 Today's session in our professionalism workshop is going to be presented by Professor Robert Topel, who is the Isidore Brown and Gladys Brown Distinguished Service Professor in Urban and Labor Economics at the Booth School of Business here at the University. Bob kindly came over, sort of as a pinch-hitter for Kevin Murphy, with whom he wrote this paper, Professionalism and Economic Approach, because Kevin, believe it or not, was called to jury duty this week, and having missed the jury on Monday and Tuesday, they designated him for Wednesday. So last night at 11 o'clock, when Kevin found out, he sent me the note and he sent Bob the note. So Bob is pinch-hitting and stepping in. Those of you who have heard Professor Topel and Professor Murphy speak to us in previous years about the returns and value of research in cardiovascular disease know that their line of inquiry is extraordinarily creative, innovative, and insightful. Just to summarize those results, for those of you who may have missed it, a 30-year investment of 50 or 60, maybe 70 billion dollars in cardiovascular research has over a generation returned about a trillion dollars in economic benefits, a staggering roughly 20-to-1 return on biomedical research. But that's not what they're going to be talking about today. Bob is going to talk about an economic approach to professionalism, and I appreciate you coming. Thank you so much. Thanks for having me. I got called to jury duty a long time ago, and I went down to a big group of people and they handed me the little card for your background and said, what do you do? I said, well, I'm an economics professor at the University of Chicago, and do you know when he judges? I said, yeah, I used to brook and pose, and I know those guys, and do you know any police officers? Yeah, my dad was a cop, and have you ever been the victim of a crime? Well, I live in Hyde Park. I was broken into a, this is back when I lived in Hyde Park. And so I went up, I handed it to the bailiff, and he looked at my card and he said, you can go. They'll never call you. Now the better story, though, is that Easterbrook, for those of you who may not know who Frank is, he was a professor in the law school, and now he's a federal judge with Posner on the, what is it, the seventh surrogate, got called for jury duty. And the plaintiffs ran out of challenges. So they got Frank on the jury. You know, it's one of those things where they could probably go back and say, well, no, we'll take that other guy. Anyway, so Mark asked us to talk about professionalism. I guess whenever Mark asks us to talk about something, we have to say, oh, now what do we do? And so it's an economic approach, because that's the only thing that we know how to do. And this is with my friend and colleague, Kevin Murphy. So I should get out of the way, I'll sit. We think about professional labor markets, actually I'm going to tell you in a little bit that amateurs in professional labor markets are professionals by our definition, even if they don't get paid. So we're going to distinguish professional labor markets from the other kinds based on whose reputation is at stake or who stands behind the product. And in most labor markets, the good or service is associated with the firm. We talk here about General Motors or Toyota, but I go down to Applebee's and have a meal. I don't know who cooked it. It's an employee who's been monitored by the folks and run the joint. But in a professional labor market, people buy a service from this person called a professional or from some combination, this is important, of a firm and a professional where individual reputations and qualifications might matter. So the example I give is that UFC hospitals or this hospital where I used to be on our board, advertises both the brand of the hospital, but they also advertise the individual identities of the docs who will fix you up if you come to their hospital. So both things matter. So both reputations are important. Now we want to distinguish between professionalism and self-employment, although they can be closely related. In self-employment, the provider of a good or service is the residual income recipient or has the residual rights of controls over the assets of the organization. In professionalism, a provider is held responsible for performance or is the individual producer whose reputation is relied upon by the buyers. Now absent some buyer uncertainty about the quality of what they're getting or on other aspects of the product, professionalism in our sense has little meaning or importance. So it's all about who's delivering this and there has to be some uncertainty about what it is, otherwise you don't care about where you got it, who you got it from because you know exactly what you got. So that's our definition of professionalism. It's not 0-1. In a lot of occupations that I've mentioned, it involves a combination of professionalism and more traditional firm-based reputations and a little later I'm going to talk about a little model of competing ways of supplying the same thing in either a professional way or a firm-managed type of way. But traditional firm-based reputations can matter in combination. So colleges where the university's reputation matters but so do the reputations of the faculty, architectural firms, law firms, hospitals, you can think of a lot of examples. So according to our definition, athletes and professionals because the quality of what you get depends, is imperfectly observable but it depends a lot on the reputation of the individual providing it. But even then amateur athletes are in the usual sense, are professionals in our sense because the quality of what you're getting depends on the reputation of the individual. Construction contractors if you've ever had to build a house, depending on the reputation of the contractor, who's in the business of monitoring the workers. So the contractor is a professional but the individual workers are not. Doctors, lawyers, architects, actors, chefs, Charlie Trotter is a professional. The chef at Applebee's however is unlikely to be deemed a professional by our