Richard Susskind and Jonathan Zittrain recently sat down for a conversation on online courts, lessons from the COVID-19 crisis, and how we might move forward. The conversation was moderated by David B. Wilkins, faculty director of the Harvard Law School Center on the Legal Profession.
David B. Wilkins: I want to thank Richard Susskind and Jonathan Zittrain for being here for this special Speaker’s Corner of The Practice. Our goal here is to have a conversation. This all started because Richard wrote Online Courts and the Future of Justice (Oxford University Press, 2019), which was, as he always is, prescient, coming out when remote courts were just theory. Fast-forward to today: online courts are everywhere, and he’s written the lead story for this issue of The Practice about this current moment. While Richard’s article speaks for itself, today I want to engage with both of you on the big macro issues at play. And I’m going to start by highlighting a news story that you both have probably already seen about Texas holding its first online jury trial. Now, it’s not really a jury trial. It turned out to be one of those summary jury trials they do to try to get cases to settle. As others have reported, there were issues with the rollout, to say the least. But it is being looked at as one possibility in an era in which we don’t know when we will be able to open physical courts, and one in which there’s a huge backlog of cases building up.
Do we adjourn cases, often incarcerating people indefinitely, or do we find other methods of conducting serious criminal trials? – Richard Susskind
Richard, I want to start with you because when you spoke in my legal profession class a few months ago, right as the pandemic was hitting, you said, “Well, the one thing I’m not saying is that we should be holding online jury trials.” And yet, here we are. I wonder what you think about that today.
Richard Susskind: I did indeed say that. I remain very cautious about criminal hearings being conducted remotely. Around the time we spoke about it, the advice I gave here in the U.K. was that if the virus is with us for a few months, rather than years, I would prefer that we hung on until we could return to traditional physical hearings, especially for serious crime. What has transpired, of course, is that the virus is unlikely to disappear in a couple of months. It’s going to be with us for some time yet. And so we’re in the land of undesirable alternatives. Do we adjourn cases, often incarcerating people indefinitely, or do we find other methods of conducting serious criminal trials?
There is a great public debate in England about having trials without juries and, for instance, using professional judges instead of juries. My concern with the virtual trials remains, however, which is that we have no empirical evidence at all of how this will work. And so we are faced with the possibility of introducing, as I think people would rather glibly say, trial by virtual jury without having any sense of whether or not this will preserve the values and principles that, after all, are so dear to us in holding jury trials in the first place. I have a conceptual difficulty with the idea of simply dropping the current process into Zoom, as it were. The real power of technology should be to enable us to meet our fundamental objectives in running jury trials in new ways—using the power of technology to deliver court service differently and not simply to automate how it is we currently conduct these trials.
Wilkins: I would love to get your reaction, Jonathan, about Texas in particular, because it wasn’t a full-blown jury trial. It wasn’t a criminal trial at all. There was a kind of public part of the jury selection on YouTube, but then there was a closed Dropbox for evidence. So, there is a technical part at play. What I’m really mostly interested in, though, is how you think this fits in with broader ideas of the legitimacy of the court system and “having your day” in court.
Jonathan Zittrain: Let me start by saying thank you, Richard, for your farsightedness, not just on this topic but on many others touching the profession over the years. I love how your books go from being a speculative crystal ball to a pretty good history of what ended up happening in the profession!
I think with respect to COVID and the court system, there’s a Maslow’s hierarchy of legal needs and we are hitting the base of the pyramid. We need functional decisions made by decision makers of stature and insularity from pressure. I’m describing the idealized version of judges. And if the judicial system has to function over a telephone—or whatever technology is needed to present evidence and render decisions—half a loaf is better than none if the alternative is just to, say, confine a criminal suspect indefinitely. But that is, of course, no way to rearchitect a justice system. It’s just a way to try to keep things going as much as possible with baling wire and twine.
One way to look at the outputs of the court system are decisions. How do we most efficiently achieve those outputs? – Jonathan Zittrain
Richard, you ask the provocative question in your work of whether court is a place or a service, and I think you suggest there’s no right or wrong answer to that question—but that if there’s a spectrum between the two, it’s more of a service than it is a place. I agree with you, but it’s worth asking, as even within the university this spring, as we were hastily evacuated like out of Pompeii with the ashes chasing us from our physical locations, what it is that we draw from the place in service of the service. Because many courthouses—the physical structures—are older and have not been recently remodeled. They somehow evoke an almost churchlike atmosphere; they seem built to instill a certain sense of awe as a litigant, as an attorney, maybe even as a judge.
