New Ideas About How Judges Think

Political scientists and legal scholars don’t necessarily have the same perspectives when it comes to the study of how judges make decisions. Legal scholars tend to take a more internal view of judging, in which constitutions, statutes, regulations, and precedents drive how cases are decided. Political scientists usually take an external view of the drivers of judicial decisions — such as a judge’s ideology, political influences, and group dynamics within the courts. These views have remained generally discrete within the academy, with neither field borrowing much from the other in order to deepen an understanding of how judges work.

To bridge the divide, a leading political scientist and a leading legal scholar came together to collaborate and learn from one another. More than a decade later, and in collaboration with several other colleagues from their fields, their efforts have resulted in the publication of a treatise that integrates both perspectives: Judicial Decision-Making: A Coursebook, written by Barry Friedman, the Jacob D. Fuchsberg Professor of Law and Affiliated Professor of Politics at New York University (NYU) School of Law; Margaret (Maggie) H. Lemos, the Robert G. Seaks LLB ’34 Professor of Law at Duke Law School; Andrew D. Martin, chancellor of Washington University in St. Louis and a professor of political science and law; Tom S. Clark, the Charles Howard Candler Professor of Political Science at Emory University; Anna Harvey, professor of politics and affiliated professor of data science and law at NYU; and Allison Orr Larsen, Alfred W. & Mary I.W. Lee Professor of Law and associate dean for research and faculty development at William and Mary Law School.

The casebook is designed as a teaching tool for a range of audiences — from undergraduates to law and graduate students to practitioners in the political and legal professions. It tackles such topics as whether the identity of the judge matters in deciding a case, how different types of lawyers and litigants shape the work of judges, how judges follow or defy the decisions of higher courts, how judges bargain with one another on multimember courts, how judges get and keep their jobs, and how the judicial branch interacts with the other branches of government and the general public. The authors hope it could also serve as a treatise for anyone interested in learning about the influences on judging.

In spring 2022, three of the book’s coauthors — Friedman, a leading authority on constitutional law, policing, criminal procedure, and the federal courts; Lemos, a scholar of constitutional law, legal institutions, and procedure; and Martin, whose expertise includes judicial politics, quantitative political methodology, empirical legal studies, and applied statistics — discussed why they wrote the book and some of their findings with David F. Levi, director of the Bolch Judicial Institute at Duke Law School and Judicature’s publisher. Their conversation, edited lightly for length and clarity, follows.

---

David F. Levi: Thank you all very much for making time to discuss your excellent book, Judicial Decision-Making: A Coursebook. Let’s start with why you decided to write a book on judicial decision-making, and what makes it different from other books in the field?

Barry Friedman: This book began as a joint venture of Andrew’s and mine. Long ago, we met and decided to spend a week at NYU where we taught each other some of the basics of our respective disciplines as best we could. We felt that each discipline had an enormous amount to learn from the other, but nobody was really trying to do the hard work to go beyond a superficial conversation. That is how the book was launched.

Andrew Martin: Leading up to that time Barry and I spent together in New York about 15 years ago, lots of purported interdisciplinary conversations between political science and legal academics occurred that weren’t actually very substantive. The existing conversations lacked real engagement. We both viewed that as a missed opportunity. And so Barry and I began, just the two of us, spending time learning from each other’s perspective, and ultimately this book was born from that.

Barry, of course, knows the law curriculum. I know a little bit about it, but not much. I knew what was being taught to political science students, and the real dearth in available teaching materials was clear to us. There weren’t, frankly, enough good political science texts that taught law in a meaningful way, and within the legal academy there hadn’t been real engagement with a broad swath of the social science literature. Our goal was to bring these two things together. It took us many years to complete, but we finally got the project over the finish line.

Levi: I think you’ve succeeded rather well with this idea that you have these two perspectives — one internal, one external — in discussion with one another. I am not aware of anything else quite like this. I’m wondering: Judges often write about judicial decision-making, and it’s notable that you’re writing about judges but don’t have a judge within your authorial group. Did you consider whether you should have a panel or a coauthor who would actually be the guinea pig?