standards. So now in professional markets as most of us recognize the professional organizations arise. So the American Medical Association, the Professional Golfers Association, the Association of Architects, the American Economic Association, reputations and quality uncertainly provide a natural role for those professional organizations because they provide a certification role for the individuals who are allowed in. Now once that certification role or that certification reason for the existence of a professional organization takes hold, the reason they were created that they have incentives that go beyond the reasons they were created. So rent seeking and to an economist rent seeking means we're trying to get stuff that from political activities or law being or enactment of laws, whatever, transferred to our members. So it's often the case that professional organizations try to restrict entry into the profession or create laws that restrict entry into the profession to reduce competition. If some of you have been following the news over the last year or so, there was some controversy about professional organizations seeking to restrict supply. One of the great examples was interior decorators in Florida and a woman was precluded from practicing interior decorating because she had been practicing without a license. And in testimony before the Florida legislature, the representative of the interior decorator said such things were necessary to prevent people from getting bad color schemes in their offices. I suppose if such restrictions had been in place back in the 1970s, we wouldn't have gotten avocado shag carpeting or something. So then one policy issue is what then limits, given the incentives of professional organizations to try to transfer resources from the rest of us to their members, what limits or rules should be imposed on what professional organizations can do and what they cannot do. We're not going to spend a lot of time talking about that, but it's something that could be part of a broader conversation about these issues. Now it has a lot of important implications for several things. How should professional firms be organized and how should they be managed? They be managed like regular firms, should they, as I think you all recognize in the professional firms that we're associated with, they're managed in much different ways. The nature of the principal agent problem within a professional firm is important because when I say principal agent for all the docs and medical school people here, I mean that somebody is a principal and they're charged with monitoring the efforts of agents who are producing output on the principal's and the customer's, therefore, behalf. So that's like the manager and the relationship with an employee, but it's also here a relationship with a professional in the organization. An important issue is competition between professional and non-professional forms of organization. And I'll get to examples like that in a minute, but you can think of individual one way of supplying medical services is to go see your individual doctor on whose reputation you depend and that's a professional way of providing medical service. On the other hand, there are a lot of medical services that are offered purely through clinics where those are basically managed firms and you've signed up for whatever the plan is and you go in and you get what they did. So it's like getting your meal at Applebee's. You don't know what doctor you're going to get. You're going to get who's available in that particular day. And often that those things can coexist at the same time. And as you go up through the complexity of the type of treatment you're going to be getting, professionalism we all expect is going to get more important as you go up to more and more complex procedures. Another important issue is the rewards to talent and the allocations, allocation of risks. So what leads to professionalism? In a lot of cases, it's the production process that's important. So the key element of the product are determined by individual performance. That's not ex-anti-observable. For example, surgery or home building. So you have to rely on the reputation of the individual supplying. There's persistent heterogeneity in individual performance. And production is sort of intrinsically individual. It's their signed art, research papers, and the like. All are in art by our definition professionally supplied. Observability is a key. The identity of who produced the good or the service is it's observable. It's important who did the operation, who was the chef, but not necessarily the quality ex-anti. So you're relying on who did it and their reputation as communicated by other users who have experienced the person's output in the past. So another important issue is how the individual, how do you identify the individual contribution to a jointly produced product? Which is imperfectly observable. Who scored the most points on the team? Those are professionals, but it's a jointly produced product. Who led the research project? Those are professionals, but it's a joint product. And you have to be able to ascertain what individual contributions were. Now, there's often going to be a choice between a professional and non-professional market structure. And I talked about a clinic already. In that final consumers go to a clinic, they don't rely on individual identities of the docs. But that can co-exist in competition with an individual practitioner, where I choose my doctor based on his individual reputation. So that's a professional way of supplying that service. And there can be hybrids. And I've already talked about this, hospitals have reputations sort of the docs within them. Consultants in law firms, sometimes people hire firms. Sometimes people look to hire specific professionals within those firms. So if you've got a case, you might say, well, the kinds of guys I want, I don't know exactly who they are, are at Skadden Arps. Because they happen to be good at whatever the issue is. But also I may know the particular guy I want to hire at Skadden. So it can be some