Now, how much of that happens at all anymore is its own question, and I don’t mean to valorize the status quo. But there is another question in the extent that that “place-ness” was intended to—and in its best days, I would argue, did—serve that function of taking people and saying, “This is something special. This is something that has gravity to it. Life and death or other momentous decisions are going to be made and people’s lives are going to be affected and the power of the state is going to be exercised.” The question would be how or what pieces of that should be carried over to online forms.
Maybe that sense of appropriate awe can be evoked in Zoom or on the phone or whatever technology we use. I don’t know that I have answers for that, but it does neatly augment your question. One way to look at the outputs of the court system are decisions. How do we most efficiently achieve those outputs? Right now, there is a shortage of timely ones. They take too long, and not enough people get their day in court. On the other hand, are people willing to endure litigation so that if they lose—as one of the parties is likely to lose—they merely feel like they were misunderstood but that otherwise it was a fair fight and they’re ready to move on?
Wilkins: As Richard knows, when he came in to talk to my class back when we were just moving online, one of my students wrote a very interesting response paper (which she is working up into a fuller paper) about the symbolic power of the physical courthouse and the courthouse as an idea. And since we just finished the Supreme Court’s session during which people want to have their say right on the courthouse steps, how do you reconcile the court’s role as place and all that is associated with that with their more instrumental roles of producing resolutions or decisions or outcomes?
Susskind: I think there are two ways of having this discussion. There’s the very practical question of how it is that you sustain a justice system when physical courtrooms are shut. And that does take us into the realms of choosing, as I say, between undesirable alternatives.
But a better line of debate, in light of this massive unscheduled pilot that we’ve been conducting while remote courts have been used in extreme circumstances, is this: What is it that we might want to preserve from this period? And in what circumstances do we think we simply must go back to the physical courtroom? I think it’s better not to be forced into the assumption in our discussion that there have to be remote courts.
For me, the interesting question is—and I’m speaking to judges all the time about this—if you had to start with a blank sheet of paper and engineer a justice system from scratch, knowing what we know about the way in which some cases, hearings, issues, and so forth can be disposed of on a remote basis, would you build an artifice that was very different from the one that was designed in England around 150 years ago and, more distantly, 900 years in the past?
In light of this massive unscheduled pilot that we’ve been conducting: What is it that we might want to preserve from this period? And in what circumstances do we think we simply must go back to the physical courtroom? – Richard Susskind
I tackle this from the point of view of “justice.” Justice is a concept that has always fascinated me, but in this context it also amuses and slightly perplexes me. Both proponents and detractors of online and virtual courts invoke justice in their arguments for and against nonphysical courts. Of course, they do this because they’re invoking different conceptions of justice. In my own work, I’ve tried to identify precisely what conceptions of justice we’d like to see instantiated in a court, whether or not that’s a physical court or a remote court. We want the decisions to be fair. We call that substantive justice. We want the process to be fair. We call that procedural justice. We want our system and individual cases to be transparent. We call this open justice.
We also hanker after access to justice. There is little point in having a court system that’s wonderful for the very small number who can afford it but inaccessible to the great majority. We call this distributive justice. If we regard court service as a social good that we want to be made widely available, as a matter of distributive justice, we should want our system, I think, to be more widely accessible.
A key issue in England, but one that is discussed more widely, is also proportionate justice—relating to questions of cost, time taken, degree of combativeness, and so forth. We should want the process of settlement to be appropriately balanced with such considerations.
Finally, we want sustainable justice. There is no point in developing a whole suite of online services if that would, for example, greatly reduce the number of lawyers and judges coming into and need in the profession, perhaps because online services do not offer the careers they are after.
We should also want our system not just to resolve disputes between individuals but to settle and to clarify the law for the future. This is enforceable justice.
Given all this, I find myself pulled in a whole bundle of directions. I think if I was a policy maker, distributive justice trumps almost all else, assuming the decisions are fair and the process is fair. But when I put on my hat as a former jurisprudence teacher in Oxford, it worries me that this may not provide the platform that concerned your student, or it could perhaps whittle away, as some worry, at the effectiveness of the common-law system.
At the end of the day, we need to have a public debate on this issue that goes beyond people saying, in a slightly superficial way, that remote courts can’t deliver justice. I think we have to unpack these different conceptions of justice. To start, let’s engage students and the finest young minds not just to think about these issues but to go out into the world and be involved in changing how it is we currently resolve disputes.
