Russ Roberts

Gillian Hadfield on Law and Rules For a Flat World

EconTalk Episode with Gillian Hadfield
Hosted by Russ Roberts
Rob Reich on Foundations and P... This is a Job for Super Regula...

Rules%20for%20a%20Flat%20World.jpg Law professor Gillian Hadfield of the University of Southern California and author of Rules for a Flat World talks with EconTalk host Russ Roberts about the ideas in her book for regulating the digital future. Hadfield suggests the competitive provision of regulation with government oversight as a way to improve the flexibility and effectiveness of regulation in the dynamic digital world we are living in.

Size:30.9 MB
Right-click or Option-click, and select "Save Link/Target As MP3.

Readings and Links related to this podcast episode

Related Readings
This week's guest: This week's focus: Additional ideas and people mentioned in this podcast episode: A few more readings and background resources: A few more EconTalk podcast episodes:


Podcast Episode Highlights

Intro. [Recording date: August 28, 2017.]

Russ Roberts: Gillian Hadfield is the author of Rules for a Flat World: Why Humans Invented Law and How to Reinvent It for a Complex Global Economy, and that book is the subject of today's conversation.... So, let's start with the central idea of the book. Why do we need to reinvent law for our complex global economy? What's missing with what we have now?

Gillian Hadfield: Well, the reason we need to reinvent is first that it's critical infrastructure. So, it's not just something we layer on top of our market economy to maybe respond to market failures; but it's the basic platform on which humans have always built sets of rules; they've also built whatever relationships are creating value in that society. And the legal infrastructure we have today, which is what everybody thinks of when we say "law,"--law made by governments, nation states, changes when you cross boundaries, jurisdictional boundaries--mostly done through closed professions and public production. That's not working very well. It's too expensive; it's too complex. And, surprisingly, it's just not solving the kinds of problems that we face in light of globalization and technology and our increasingly diverse global environment. So, the reason we need to fix that is it's not doing what it's supposed to be doing. Most ordinarily people can't access it all. And even for large global corporations who can afford the best of the best, what they will tell you is it's not really doing what they need to do to operate in a global, complex environment. And it's certainly not well-adapted to managing changes that are coming down the pike, in terms of artificial intelligence, new technologies, and so on.

Russ Roberts: Explain a little bit by what you mean by 'legal infrastructure,' which is really the center of the book. I think most people probably misunderstand to mean simply things like how the courts work. But you mean something much broader. Talk about that.

Gillian Hadfield: Yeah. So, I use the term 'legal infrastructure' because I really do want people thinking in terms of infrastructure--the thing that you're building on and the thing that you are taking for granted. It's shared capital. And I mean something more than the legal system, as you say. So, most people, if you say, 'law,' to them, they will think, as you say, 'Laws--rules--are made by legislatures and regulatory agencies. They are adjudicated in public courts, state courts; enforced by the state.' And we talk about different legal systems like common law systems and civil code systems, Continental, Anglo-American. And what I mean by legal infrastructure is that stuff--that stuff we're currently using the make and enforce our rules. But I also mean the practices, the norms, the beliefs that structure it, the cost of accessing it; the ways in which our markets for legal services and legal rules are organized and how well they work. So, you know, not just, you know, the fact that you would get legal advice, you can only get legal advice in America from a licensed attorney in the state, but I'm also interested in: What's that lawyer going to tell you? How are they educated? What will they bring, what human capital will they bring to resolving the problem you want to address with them? How much will it cost? And so on. So, I'm trying to think of the entire mix of both the design stuff, and the emergent stuff, in the same way that our traffic infrastructure is defined, both by the roads we've built and how the traffic develops over time and how people use it and what people do with it.


Russ Roberts: Long-time listeners here will recognize the distinction that Hayek made between law and legislation. For Hayek, legislation was what governments and legislative bodies did. Law, he reserved that term for the expectations people had about the rules of the game. And you are making a slightly different set of distinctions. And I actually--I want to try to get to one in a little bit that I found extremely useful and thought-provoking. But the other point I want to get to that you make is that we should be thinking about the rules by which we make the rules. Or the laws by which we make the laws. And how we're so used to the current system--which is centuries old here in America--that we can't even imagine really breaking out of that box. And, talk about why that's important and why that should be a focus.

Gillian Hadfield: Yes. And I find this especially true with economists and political scientists--we define 'law' as the stuff we know now, the system we know now. Like you say: governments producing it and states enforcing it. That's not even actually a very good description of what, how a lot of our rules are created and enforced today. But it's what dominates people's thinking, so we talk about changing the law; we talk about what will we lobby Congress to do, or what regulations will we propose or oppose. And the reason it's so critical to think about, to recognize that's not the only way to make law--law is the way that, actually my colleague, Barry Weingast, at Stanford and I have been thinking about this for quite some time: We really define law as, you know, a set of classifications that a community has landed on. So, everybody knows that everybody knows that it's okay to not do what you said you were going to do if you didn't shake hands on the deal. Everybody knows that everybody knows that shaking hands on the deal is what's required to make it enforceable. And, there's lots of ways you can do that, for the community to do that. Law is this set of rules that we have about how we do that. And, the reason we need to think about what the rules are for making the rules--who can make them, how can they make them, what can they look like, how are they enforced--is that that's the production system for our legal infrastructure, and we spend a lot of time talking about what rules we think we should have, and almost none talking about, 'How well is the production system working?' So, I want to get people thinking about: What's the process by which all that legal infrastructure is being produced? That's what's affecting behavior on the ground. And we need to think about that production system, just like economists would think about production in other settings.

Russ Roberts: It's a great point. And I think the way most economists think about is: The regulatory structure that we have is whatever the legislature produces. And then, whatever the administrative state then implements. And maybe it's not always enforced--that's, I think, the economists' contribution is to remind people that not all laws are enforced. Which is a very important point, because I think it's often assumed that, 'Well, it's against the law, so it doesn't happen.' But, of course, it does happen often, especially in other countries but even in the United States. Legal--enforcing laws is expensive, and a lot of times there's no stomach for enforcing certain laws. And so, we end up with a very complex set of rules, some of which are on the books--say, the one that Don Boudreaux likes to talk about when he's been on here is, the speed limit is not 55 [miles per hour]. You can go 57, for sure. In some places you can go 61. 62, you start to get into some risk of being in trouble for speeding. But, there is this complex interaction between what's legislated and then the social norms that develop around it.

Gillian Hadfield: Right. No, and so this phrase that a lot of people use--Roscoe Pound introduced the idea of this difference between law in the books and law on the ground. And, part of talking about legal infrastructure rather than legal systems is to emphasize that what we're really interested in is what's happening on the ground. As you say: What are people enforcing? What does your lawyer tell you is the actual rule? Like, you just said with speed limits, right? Well, the actual rule is not 55. Although they are allowed to ticket you for going 56. But, everybody knows that everybody knows that that's not really not what's likely to happen; and therefore you can evaluate your risks of, you know, heading down the highway at 62. Our legal production system is a very, very closed system that lawyers are particularly prone, I think, to the view--even though they are well aware that things are different on the ground--we have this strong view that, you know, you have made law when you put it down on a piece of paper. And, as you point out, that's not true even in advanced settings like the United States. And it's really not true in countries where we recognize that they need more rule of law, to support better economic growth and other goals. So, yeah: absolutely. We need to be really thinking about what's happening on the ground


Russ Roberts: And I want to come back to this closed aspect of the system, because that's really interesting. But I first want to read a quote about this norms, rules, interaction that I think you make extremely well and I hadn't thought about before. It goes like this--it's a quote from the book:

Social norms are not the product of an enterprise. They are organic. They emerge and adapt to processes that may be hidden and hard to control. There may be social entrepreneurs who try to establish a social norm by preaching from pulpits or billboard or blog, for example. But there is no single recognized process that everyone understands is the way that a norm becomes the norm. It is when norms become the subject of a deliberate project intended definitively to designate a particular set of rules, including the rules for making the rules, as law, that social norms can become law.
There's a lot there. It's a very rich quote. But, the idea--you know, I talk a lot on this program about emergent order and about how norms emerge and how they are hard to control. And you can urge people to do something and it may make no impact whatsoever. Or, there can be a social movement that changes the way people look at something like smoking. Or attitudes of various kinds. But your point that, until it is when norms become the subject of a deliberate project--meaning, they are not emergent in the same way: there is some explicit process that people recognize as the rules for making the rules. Which is not true of social norms. Social norms--the idea of wearing a hat; the idea of greeting someone you don't know well by their first name--all these changes that have taken place in our culture that are constantly changing--the way language and social interaction evolves--those are under no one's control. And everyone understands they are under no one's control. And yet, they also understand that they change, and that you have to pay attention. But, law, the point you are trying to make here, which I think is an interesting--it's a different distinction than the Hayekian one--you are arguing that law is when a social norm has an established, public, objective process, transparent process, for how the rules get made. Legislation being one of those, but maybe not the only one.

Gillian Hadfield: Right. Right. And it may be transparent, but not necessarily so. I think what's really critical--

Russ Roberts: Not so transparent--

Gillian Hadfield: Right. So, I like to say, 'Look, it's when'--and, the concept that really matters here is that of common knowledge. And that is: So, in the emergent setting, in the social norms setting, we have common knowledge that everybody knows--that everybody knows that men don't wear dresses to work. And, when we are talking about law, we are saying: The source of the rule is an identifiable institution or person, and everybody knows that everybody knows that that is the institution or person. It's the king. It's the counsel of elders. It's the legislature. It's the community council. It's what we voted on at the camp the night before. And what's critical, if it really is law in the sense that I'm using it, everybody knows everybody knows that that's now the rule. And the capacity for deliberate content, which is I really think the important thing to emphasize about law--not the enforcement mechanism. Let's assume the enforcement mechanism is exactly the same as in the social norms. Which is to say, ordinary people deciding, 'Hey, I'm not playing with you because you don't follow the rules.' What's really critical about law and the reason it's such a critical human invention is that when you have transferred from the world of 'all your norms are just emergent' to the capacity to now say, 'Hey, everybody, we're going to have a meeting. We're going to choose the rule. It's now okay for men to wear dresses to work.' Or, for people to get out of contracts when there's been a significant change in circumstances. Whatever it is. We can go to sleep one night with the old rule. We can wake up in the morning and have confidence that everybody in the community knows that that's the new rule and that that's the basis on which we'll be moving forward. And that's really critical in terms of the capacity to adapt to changing circumstances, changing knowledge, complexity.


Russ Roberts: Just as an aside, and I don't think you talk about this in the book: Even when you, even when everyone wakes up the next day and knows there's a new rule that says, 'Contracts are not 100% enforceable, because there may be,' let's say, acts of nature that everybody agrees on. That, that's now known to be a reason that you don't have to uphold your end of the contract. It's still going to be a question of what an act of nature is. And I just want to mention that in Law, Legislation, and Liberty, Hayek argues that what judges should do when they interpret a dispute over acts of nature, is try to figure out what people had in mind; and what their social norms were underlying that legislated contractual requirement. Because, inevitably there is still uncertainty. There is no way, and there can never be a way, that any law, rule, legislation can cover every case. And so, almost by definition--not 'almost.' By definition there has to be interpretation at some point along the way. And, a mechanism for who does the interpreting.

Gillian Hadfield: Absolutely. In fact, I think Weingast and I have been really been emphasizing that we think you start to see the emergence of formal institution that has that capacity. As if to say, when I say, 'Announced of rule,' it's forward looking, as you said. Acts of nature are a valid excuse for not performing your contract. That, that, you know, we can announce that. As you point out, that's going to require interpretation in various settings. There's going to be ambiguity in what's an act of nature. And, it's precisely the pressure and the creation of ambiguity in what the rules are, that we think generates the demand for, and then the production of, institutions that perform that role, where everybody knows, everybody knows, 'That the institution.' We can argue about it, but we've now we've now set up a country system. Or, we've designated a smart member of the community, whoever--an arbitrator--

Russ Roberts: It you could agree on it,

Gillian Hadfield: However: an arbitrator--

Russ Roberts: Yup, if you'd agree on it, an arbitrator.--

Gillian Hadfield: That's right. We can set up whatever the institution is, we know that, when we go to that ambiguity resolution institution, what that institution or person says, that's it. We stop arguing. I may still think it's the wrong answer, right? I hate the fact that I lost the argument. But, I do know that everybody knows that everybody knows that that was the answer, and therefore I can predict that that's how people will go forward. If I don't pay up on my contract after the court had said, 'Sorry: that's not a valid, that's not an act of nature, that's not an excuse of performance,' then I know that the community will say, 'Ah, she breached her contract. She's less reliable as a business partner,' or somebody we're not going to do a deal with. And that ambiguity reduction is absolutely central to the reason for the emergence of law and the function it should be performing.


Russ Roberts: So, I want to probe into that a little bit, because in the middle of the book you spend a lot of time arguing that our current innovations and technology aren't served well by the current set of rules. And it's not enough to just change the rules. We need a better way for how the rules get set and allow more innovation in that process. We'll get to the details of what you propose in the last third of our conversation. But I want to try to set the stage for that and challenge your view a little bit, and try to get you to clarify it a little bit. So, we recently had Benedict Evans on EconTalk talking about autonomous vehicles. You mention them in a number of places in the book. It's pretty clear that the regulatory legal landscape for an autonomous vehicle is uncertain right now. You have people pouring literally billions of dollars into these innovations because they assume that will get "fixed." They assume that we'll "figure out" how to get there from here. And one of your arguments is that the current system doesn't handle these kinds of innovations very well. And what I want to ask you is: Why? I want to push you on why. Because, it dawns on me that part of the reason that it doesn't solve these systems, these problems so well isn't just because our legal system was set up in 1750 or 1850 and comes from England or whatever. It's that these are really complex, unusual, and not easily-defined situations. One example obviously being liability. You talk about this: when the autonomous vehicle hits a pedestrian, whose fault is it? Is it the program or is it the corporation that made it? Etc. So, talk about why those kind of problems aren't served well by our system and why we need a different way to set the rules.

Gillian Hadfield: Well, I think the important thing about--the autonomous vehicle is a great example, because what we're looking at the development of there is, it's not just that it's an autonomous car that used to have a human in the driver's seat and now it's got a computer. What will be [?] development of, of course, is an entire connected system. Those cars are going to have sensors in them; you're going to be talking to your cars; they're going to be talking to weather, they're going to be talking to traffic; they're going to be reading the environment. There's going to be a changing number of autonomous and human-driven cars on cities and streets; it'll go place to place to place with different sets of rules. So, we want to think about that as a complex, adaptive system in that formal sense of complex, adaptive system--meaning, there's lots of interacting agents making lots of decisions, and it's very difficult to know where it will go next. That's a really new type of problem for us to be figuring out how to regulate. So, the critique I'm making about our existing systems we're solving that, is that they are currently dominated by--we have technical experts, of course, who are playing a role. But dominated by the legal infrastructure we have available, which is: Lots and lots of lawyers involved in the development of policy, the drafting of policy, the implementation of policy, advice about legal rules. And we have a particular kind of technology of regulation. Our technology is: smart people will sit down, figure out what the rules should be, write them down somewhere; lawyers will advise about them, people will make decisions. There will be potential enforcement actions, public and private. And penalties imposed for failing to follow the rules. Well, you know, to regulate complex traffic systems with lots of machines in them--maybe we need technology to do that. Maybe we need AI (artificial intelligence) that is actually performing the function of regulation in constant data exchange with what's happening on the roads and maybe the capacity to directly intervene in the way a machine learning algorithm is developing or the way a particular vehicle is performing or sensors are performing at any point in time. And that's the kind of thing that our existing approach to building a regulatory structure just really can't do. So, the reason I'm emphasizing we need to be thinking about our production process for rules and regulation is: Where are we going to get the new ideas for how to regulate? Where are we going to get the new approaches to regulation?--that I think we'll need for a much more complex environment.


Russ Roberts: So, I want to take you down a little thought experiment, then. One of the challenges of the autonomous vehicle is it's really expensive to develop 5 or 6 different sets of roads. So, we start off, perhaps incorrectly, but we generally start with the assumption that autonomous vehicles are going to use the public roads, not private sets of roads developed by Tesla, Uber, Apple, Google, etc. And one way we could regulate this market is to give all those different providers, say, of autonomous transport the authority to set their own rules--knowing that they're in competition with each other: so they are not going to just set rules that make them rich, because if they do that, they are not going to have too many customers because they have competitors. And then, you would argue, if I understand your book correctly, is that the government could impose some meta-rules on those regulations. Or, they could even--which is the main thrust of the last part of your book--we could imagine private rule-makers there weren't Google, Tesla, Apple, etc., that were regulated by government. So, using that setting, perhaps, try to tell us how a different model would work for the production of rules.

Gillian Hadfield: Yeah. So, I want to really distinguish the proposal I'm making, which I refer to as super-regulation or competitive-approved private regulators, as a regulatory option. Not all regulation can work[?] this way; and one of the things I'd like to see economists and regulators working on more is, where could this work? But it's different from self-regulation in the sense of Tesla, Google, Apple coming up with their own rules maybe subject to some government oversight. I really think we need to be looking at the development of basically an industry that's attracting investment and attracting good engineers and more that's dedicated at the problem of figuring out how do you regulate in a complex environment, and being able to solve those services. Now, the fact of the matter is that the Googles and the Teslas and the Apples thinking about this autonomous vehicle world are already in the business. They know they need a regulatory environment. They are already in the business of participating in the building of that regulatory system, partly through lobbying Congress and legislatures and governments around the world. But also through participating in private efforts of standard setting[?]. And, what I'm saying is, we already know we need the private sector much more heavily engaged in the problem-solving of figuring out better regulatory regimes. How can we bring that within a public-law framework that says: Okay, you can figure out how to regulate; but here are the publicly determined, accountable criteria that that regulation has to achieve. So, very simplistically, in the case of the self-driving car, there's a tolerable level of accidents on the road, and congestion levels; and you've got to hit those publicly-determined outcome criteria; but we're not going to tell anybody how they achieve those outcomes. We're going to let private regulators who develop systems that then private companies have to purchase into--we're going to let those private regulators figure out better ways to achieve those, those objectives.

Russ Roberts: Expand on what you mean by a private regulator, because for people who haven't read the book, that might sound like a weird concept. And, you can move away from autonomous vehicles if you want to talk about a different setting to make it clear. But, what you have in mind is sort of a middleman of regulation, that would, as you say, attract investment, charge for its services, and be competitive. It's not like a private version of the EPA [Environmental Protection Agency]. It's one of many monitoring or regulating bodies that would be distinct from the corporate world but also not governmental, that the governmental process would oversee. So, try to explain that a little more clearly.