hybrid. So one question is, what determines the choice between a professional and a non-professional model? What are the advantages and what are the disadvantages? So here's a purely professional structure. The advantage, one advantage, is that it relies on customers rather than managers to assess performance. So it's not that the manager is saying, okay, you have to do these things. And go through this set of activities, and that's how the output gets produced, like what on the assembly line at Toyota say. Customers are likely to be the best judges of what matters. When they can evaluate X-Post and communicate to others, they're evaluation. Because that's what generates reputations, that flow of information. That eliminates the agency problem of managers where the manager has to be monitoring individual outputs because by relying on reputations of the professional, it internalizes the costs of their actions and the benefits of their actions and promotes more efficient provision of the services. The disadvantages are that firms may have a lot more knowledge about the quality of the product being produced, or they may have a better ability to measure that quality because they're in the business of measuring what their people do. So monitoring quality at the firm level can prevent low quality products. And information transmission between customers may be poor, which allows low quality to persist and equilibrium. So those are some of the advantages and disadvantages that we see. Now competition, there is going to be competition between more professional and less professional forms of organization. Again, my doctor versus the clinic. And then the competitive advantage of each are going to determine the winner in any given market setting, which will determine the dominant form of providing the good or service. Now sometimes it's going to be entirely homogeneous so that the way of delivering the service is going to be a purely professional. Sometimes it's going to be a hybrid and sometimes it's going to be purely through firms. So, and then in a point I've already partially made, one model may attempt to gain competitive advantage, like a professional organization seeks to restrict entry of non-professionals. Which can have both efficiency enhancing because it improves the average quality of the product and the information that people have. But also it can have efficiency harming aspects because it tries, it's an effort to drive up the price or the earnings of the people who are accepted into the professional organization. So for a long time the AMA has been accused of trying to restrict the supply of medical school graduates in order to keep doctor's salaries higher. Now markets can transition between being more or less professional or transition from a professional structure to a non-professional structure. Now, we had an easy time thinking about examples where professionalsism has become less important. So you think about pharmacists. Used to be you go down to the neighborhood drug store and the pharmacist often was the owner of the drug store. And you relied on his reputation for knowing what to give you and making sure that they didn't give you some combination of drugs that was going to be dangerous for you. And that service has become much more mechanical. Now I go to Walgreens. Or at least when you go to Walgreens you got the plaque of the guy on the wall and you say okay well he went to University of the Pacific Pharmaceuticals Pharmacy School or whatever or else you send your stuff to Medco. And they got an army of guys in there that's a firm. You don't know who these people are and they're managed by those folks and that's become a much more mechanical thing. So that market has has transcended sort of from being a professional labor market to a much more firm based market. So professionalism has become less important. Same thing for tax preparation. Used to rely on the reputation of individual accountants. Now a lot of people just go to HR block. Okay, now we sat there, we said, we're making this thing up. So okay, now we need a bullet point on examples of increases in professionalism. And it kind of set blank for a long time and so the lack of ready examples actually tells us something about the evolution of technology for achieving the supply of these goods and services and markets. It seems that firms have displaced individuals as efficient providers in a lot of contexts. So that it suggests this evolution of technology or that the evolution of the technology of monitoring makes professionalism less important. It's kind of more restricted to the high end of the market in complex procedures. So to go back to the sort of the locally supplied good. Used to rely on the reputation of the local restaurant and the chef who worked there to decide whether you wanted to eat there. Now people go to Applebee's or they go to Chili's or something like that. Had that part of the market where professionalism has been displaced. Whereas in some other contexts, maybe professionalism has become much more valuable in what looks like the same market. So Charlie Trotter, until he decided to close and go back to graduate school, ran a restaurant that revolved around his reputation here in Chicago. And then he ran them in Las Vegas and God knows where else. Where they could rely on his reputation for drawing people in. So as technology has evolved, it's both created more opportunities for firms to displace professionalism. But at the very high end, because you can transmit these reputations much more broadly than you did before, Charlie Trotter's gotten more valuable than an individual's chef would have been, say, 30 or 40 years ago. It's kind of the economics of superstars if you well, if you're familiar with that notion. So we decided we'd have a model because we could. And so here it is. We're gonna have two forms of organization. You can be a firm or you can offer a service through professionalism. So there's two levels of output. Output can be high quality, so it's that some value V that's a positive thing. Or it can be low quality, in which case there's a