Wilkins: Jonathan, I’m going to throw yet another grenade into the mix, because another thing that has changed since Richard wrote his book is that we are now in a deep discussion about race and justice, and the implications of the justice system on racial justice. At the Berkman Klein Center, you have been thinking about this issue, and its intersection with technology, with your algorithms project around bail and sentencing. I wondered now, as we are in this critical moment in which people are asking to rethink the justice system in fundamental ways, what does the intersection of technology, the courts, and the justice system tell us?
Asking a model to make predictions about what a judge would do based on past decisions risks freezing a status quo by simply trying to predict what a judge would do if they were still there. – Jonathan Zittrain
Zittrain: That is a vital, difficult, and completely urgent, if indeed not long overdue, question to take up. One way to approach answering it is by responding to some of what Richard put on the table in his last comment. Richard spoke of the ways in which the purpose of litigation maybe isn’t only for the parties involved but also for the system as a whole. This holds especially true in a common-law system, which is looking to evolve through a series of reasoned decisions. There is additionally a concern, however, that sophisticated commercial parties opt out of the system altogether. They’re just going for their own private alternative dispute resolution, possibly in a simulacrum of a courtroom with an American flag and a bench and all that if they want it, or they can teleconference. They just pick what they want. But then you’re left with a series of decisions that are somewhat untethered to the state of the law. If it’s like, “Look, we’re just going to let the arbitrators decide,” you end up with, in a distributive-justice context, judges who are paid by the litigants in one system that is totally market driven, and then you have another system that only contains the people who can’t afford to opt out of it on the civil side. And in this scenario, common law no longer evolves with some of the most difficult and interesting problems put before it.
I think the bridge from the arbitration example to your question about race is in some of the issues we’re hitting as we look at the introduction of AI to various pieces of the judging process. Asking a model to make predictions about what a judge would do based on past decisions or about what somebody would do in the case of, for instance, a defendant who wants to post bail, risks freezing a status quo by simply trying to predict what a judge would do if they were still there.
Now, if you fast-forward 10 years, you get a system trying to predict what a judge 10 years ago would have done—there is no evolution. And if you take a system that is as hobbled, even broken, as ours is right now and make that brokenness more efficient by using machine learning to make predictions about what the broken system would do—to set it and forget it—well, that seems like the worst of all worlds.
Of course, if you’ve got powerful legal help on your side, you may have the wherewithal to ask for a more artisanal judgment and plead your case. But for the rest, it will be more a matter of cranking the case through the system. I know that Richard has, at times, proposed trying to sift between cases that might be more readily amenable to disposition more quickly. This would be kind of an ADR-and-triage approach within the system. A court officer could help the parties understand their respective rights. And sometimes judges will do it with a status conference and say, “Do you really want to take this to trial (hint, hint)?” And that might be a way of devoting the scarce resources of the decisional system, and the human aspect of that, to the pieces that most cry out for it. I think that’s very promising as a way of trying to sift.
And, just going back to this place-or-service dichotomy, I can’t help but put on a different hat too—that of a librarian, wondering if local libraries might be an interesting kind of hub-and-spoke system where you can have designated rooms and spaces within those libraries, which are far more numerous than courthouses that could produce a hybrid system. (Editor’s note: For more on the evolution of law libraries specifically, see the March–April 2019 issue of The Practice.)
What’s become very apparent in the last few months is how little data is available about our past system. – Richard Susskind
As Richard says, what are the elements of the system that we want to carry over? What are the ones that we are so anxious to reboot? And here is our opportunity to do it, given the absence of confidence in the system, the crisis that it has hit around race, and the exigencies of COVID, and the combination of these factors, which are both cataclysmic and awful. This could be the stimulus for ultimately rejecting and rebuilding what we currently experience as the status quo.
Wilkins: I can’t help but think about my dearly departed colleague, Professor Frank Sander, who had said 40 years ago that we “should fit the forum to the fuss.” He had this idea of the multidoor courthouse in which we would channel different kinds of disputes to different kinds of places, which is a very appealing idea, until it runs up against questions of equity that are prevalent today. (Editor’s note: For more on Sander and alternative dispute resolution, see “Separating the People from the Problem.”)
Richard, I’d love to get your thoughts on this. How do we possibly measure whether we’ve gotten it right? How do we evaluate what we are doing and if it is working and producing the desired outcomes?