Gillian Hadfield: Yeah. Sure. So, we're thinking here--it could be a nonprofit entity; it doesn't have to be a for-profit entity. We already have lots of nonprofit entities that produce regulation right now. So, FINRA [Financial Industry Regulatory Authority] regulates broker-dealers. That's a private nonprofit entity operating under the supervision of the SEC [Securities and Exchange Commission]. And the difference between the model I'm talking about and the FINRA model is I'm talking about: Okay, we've got FINRA, but it has a monopoly. So, it's really operating as if it was a government. And what we see is it behaves like it's a government.

Russ Roberts: Yeah.

Gillian Hadfield: It's like it's a government agency. So, what I'm trying to suggest we want to do is we want to say, 'Okay, it's okay, it's appropriate for the political realm to say: here are the criteria for regulating--here are outcome criteria; here are things we want from regulation,' say, in the financial sector. And, you know, there's various ways in which we are going to, as a government now going to verify that our private regulators are achieving those outcomes. We can audit them; we can sample particular cases; we can talk to the regulated entities; we can talk to the consumers or customers, businesses that are dealing with the regulated entity. But the ways in which that's achieved, whether it's achieved through written-down rules or adjudicated ex post, or it's achieved through auditing, or it's achieved through constant data-monitoring--whichever way it's achieved, individual providers of that service can compete for the most effective ways to achieve those objectives. So, the difference here is--so, it's important to recognize, we already have lots of private regulators in the business. They tend to be monopolies. Or they are voluntary, so that companies can choose or choose to or choose not to comply with their standards. But, the--I'm really emphasizing all three pieces of that. So, it's a private regulator. So, that's a profit or nonprofit that's in the business of developing regulatory schemes, regulatory services. It's approved, in the sense that it's licensed by the state--that it's achieving the objectives that are politically set. And, it's competitive. There are multiple, um, private regulators that are competing for the business of providing regulatory services. You have to be licensed to participate in that market. But, once you are achieving those objectives, you can explore, experiment, invent better ways of doing it.


Russ Roberts: So, the analogy that comes to my mind--and you can tell me if this is a good analogy or not--is, kosher food. There are a lot of different agencies that will certify that food is kosher. They have slightly different standards, but on the big things, they don't disagree. No pork, just to take one example; pork's out. So, if a kosher certifying agency says that this product is kosher, barring fraud or an accident, you can be pretty confident as a consumer it doesn't have pork in it. There might be other differences in the level of kosher that is kept by particular agencies. And, consumers learn about what those differences are and choose products accordingly. Those agencies charge the firms for their certification. It's very competitive. It's overseen by the government in some dimension in that if you are caught, if fraud, if you are defrauding your, if you are defrauding kosher consumers, I think kosher consumers can take you to court. Not quite what you have in mind. But it's something like that, I suspect. Is that a fair example?

Gillian Hadfield: Yeah. Yep. That's an example. So, that's a certification regime. We have lots of examples of certification regimes: organic certification. And they are under different levels of government oversight. I don't know enough about the way about the kosher certification system works. Sometimes what happens with the government oversight is the government is still in business saying, 'You have to have the following rules,' or 'You have to--your rules have to be approved by our agency.' So that's how FINRA operates, for example, with the SEC [Securities and Exchange Commission]. And I'd like to see more of the regulation of the regulator focused on the outcomes: Are you achieving the objectives that have been set for this type of regulation and the reason that we're regulating. So the only difference between, but really other than that, the only difference between what you are describing there and what I'm thinking of here is, that's a voluntary regime.

Russ Roberts: Correct.

Gillian Hadfield: Somebody can decide: I'd like to be certified kosher. I'd like to be certified organic, or not. And then there's a way of doing that. So that's solving an information problem in the market. And the difference here is I'm saying, 'Okay, let's take some of the features of that system, and now: Can we use that to think about how we regulate workplace safety? Can we use that to think about how we would regulate the use of machine learning in loan agreements, in loan decisions?' Or autonomous vehicles, as you said. So, it's really taking--there's features of this system in place all over. There's much greater role than I think people appreciate already for private providers of regulatory regimes. In a lot of cases we leave that to the actual regulating entity itself. Which I think is problematic--

Russ Roberts: It's nuts--

Gillian Hadfield: Or, it's voluntary. Or, it's not really supervised. Or, we're not really holding it accountable to publicly-set criteria. And I think this is a critical point. So, I use the example of the right to be forgotten. You know, Europe has introduced a directive or decision out of the European Court of Justice in litigation with Google saying that search--that people have the right to be forgotten by search engines. So, you have to be able to petition the search engine to stop producing old or invalid or false information about you in searches. So, that's a very elaborate right to have created. But governments are not capable of actually implementing that regulation. And so, it's been handed to search engines like Google to do the actually adjudication of that. And I think at some point last year, Google was reporting that they were adjudicating about 500 of these a day. Okay. So, the reason it's being handed to Google is the government can't afford it, doesn't have the technology. I'm quite sure Google is not stepping up to hand them all their access to all their servers--if they can find the stuff that [?] supposed to be in there.

Russ Roberts: It's problematic for a different reason.

Gillian Hadfield: So, it's handed to a private regulator. Now, it's handed to Google to regulate itself. I'd like, you know, for there to be an alternative, where it's, you know, Hadfield, Inc. or Roberts, Inc. or whatever; and that's, 'Hey, I've got a lean, effective way of implementing the objectives of that regulation--people being forgotten on the Internet. And I'm now selling that service; I'm demonstrating to governments around the world that I achieve their objective. And now I'm competing to sell that service to Google.' Google's being told you have to purchase a service like that. And you have to purchase one that political bodies have certified, licensed, achieves the objectives that we're looking for.