loss relative to V of L, okay? Now the chance of producing V, high quality stuff, depends on the talent of the individual producer. So to keep things simple and no fancy distributions or any of that stuff, talent's either high or it's low. Either you're good at this or you're not. And high talent individuals always produce high quality stuff. So they always give you V, value V per unit of output produced. So the surgery is always worth V, all right? Then there's a fraction of low talent people, and that's that little Q down there. So that's a share, some number between 0 and 1. Say they're 50% of the population is low talent. And low talent individuals, they have a probability of messing up. So they've got some probability P that the value of output falls to V minus L, okay? So that, and then individuals have careers, they work for T say 30 years or something like that. And then the market conditions are the following. Individuals have a choice. They can be in this industry where they're producing whatever this good is that has high and low qualities and a mixture of talents. Or they can go to an alternative industry where everybody can just get W. Why would they just get W in this? Why would everybody, even the highs and the lows get it because it's easy to tell who's who or individual performance talent isn't so important in that industry. So all information is public. There's no asymmetric information. So I don't know if I'm the low talent guy, I don't necessarily know I'm a low talent guy until I get to observe my own output over time and see how many surgeries I screw up or whatever it is, okay? How many putts I miss? And then all this, so that's what we mean by there's no asymmetric information in this world. Individuals earn their expected value of their marginal product. So when I say expected value, I say, I have an assessment of whether you're high or low based on what I've observed you do in the past. And people get paid the expected value of that output. If they're in this professional market, alternatively they go over and they work for W somewhere else. All right, now a low talent worker has expected output of this thing WL here. Which depends on the probability they make some mistake and the magnitude of the loss if he does make a mistake. And since a high quality worker always produces V, we know that the value of his output is just equal to V. So that's the distinction between the two types. Then the expected output of a new worker, we don't know who he is yet. We don't know if he's a high or a low talent guy when he gets in. So the expected value of what he can produce is W at time zero. You haven't observed anything about the guy. All you know is with some probability, what did I call it, Q, he's a low talent guy. And if he is a low talent guy with probability P, he messes up and produces V minus L. So when a new guy comes in, he's either a high or a low. But the expected output is of convex combination of those two things. All right, so now in any equilibrium, we gotta have that the value that, in order for this thing to make sense, you gotta have that the value of what the low produces less than what they can get somewhere else. And the value of the high produces more than what they can get somewhere else in order to keep people in this profession. Now that's gonna allow people to sort. The known low talent guys, once you've observed a guy over time, and you know he's low talent, because he's messed up a few times or something like that. Actually, in this model, once he's messed up once, because there's zero probability of making a mistake if you're high talent. Once you've made a mistake, you know you're a low, you might as well go work somewhere else, all right? So, just like here, so this allows the workers to sort. The known high talent workers will stay, the guys that you figured out are lows will leave. So workers with a sufficiently high chance of being high quality stay, okay? All right, which means you haven't made a mistake yet. So, a low talent person has lower expected output than a new entrant. So if realized output is of low quality, this is what I was just saying, the individual leaves and goes to work for W somewhere else, they get out. And if a person produces high quality output, they continue to produce because it's still likely they're gonna be, they've still got a chance of turning really out to be a high quality guy. And then they sell the market assess, sell since the market assessment of their talent increases over time, okay? Because given what the cohort you were in, more and more of the low talent guys have been sorted out. And the longer you've survived producing V every period, the higher the probability that you're actually a high quality individual. So your expected output is going up. So the expected quality for workers that have produced high quality output for t periods, little t, is that ratio there that you can admire at your leisure, okay? And what it's doing is it's going up over time, because the guy hasn't made a mistake yet. So the expectation of his output is more and more likely to really be a high quality guy. Whereas the earnings of a worker with, and excuse me, not whereas. The earnings of a worker with experience at time t will then be that convex combination there, which is what's the probability that you're a high quality guy times V? And what's the probability that you're a low quality guy times L? So their earnings will be going up over time because they're paid their expected output. So the fraction of initial workers that remain in the pool after t years is that another convex combination, which is going down. Because you've got a pool at the beginning that was a mixture of highs and lows, and then the lows are leaving as they realize that they're lows and they go work somewhere else. So the exit rate is given by that other number that's not essential for you to understand all the details up. So equilibrium, what does it require? It requires that the expected lifetime wage of a new entrant into the profession be equal to the alternative because they can always get W. Can