ProfessorBainbridge.com

Behavioral Economic Analysis of Law

Law and economics remains the most successful example of intellectual arbitrage in the history of corporate jurisprudence. It is virtually impossible to find serious corporate law scholarship that is not informed by economic analysis. Even those corporate law scholars who reject economic analysis spend much of their time responding to those who practice it.

As with any model claiming predictive power, law and economics rests on a theory of human behavior. Specifically, neoclassical economics is premised on rational choice theory, which posits decisionmakers who are autonomous individuals who make rational choices that maximize their satisfactions. Critics of the law and economics school have long complained that rational choice is, at best, an incomplete account of human behavior. The traditional law and economics response to this complaint is that rationality is simply an abstraction developed as a useful model of predicting the behavior of large numbers of people and, as such, does not purport to describe real people embedded in a real social order. Until quite recently, moreover, empirical research tended to confirm that the rational choice model of human behavior is a good first approximation of how large numbers of people are likely to behave in exchange transactions.

Over the last 10-15 years, however, a new school of economic analysis has emerged that challenges the rational choice model precisely on its predictive power. Empirical and laboratory work by cognitive psychologists and experimental economists has identified a growing number of anomalies in which behavior appears to systematically depart from that predicted by rational choice. A good example is the so-called status quo bias: All else being equal, decisionmakers favor maintaining the status quo rather than switching to some alternative state. The status quo bias can lead to market failure where decisionmakers’ preference for the status quo perpetuates suboptimal practices.

My thoughts on this topic were prompted by hearing my colleague Russell Korobkin -- a member of the Volokh Conspiracy (albeit a rarely seen one) -- give a presentation to a faculty workshop on behavioral economic analysis of form contracts; his presentation was based on an excellent paper, by the way, which he has posted to SSRN. (Strongly recommended.) I got to thinking some more about behavioral economics after having lunch last Friday with my colleague Victor Fleischer of A Taxing Blog. Behavioral economics is something of the flavor of the month in legal education, especially on the part of those with an ideological disposition against the free market connotations of neoclassical economic analysis. After a brief flirtation with it myself, I have become much more skeptical, for the reasons developed at length in the Extended Post below.

Empirical demonstrations of this decisionmaking bias have focused on the so-called endowment effect. Subjects commonly place a higher monetary value on items they own than on those that they do not own, even if the two items have the same market value. Accordingly, subjects must be paid more to give up something than they would be willing to pay to acquire the same object. The classic demonstration of the endowment effect variant of the status quo bias was a laboratory experiment in which students were initially endowed either with a coffee mug or six dollars cash. