch it around this exact question of this is a way to disempower the court to prevent it from hearing cases in the first place. There are a lot of implementation questions that have not fully been worked out, including for example,

33 how to deal with extremely complex statutes, where there may be individual rights holders who want to enforce their rights against, say, an administration that chooses not to do so. And there hasn't been as much work on how exactly in practice that approach might work. So I think there are proposals there that still require more work before they would be live.

Larry Bernstein: Larry Bernstein, another question for you. In contrast with the United States, the United Kingdom has a court system, but it doesn't have a constitution and Parliament reigns differently than Congress here. I'm wondering recently with regard to Brexit, the court for the first time really took an aggressive role as to undermining the desires of Parliament and they were quickly slapped. I'm wondering how you think about relative judicial power in the US and in the UK, particularly given that the system we created was designed with the UK in mind. Larry, can I start with you on that one? Because as I read your book this last week, you highlighted the approach that the UK has in helping to design our own system.

Larry: Well, we weren't copying their system by any stretch. It was our background experience, so it influenced what we did. But there's a, really, it's theoretical, but it's a profound difference between the UK and the US. The notion in the UK is that sovereignty is in Parliament because of the way it represents. Whereas that was never the theory in the US, we're based on the idea of popular sovereignty, sovereignty is actually in the community at large, and that changes all of that. So the reason Parliament was supreme over any court was because it was actually the sovereign, whereas in the US the people are sovereign, and that's why the people are sovereign over any court and all of the branches of government are just our agents. And that's what's so peculiar about the idea of judicial supremacy is it basically says, we the principles have nevertheless surrendered to one of our agents final say over what is in fact the most important part of our laws.

Larry: And that question wouldn't get framed that way. In the UK, you have the additional problem, which Brexit presumably will solve going forward. But the UK, as a member of the EU supposedly had also shifted sovereignty to EU laws that had to be enforced even over domestic laws, and that was the big controversy and who would have an interpretation over that. So you ended up with some notion that there might be a principle of judicial review in England for the first time as a result of its membership in the EU.

Larry Bernstein: Okay. With that, why don't we head to notes of optimism as it relates to the Supreme Court as a part of our conclusion. Larry, how would you optimistically review-

Larry: [Laughter] I find it really hard to be optimistic about the Supreme Court right now. I think it'll be really interesting to see how it plays out. And the one thing I would say is new justices usually take a few years to get their feet under them. So you've seen that with, even with Kavanaugh and even Gorsuch on an occasional vote, is they get more and more confident, they will become more and more whatever they're going to become. The amount of scrutiny that goes into choosing judges today compared to when Blackman was chosen or Brennan was chosen is so great. But I think we can count on them all being fairly true to what we expect in terms of their ideologies. And so it's going to be interesting only as

34 long as the country remains closely divided politically, there was huge space for the court to just take off because there's no capacity really for pushback.

Larry: All the instances where you got pushback were instances where actually you had unified control in the political branches that was very much at odds with where the court was and when the court tried to push its ideology, it got pushed back. We're not looking at that situation. So there is really a lot of space for the court to continue to expand its power and I would expect them to do that, and I would expect them to do it in a very conservative direction.

Mark: So my note of optimism is only that the issue of court reform has been placed on the political agenda in a way that I think makes it likely to be a sustained part of the discussion going forward. What that basically means is that while Republicans in the past have focused very closely on what the courts could do and Democrats were not as concerned about it, now Democrats are concerned and something might happen as a result of that.

Ganesh: My note of optimism is similar to Mark's and it's that court reform is being discussed and debated in a way that it hasn't been in a very, very long time and with a lot more seriousness than it has been in a long time. And that's an awareness that's tied to an awareness of some of the major structural problems in our Democratic system that have to be addressed at a structural level at core issues on voting and related topics. And I think that that is a major positive, because those have been problems for a long time, but they have not received the attention that they're receiving now.

Larry Bernstein: Okay. With that, that ends today's session. I just want to give a plug for next week speakers. We will have Stanford law professor Michael McConnell speak about his new book, The President Who Would Not Be King: Executive Power under the Constitution. Andrea Kramer will talk about gender bias. Andrew Hussey will be speaking from Paris about his book, The French Intifada, as well as Ayaan Hirsi Ali will talk about the recent beheading of the French schoolteacher in suburban Paris. And then finally, Maria, Sobolewska will discuss her book on Brexit. All right, with that, that ends today's session. I want to thank our speakers so much for their participation in a very interesting discussion, and my co-host Rick Banks as well. Thank you so much, you're welcome to disconnect.

PART 4 OF 4 ENDS [02:01:19]

35