M. Todd Henderson, "Lawyer CEOs"
Does legal education matter? In this lecture, Professor Todd Henderson presents some data on this question, using the behavior of corporate executives as an instrument. Looking at the 10% of large, public company CEOs who are lawyers, the talk tries to determine whether CEOs trained as lawyers act differently than CEOs trained in other ways. Do lawyer CEO firms get sued more or less or the same as other firms? Do they manage litigation differently? And, if they do, what is the impact on the bottom line? There is a burgeoning literature on how personal characteristics, from physical traits to birth order to education, impact CEO decision making. The lecture discusses this literature as well, and situates legal education in it.
This Loop Luncheon talk was presented on May 4, 2018.
Transcript
Announcer: This audio file is a production of the University of Chicago Law School. Visit us on the web at www.law.uchicago.edu.
Dean Miles: Good afternoon, for those who haven't met me, my name is Tom Miles, I'm the Dean of the Law School, and it is my pleasure to welcome you to our Loop Luncheon 2018. And I especially am pleased to welcome all of you to Reunion Weekend 2018, and it's terrific to welcome all of our celebrants and have you here. I know many of you have traveled from distant locations to be here. I was just talking to a few people who are, I think the farthest is come from Zurich to be here for reunion. Welcome!
Dean Miles: And I know many of you have come just from a block away at your firm, so welcome as well. We have over 850 registrants at reunion this year, so we're really thrilled to have so many of you back to join us, and it's a great pleasure to kick things off with today's speaker. Today's speaker is Professor Todd Henderson. Professor Henderson is the Michael J. Marks Professor of Law and the Mark Claster Mamolen Research Scholar. Now, Professor Henderson is a graduate of the law school, and in fact he is our special guest for many reasons, but one of the reasons is that he's a member of the Class of 1998 and he's celebrating his 20th reunion this weekend.
Dean Miles: Now, in addition to being an outstanding and really incredible student who graduated with high honors and Order of the Coif, he was an editor of the Law Review. One of his great distinctions when he was a student was that he was captain of the intramural football team and captain of the all university champion, all university champion intramural team. And in the 20 years since he graduated, he has an enviable set of accomplishments. He first clerked for Judge Dennis Jacobs of the US Court of Appeals in the Second Circuit. He then worked as an associate at the firm of Kirkland & Ellis. He then moved onto McKinsey and Company where he was a management consultant specializing in advising, telecommunications, and high tech clients on business and regulatory strategy. Then we at the Law School were very fortunate to get him back as a member of our faculty, and since joining Law School as a faculty member in 2004, he has been a prolific scholar in areas of corporate law and securities regulation and, in the true interdisciplinary spirit of the University of Chicago,
Dean Miles: he has been a frequent collaborator with financial economists on topics such as insider trading in corporate governance. He is a coauthor of the leading case book in securities regulation, and just last week he published his most recent book "Outsourcing the Board: how board service providers can improve corporate performance." So this came out just last week by University of Chicago Press, available at your local bookstore and on Amazon. So in addition to his scholarship, Professor Henderson has served as a member of the adjudicatory council of FINRA, the Financial Industry Regulatory Association. He is in incredibly high demand as a speaker and visiting professor at universities across the globe from the University of California, Berkeley to the University of Genoa in Italy. At the Law School, he has of course taught Business Organizations and Securities Regulation, but he has also taught an incredibly wide range of subjects from torts to banking regulation to American Indian law. So we are thrilled to have today Professor Henderson to speak to us about CEOS in large public companies who are also lawyers. So please welcome Professor Henderson on Lawyer CEOs.
Todd Henderson: I'm glad to see Tom got the email from my mother on that introduction. Thank you all for coming, and most of all, thank you for supporting our Law School. Five years ago, I gave the remarks at the Loop Luncheon on the occasion of my 15th Reunion. I'm not a mathematician, but I think that then means this is my 20th, and I'm glad to be here and celebrate that and glad you all are supporting our beloved institution. I've done a lot in the last 20 years as Dean Miles said. I've had one wife, three kids, four jobs, taught 15 different subjects, and written more articles and books than I can remember. I've lost long track... lost track of where they are or what I said in them. Most of the ideas, of course, were wrong, but being right was really never my ambition. It's far too difficult to be correct about anything since most human affairs are far too complex for any one person to have anything meaningful to say about them, let alone be wise or knowledgeable enough to direct the lives of other people.
Todd Henderson: We would all be much better off if those who think that they know exactly how the world should be run and designed took heed from the 16th century French philosopher Michel De Montaigne, who wore a medallion around his neck that said "Que sçay-je?", what do I know? My topic today strives for this humble approach. I'm not here to tell you how President Trump is wrongheaded on steel tariffs or immigration policy or how he's right on foreign policy, taxes, and deregulation, although I think all of those things. instead I have a much narrower goal, I aim to offer some evidence on a question that is relevant to my job as a teacher of law students. Does legal education matter? Or more particularly, what influence, if any, does being a lawyer have on the decision-making of people who become CEOs of publicly traded companies? The question of the impact of legal education on individuals is obviously relevant to most of us in this room.
Todd Henderson: We went to law school, we invested in a particular type of human capital, and we've deployed those skills in a range of fields from government to business to family life. We believe legal education has value, or at least we did when we agreed to go to law school. When we pass around the collection plate at the end, we'll see if you still feel that way. Legal education is also relevant to some other people in this room who are not lawyers. For better or worse, lawyers have a great deal to say about the way the world is run. Mostly Tony Kennedy. Many elected bureaucrats and government officials are lawyers so are judges and prosecutors and as we'll soon see about 10 percent of CEOs. But there's another reason for you to care about this question for some of you aren't lawyers. You non-lawyers are here to spend part of your day listening to a lawyer talking about educating other lawyers, that my friends is what we call unconditional love.
Todd Henderson: For instance, I have the great pleasure of having my father in the audience. As it happens, he was here five years ago as well. Apparently, retirement means time is worth a lot less than it used to be. In any event, it's a delight for me that he's here and all of you. Now, let's turn to the question at hand. Do lawyers CEOs act differently than non-lawyer CEOs? There are two hypotheses about CEOs that we can test. The first possibility: CEOs are CEOs. In this theory, the market for CEO talent identifies the CEO trait, whatever that is, and this trait makes people act like CEOs, not like lawyer CEOs or MBA CEOs or college dropouts CEOs or anything else. The second possibility is that CEOs with legal education are different, in other words, law school changes individuals' human capital in ways that manifest themselves in corporate decision making.
Todd Henderson: To get some purchase on this question, I'm going to look at a specific area of potential CEO decision making. Are CEOs with legal training associated with more or less litigation? If they are, which way does causation run? Do lawyer CEOs cause changes in litigation or do changes in litigation cause a firm to choose a lawyer to be the CEO? It could be that a lawyer CEO takes the rudder and she trends the sales in a way that reduces the firm's legal exposure, in which case we'd expect businesses run by lawyers to be associated with less litigation. Or it could be that a company hires a lawyer to be the CEO when they are exposed to a big new legal risk, in which case we would expect businesses run by lawyers to be associated with more litigation. Before we get to the data, some background. There is a vast and ever growing literature on the influences on CEO decision making and governance, and let me give you a few prominent examples.
Todd Henderson: First, there are numerous studies, only a few which I'll mention here, which show the influence of personal traits of the CEO on corporate decision making. Steve Caplin, my colleague from the Booth School, led a team of researchers that examined the relation between CEO personality traits and firm performance. Their data were the comprehensive personality tests given by private equity firms to choose leaders of their portfolio companies. They found that aggressive characteristics: execution, resoluteness and overconfidence were positively correlated with buyout success while softer skills: teamwork, empathy, and openness were negatively correlated with performance. The punchline: if you're picking the CEO, a jerk that get stuff done is better than someone who feels your pain. In another study, researchers looked at whether good looking CEOs do better than ugly ones. They gave nearly 700 CEOs an attractiveness score based on their facial geometry and then they compared the market reaction to the announcement of that person as the new CEO.
Todd Henderson: Shocker, the market likes beauty. Firms that hire beauties had greater returns during the announcement window than ones that went with ugly ducklings. Interestingly, they found this was only true when the press release contained an image of the new CEO. Suggesting that causation runs from beauty to performance, not the other way around. In research coming out of Sweden, we learned that height is positively associated with good corporate performance. I especially like this result. The researchers found that intelligence and height at age 18 in the top one percent of the distribution were significantly correlated with becoming a CEO and with firm performance. Another study in what we might call the "What If Todd Were a CEO" category researchers found that Republican CEOs earn greater profits for shareholders. Tall check, beautiful check, conservative check. If anyone out there needs a CEO, I'll admit I'm open to considering my fifth job since graduation.
Todd Henderson: Note: I didn't discuss the paper finding that jerks are better CEOs. It seems no one is the perfect CEO candidate, but three out of four ain't bad. The second group of studies in this literature show the impact of professional experiences on corporate performance. For instance, in one that my dad, West Point Class of 1962, will like: CEOs with military experience were less likely to engage in corporate fraud and performed better during industry downturns. Financial experience matters too. A series of papers show that CEOs with degrees or work experience in finance carry out more sophisticated financial policies and investment policies that are less sensitive to firm cashflows. When faced with a cash crunch, a CEO that studied marketing at Kellogg might be forced to leave money on the table turning down NPV positive projects because of the lack of financing. While a CEO that studied finance at Booth would deploy sophisticated financial engineering to get the job done.
Todd Henderson: Turning now to the specific subject at hand, there is evidence that law training matters as well. On the south side of the Midway, we generally instruct students that risk is something to be mitigated. After all, in the first year, we teach students about packages exploding on train platforms, people with eggshell skulls, and consumers who lose all their belongings on installment loans because they missed one payment on one item. That is Palsgraph, Vosburg, and Williams v. Walker-Thomas Furniture Company, just to refresh your recollection. In upper level courses, the curriculum is run through with market failures, externalities, and discussions of human decision making heuristics that could make even the most free market free marketeer into a Marxist. Is it any wonder lawyers are risk averse? Frank Blake, the former chair and CEO of Home Depot who graduated from a lesser law school in Harlem, said law school consists of taking normal people and getting them to worry about what no sane person should worry about. But lawyers' risk aversion may not be our fault. To quote Lady Gaga,
Todd Henderson: Maybe they were just born that way. Before they even show up at graduate schools, students may sort into schools and therefore professions based on their risk preferences. The Midway Plaisance maybe like the sorting hat in Harry Potter. Risk averse people draw House Gryffindor and they go south of the midway to the law school where they become courageous defenders of truth, justice, and the American way. Risk preferring people may draw House Slytherin and they go to the Booth School where they spent two years golf and sunning themselves in the glass enclosed quadrangle of the Harper Center on a mere waystation on their way to becoming masters of the universe. There is some evidence to support this sorting hat theory. In my current IL elective, Elements of Business Law, I offered the class the following choice to illustrate the concept of risk preferences in decision theory. Here's the choice: $10,000 in cash or a 50/50 chance of $20,000 or zero. These are equal in expected value terms, and a risk neutral person should be indifferent between them. Of course, risk takers would go for the coin flip and the risk verse would take the 10 grand. All 56 law students preferred the sure thing. I'm certain the result would be different with Booth students. In fact, I may, I suspect many may prefer the gamble just because of the thrill of the risk.