Russ Roberts: So, let's talk about a marketing challenge that you have with this idea. Which is that, when I was reading your book, I thought, 'Well is kind of interesting,' and I really liked the idea that if there was this intermediate layer of private regulation, there would be more innovation. Like we were just talking about with the Google example. But my general thought would be--and you can tell me if I'm wrong--that, this is not appealing to, really anybody. For hardcore libertarians--of which I am one, but I am sympathetic to some of what you are pushing for; but most hardcore libertarians are going to say: Oh, well, all you are adding is another layer of regulation; and that just pushes it farther from accountability. And the regulated company will get cozy with the private regulator. It's going to hard for there to be competition. There's going to be economies of scale. And then, that's on the free market side. On the regulatory, pro-whatever you want to call it, pro-government or interventionist side, they are going to say, 'Oh. Well, this is just a mistake. We just need to make government better at regulating. We don't need this.' So, does anybody like this idea? I mean, it's interesting and creative and clever, but does anybody like it?

Gillian Hadfield: It is a challenge. Yeah, yeah. Well, I'm working on it. Yeah. So, the point is that both positions that you described I would say are fantasies, in a complex environment. So, Fantasy Number 1 is that we can have this thing called free markets without rules. Well, there's no such thing as a free market in that sense. All markets are constituted by rules. And nobody would play in a market that had--well, you couldn't even define a market, right?--you can't define a market if you can't define who owns what, what are the criteria under which x is transferred from A to B for a price. Everything we talk about as economists is built on a pretty thick environment of what are taken-for-granted rules about the way the game works. Why don't we see--I think everybody agrees that one of the most important reasons we don't see growth in poor and developing countries at the rates we think there should be, why there were challenges, [?] challenges still for transition countries, is a lack of what now people call 'governance,' or the rule of law. You can't have markets without basic legal infrastructure. So, that's Point Number 1. This is not just about interfering in markets and developing rules to interfere. It's about building the basis of that market. So that's fantasy.

Russ Roberts: So, let me concede that. And I take the point. I think some people use it as a trump card to say, 'Therefore all regulation is required.' And that's obviously not your point. So, what your point is, which I really like, is: There's going to be some regulation. But then you have to make the case that your private regulator is better than the government.

Gillian Hadfield: Right.

Russ Roberts: So, you are adding this other layer. There's going to be uncertainty about how the government enforces--I don't know. It's--I think it's still a tough sell to a free marketer. Or just some of us like it. I kind of like it.

Gillian Hadfield: Okay. Well, I'm getting there.

Russ Roberts: Yeah.

Gillian Hadfield: So, let me just go to Fantasy Number 2, the idea of 'Just government can do it,' and say: Look, government can't. I like government. I think government should be doing lots of stuff. But they are not the type of entity that we think of as well-adapted to innovate. They don't produce technology. They are not going to be in that business. So, I'm partly saying: Look, we need innovative approaches to how to regulate complex things. We need jurisdiction jumping solutions for regulating complex technologies, because it's a global market, it's a global system, it's a global internet, and our existing governments are all territorially based. We need something that's going to jump those boundaries well. And so, the fact that we can--the idea that, 'Oh--well, the private sector can produce a government,' can produce it--I think that was the mistake of the reinventing-government effort that said, 'Well, let's just get government to behave more like the private sector '--I'm[?] saying, 'Well, no. It operates under a totally different set of rules and incentives. For good reason, to achieve different objectives. So, is this layering on? Is this more regulation? Oh, my God, is this, you know, we are going to double, triple, quadruple the amount of regulation that we have? The point here is that I'm aiming at a real shift in what we think of as government regulation. Right now, we have, almost all our regulation through government is command-and-control style: You know, this is the technology you have to use; this is the behavior you have to engage in; this is how often you have to, you know, review your transactions; this is what you have to be able to prove to us. As opposed to outcomes. Right? You can't have more accidents than this in the workplace. You can't have more suspicious transactions than such-and-such in your banking system. So, it's shifting that to the idea that government is getting into the role of establishing the outcome criteria. Those are the politically-determined outcomes that we require. And, getting out of the business of how are we going to produce that. So, that's actually a very different and potentially smaller role for government. It's not a less important role for government. I'm not in the business of trying to kill off governments. This is a different role for government. And I analogize it to the shift from central planning to a regulated market-based economy. You know, we used to have, 40% of the world or more, under a regime where governments decided which apartments to build and what dresses to make for this season and what foods to have on the shelf and how much steel to manufacture. And, you know, that's a system, for, you know, lots of reasons; but I think an important one is economies became so complex, that's very difficult to continue to operate. And, what we have as an alternative is: Okay, government gets out of the business of deciding how much steel to produce; but it does get into the business of making sure that the private providers who are deciding how much steel to produce are doing so through a competitive system. So, we have antitrust law. We have fraud law. We have--we might have rules in different environments that say, 'You have to make it available to poor people at this price,' with a subsidy, or whatever. So, we've changed the role of government. And I don't think that's necessarily more regulation at all. The fantasy on that side is the idea that somehow we can just have less regulation. I think that's just a very unhelpful conversation to be having. We can have differences, political differences, about how much we should try and change the outcome of whatever markets we construct. But the real argument should be about how we are going to produce the basic rules of how those markets operate.


Russ Roberts: So, one of the things I don't think you talk about in the book is Public Choice--in that, the incentives facing the people who currently making the rules are, I assume, somewhat in their favor. Meaning, the reason we don't have your world isn't just because they haven't read their book. They may not like that world for themselves so much, the people who right now are the gatekeepers to the rules. So, for a politician to give up control over the current structure, the way the current structure is set up, may not be in their self-interest. What are your thoughts on that?

Gillian Hadfield: I think that's exactly right. And I do think that's another reason to be thinking about this. One of the reasons I think it could be helpful to focus much more on outcomes rather than on the detailed form of regulation is that it's potentially a much more transparent approach to regulation, in the sense that we're having political debates--politicians are taking political stands--on what should the outcome variables be, what should the metrics be, for deciding what's effective and what's not effective in regulation. Rather than having 2000-page statutes produced, or bills, produced through a very non-transparent process of lobbyists and civil servants and, you know, politicians and other kinds of money and their, the role for corporations that are lobbying about their own regulation and be affecting them. I think that's a very non-transparent process; and I don't think our systems work very well at all in that context. So I spend a fair bit of time in the book talking about, you know, the way in which our regulatory instruments, our legal instruments, become so complex. Who really knows--does anybody know what's in most of these statutes or bills that are being voted on? So, I think it can be more transparent. But I completely agree with you that public choice is a big obstacle here; and that's also why I talk about the market for lawyers. Because, you know, one of the groups that I'm saying needs to not be so fully in charge of how our legal infrastructure is designed and developed are lawyers--who, kind of have too much sway over how we think about these things. And, yeah--that's usually--this could never happen, because lawyers wouldn't let it happen--a response that I get.