Maggie Lemos: We had endless conversations about coauthors over the years, in part just because it took us so long to finish the book that it was always a moving target. We ended up with a group of well-informed legal scholars and political scientists. We’ve gotten terrific feedback from judges in the judicial master’s program at Duke. I’ve taught chapters from the book to that group, and they were wonderful sounding boards to bounce things off. They gave us lots of good insights from their own experience, and it also was really useful to see what resonated with them and what didn’t.

We found a lot of interest among judges who are teaching classes like this. Not many judges teach, but there are a handful. Anthony Scirica (a senior judge on the U.S. Court of Appeals for the Third Circuit) uses our book in teaching at the University of Pennsylvania, and we’ve heard from some other judges who are interested because it’s really hard to find any kind of materials that talk about courts and judges in a sophisticated way, in a way that’s not just one caricature or the other. Like many coursebooks, ours is one that I don’t think anyone could possibly teach in one semester, and so you can pick and choose different kinds of focus. The folks who are teaching — including the judges — probably teach wildly different classes, but that was one of the things we wanted to do with the book. We wanted it to be used in lots of different ways by lots of different people across different disciplines.

Levi: That’s an interesting point you just made. I was going to ask you who the audience is for the book. It seems that you can tailor your selections from the book depending on whether you’re teaching undergraduates or law students or political science students or others. Is that a fair way to put it?

Lemos: Yes, that’s right. The three of us have taught it to different audiences. Barry and I have both taught classes with law students, with some political science undergraduates in the room. It was really interesting, in my experience, to have political science students who didn’t know law yet but who were learning different kinds of things about courts. I also use parts of the book in my civil procedure class, teaching only law students. But other folks use it to teach grad students.

Martin: You’re right. This is a book that was written for many audiences, and you’ve got to choose carefully what you want to focus on based on what that audience looks like.

Friedman: I just want to make a couple quick points.

First, just for the record, I’ve taught it to all kinds of students in the room at one time: LLMs, JSDs, PhDs, undergraduates, law students. One thing I noticed in those experiences is that the very best students often are the undergraduates who work much harder and are very excited to be dealing with law.

And second, I want to stress that we didn’t write the book just as a teaching vehicle. We also hoped to define the field a bit, because some social scientists look at courts just as institutions, and some legal folks consider the social science as just politics. We wanted to write something that showed the depth and breadth of what the field could accomplish and the kinds of questions it could take up. And so, though it’s great to sell books and it’s fun to teach out of the book, we aspire as much to have it be used and read by people who work in these fields to understand the sorts of questions they can tackle. And, as we’ve said, it also is a great learning vehicle for interested lawyers, judges, and social scientists.

Levi: You’ve implemented your vision wonderfully well. We have a lot of decision-makers in our society. Managers of baseball teams make decisions all the time, as do CEOs. So what’s different about judicial decision-making? What makes it worthy of study and what distinguishes it from other ways in which decisions are made?

Friedman: What’s fascinating about judicial decision-making is that it’s both deeply structured with great formality in the way that cases are presented and decided, but, when it comes down to it, it also involves an enormous amount of discretion. That odd mixture of formal structure and discretion makes it pretty fascinating.

Martin: I view this from the political scientist’s perspective: What makes it different to be a decision-maker in a legislature or a decision-maker in the executive branch? The biggest difference I see, and what makes judicial decision-making so complicated, is that it’s not just about disposition. Judges have to explain publicly what they are doing and why. Providing a rationale is really important. Imagine members of Congress all having to agree on why they were voting for a particular piece of legislation. Nothing would ever get done. The fact that judges have to state reasons is significant. CEOs sometimes have to state reasons, too, but often those reasons are just pretext. When judges state reasons, those reasons have consequences for the world, for future litigants, and for judges down the line. To me, that’s what makes it really different from just about every other decision-making context.

Lemos: All decision-makers have to operate within a particular context and institutional structure. We all believe those things shape how they behave and what kind of decisions they make. So one of the things we really were focused on in the book is just trying to trace the different things that structure and influence judicial decision-making. The book walks through all of those structural influences: how judges are situated vis-à-vis other judges, whether they’re on collegial courts or deciding by themselves, how they’re situated in the judicial hierarchy, what their relationship is with the other branches, and what their relationship is with the public.

Levi: I think each year about 1 million cases are filed in the state court systems, about 100,000 in the federal courts, and about 75 are decided by the Supreme Court. There are lots of state cases, many of them quite repetitive and dealing with collection-type matters. Compared to the whole body of cases, momentous constitutional cases are a small percentage even though they may capture the lion’s share of our attention. I’m wondering, given that general landscape, where your focus is. The field of inquiry could be very broad, because judges make lots of decisions — from whether to give a party an extension of time or to give a defendant a certain sentence, all the way to deciding Brown v. Board of Education. So what did you focus on within that large universe of decision-making?

Friedman: What you just described, David, very often defines the nature of study done by academics. It’s easy to focus on the Supreme Court because there are nine justices in one court and just 75 or 80 cases (and plenty of time for vacation). It’s hard to focus on state trial courts. It’s hard to get the material, to generalize, and to do the work. I think we suffer from some of the challenges, but we really wanted to try to address as broad a swath as we could. I don’t know that we were successful but we tried.

I wish we’d done more with private law because I actually think it might be very interesting as contrasted with public law, but so much more of the judicial decision-making scholarship actually is around public law. There’s only so much you could do with the countless state trial courts, but we tried to generalize — including, for example, in the judicial selection chapter, where we are trying to take account of all the different ways that judges get selected.

Lemos: Part of what we’re trying to do is remind the reader that the book is called Judicial Decision-Making, but lots of different kinds of judges serve on lots of different kinds of courts. And as I said earlier, if it matters how the decision-maker is situated institutionally, then we shouldn’t think of decision-making at the U.S. Supreme Court as the same thing as decision-making in a state trial court. So we include lots of reminders of that throughout the book. Where I think we all wish we could have done more is to actually dig into what decision-making looks like and how it is, in fact, different in these various sorts of settings.

Writing a coursebook, I think for us, was a different kind of writing experience. We were trying to bring together what’s already out there and use it to teach folks and maybe raise some new questions.

Levi: Andrew, it seems that political scientists have been somewhat focused on the Supreme Court. I don’t know if that’s a criticism necessarily. It’s just where they often put their energy, recognizing that there is a huge amount of interest in the Supreme Court. It strikes me as potentially misleading to the extent that people reading that scholarship would think that you could generalize what the Supreme Court did to what, say, a U.S. district court would do. They seem like different jobs.

Martin: I’d say that’s a somewhat fair statement. It is certainly the case that most of the early positive political science looking at courts did focus on the U.S. Supreme Court, including almost all of my early work. But over the last couple of decades there’s been a renaissance, particularly as more and more datasets have become available. Some folks are doing some great work on intermediate appellate courts and state supreme courts as well as lots of interesting work on the federal district courts, some on state trial courts, too. High-quality empirical studies of the lower courts have blossomed over the last 20 years or so. One of the things that we hoped to accomplish in our book is to highlight some of that scholarship.

Levi: Consistent with your point, I’m aware that it’s standard for law firms to obtain an incredible amount of detailed information about what trial judges have done in the past, how long they take to reach a decision, and how often they decide in favor of this kind of litigant or that kind of litigant.

In your first chapter, you start with Brown v. Board of Education. That must have involved a decision of sorts regarding starting there and showcasing your insights in the context of one case. Why did you select Brown and what did you hope to accomplish in that chapter?

Friedman: Starting the book with Brown was one of our most difficult decisions. It’s an old case, and there’s certainly a claim that whatever’s newest is most interesting. But I probably dug in more than anyone on Brown because I felt that with Brown we had public information that was unavailable for other decisions, which would better allow us to show how the process works.

You have to persuade an audience pretty quickly that the kinds of things you’re going to talk about — bargaining among the justices or the reactions in a fight between lower courts and higher courts — actually happen. With Brown we had all of that. And so not only is it an icon, but also we just felt that we had the materials that we needed to tell the story.

Levi: In the next chapter, you address what you call the limits of law, and you say that there are roughly two kinds of judging: legal formalism and legal realism. Can you explain what you mean by those terms?

Lemos: That chapter is in a sense a setup and some training ground, particularly for undergraduates who might not have as much familiarity with the law. I have stopped teaching that chapter, at least most of it, to law students because they don’t need quite the same introduction. But it’s a way of setting up these two opposing views in a caricatured way: legal formalism being, roughly, the idea that there is this thing called law, which spits out answers. The other side is a caricature of the social science view of legal realism: that the thing called law doesn’t exist at all, and it’s just whatever the judge ate for breakfast. Our goal in laying out those two extremes is really just to knock them down and say, of course, it’s neither of these things.

Actually, some easy cases do have answers that get spit out. Then there are lots of other cases where there are boundaries, and the law creates a framework, but within that framework there are still open questions and discretion.

Levi: I’m wondering how easy cases, hard cases, legalism, formalism, realism — how does that relate to how a judge might decide cases involving constitutional law? In that field, we now have originalism and living constitutionalism. We have political process theory. We have pragmatism. That field has become very theoretical in the last 30 years or so, and a lot of attention has been focused on that. How do those theories fit with this broader template you have here?

Lemos: One of the things that was both exciting about working on this book — and also frustrating when framing it — was that the book is completely trans-substantive. It’s about the decision-making process in all sorts of cases, and, as you emphasized earlier, David, in all kinds of courts.

In statutory interpretation and in constitutional law, there are methodological questions about what theory the judge should apply. What method does the judge purport to wed herself to? Can that be constraining? How do those kinds of questions interact with the tradeoff between mechanical decision-making and pure discretion? We do touch on those issues, especially as they relate to questions about constraint and about the epistemic challenges of judging, but we don’t get into them all that deeply since each could be (and in many cases already is) a book in itself.

Levi: A proponent of originalism would say, perhaps, that originalism minimizes discretion. This tends to address some of the criticisms of pragmatism or legal realism, as if to say, “if you don’t want my breakfast to determine the outcome, then it’s best to undertake a historical inquiry into what other people thought.” There’s obviously a rejoinder to that.

I want to ask you about artificial intelligence in judging. If you went to Estonia or you went to China, you very well might have an algorithm giving you a judgment, rather than a judge or a human being. Assuming we could design the algorithm so that it wasn’t biased and that it was totally transparent — even that it never has and never will eat breakfast — how would that work? Did you consider AI judging? Twelve years ago, when you started this project, it didn’t exist. But now it’s a big deal.

Martin: I’ve never had a conversation about AI judging with Maggie or Barry or, for that matter, with any of our other coauthors. But thinking about it really quickly, the big question turns on, what do you want to train the AI to do? Probably the easiest thing to do would be to train an AI or algorithm to do what judges have done historically, replicating the way in which judges have made decisions in the past. That, of course, raises all sorts of interesting questions about whether those were the right decisions and whether they were done in a bias-free manner and the like. I think the evidence in this book would suggest that hasn’t been the case historically. So the question remains: What do you want to train an AI judge to do? And that seems to me to be a pretty hard question.

Levi: I think it’s an interesting question, though, because this is actually happening. Today, all kinds of judicial decision-making models don’t really involve judges. They are dispute resolution techniques, and they are computerized, and sometimes they have this AI capacity so that you have a computer that can learn over time. It’s not just a decision tree. The AI judging is more complicated because apparently there’s a capacity for judgment built into the machine, and it can weigh certain kinds of evidence against other kinds of evidence. An AI judge would have instant recall of everything in the record, and then might also someday down the road be able to integrate and apply every scholarly article in the area.

What do you think is your ideal judge? What do you think judicial decision-making should look like?

Lemos: This interacts in an interesting way with the point that Andrew made at the beginning of this conversation about rationale and disposition. In an AI world, I don’t know what the rationale looks like. I’m guessing it is gleaned from the dispositions. I think it could be a really interesting foil for us in thinking about real live human judges.

Levi: I think what we’re calling easy cases, maybe you could clarify this for me, are probably ones that could ultimately be decided by a machine pretty quickly and pretty cheaply. The big appeal of AI judging is that it increases access to justice. If we can make it easy for people to go online and get a decision at a very low cost, then that’s a win for the system. Of the million cases in state courts, many of them can be quickly decided once the issue comes into focus. What do you all consider to be an easy case? What do you consider to be a hard case?

Lemos: The way you just put it isn’t far off. We think of an easy case as one in which reasonable minds wouldn’t disagree, where the law machine really does spit out an answer that pretty much anyone who has the right inputs would reach. And a hard case is one where that’s not true. I think it’s Judge Harry Edwards who adds a distinction between very hard and hard cases: The hard cases are those with a couple of plausible answers, but if you work really hard and think about it, one is better than the others. And then the very hard cases are the ones where the law just leaves you in equipoise or something like it.

AI could be a useful teaching foil for the book. I think it does have to do with the capacity to decide things in a mechanical way.

Levi: In chapter three, we enter your wheelhouse, Andrew. This is the chapter in which you look at the identity of the judge and how the judge’s preferences, biases, or demographic characteristics may affect the judge’s decisions. It would be helpful to our readership if you would explain. Is this the attitudinal model or something else?

Martin: This is not the attitudinal model. One purpose of this chapter is to teach some of the basics of research design on how to conduct an empirical study. A lot of what I’d call elementary aspects are in here. The attitudinal model, though, is a theoretical model developed by Jeffrey Segal (Stony Brook University professor of political science) and Harold Spaeth (the late Michigan State University professor of political science), which basically says everything that happens on the U.S. Supreme Court is all about politics.

And frankly, from my perspective, that model is a bit of a straw person. The attitudinal model can’t really come to terms with the Court even today, where 40 or 50 percent of the cases are decided unanimously. And so I don’t believe there’s anybody on the planet, maybe Jeff Segal — Harold Spaeth is no longer with us — who actually believes that the attitudinal model is the right explanatory model for what happens on the U.S. Supreme Court. In a sense, the attitudinal model is the political science baseline and everything else builds upon it. The idea that judges have preferences and that those preferences sometimes enter decision-making is consistent with the attitudinal model, but it’s also consistent with other positive models of judging.

Levi: Could you just explain how the image of judges as politicians in robes might fit with other models?

Martin: Sure. The attitudinal model is narrowly defined. It actually removes all sorts of interesting parts of the broader decision-making process. It’s the idea that the justices are nothing more than politicians in robes. That’s consistent with other models of decision-making, including any rational choice model in which policy-seeking behavior is something that the justices are trying to do. The attitudinal model has frankly been caricatured both within political science and within the legal academy. From both a theoretical and empirical perspective, the literature has moved on quite substantially over the last 30 or 40 years since the theory was first published.

Levi: Many judges, at least judges that I know, don’t think of themselves as being politically driven. In fact, they’ll often say, “I don’t agree with half the decisions that I reach. They’re not my preference.” Whereas that’s not the kind of statement that political scientists, at least some of them, would think was worthy of any kind of credence. They just have a different view. They may think a judge’s saying they are “applying the law” is just self-deception.