Todd Henderson: Existing research doesn't permit us to differentiate between the human capital and the sorting hat story. Whatever their risk aversion, we observed maybe due to either, but there is research on the impact that lawyers have on firms. One paper found that elevating the general council into the top five in the executive team is effective in curbing regulatory noncompliance and monitoring failures. Another found that if the general councels represented among the top executives this leads to more accurate earning for class and less insider trading. It's not just general counsels in one that my friend and sitting in the back, Eileen Camerick, Class of '84 will like, lawyer directors have a positive influence too. Directors with legal backgrounds on the audit committee are associated with higher financial reporting quality. Most on point for what I want to discuss, recent research suggests directors with legal education are useful in monitoring executives, managing litigation and reducing regulatory costs.
Todd Henderson: Okay. So what about CEOs? Does it matter whether they're lawyers or not? One might think the CEO is above litigation matters, setting the vision, picking the right lieutenants and managing external stakeholders and the like. And what matters more is whether the general counsel has an important role or there's directors on the board. After all, one may think that all CEOs are risk takers, pushing the envelope, and having other lawyers watch them is more important than whether the CEO is a lawyer themselves. CEOs may be CEOs, whatever their educational training. We all had classmates we believed were like this, that they ended up in law school by mistake. They were destined to be business leaders, not lawyers. One of my classmates closed the acquisition of his first company in the middle of our Civ Pro I exam. When he walked out in the middle, our hearts were aflutter with possibilities.
Todd Henderson: Was it a panic attack? Was he giving up? Was he the smartest person in our class? Then he would end up as one of the richest people in the Midwest did not occur to us, at least at the time. It didn't occur to me until a few years ago I saw his name in Crain's magazine. Back for a visit recently, he told our students he felt pressure from his immigrant parents to get an advanced degree and an MBA just didn't count. He chose law school because the opportunity costs were lower since it's a year shorter than medical school, but all along he was destined for the board room, not the courtroom. In this research, I asked the question of whether law school had any effect on people like him. Did law school impact their way them in ways to change decision making? I looked at the period 1992 to 2012 and the data includes about 3,500 CEOs.
Todd Henderson: About 10 percent are lawyers. We don't have enough lawyer CEOs to divide them into more fine grained groups based on degrees of experience. Lawyer here simply means they have a law degree. Some may have gone straight from law school to the corporate world. Others may have more extensive legal experience. We'd love to try to unpack these just being a prosecutor or a defense lawyer, being in house versus government, do these matter, but our data sample is just too small to test these. Our litigation data is all federal lawsuits filed during this 20 year period. There were a lot of them. About 154,000 suits over 20 years, meaning on average, each firm is sued twice per year in federal court. We don't have data on state suits, they were excluded, but we don't really have any reason to believe they're different along the dimension we care about. So we're pretty confident in our results. Of these suits, about 30 percent were settled or lost... settled and about two percent were lost.
Todd Henderson: The rest were dismissed or dropped out of our sample. The economic impact on firms is large. Average cash cost for the ones that we have data about $2,000,000, but that vastly underestimates the cost of litigation. Managers are distracted by lawsuits and reputations both personal and corporate may be impacted as well. These additional costs are reflected in the stock market's reaction to when a lawsuit is filed. For the average company in our sample, the stock market drop is minus point one three percent. That's 8 million bucks. The market losses are much bigger for certain types of suits. The filing of a securities class action typically results in a stock price drop of 15 percent. Okay. So there were a lot of lawsuits, a lot of lawyer CEOs. What was the effect on lawsuits from having a CEO? In our baseline results, we simply compared companies with lawyer CEOs and companies with non lawyer CEOs and ask which one had more litigation.
Todd Henderson: We find that firms that have lawyer CEOS are associated with less litigation. The differences were not only statistically significant but also economically significant as they differ by a factor of two. Companies run by lawyers are sued a lot less often than companies run by non-lawyers. At this point, we can't say anything about causation. It may be companies that don't get sued are more likely to hire lawyers as CEOs since these companies might be risk averse for other reasons, but we'll look at this question below. We can say that in the run of cases, however, lawyer CEOs are associated with less litigation and therefore rule out the alternative story that companies that are in trouble hire lawyers to be their CEO's.