Russ Roberts: But, we can hold their feet to the fire a little better. And we can put political pressure on them to give up some of their power. I want to turn to that, because I have a--there's a very interesting part of the book, and particularly since we recently had an episode with Christy Chapin on the American Medical Association and how it shaped the evolution of medical care in the United States and our health care system. Talk about how the American Bar Association constrains our current legal infrastructure. Because, even someone like myself who is somewhat cynical about the American Bar Association, learned some unfortunate things about what they are able to do. So, talk about what they do and have done, historically.

Gillian Hadfield: Yeah. So, let's go back--if we go back to, say, the 19th century, in the United States. Nineteenth century is wide open. Anybody can practice law. In fact, a lot of states deliberately put in place legislation that said anybody who wanted to practice law could do so. There were no--very few law schools; there were no requirements of the Bar. If there were any Bar exams it was a couple of questions with the judge to be admitted to the Bar. So, it's a very open system. Now, the American Bar Association forms at the end of the 19th century, would be--the important goal is saying, 'Look, we need, we're facing the transformations of that economy, the industrial economy. We need more uniformity. We need better law. We need smarter law. It's important.' So, they began a 20-, 30-, 40-year process of what I argue is sort of creating a platform that really worked very well for the 20th century, much more uniform, high-quality legal infrastructure that allowed national businesses to emerge because they had some confidence about what was happening place to place: you didn't have conflicting jurisdictions and so on. But that process of--and the American Bar Association, it's important to remember, is just a trade association of lawyers. It actually doesn't have--well, it has only one formally recognized role, and that's it is the entity that is recognized for accrediting law schools in the country. And you have to be accredited if you want Federal funds and most states, Bar associations, have required accreditation to become a member of the Bar. But, what you end up with is, the American Bar Association coordinating Bar Associations in each of the 50 states. And it's the Bar Associations in each of 50 states that create a bunch of rules that say, 'Hey, there's really only one way to be--you have to have a law degree and be licensed as a lawyer to provide any kind of legal goods or services in the market.' Which is a very expansive definition. You have to follow what we'll call our ethics rules in providing those legal services. But the ethics rules--some of which are truly ethics rules and really very important and long-standing about, you know, fidelity to courts and--

Russ Roberts: confidentiality--

Gillian Hadfield: confidentiality. Conflicts of interest, and so on. But the really important regulations in there, and they are regulations--they are economic regulations. And they are economic regulations that say, you know, 'A lawyer can't be in a partnership or a business entity with someone who is not licensed as a lawyer.' They can't be--any kind of equity funding from someone who is not licensed as a lawyer. They can't be employed by a corporation that is controlled by anybody who is not a lawyer. The y can't enter into contracts with people who are not lawyers or businesses that are not run by lawyers in which they share profits or revenues. So, just basic incentive contracts, you can't enter into those. And, I think that's a tremendously problematic set of rules for the prospect of inventing the kinds of responsive, innovation legal infrastructure that I think we need for a much more complex environment, or even solving a problem that we've had for decades, which is reducing the cost of law so that more people can actually make use of it, make use of legal services. So, I think, that economic regulation that is overseen by bar associations and under the leadership of the ABA [American Bar Association], but the ABA proposes rules and bar associations have to choose to adopt them.


Russ Roberts: What's it like in other countries, though? Do they have cheaper legal services? More effective legal services? Or does every country tend to have a national organization like this that has restricted innovation in the provision of legal services?

Gillian Hadfield: Well, the American model has been influential. But, one of the things that is really interesting about the cost of legal services in this form--it actually seems to be, it's inversely related to or, you know, directly related to the level of development. The simpler a society is, the more likely it is that more people have easy access to legal help. And, I think--and the lack of ability to access legal services is because the complexity tends to grow along with the cost and complexity of the economy. And, what you'd like to do is say you want really smart legal structure, but smart doesn't have to be more complex and expensive. I think one of the important models to know about here is actually in the United Kingdom, where there are many fewer restrictions on the practice of law. So, in the United Kingdom--and this has been true forever--and that is that anybody can provide legal advice or help draft documents--wills, contracts, etc.--in England and Wales. So, that's never been restricted. And it is restricted to licensed lawyers in the United States. And one of the things that means is, you know, unions can provide legal advice. Community organizations, volunteer bureaus, entities that form, profit- or nonprofit-based, that will provide some legal help, maybe under charitable, on a charitable basis or under contract. So, the United Kingdom has always had a much less restrictive legal environment than we have. But, they've also made changes since 2007 to introduce multiple professions and to really try to take on this problem of making sure that the rules are not being decided by the trade association of lawyers. So, they've deliberately separated out the regulatory function in professional organizations from the trade association function. So, it's--yeah, we are so immersed in it in the United States that it's really important to recognize it doesn't have to look like that. It will look more like medicine. If we look more like medicine, we would--we, law--would have people with different kinds of life[?]. First of all, you wouldn't need a license for everything legal. And we could also have many different types of licenses. Our license is very expensive. It's a 3-year graduate degree, after a 4-year undergraduate degree. And, you know, that's a lot of education. It filters out a lot of people. Well, some of those people could be licensed to provide some basic services--help understanding what a subpoena is, help filing the documents in a family law case. You know, basic assistance to small businesses trying to manage their compliance with local regulations.

Russ Roberts: The way nurses and other medical practitioners--I think the medical field is way to over-regulated, in terms of licensing. But at least there's some variability.

Gillian Hadfield: Right. If medicine was regulated the way law is regulated, every single person who performed any kind of medical care would have to have an M.D. [Doctor of Medicine]. You couldn't have a phlebotomist who is licensed just to take blood. Or, you know, [?], licensed, its relatively minimally intrusive licensing, I think. There's practitioners, chiropractors, pharmacists--all of the--the way law is organized, every single one of those people would have to be a full-fledged M.D. And, you know, law--as I say, a chunk of this doesn't need to be licensed. Which is not to say that a certification system wouldn't still operate, as you mentioned earlier with, say, kosher food, organic food--it might still be, it's still the case in the United Kingdom: anybody can provide you with legal advice; but if you've got the money, you are going to somebody who has been qualified and [?] got lots of great experience as a solicitor to give you that advice. But, if it turns out that the former union rep [representative] who spent 30 years in the union knows a lot about employment and labor law, [?] he's a helpful person to have on the line for 80-90% of the people who have a question about their rights at work. And that's feasible in the United Kingdom. It should be here, too.