Susskind: I want to start answering this by stressing how important it is that we manage to capture more data about our court services and systems. What’s become very apparent in the last few months is how little data is available about our past system. On asking some quite basic questions about volumes and throughputs and timings and so forth, it is clear we just simply have not been capturing that data in most jurisdictions. One thing I always say is if we are reconceptualizing, redesigning, and rebuilding our systems, please, please, please bake into these new processes and systems the ability to draw out as much data as we can so that we can slice and dice and analyze and evaluate the performance of these systems.
It’s also quite interesting to ask the question, even if we have lots of data, how indeed is it that we measure the performance of our court services? There’s an inclination amongst many that we should assess our court’s performance on the basis of some kind of user-satisfaction assessments. You’ll see in my book I point to the poster child in the world of online courts, which is the Civil Resolution Tribunal in British Columbia. And if you have a look at the surveys they’ve done of their users, there’s a very high degree of satisfaction. And I often hear judges saying that one of the tests of the justice system—I think Jonathan was hinting at this—is that when parties do not succeed, if they nonetheless felt they were treated fairly, then that’s an indication of a system that’s working well.
Ignorance is not bliss, but I’m mindful of a version of Goodhart’s law: when a measure becomes a target, it ceases to become a good measure. – Jonathan Zittrain
But one of the difficulties here is that we shouldn’t confuse user satisfaction with genuine insight into the substantive operations of the courts. A court service is, as economists would say, a “credence good”—the recipient of the service does not actually have the knowledge to evaluate its worth fully, comprehensively, and reliably. So, you may well have individuals emerging from an online court process in which they were treated very warmly and with empathy, and in which they feel they’ve managed to be heard, but a lawyer worth his or her salt might say that substantively there were three or four major legal errors made there. So did the court do well? In this sense, it isn’t sufficient simply to undertake market research into the levels of satisfaction of users of online courts as compared with conventional courts. It seems to me that we need what I call a justice test where we can measure the performance of our courts relative to the different conceptions of justice I laid out earlier.
We need to begin under the headings of each conception of justice to gather data and to start evolving tests. None of this will be perfect, but as Voltaire says, “The best is the enemy of the good.” Just because we cannot put in place a perfect set of justice tests doesn’t mean we shouldn’t try to capture as much data as possible and have some rough-and-ready measures and mechanisms for evaluating progress.
Wilkins: Well, you’ve warmed the heart of an academic research center director—and one who does empirical research on the legal profession. But in addition to understanding a new way of evaluating courts, we also need a new way of training lawyers to be the participants in the court system. And you said as part of your sustainability idea, Richard, we need to have something that will be sufficiently attractive and teachable that we can train lawyers to do it. Jonathan, as a law professor, how do you think about teaching students to operate in this new world? What is it that we should be teaching them?
Zittrain: It calls to mind Roger Fisher and Bob Mnookin’s framework around negotiation, where I think the colloquial understanding of what it means to be a skilled negotiator would be about how to make sure any surplus on the table goes to you and not to the other person—the “art of the deal.” And of course, what you learn, if you’re learning from the Fisher–Mnookin tradition, is there are ways to make the surplus bigger. Are there tradeoffs that would in many instances make you want to come back and interact with the counterparty again? Are there things that are sustainable in a bigger picture as opposed to just in that one moment? As we train our law students in the art of argument and pushing boundaries and working the levers of power that otherwise seem obscure and complicated, we must ask: How will they adapt if we can surface the kinds of metrics that Richard was talking about?
For our students, we need to give them space and context in which to reflect on what it means to be a lawyer, not just as a zealous advocate for your client but as a professional and as an officer of a court and a system that you would ideally have reason to want to uphold and serve. – Jonathan Zittrain
Ignorance is not bliss, but I’m mindful of a version of Goodhart’s law: when a measure becomes a target, it ceases to become a good measure. And lawyers are really good at the kind of gaming those measures-as-targets invites, which is not an argument against measures but a caution of how hard it is to use them. It’s even a philosophical question, to me, when I think about the way in which a typical jury delivers a typical verdict where they speak the verdict and by design offer no reasons. How much would we want to know about how those juries worked everything through and how much all the work of the litigators was reflected? And tying it back to AI, researchers and then commercial litigation support vendors may soon have enough weight to the models that they could try out different arguments and weightings of evidence and everything before going to trial. And with a virtual jury—as they do already, where if you have enough money, you convene these Potemkin juries before you go into the real court—you can see how to hone your argument. But that all then starts to have a feel of artifice about it.