Todd Henderson: Okay. This is where it gets interactive, so here's where the fun starts. I passed out some handouts that you have there. Don't be intimidated if you haven't seen one of these before. I'll take you through it, but if you turn to page two, you'll see this result broken down by types of litigation. The last row on that table is the difference between litigation and firms with lawyer CEOs and those with non-lawyer CEOs. Start in column three and move to the right, and you'll see the difference in the number of suits between these two groups in categories ranging from antitrust to products liability. If you've never looked at an economics paper before, this table may look strange, but we're looking for two things. First, the sign of the difference. If there is a minus sign, that means there's less litigation in lawyer run firms, and second, whether there are little stars or asterisks, the more stars, the more confident we are that we can reject the alternative hypothesis that this is attributable to chance.
Todd Henderson: Econometrics is more art than science so you should be skeptical when you see data like this. Only if the intuition makes sense should you give these stars any credit. It's just too easy to manipulate data. We've tried our best to be above board and transparent about our hypotheses and testing methods, but so far you haven't seen much. Back to the table, you'll see there are minus signs in all but one category, products liability, and there are stars. This means the difference is not likely a fluke in the data, although the sign is negative for environmental suits, we don't see any stars, meaning we can't say anything about that one way or the other. There aren't very many environmental law suits and that's why we don't get the statistical power to get stars. What about products liability? Well, frankly it would have been much cleaner had this come out the other way, so the positive sign more litigation and the stars.
Todd Henderson: That's a bit of a problem that we have to address, but I don't think it's fatal for two reasons. First, products liability may be cases where they're relatively more immune from the CEOs meddling compared with other types of cases. It's not difficult to imagine a new lawyer CEO coming in and reducing the incidence of antitrust violations or employment discrimination, but having less ability to impact product design at least initially. Second and related, this may be an example where reverse causality has more explanatory power. Perhaps there are firms that have a large amount of products liability claims in which a lawyer is hired to be the CEO specifically to manage the claims and the aftermath. This was true in a lot of the pharmaceutical litigation involving Vioxx and other companies, for instance. If true, and there is some evidence as I said, to suggest this, this would be akin to the work of my economics colleague and Freakonomics book author, Steve Levitt, who found higher crime rates are associated with more police.
Todd Henderson: Obviously the police aren't committing crimes to explain the data, although police are committing crimes, but the causation runs the other way. Police are a task to high crime neighborhoods. They're not the ones causing most of the crime. At this stage, we can say, except for products liabilities, lawyers are associated with less litigation. You can see that in the second column, but this may not be because they have a law degree. There could be something else that is driving the result that's correlated with having a lawyer as a CEO. If conservative firms are less likely to be sued and more likely to hire a lawyer to be the CEO, then we would see a negative association between litigation and lawyer Ceos, but not because of anything the CEO did, but because of something that predated her arrival at the company. Economists have developed techniques to address this problem. Econometricians use a technique called a regression to isolate the impact on one thing called a dependent variable from other things called independent variables. We write an equation in which the dependent variable is estimated based on the independent variables. On the left side, we put the dependent variable. This a variable we're asking about. In our case, the amount of litigation, and on the right side we put everything we think of that might affect the amount of litigation. There's a big literature that estimates litigation based on firm characteristics: markets, firm size, market to book ratio, leverage, profitability, stock return volatility. All of these things have been shown repeatedly to influence the amount of litigation.
Todd Henderson: The traits of the CEO may be correlated with the amount of litigation too. CEO age and number of years on the job may predict litigation because older CEOs with extensive experience may be better at managing litigation, and of course the most important variable for us on the right side is the thing we're adding to this equation, which is a binary variable is the lawyer a CEO or not.
Todd Henderson: So we write an equation with all these. We ask a computer to solve the problem and by solution it estimates the influence of particular variables on the thing of interest, the amount of litigation. That output of that calculation is on page three. The independent variables appear in the first column. These are the things we're testing to see whether they influence the amount of litigation. The other columns are types of cases. The number in each column is a coefficient for each independent variable and this is the best estimate of the solution to this problem, that's the influence of every factor. Some variables drive more litigation, those with positive signs, and some drive less litigation, those with negative signs. Some are important, those with stars, and some are not very important, those are the ones without stars. Look at the top row. That's lawyer CEO. How important was this variable in deciding how much litigation a firm would be engaged in, and look for two things, signs and stars. In the second column, all cases excluding products liability, the sign is negative, and there are three stars. CEOs are associated. The fact that you have a lawyer CEO is associated with less litigation, even correcting for all these other things, and the result is very unlikely to be explained by chance.
Todd Henderson: Good for us. There are other stars and signs of interest in the table. Look where the second row intersects the second column. Being a big company, log of total assets, means you're much more likely to be sued. This is the theory of deep pockets. Another example. Look at the impact of volatility and securities fraud. A very strong result, a positive correlation between volatile stock returns and being sued, not a matter of chance. Or take tenure and civil suits. The results suggest that longer serving CEOs are associated with less litigation. Experience matters. What do these numbers mean exactly? Well, that's a good question. Look at the number called a coefficient for civil suits. These are mostly employment discrimination cases and the Coefficient there is -0.3. That means nothing. If you don't know the average number of suits, that's the reduction caused by having a lawyer CEO, but you need to know the baseline how many times our firm sued on average.