Russ Roberts: Do we have any evidence that it's better in the United Kingdom, for everyday people in terms of their access to those services? I assume it's cheaper. Is it? Do we know?

Gillian Hadfield: So, this is a point I make throughout the book: We have very little systematic data about how well our legal systems are working. I think that's another consequence of the fact that it's kind of fallen outside of our usual ways of regulating and left to the professions. And, it's tricky to explore this in the United Kingdom, because they've also had phenomenally generous legal aid, historically--60-70% of the population qualifying for legal aid in all kinds of matters: family matters, housing, whatever. They've cut back substantially: It's now probably, I think 20-30% of the population that can access legal aid. But, one of the reasons they were able to have extensive legal aid is because it was less expensive to provide these services. You didn't have to give the contract to a solicitor working in their own little office. You could give it to a company that was focused on housing benefits cases, for example. One other piece of data we do have is the fact that surveys indicate that people in the United Kingdom facing legal problems--which nobody thinks of as legal problems, by the way: they think of them as housing problems, family problems, misdemeanor problems, small business problems, and so on. A much smaller fraction of people in those circumstances manage those without some legal help; and a much smaller percentage do nothing. Studies in the United States, which again there's not enough good studies, but U.S. study says that people face in those kinds of situations as much as 30-40% just don't do anything in response to them. And that percentage in the United Kingdom is probably between 5 and 10%. The same thing is true in the Netherlands, where legal insurance companies are able to have lawyers on staff who work with people who hold the legal insurance. So, if you hold legal insurance and you've got a problem with your neighbor, you call up your legal insurer and they can put you directly in touch with a lawyer who can give you help with that problem. And, I think 40, 45% of the population in Netherlands has an insurance policy like that. So, it's a very different environment.


Russ Roberts: Let's close with an example that I found very eye-opening as a reminder of sometimes how hard it is to get outside one's normal way of thinking. You make a thought experiment about what if librarians who have degrees in, after all, in library science, were charged with coming up with a search engine. You'd think that that's their expertise: they're good at looking stuff up. And yet, the solution that we have now, which is pretty amazingly great, on, at least the dimensions of finding stuff, of search engines, was not developed by librarians but by computer scientists. Talk about that thought experiment and why it leads you to the importance of being innovative in how we think about making the rules.

Gillian Hadfield: Yeah. So, I think the--what do we know about the ways in which innovation happens, the environments in which innovation happens? We know that they are environments with a lot of diverse participants, lots of different points of view, lots of different ways of thinking about problems. You have a rich, diverse idea pool. You have people who willing to take risks and capital that's willing to take risks. And, if you are really looking for something that's innovative, it's not coming from the established providers. Clay Christensen has taught us a lot about that, I think, with the innovators' dilemma. And, so, the thought experiment there was to say: Look, if you'd left it to--if you had the kind of rules in library science that we have in law, you would have said, 'Look, this is our turf. We are the ones who are responsible for organizing the world's information and helping you find it. And anybody who wants to provide that service needs to be licensed as a librarian and trained in our ways of doing things.' It would have been--okay, we need to get better at classification--

Russ Roberts: It seems reasonable--

Gillian Hadfield: cataloguing. It seems very reasonable. Right. But what we know is that the truly innovative move generally comes in from the outside, with somebody who hasn't been steeped in that problem, saying, 'Huh. How come we don't do it this way?' And, Sergey Brin and Larry Page, they weren't actually looking to figure out how to find information. They were saying, 'Gosh, we have this capacity now with the Internet to access all of this information, and when we go looking and we just do keyword searches, we find a lot of junk. I really just want to know where's the good stuff, where's the stuff that's really important?' And the development of the PageRank algorithm, to develop that, was a solution that maybe librarians--would they eventually have gotten there? I don't know. Somebody told me recently that there was more discussion of this than I realized. But, you know, the path everybody was going on--the path that Yahoo was down at the beginning of the Internet in the late 1990s was cataloguing. They hired armies of librarians to review every single website and catalogue it the way a library catalogs. Well, that just was quickly completely snowed under by the volume of stuff on the Internet. You can't have a human being reviewing everything. So, I would say, 'Look, if we regulated library science the way we regulate law, we would have said only librarians can be doing it.' And I think it's really important--lawyers need to be deeply involved in the process of innovation; lawyers are well-steeped in how rule systems actually work on the ground, and that's critical; and what kind of legal needs are out there. But, we also need software engineers and people who are great at thinking about organizations and people who are great at coming up with better customer service models and pricing models. We need all of those kinds of people involved in developing systems that work better for us. And I think opening up that process is probably the most critical move we can make.

Comments and Sharing

TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (10 to date)
SaveyourSelf writes:

~ 24:17 Gillian Hadfield says, “I really think we need to be looking at the development of basically an industry that's attracting investment and attracting good engineers and more that's dedicated at the problem of figuring out: how do you regulate in a complex environment…”

I think it is worth considering that it may be impossible to regulate well from a central institution in a complex environment. That's what is meant by “complex” in the above sentence. It defies easy understanding and manipulation.
~24:17 Gillian Hadfield, says “…here are the publicly determined, accountable criteria that that regulation has to achieve.”
Imagine a world where you cannot know what technology will exist and you have a world where you cannot imagine the “accountable criteria” that the technology “has to achieve.”
~25:00 Gillian Hadfield says, “We're going to let private regulators who develop systems that then private companies have to purchase into--we're going to let those private regulators figure out better ways to achieve those, those objectives.”

History is very clear on this—at least to my review—when the government regulates something, it moves that which is regulated towards greater monopoly power. To my knowledge, the government almost never reduces monopoly power with any action it performs—excepting perhaps breaking up monopolies or attempting to prevent price collusion. I cannot even imagine the government successfully producing a competitive regulatory apparatus. Such a thing would almost, of necessity, result in civil war, not intelligent legislation.

I'm reading this interview, not listening to it, and the second half is not yet posted, but I think it is appropriate that Russ keeps bringing up Hayek. Hayek considered this exact problem of regulation and determined, using different language, that there already exists a tool for managing complex regulation in complex human systems. It's called, "common law," and it works remarkably well.