always just opt to get W, go be a plumber or whatever it is you want to be, but you're going to be a doctor. And you're going to come in here and so the present value of what you get in this profession, this is just a supply and demand in occupational choice. The present value of what you get in this profession has to equal and expected value because everybody's risk neutral here, what you can get somewhere else. But wages in the professional environment start out below W. Why do they start out below W? Because you got a chance of being a high guy and earning more than W later on. And it's the present value of the wages over your life cycle that matter in making the occupational choice. So people are willing to go into being doctors or young lawyers or something and beating themselves to death with long hours and earning low wages because later on it's going to pay off and they can do better than the market alternative when it's revealed that they're the high quality folks. Then the exit rate from the profession is going to be relatively high at first, right at the beginning it's going to be equal to P. Because what's the probability that turns out that you're produced low output? You did it, you're gone. So P fraction of the guys will leave in the first period. And then it's going to decline over time because the mix of guys who are left is of higher and higher average quality. Now consumers are going to get both high and low quality output in equilibrium, which is why you want to have a doctor, why you want your back surgery, I guess to be done with a guy who's done 5,000 procedures. Because he's proven that he's one of the highs. So non-professional production model, this can exist in competition with profession. Firms hire the workers to produce output. The firm can evaluate the output produced by a worker at some cost. So monitoring is going to be imperfect, which is to say that you can't tell with certainty what the guy has produced. But you can detect low quality at a cost of C of delta, where delta is the probability of detection. So you're devoting resources to figuring out whether this guy is actually producing high quality output. And the more resources you devote to that activity, the more watching you do, measuring you do, stuff like that, the higher the cost is going to be. So when we say C prime is bigger, that's a derivative. It means cost of monitoring is higher, the more resources you devote to monitoring. So every time you call this hospital, you get, or the page system, you get an announcement that says, this telephone call may be monitored to ensure quality of service. That's the- I don't think that's the full breadth of it, Mark, but- But that would be an example. So, and then if low quality is detected, the firm discards the low quality output because minus L is less than zero. And it can terminate the worker. So, with no monitoring, with no monitoring the expected output is going to be that Q bar, because you just get what you get. With monitoring, the net output of the monitored workers at higher is going to be that number. Because remember, we get to discard the low quality stuff when it's produced. And since workers have the highest probability of being low talented and have the longest working lives, they're the most profitable to monitor. Since new workers, excuse me, I didn't say the word new. Since new workers are the ones that have the longest time working and the highest probability of being slugs, you want to keep an eye on those guys. So those are the people you're going to watch. So in the general case, the optimal level of monitoring then is going to decline with experience. Since the chance of the worker producing low quality output declines also with experience. Productivity in wages will rise since the chance of low quality output is reduced over time. And remember, you're paying people their expected output. So a special case is just whether you monitor or you don't, zero, one. Firms can pay the cost C, they can perfectly detect low quality in this special case, has most of the key features. The firm monitors when workers are new, they stop monitoring when workers gain sufficient experience and then you just let them do what they do. So the firm will stop monitoring at some date when the probability of a low quality output gets sufficiently low. That it isn't worth incurring the cost C to look at these guys and see what they're doing. And you can determine that level, that date T, by replacing that inequality with inequality. At that point, the cost of monitoring exceeds the loss from producing low quality stuff, the expected loss. Plus the expected value of learning about future productivity, because you gotta care about what's going on in the future too. So we're gonna let that experience level when monitoring stops. We'll call that T star, okay? So you stop watching them after 15 years or something like that. The wages of workers that remain will rise with experience. But when workers are monitored, wages are given by the non-professional wage, which is their expected output, less the cost you incur for having to monitor them. And then when monitoring stops, wages become this number here and it's flat after that point. Because after that point, you're not learning anything more about the guys. You're not monitoring anymore, so everybody's gonna get paid a flat wage. I'm gonna show you a picture of that in a second. So now you gotta choose between these two models of supply and stuff, or these two things have to compete with each other. So when's the professional model gonna dominate? Well, you learn in the professional model, you stop learning after a point in the non-professional model. So in an anonymous worker model, that is gonna dominate when the costs of monitoring are really low. So this is what I meant about the technology changing over time at the beginning of the talk. Because if technology changes over time and monitoring costs go down, then the anonymous model gets relatively more efficient. And so it can displace a model that was previously relied on the reputations of what you learn about, what