Todd Henderson: Well, that's about 0.4. A 0.3 reduction of a chance of being sued about 0.4 times per year is a 75 percent reduction in employment litigation. Similarly, antitrust suits are reduced by 74 percent. Security suits by 72 percent. Labor suits by 38 percent. Personal injury suits by 37 percent, and contract suits by 16 percent. This is a huge result. If we're correct, that lawyer CEOs are associated with less litigation, another question is whether the CEO is doing the work or not or whether there's someone else in the company doing it and it's their presence in the company is just associated with having a lawyer CEO. Maybe all we're observing is the CEO has lawyer friends and the CEO elevates the general counsel in the hierarchy because the general counsel is also a lawyer and CEO who are lawyers like other lawyers. Maybe the CEO who's a lawyer stacks the board with lawyers because they like other lawyers and that these people, the general counsel and lawyer directors are the ones that are doing the work.
Todd Henderson: To test this, we run a regression. We have, and this is seen on on the next page, we put the amount of litigation on the left side and we put lawyer CEO, important general counsels and lawyer directors. And what we find? The lawyer CEO variable is still strong. The effect of a strong GC or lawyer director, sorry, Eileen. This suggests that reduction is driven by the CEO, not indirectly driven by the presence of others in the corporation. Okay. We've gone pretty far, but there's more to go if we want to say anything confidently about a lawyer causing the reduction in lawsuits. After all, it may be there are characteristics that are associated with going to law school that result in less litigation but aren't about what you learn in law school. The result could be driven by the selection story or human capital story. Maybe getting into and graduating from law school is really hard and this is associated with greater skill and ability or a strong work ethic and that's the reduction in lawsuits may arise from a pure selection effect.
Todd Henderson: Law students may just simply be smarter than MBA students. Or consider another selection story. There is credible and persistent evidence that females are more risk averse than males. If this is true, and I don't have a strong view on this, and it's true that women are more likely to be lawyers than MBAs, then the impact we're seeing on firm litigation may just be decisions made by female CEOs who just happened to be lawyers. In other words, gender based risk aversion may be causing the reduction in litigation and also the choice to go to law school. In this story, I'm irrelevant to this outcome, and the quote from the CEO of Home Depot is not what's going on. To address this, we include additional controls in our regression: gender, several background educational backgrounds like did you go to an Ivy League school, and proxies for the talent or professional connections of CEOs.
Todd Henderson: As you can see on page five of your hand out, the results are largely robust to these control variables. The lawyer CEO variable still has stars and the right sign and the other variables don't. This isn't about men versus women. It's not about educational background or which schools you went to. It's something about being a lawyer. Okay. At this point, I hope I've convinced you there's something, some negative associated between a lawyer running a firm and the number of lawsuits at that firm, but there's a question that still should be troubling you. How can we be sure the lawyer CEO is causing the reduction of litigation and not the other way around? After all, perhaps risk averse company hire lawyer CEOs who are also risk averse, hence the attraction. Birds of a feather flock together. And it is this conservatism of firms that's causing the reduction of litigation, not the lawyer.
Todd Henderson: The problem here is ubiquitous in research. If you watch the news, you will see research purporting to show all sorts of things, none of which are likely true. People who eat dark chocolate or drink red wine or drive Swedish cars live longer. I'm not doubting the result. Presumably the researchers aren't lying, but that doesn't mean they're telling us anything useful. The world is full of true but irrelevant information. The problem is that maybe people who are rich or who exercise a lot or go to good doctors or care about their health or whatever, eat expensive chocolate, drink more red wine and drive Volvos. If this is true, other things may be, causing these other things may be causing the outcome. If you hear one of these reports and go out and eat chocolate or buy a Volvo, you're doing it wrong. Maybe the lesson should be get richer or exercise more.
Todd Henderson: Okay. In my research, we use two different techniques to see if we can tease out causation from the data. I'll describe one and mentioned the other briefly. The first is the use of what's called an instrumental variable. The second is a natural experiment. In both cases, we're looking for something that breaks the link between conservative firms and hiring a lawyer as the CEO. In the case of the instrumental variable, we're trying to find something that predicts whether a firm will hire a lawyer as a CEO that has nothing to do with the characteristics of an individual. From in the case of the natural experiment, we're looking for some external factor or shock that changes the environment we can use to test and see how particular firms respond. Instrumental variables and natural experiments aren't easy to find which is why most researchers give up before they have to.
Todd Henderson: So what I'm about to tell you is the most challenging bit. The instrumental variable we use is bit strange, so bear with me. Research has shown that companies have a preference for local talent. When choosing a CEO in one study, about 40 percent of transitions in CEOs took place within a 60 mile radius of the company's headquarters. They picked local talent. We can exploit this preference. Here's how first we add up all the number of people within 60 miles of every firm's headquarters who have a law degree who are either CEOs, directors of public companies, or in the top five of public companies. This is our local CEO talent pool. We then use this instrument to predict the choice of whether a firm will hire a lawyer CEO. In other words, given the distribution of potential CEOs that are lawyers, what is the chance that a firm chooses a lawyer to be their CEO?