[spelling of guest's name corrected, per commenter's okay--Econlib Ed.]

Martin Damgaard writes:

I believe that the type of innovation of the legal systems decried, does have a striking similarity to the development of smart contracts on systems such as Ethereum. It is actually disciplines outside law (manly coders) who through the smart contracts redefine what rules should govern in a cross national setting despite of different national laws.

Jonathan Spence writes:

Much of this reminded me of David Friedmans work "The Machinery of Freedom", or at least one of the chapters of that book.

Clearly Gilliam Hadfield seems to be approaching similar ideas. Though from a different direction as she says that she is a fan of government.

BTW those in the US may be interested/amused by the recognised position of McKenzie's friend in the UK.

This is where a defendant chooses to represent themself in court but asks another (who need not have any legal qualifications) to advise them.

We have Sheria courts in the UK. They operate "outside" the UK legal system, but they do operate.

We also have something called the Arbitration act that governs dispute resolution groups. Usually they are trade bodies, but some cover family law.

Nonlin_org writes:

Sadly, this just another ancillary (some say parasitic) profession and its representative claiming to be the center of the universe:

"it's not just something we layer on top of our market economy to maybe respond to market failures; but it's the basic platform on which humans have always built sets of rules"

No, true markets self-regulate and never need economists, lawyers, government, philosophers, etc. Market rules are developed organically by its participants and then lawyers come and claim ownership.

As far as AI, we already have autopilot and autonomous transportation which works just fine without yet another layer of lawyers, thank you very much.

The technology is coming and is not waiting for the law "infrastructure". In fact, a great benefit of new technology in general is that for a while at least it renders government and the other unnecessary "services" irrelevant.

JHM writes:

"... to maybe respond to market failures ..."

Often when people reference "market failures", they mean market outcomes that are not the outcomes that they would choose. Hence, they impose regulations.

We have an example of government "certifying" private entities regulating an industry with the financial ratings companies (e.g., S&P). That certainly worked out well.

This entire conversation dealt with imposing new and more regulations. The goal should be to reduce the number of regulations.

JHM writes:

There might be a simple solution for autonomous vehicles. Have the manufacturer be responsible for any violations or accidents. After all, the manufacturer is essentially the driver. A passenger in a taxi or on a bus is not responsible for actions of the driver.

Being a law professor, it appears that she is inclined towards complex legal solutions in contradiction of her stated goals.

Mel Thaanum writes:

"Regulatory Capture" was one of the first responses to Prof. Hadfield's program. As another commenter noted, she describes the Credit Ratting agencies who did such a wonderful job of protecting us from the wonderful new investment instruments that were at the heart of the Great Recession. Her proposals remind me of the Progressive Experts described in Thomas C. Lenoard's Illiberal Reformers.
Perhaps it would be better for the Global Economy to allow the participants to decide for themselves if the countries they trade with and invest in have laws and social structures they can live with. I suspect most governments/countries will move to compatible systems as they face the the demands of Global trade.

Eric writes:
Russ Roberts: know, I talk a lot on this program about emergent order and about how norms emerge and how they are hard to control. ... Social norms--the idea of wearing a hat; the idea of greeting someone you don't know well by their first name--all these changes that have taken place in our culture that are constantly changing--the way language and social interaction evolves--those are under no one's control. And everyone understands they are under no one's control. And yet, they also understand that they change, and that you have to pay attention.

What exactly do we understand when we say:

"those are under no one's control. And everyone understands they are under no one's control"

If we mean by this that no single person controls social norms, that is certainly true.

However, if we meant to imply that social norms happen without human control or apart from human control, that is certainly false.

Upon close inspection, changes to social norms are just as much the result of distributed human design as examples we've seen in the economy and in language.

Russ pointed out that everyone understands that social norms "change, and that you have to pay attention". Why do we pay attention? It's precisely because people use that changing knowledge to make choices regarding their own practices, including whether and how they will respond to changes others are choosing to make regarding social norms.

The key issue is one of scale. There is no monolithic or central authority. Yet it is not without authority. Each person exercises their own individual authority over their own individual practices. Like working together on a gigantic quilt, each person designs the square of their own contribution.

If some people get up on a soap box and urge a change, it might or might not happen because other people choose to adopt that recommendation or not.

Rather than being static or rigid, the changes are dynamic. "all these changes that have taken place in our culture that are constantly changing" --R.R.

Rather than being centrally controlled by a central authority, the authority and the changes are distributed across the participating population.

Rather than having a single planner, we have many concurrent peer designers.

That is the Dynamic Distributed Design process responsible for what is called emergent order in complex human systems.

LS writes:

At 59:08 Russ Roberts asks if there is any evidence that it is better for people in the UK in terms of access to everyday legal services.

Important evidence is found in the World Justice Project Rule of Law Index.

On the criteria of "people can access and afford civil justice," the US currently ranks 94th out of 113 countries. In contrast, the UK ranks 46th.

I talk about this and other WJP rankings at this link:

Ben Riechers writes:

I was encouraged by Gillian Hadfield’s comments about transitioning at least some of our regulatory infrastructure away from process compliance and toward outcomes. I’m not sure I agree with her proposal, but her description of the problems associated with our regulatory infrastructure were very much on point.

In his book, The Other Path, Hernando de Soto describes the creation of markets in Peru by people he refers to as “informals.” These were markets that were outside of the law…not to sell drugs, but to build homes, educate kids and feed families. These informal markets grew to such a point that they ultimately represented a majority of the wealth creation in Peru. Much like your guest, Hernando de Soto does not promote the elimination of government or think the rule of law unnecessary…quite the opposite. However, among the many insights the book provides is the importance of a radical simplification of bureaucracies and regulations transitioning the focus of both to be primarily on outcomes. Gillian Hadfield shared a few of the reasons in support of outcome-based regulation and there are many more reasons than she mentioned.

In my view, regulation is a bit like water, too much of it is destructive. Unlike flood water, regulations never recede. Here is a thought experiment. How many rules would society advocate for if they knew that once created, the rule could never be changed? I suspect we would quickly move to guiding principles complemented with a small number of laws.

I believe that transitioning to outcome-oriented regulation (that doesn’t diminish any of our protected rights) could unleash economic growth here in the U.S. and around the world.


Return to top