the market learns about these individual performers. So that anonymous model is gonna become more important when the cost of monitoring go down and the value of detecting quality output to be high. So the law when L is really high. So a necessary condition for the production model to survive in competition with professional model is that the expected loss, which depends on how many guys are low quality. What's the probability they make a mistake? And what's the magnitude of their mistakes? Those three numbers, when the product of those three numbers, the expected loss is bigger than C. We're just doing it as a constant cost of monitoring. If that's bigger than cost of monitoring, monitor the guys. And that's the sense in which if C goes down, then there are gonna be more circumstances in which monitoring is the efficient way to do it rather than relying on professional reputations. So the advantage of monitoring is the ability to discard the low quality output when it occurs. All right, so which models better can be examined by looking at the two expected wage profiles? In the professional model, the wage profile is gonna be always upward sloping because as time goes on and more and more guys have made a mistake, they leave and go work for W because now you know they're low quality once they make a mistake. And so the expected productivity of the guys who are left is getting higher and higher. So their wage profile in the professional labor market is always gonna be upward sloping. Whereas the non-professional labor market, it's upward sloping for a while because you're monitoring. But then you stop monitoring, so it's flat after that point. So and that's bigger than that after experienced T star that we talked about before. So here's an example. Careers last 40 years, there's a 5% chance of making a mistake. Okay, so it's not all that big of a probability of making a mistake. And 50% of the guys who enter the profession at the outset are low quality types. The value of high quality output, if you don't make a mistake, is 100, but if you make a mistake, the loss is 200 so that output net output is minus 100 if you make a mistake, okay? And with those conditions, a professional structure, trust me on this, dominates if C is bigger than 3.366, all right? So the figure shows the wage for the two models when you're indifferent, when costs are exactly equal to 3.66. And so there's what the profiles look like in the non-professional, excuse me, let's do the professional market where wages always slope up. You're learning more, reputations are getting better and better and better for the guys who've been in there the longest. They start out making less than the non-professional guys who are being monitored because their expected output is pretty low and you're not getting to discard their mistakes, okay? Those mistakes create real costs. And then the wages of the non-professional guys go up over time, but at a slower rate than the professional guys. And then at this date T star, you just say screw it. It's not worth looking at these guys anymore, they're all in the same boat. We're not going to figure out anything more about them. They get paid this higher wage, but the professional guys go buy them. So in order for these two things to coexist here, the wage profile for the professional guys has to have a higher payoff at the end. Compens offset by a lower payoff at the front end. And so those things have to have equal present value if these things are going to coexist in equilibrium, all right? So a little bit of discussion. What our analysis tells us is a professional model ought to be associated with steep wage profiles as productivity is low initially, due to lack of monitoring, but really high litter due to learning. When I say learning, I don't mean learning how to do it. I mean learning about the individuals, who's high and who's low. So there's no accumulation of talent in this thing, in this model. It's all pure selection. The low level of initial productivity generates an incentive for pre-market screening or some form of market monitoring. Because you want to figure out, you'd like to do something that says, you know, it would be better if we didn't have so many of these low quality guys starting out in the profession. So you might want to have some entry test or filter that keeps that, that keeps the average quality up, but it can't do it perfectly. And the model suggests a hybrid solution where workers start out monitored and later move to professional status. So you can have both a firm and a professional type market going on. Few words about pre-market screening. Professional markets are often associated with pre-market training or screenings like dedicated education in a medical school, a law school, or an apprentice slash professional system where people start off as apprentices, but then they become professional actors in whatever their line of business is. So you have skilled trades and doctors are the key examples that came to our mind. The excessive reliance on those, how are we going to lower quality rather than improve quality because you're keeping people out. Now professional associations arise naturally in the model, as I mentioned earlier. The earnings of the high quality guys are limited by the existence of the low quality workers. They don't like that, right, because you're paying people based on expected output. So you don't want all these slugs mixed in with you because they just reduce your own earning power. So professional associations can provide some of the monitoring that would be supplied by firms in the non-professional model. So that's kind of the role of professional associations. But as I noted earlier, that creates these opportunities for rent-seekings where the interior designers get together and they want a law that says you've got to pass a test to be an interior designer, or you've got to take 70 hours of training in order to be a manicurist in some states. So way more than 70 hours of training. So they have that incentive to limit competition generally, not just competition from low quality sources, but