For our students, we need to give them space and context in which to reflect on what it means to be a lawyer, not just as a zealous advocate for your client but as a professional and as an officer of a court and a system that you would ideally have reason to want to uphold and serve. And how do you help create a system worthy of the new generation of law students coming in to serve it—and, in turn, students worthy of that system? That seems to me the big task of a law school, and it is an artisanal one. Because when you look at what it means to disaggregate legal education—that is, can we just make it so that people can pass a bar exam and then go into court and know which rule of civil procedure to invoke at the right time—none of those sinews seem to me as easily touched upon in that kind of quantifiable way.
Wilkins: Richard, to bring us full circle, how much of that ethos of what it means to be a lawyer as a member of a public profession—with courts as temples, even if it’s a construct—ought to be preserved?
Susskind: The question I often ask is, What are we training young lawyers to become? My feeling is that most law schools are generating 20th-century rather than 21st-century lawyers. So what’s the 21st-century lawyer, it might be asked? But I say, that’s the wrong question! The question we should be asking is how in the future will we be solving problems to which lawyers are currently the best answer? How in the future will we be solving problems to which our court systems are currently the best answer? We should look at the outcomes we want of our lawyers and courts and ask whether or not, in a digital society, these might be delivered in very different ways.
I was once asked to speak to 2,000 neurosurgeons, and they asked me to be controversial. My opening line was, “Patients don’t want neurosurgeons.” Gasps in the audience. I said, “Patients want health.” I went on: “Isn’t it the case that, 50 years from now, we’ll be looking back and thinking, ‘It’s unbelievable we used to cut bodies open,’ because the future of health care is noninvasive.” To suppose the alternative notion, that the future of health care is robotic surgery, is to perpetuate the notion of surgery as the end in itself. And I worry a great deal that we similarly view law and lawyers, courts and process, customs and traditions, as intrinsically valuable, as ends in themselves. Whereas I argue, in fact, I think they’re instrumental values. They are there to produce a set of outcomes, to uphold a set of values and principles, that we haven’t yet articulated well, perhaps because we been focused more on the process and less on the outcome.
The question we should be asking is how in the future will we be solving problems to which lawyers are currently the best answer? – Richard Susskind
I see a world emerging quite rapidly, in which our machines are becoming increasingly capable, and in which both legal service and court service will be dominated in a variety of ways, not necessarily by entirely autonomous systems but by systems that take on more and more tasks that we used to think required human beings. So the question I ask every student to ask is, “In your future, what are you going to do? Are you going to compete with these systems or build these systems?” By competing with these systems, you would acknowledge a role for online courts, AI, and all the rest of it, but you might feel there’s something intrinsically important in the traditional way that lawyers work. Not only that but you would say that human lawyers can empathize, can relate to their clients at a personal level, can show creativity and imagination, and can exercise judgment in ways that will never be replicated by a machine. And therefore you might conclude that you are going to devote your career, essentially, to competing with the systems that are emerging into the legal world. And my response to that is, “I can’t say you’re wrong, but I wish you well with that, because almost all the research of technologists and economists is pushing in a different direction.”
But where’s the real excitement here? I think it is building tomorrow’s justice system. And I believe that the mission of lawyers of the future will be to become the main architects and engineers of the next generation of legal and court service. This is not just some superficial ideas about technology. This work takes us to the heart of what it is to live in a society where we can take pride in justice under the law. And what it is as a human being to serve others in a way that is focused less on perpetuating the old ways of working and more on the outcomes of those we seek to help.
Wilkins: All I say is that I can’t imagine two better people to help us to rethink these deep questions at this pivotal moment in time. I agree that this is an important moment. It’s been building for some time, as both Richard and Jonathan have helped us to see. The combination of the recent events has put it now into sharp relief, and the challenge is for all of us to try to rethink those deep questions about what it means to be human, to be fair, and to be just. I can’t thank you enough for giving our readers an insight into both the complexity of those questions and also some very, very helpful ways to think about them. Thank you very much.
Richard Susskind is president of the Society for Computers and Law, technology adviser to the Lord Chief Justice of England and Wales, and chair of the advisory board and visiting professor at the Oxford Internet Institute.
Jonathan Zittrain is the George Bemis Professor of International Law at Harvard Law School, vice dean for library and information resources, and faculty director of the Berkman Klein Center for Internet and Society.
David B. Wilkins is the Lester Kissel Professor of Law at Harvard Law School, vice dean for Global Initiatives on the Legal Profession, and faculty director of the Center on the Legal Profession.