Todd Henderson: Then we take this result and use it to predict the filing of lawsuits against these firms. That made no sense, so let me try to explain it again. In the first stage, we assign a probability to a particular company that they hire a lawyer as a CEO based on how many lawyers in the geographic area are plausible CEO candidates. Each company is assigned a percentage chance, and we don't care whether they actually do hire a lawyer to be their CEO. In fact, we don't want to know because if we did know that would then create the problem of the link between firms and picking the CEO. We then take the probability of a firm hiring a lawyer to be their CEO and see whether or not this probability affects the amount of litigation against that firm. Consider two cities: Chicago, lots of lawyers, lots of potential lawyer CEO candidates and Indianapolis, where there aren't as many. In the first stage, we give every Chicago from a probability. They hire a lawyer as the CEO, and then we look at that litigation against Chicago firms and see whether it's less than Indianapolis firms because there's more likely to be lawyers and CEOs. If you look at page six, if I haven't lulled you to sleep already, you'll see that the results.
Todd Henderson: The first is the coefficient in the top, the coefficient of a lawyer being a CEO. It's low, but we can assign it differently based on different geographies. And then when we test that in our regression, we find a negative effect on litigation from where your company's headquarters is and how many potential lawyers there are in that city. That's a pretty remarkable finding. We show the number of lawyers in a local labor market for companies is correlated with less litigation against firms headquartered in those places. This is pretty good evidence, at least from where I stand, that lawyer CEOs are causing the reduction in litigation not that firms that face less litigation are hiring lawyers.
Todd Henderson: Briefly. Let's look at the natural experiment. We looked at the passage of Sarbanes Oxley. Sarbanes Oxley dramatically increased litigation risk for firms and some types of areas, and we looked at how the stock market reacted to firms after the passage of stock of Sarbanes Oxley based on whether they had a lawyer CEO or not. And consistent with our theory across a large number of firms, the market reacted very positively to firms that had lawyer CEOs and not positively to firms that didn't have lawyer CEOs. An exogenous shock that changed the litigation environment and the market like lawyer run CEOs better in high litigation environments after the passage. This is all I can say about causality. It's pretty good but not perfect. I have two last things to say. The first one is about the mechanism and the second is the one that matters the most, which is does any of this matter? in terms of mechanism,
Todd Henderson: I can't say much. We don't have ability to test this. There's two possible theories. The first one is the lawyer CEO shows up on the first day and they just see all sorts of stuff the company is doing that are illegal. Please human resources, stop asking female job candidates if they're going planning on getting pregnant. Hey, head of sales, please stop sharing our price and term sheets with our competitors. These are easy no brainers and maybe the CEO trims the sales. There's an alternative theory which is, when the CEO is a MBA, the head of HR, the head of sales, they push the boundaries, and when a CEO comes into the corner office, they think, oh, I better trim the sales because she's watching what I do and I'm going to be evaluated on a different metric. It could be either of these things and we have no ability to test these.
Todd Henderson: There's a lot of other questions we'd like to test. I'll mention one and then I just want to say something about firm value, and then I'll shut up and take your questions. Uh, we looked at whether going to a top school matters. So we looked at where CEOs went to law school. It had no effect. So whatever it is that is driving the effect we observe, you can learn it at the Toledo College of Law as well as you can at the U of C. Okay. There's one topic left to discuss and that's whether reducing litigation is a good thing. The answer is obvious. Being sued is costly and therefore reducing it as a good thing. Right? Wrong. You went to the University of Chicago, you know that that's not the right answer. Whether reducing litigation is a good thing, depends on the costs of reducing it. The optimal corporate policy would be to spend $1 to reduce litigation until the return from doing so yield $1 in savings.
Todd Henderson: No rational person would spend $2 to avoid a dollar in costs, litigation or otherwise. To test whether firms are behaving optimally, we can look at the valuations of firms run by CEOs with legal training, and the results are on page seven. If you haven't already crumbled your pages up and thrown it at me or someone else. We use a metric called Tobin's q, which is how the sort of market value of a company compared to its book value. If you compare two firms with the same assets, and one is a higher market value, that suggests more value is being added by the CEO. Firms run by lawyers CEOs are less valuable than firms run by non lawyers. The sign's negative and there are stars. One potential explanation for this result is that lawyers pursue risk management through more conservative firm policies at the expense of growth.
Todd Henderson: We present a bunch of paper, a date in the paper that is significant with this. Lawyer CEOs invest less than R&D. They have much less volatility. They are more conservative than non-lawyer CEOs and their conservatism is not limited to litigation. Lawyer CEOs are sale trimmers, and this has good effects and bad effects. But our story is not quite done. Back to page seven, there are three other rows to consider. These are, these are. We broke out industries where there's lots of litigation risk, so high litigation risks industries, industries with lots of growth potential and pharmaceutical firms. Prior work tells us predictors of these are things that predict being sued and we can sort into high and low risk being sued. Pharma companies are particularly susceptible to litigation in a variety of areas. The rows of interest are the interaction between lawyer CEO, and these three sets of industries.
Todd Henderson: Those are in the three rows after lawyer CEO. In each case, the sign is positive and there are stars. In other words, in these companies, high growth, high litigation risk, and pharmaceutical companies, lawyer CEOs create value. The gains from reducing litigation are greater than the costs. While this result should be interpreted with some caution, it's consistent with our prior finding that the value of the lawyer CEO comes from active litigation management. Okay, I'm done, but let me summarize quickly my three main findings. My research shows lawyer CEOs not only reduce the frequency of most types of litigation, but also their severity. It shows this result is at least partially causal rather than driven by pure passively by lawyers CEOs mapping onto firms with low litigation risk. The result is not driven by omitted variables like the CEO talents or the presence of other parties with litigation or legal training like the firm's general counsel or lawyer directors. The results demonstrate that the reduction litigation is consistent with the implementation of more cautious risk management in firm policies, whether it's earnings management, management of analyst expectations, or increased oversight. At the end of the day, risk management pays off in a subset of firms with high litigation risk and high growth firms, but leads to higher lower firm value in all other firms. It seems as an all things, hiring a lawyer sometimes pays off and sometimes doesn't. As for all of us and the question we started with, law school does seem to matter. Thank you.