Martin: If the statement you’ve made about what judges think were true, there’d be implications in the data. The fact is, though, that we see these very strong statistical signals that there’s something going on in judging — at least some judging — other than just calling balls and strikes. There are oftentimes correlations between judicial decisions and things like the party of the appointing president or, for state court judges, partisan elections.

Levi: In one sense, that wouldn’t be surprising, because, for example, if you think about a state court judicial election, sometimes judges will present themselves during the election as being, for example, tough on crime. Or maybe there will be another issue where they’ll say they’re sympathetic to certain kinds of claims or where the governors or the appointing authority will be very open about what they’re looking for.

So then you’d expect certain kinds of judges who would act consistently with the way they presented themselves to the electorate. In fact, you might even go further and say that they should be consistent because that’s — at least in an elective system — the premise of their election. However, that’s not the model of judging that many of us have.

In chapter four, you look at how the docket is constructed. The chapter is not directly about judicial decision-making but about how these decisions come to judges. And I thought that was a very subtle, sophisticated point. What were you trying to accomplish in that chapter?

Friedman: I had nothing to do with that chapter, but it’s my favorite chapter in the book. That’s all Maggie’s work, and I will defer to her. But I do want to say that the chapter exemplifies everything that’s right about the book. It takes not just one literature but a set of literatures, and puts them together to allow us to understand that decisions and the directionality of those decisions are not accidental but constructed by the system itself.

Lemos: David, I think the motivation goes back to your question about different kinds of decision-makers and how judges are distinctive. Certainly one of the ways that judges are distinctive is that they don’t get to choose what decisions they make. To some extent the same is true of other decision-makers, but it’s easy to forget that aspect of judicial decision-making and to focus on the individual judges who are making the calls — as opposed to thinking of them as part of a system that brings cases to them in a non-random way, and that shapes how the issues look when they finally get to the judges.

One of the things we try to draw out in that chapter is that a legal question isn’t presented in a vacuum. It’s presented in the context of a case with people and a certain set of facts. That context really matters, and it isn’t up to the judges. The judges are not powerless, though; they can do various things that make it more or less likely that certain cases will come to them, and they have tools for avoiding decisions in cases they want to avoid, and so on. So one of the things we try to tease out is how judges are both reliant on the system to bring them cases and also are players in that system themselves.

Martin: This chapter also includes a really important methodological point, which is that because a given court’s cases aren’t randomly sampled from some broader universe, any statistical analysis you do has to take the mechanism of the whole system into account. Otherwise, you can come to erroneous conclusions. That idea of selection and selection bias runs all the way through this chapter, and it is important to think about that when looking at the empirical studies.

Levi: We could go into much more depth on many of these points, but I thought it would be good to at least say one or two things about each of the remaining chapters because they’re all so interesting. Chapter five is about judicial capacity and about how judges decide extremely complex matters. You also talk about the effects of limited resources and the use of adjuncts and law clerks and that sort of thing. What would you say are the takeaways for this chapter?

Friedman: Judging is not an exact science. It happens in conditions of uncertainty and it happens in conditions of limited resources. And so the question is, how does that affect the kind of outcomes that we get and what can we expect given those conditions? Caseloads limit the amount of time judges can spend on decisions; uncertainty limits what a judge could accomplish even if there was unlimited time. The effects of this are seen widely in the judicial system in terms of things like deference rules (that allow judges to defer to another decisionmaker), or motions practice that puts what might be determinative burdens on one party or another.

Levi: In chapter six, you call it judging in a hierarchical system, and one of the questions you ask is, “Well, why do judges follow precedent?” What’s the answer to that? Do you think they’re scared of being reversed?

Lemos: That question seems like such a silly question to many students at first: “Well, of course they follow precedent because they’re scared of being reversed.” But we think at the end of the day that is an incomplete and unsatisfying answer given how unlikely any given decision is to be appealed and reversed and how ambiguous the effect of a reversal is on the judge. It’s an example of the kind of question that seems easy at first and then, when you dig into it, it actually turns out to be really interesting and complicated. We discuss both the naïve legal model that says judges follow precedent, and political models that see lower courts trying to buck precedent, with higher courts trying to get them to comply. We try to show the impact in terms of the various rules that do more (or less) to constrain lower court discretion.