competition generally. And that's where there's this trade-off between the efficiency aspects of professional organizations and their monitoring and the inefficiency aspect where they just want to transfer resources to their members. Firms are often important in professional markets and they play something of a different role since they have to deal with the different, the reputations and the interacting reputations of their professional employees. And that creates a different set of monitoring and management issues than you would encounter in non-professional markets. And that's the, that's what I have to say. You got it. Very different material from that which we normally hear. Well, when you know the saying is when you got a hammer, everything looks like a nail. Well, one of the questions on the original title was which professions need professionalism the most? Fewer than before. Fewer than before. I think that's, I mean, that was like the most, I hadn't thought about it before. But when we were going down those bullet points and said, OK, where are places where you can obviously transit from one type of supply, one type of organization or industrial organization to another? And we thought, oh, well, we'll make this bullet point about the ones that have become less professional. I was the pharmacist and stuff like that. And then we paused for a long time over the ones that have become more professional. And we think it really has a lot to do with the technology of monitoring and the efficient way of organizing firms. It's become one way of providing the service has become less costly than it did before. And it's winning out in competition with a model that depends on so much on individual reputations. Well, let me take it right to medicine. Does medicine run the risk of following the pharmacists into a firm type corporate model? Or is there a way to retain professionalism in medicine? Yes, an economist would phrase that differently. Because you said it's like, well, if this happens, it would be bad. All right? And in the competitive difference. No, I didn't say that. Well, you implied it. I did imply it. OK? And it might be bad for like you. But it might be good for consumers because it's a lower cost way of supply and high quality stuff. It might be bad for some consumers. Like if for reasons other than fixed cost reasons or something like that, all the firms became non-professionals. So I relied on them to decide who should be treating me rather than relying on the reputation. Some consumers will be worse off because they'd rather rely on the individual reputations. But in general, the most efficient way of organizing and supplying these services is going to win out with competition. But again, given the importance of, let's think about the difference between hands-on surgery and robotic surgery. I'm going to really simplify. A monkey can do robotic surgery. So it used to rely on the reputation. How many times has this guy done this thing? And that's a purely individual reputation. You want to know who he's treated, what the outcomes have been, stuff like that. Whereas if a robot's going to do it and somebody's just going to run the robot, then that moves to be a firm organization. It's a much more efficient way to supply the same thing. So technology can displace the importance of reputations in a professional market. I thought a profession is something that it has to be training for a certain period of time, usually a year to several years. The curriculum arranged by people who are equally professional and knowledgeable about it usually has an exam and a certification. And it is something that other ordinary people cannot just know it in a very short time. You have to be trained for. And because I cannot evaluate somebody else's profession, then the trust, you call it reputation. But the trust is important when somebody is a doctor and engineer that I know that this person has had certain qualification. But the length of it, how long this doctor operate 5,000 or 2,000, I have no way of having. So professionalism as a simple way to mean that it's required knowledge in a certain period and education and exam and somebody certified them. And it is often managed by, not by government, telling you what to do and what not to do, but by professionals themselves, say the standard. They produce a standard of professionalism. What is professionalism in medicine? So if somebody, another doctor is not a professional, I know it, but you may not know it. But now you bring it at the level that the consumer are going to evaluate. In other words, you see what I mean? It looked like you're saying what the production is and how you are going to do better now. But people could never know how professional a person is unless they are professional themselves. Actually, I don't disagree with anything you said because I think in different language, you were saying the same thing that we were saying. So when I say that the consumers know something, I can go to, I realize that I'm not qualified to evaluate the reputation of so-and-so. So I ask Mark because I know he can do it, but it's the same thing. Now I've got somebody who's my agent in evaluating the reputation of somebody and deciding to whom I should go. And my experience is in a lot of these procedures. The amount of time you've been doing this and part of its pure experience and part of its, well, you've been successful in this business for a long time, so you must be an H rather than an L that that really matters to people in evaluating the reputations of the docs that they go to. Now you referred to the certification process that, and you go through this educational process and the knowledge is supplied by other guys who've already been in the profession and so on. And to an economist, we look at that, we say, well, why do they do it that way? So why do we have the certification process? Well, the certification process is to certify that these guys have a certain portfolio of skills and are of a certain quality. And if their quality was so easily monitored, we wouldn't need so much of that certification. So that I think what you're saying is entirely consistent with what we are saying about the role of screening individuals, because remember we said it's important to screen people coming in. And I'm not saying that the whole role of professional school is to screen because you teach them useful stuff, but it was a model. It emphasized the key aspect, which is the information about quality that gets transmitted via reputations in professional markets. One of the problems with the non-professional model for applying it to medicine is the notion of trying to define what quality is in the practice of something as complex as medicine. So that, for instance, making a single mistake once is an inadequate way of measuring it. So you've got that kind of simplifying assumption in the model. The question of whose perspective should be taken becomes complicated. We have recent publication of some evidence that patient satisfaction, consumer satisfaction, is very poorly correlated with actual outcomes. And the issues of even the assumption you have that the knowledge of the person will be stable at a certain point and their ability to perform will be stable is probably an incorrect assumption as well because within something like medicine, there have to be lifelong learners so that the assumption you had in your model about people reaching a static point while the monitoring technology might change, so does the technology of producing the craft change? So I'm wondering whether in light of all these obstacles, you really think that moving to a non-professional model is ripe for medicine at this point where you think the obstacles are still so difficult that it really is effectively not gonna be something we can do despite the fact that there are a lot of people who are asking to have it done. Okay, well, I don't know who's asking to have it done, but my answer is, I'm not claiming that medicine should or would or will move to a non-professional model. My answer is sort of, I only work here. And I look at what I see and we build them out. Now some of your comments about, this is what economists hear all the time, I didn't like that assumption, that was unrealistic. And as my co-author once said to his students, if you wanna understand the world, you gotta like ignore some stuff. And all the stuff you want us to put, I can make this model a whole lot harder in a little bit, only a little bit more realistic. So you say, oh, you only make a mistake once and you're gone. Well, obviously we don't believe that, okay? So we've got high quality guys who make mistakes less often than low quality guys, all right? Then you're learning is a little bit more complex because when I see you make a mistake, I'm, well, it's higher probability that you're a low quality guy, but I don't know. So, and I build into this not only that these guys are high quality and they're always high quality, but that quality goes up with experience. So there's learning by doing and people just get better at stuff. So I can make this model a lot harder, but our point here is to really focus on the stark stuff about the importance of reputation and learning about who's who in these competing models. So everything you say we could build in and some of them would make it, perhaps a more insightful model and some of them wouldn't, okay? I didn't realize it, but in 1970, when I was the chief resident, I brought Milton Friedman over from the business school to do grand rounds and Professor Friedman did a grand rounds arguing for the non-licensure of physicians. The physician should not be licensed and he said that in place of individual physician licenses, meaning the state certifying that the person had a certain level of competence so that they could be trusted, you'd have to rely on institutions and Professor Friedman envisioned, there would be 20, 25 great Mayo clinics, Cleveland clinics around the country, who University of Chicago, who would employ people who had all the background, all the training, all the right credentials so that you would not never rely on a particular surgeon or a particular internist, you'd rather rely on the reputation of the institution to assure you that the individual would be skilled and competent. I never saw it until today as sort of this argument against professionalism and in favor of what's the alternative. Just monitoring in firms. Firms, firms, yeah. But it clearly, that's what he had in mind. Well, no one ever won an argument with Milton, even when he was wrong. Yeah. Okay, well let me disagree with Milton then because the model that he's talking about obviously exists. I'm gonna go to University of Chicago or I'm gonna go to Mayo clinics or Cleveland clinics and I know the average quality of the guy there is better than down at the local in and out hospital but once you get in there, you just have to say give me anybody you got because for some things you really, really care about the relative qualities within the place and reputations matter. The professional standards in high prices enforced by the medieval craft guilds eventually gave way to the benefits of free market entry. I can't help but notice a modern trend that seems analogous and that is medical tourism, people going to other countries to get healthcare. What do you think will guide the economic equilibrium in this market? Well, you can already get your radiology done in Delhi I guess because it's cheaper, okay? So even if you had the same credentialing, it's just cheaper to hire a doctor over there. So there's in a world where prices are high and other people can supply the stuff, similar quality stuff at much lower cost, there's gonna be entry because market's a bore a vacuum and then you're gonna have, you're gonna see the professional organizations I would guess raising a big hullabaloo because it's undermining the earning power of their members. But in a lot of cases that hullabaloo is gonna be anti-competitive in the sense that it's, if successful, a little harm consumers. I wanna thank Bob for coming over. Thank Kevin for helping to write the paper. Thanks, Gary. Yeah, let's get Kevin over.