Todd Henderson: We have 13 minutes and I'd love to hear some questions. Yes, please.
Question 1: Is your data only based on what suits filed against the companies or does it also take whether the lawsuits were filed by companies where the was a lawyer CEO?
Todd Henderson: Great question. So the question was, are we only looking at firms playing defense? What about when they play offense? We collected the data. We're in the midst of looking at it. The sneak preview, the trailer version, although there will be no explosions, is they sue less often too. It reminds me of the quote from Elihu Root who famously said half of a lawyer's job is telling their clients they're damn fools and should stop. And if the lawyer is their own lawyer, I think experience probably tells them, you know, like going to war, you don't want to do it. So that's the preliminary version. We're teasing out whether or not there are any other effects to look at. So offense and defense.
Question 2: [Second question mostly inaudible]
Todd Henderson: Unfortunately, so yes. And the fact that there is a negative association and there is causation leads us to think that in all areas except products liability, that's not true across 3,500 firms. We looked at some examples of transitions and you know, it's in the paper if you're really bored someday and you want to read it. We looked at some transition issues. So how did litigation change for one firm going from a non lawyer CEO to a lawyer CEO. And we didn't find lots of anecdotal evidence to support that theory. There were a few that happened mostly in products liability cases and mostly involving pharmaceutical companies. So I forget if it was Merck or Bristol-Myers or whoever it was, one of them after the scandal hired a lawyer CEO and the other didn't, and their performance really was affected by that. It was a good idea to hire a lawyer CEO to deal with Vioxx Litigation. Honestly, when I sat down to work on this, that was my hypothesis that what we were going to see was the Levitt result with police, more lawyer, more litigation, but not because the lawyer was doing anything that caused the firm to be sued, but they were brought in to clean up the mess. We don't find that.
Todd Henderson: Yeah, please.
Question 3: Within the 3500 firms over a 20 year period, there must be a lot of firms that some of the time had lawyer CEOs and some of the time did not. Did you try to run the correlations within those firms to see if there was a change in litigation?
Todd Henderson: We did and we don't report the results. We didn't see any dramatic change. We also looked at the periods, uh, after, uh, the CEO had been in place for a few years to see if there was a change around, you know, a sort of an equilibrium result. Uh, and we didn't find a significant change in the way that the prior question suggested. If anything after a firm hires a lawyer CEO, litigation tends to fall down and then tends to rise when they choose a non lawyer CEO. But there wasn't enough, there weren't enough results or statistically significant results to present it in what is effectively a finance paper. If this was a law review article, we would have made a lot out of these few examples and then of course given advice in section five of the paper on how companies should be behaving and why they should change all that they're doing. But being a finance paper, we were more humble than that.
Question 4: I've got two questions for you. The first one is, I noticed that you... did you perhaps talk about how you view the R squared results? The second is what your gut is on the leverage correlation? I would assume that highly leveraged environments would bring out the strengths in lawyer CEOs, you know, being able to view documents and attention to the details. And there's a negative correlation there. I just wonder what your thoughts are on that.
Todd Henderson: Okay. So, um, with respect to leverage, I'm just at the regression results here. So there's a negative association between leverage and some lawsuits. You know, I haven't given it a lot of thought. Oh, firm value. I see. So the, the, what is the question again? Whether or not...?
Todd Henderson: Oh, that could be. And that's an interesting hypothesis and certainly we could do more to tease out that, to look at the financial leverage and do more accounting work and look at how CEOs respond in that environment. I'm not, I haven't looked at that result. So it's hard for me to, to come up with a good answer off the top of my head. It seems like a plausible hypothesis that we could test. It may be though that for highly levered firms, performance really matters more. I mean, I think that's the idea when private equity firms buy a company and they increase the leverage. The debt is discipline, and they're really pushing for extremely good performance. And the idea of being a conservative lawyer CEO, I don't associate it with risky, highly levered firms. I associated with more like managing pretty stable companies. But we can certainly test that. About the R squared's,
Todd Henderson: I think I might do a phone a friend here and call in the Dean here who is actually a pretty way more sophisticated empirical scholar than I am. He mentioned the partnership that is happening in law schools. And one of the interesting things about this paper, so I apologize if you were bored or lost, is this is sort of the way of the future of legal scholarship. For a long time, it's just been, you know, here's my pontification on the world. There's a lot more data being used and I'm not teched up enough to know the sort of uber-level details. This is a partnership, and I have a co-author who's a finance professor. So what should I make of R squareds that are 0.1 relative to the power of the statistical tests? I'm going to punt to the Dean.
Dean Miles: I mean, my reaction to the R squared's is that... they're pretty good.
Todd Henderson: Okay. There you go. Straight from the horse's mouth. They're pretty good; they're not great.
Dean Miles: I have a question while I have a microphone. Dean's privilege. Given your results, do you have an investment strategy for recommending basis?