Levi: Then you look at decision-making on multimember courts. You say a lot about how judicial decision-making in that context is quite a bit different than when you have a single judge. Andrew, can you talk about the multimember courts from a political science perspective?

Martin: Yes. I think the big takeaway from this chapter is that whenever you move from a single decision-maker to multiple decision-makers, it introduces intergroup dynamics. It turns out empirically that there’s lots of evidence that this stuff is going on and ultimately that these dynamics shape how cases are resolved. We had exactly that behind-the-scenes bargaining in Brown, and we know it affected the remedy in the case, which allowed segregation to persist far too long. You also see the oddity of what we call “pivotal concurrences,” in which some justices or judges will join a majority opinion and then write something else in a concurrence. How are we supposed to understand what is the law in such a case, if the concurring judge’s vote was essential to the majority?

Levi: In the next chapter you look at separation of powers and its connection to statutory interpretation and constitutional interpretation. You also talk about some of the responses that might be made to court decisions by the other branches — for example, a legislative body engaging in court packing. Again, what do you hope to achieve in that chapter?

Friedman: So this chapter is particularly apt in the world in which we live because it starts by talking about how, in statutory interpretation cases, the judicial word is not necessarily the last word. There’s evidence that courts are aware of that and there’s an elaborate tango that occurs between legislative bodies and courts on matters that they both care about. We go on to point out that even though the legislature can’t really overturn a constitutional decision, legislative power still can affect the content of constitutional outcomes.

We are living at a time when there’s lots of discussion about legislative bodies taking action against courts — things like jurisdiction-stripping or court-packing — as a way of curbing the discretion of the court itself. There’s obviously a debate about whether this sort of action is or would be appropriate. But one of the points we make in the book is that even if this doesn’t happen, the threat of it happening can constrain a court. We point to examples of times when this has been so. And even though we’ve not seen much of this lately at the federal level, we certainly have seen it at the state level.

Levi: You then look at judicial selection, and we talked about this a little bit already. There are a lot of different methods of judicial selection, both state and federal. You talk about judicial independence and judicial accountability. So how do you see the relationship between judicial independence and the different selection and tenure models?

Martin: Our ultimate conclusion is that there’s no perfect system — that every method for judicial selection and retention that we have has good aspects and bad aspects, but each of them is going to provide different levels of accountability, different levels of independence. And again, you can empirically see some of the implications of the types of judges and the type of decisions that you get depending on the system you have. One of the more depressing examples of this is studies that show criminal sentencing to be harsher among judges facing re-election.

Levi: In the last chapter, you look at public opinion and judicial decision-making and ask, “How should we feel about the possible impact of public opinion on judicial decisions?” So, how should we feel?

Martin: I think we’re puzzled, right, because it is the case that there is an impact. And whether this is shown through the deep historical analysis that Barry has done or some of the empirical social science, we do find, at the margin, the Court responding to public opinion — sometimes in the short term, sometimes in the long term. It’s also the case that we know that the courts can affect public opinion as well, and we’ve seen some examples of that. But the thing we still haven’t figured out is what the mechanism is. How does this work? Is it judges who are sitting there reading the newspaper and changing their opinions? Probably not. Is it just through selection and retention? Is it just through judicial turnover? Is it through the nascent threat to punish a court should it step out of line? The mechanism is something we don’t quite understand yet, and I think it remains an empirical puzzle.

Levi: I think you’re right. It’s so hard to know. We have such a big country and so many different points of view.

I will just end by saying this is such a comprehensive and interesting book, and it’s more than a course book. It’s also something of a treatise. It’s lively and written with such a nice style and a certain panache, and obviously you’ve had a lot of fun doing it. It’s an important contribution to the field. I think it will find many audiences. Thank you for talking with us about it.