Todd Henderson: So that's a great question. I think the answer is, so I don't get in trouble with FINRA, I don't want to be passing out investment advice. My advice to all of you don't ever invest in anything other than a low fee mutual fund. But other than that, and by the way, in fact, Fidelity issued a free mutual fund that's being paid for by money they make from securities lending. So it's a zero mode fund, which is I think, the way of the future and pretty awesome. To answer your question though, yeah. So it's possible that if you see a firm that is not in a high litigation area industry and they hire a lawyer CEO, sell. That person is on average going to destroy value. I see. My neighbor has a question.
Question 5: How did you deem who's a lawyer? Anyone with a law degree or a JD MBA? Or a certain number of years in practice?
Todd Henderson: Yeah. So just uh, you went to and graduated from law school. We would have loved to look at. Did you have a JD MBA? Did you work in government versus private practice? Were you a prosecutor, defense lawyer? Did you go in house? There's lots of nuances that we could, you know, how is this correlated with years of professional experience? Lots of interesting questions we could ask. With only about 350 lawyer, we couldn't get the data and Dean Miles could tell us exactly why and the statistics of that, all that. But we just couldn't be confident with our answers. We did look at things like education. Did you go to Columbia or did you go to Kalamazoo and we didn't see a result. We looked at some other things, again, with smaller data sets, like did you work as an in house council, and we just didn't have the statistical power. We didn't. We didn't find a result. So when we got a null result, but we couldn't be really excited about the null result because we didn't have the... there were no stars.
Question 6: One question I have, that maybe follows up, in terms of pool of applicants that corporations look at for the CEO position, it would seem to me that they would come from corporate type environments that were large law firms as opposed to say pool of lawyers who may be more risk takers, plaintiff's lawyers, people who are perhaps more entrepreneurial in their approach to practicing law. And I just wondering if that kind of skews the data set towards more conservative value.
Todd Henderson: Great. So the question was whether or not the lawyer CEOs are a skewed data set with respect to lawyers generally. And that must be true. It's probably the case that our risk preferring, like if we rank ordered the current law school graduating class from the most risk preferring to the least, you'd end up with business, plaintiff's lawyers, defense firms, and then you know, I don't know, government. That probably is a pretty good spread, and while we're pulling from are the lawyers who are going into business, the most risk preferring, we're not going on the defense counsel. So I think my pool is the most risk preferring, but they're not plaintiff's lawyers, but I think the plaintiffs' lawyers are probably less have less risk preference than the people who go into business, but maybe not. The people who go into business, in house as lawyers and become CEOs are probably risk takers, more risk takers than people who are plaintiffs lawyers, would be my guess. But our evidence suggests that compared with people who go and get an MBA, the most risk preferring lawyers are more risk averse than the people who go get MBAs. And as, I mean in my class, I observed this all the time. I ask them all kinds of risks questions. These are about the most risk averse people you could possibly imagine. Then we spend three years telling them all the horrible things that can happen in the world.
Todd Henderson: We're doing a double whammy. Okay. One more question
Question 7: Okay, so litigation is very public. Scandals are very public, and leverage is very public. As someone who has practiced in house for most of my career, what is done in house is often not public. It's not on television. ... You are never going to know it. So I am curious whether you gave any thought to the fact that a lawyer CEO, their actual... the minute they step into the company, the way they think, their strategies that change the company, you know, just changing a lot of policies and things that you cannot measure it or see it.
Todd Henderson: Great.- What do we do about the invisible things that happen inside of firms? So we're at a huge disadvantage. We're observing only the things that we could observe. I can't see stuff that is invisible. So you're, you're of course you're right and you have a lot more experience at this than I do. So I as a researcher, I'm assuming there's all kinds of stuff happening at firms that I can observe. Just like there's all sorts of things that are happening inside of human beings that we cannot observe. As researchers, all we can do is look at the stuff that you do and what we're interested in is metrics litigation get involved in when the stuff that you're working on that secret goes wrong and there ends up being a lawsuit that's filed. I observe that and then I observe how much the market values your firm.
Todd Henderson: Again, it may be a great place to work and the lawyer CEO creates an environment where everybody is Kumbaya and feels really great and it's all great, but at the end of the day, if it doesn't make any money, it's not really worth a lot to the external shareholders, to the stakeholders, so those are the two things I'm looking at. That doesn't mean that's the only two things that matter in the world. Creating workplaces that are happy places for people to work at. Great. Start a company and go and do it. I just can't. I don't have good data about that. We could go around and ask people what their happiness is working here or there and see if CEOs are correlated with average employee happiness or some other things. So we're limited in the things that we're looking at. I don't mean to suggest that's all that matters in the world. If you're starting a company and you want it to have some other values, maybe social value in the world, like doing social justice through your firm. Maybe hiring a lawyer is fabulous and it's always the best thing to do. That's just not what we were looking at for the data. Okay. Thank you all for coming.
Dean Miles: Thank you very much. That was a fascinating talk. A talk that was was more than just pretty good. It was a real pleasure to welcome everyone to Reunion 2018 and to our Loop Luncheon. Thank you all very much for coming. For those of you who are here for the reunion, I look forward to seeing you all throughout the weekend. Again, welcome and please join me one more time in thanking Professor Henderson.
Announcer: This audio file is a production of the University of Chicago Law School. Visit us on the web at www.law.uchicago.edu.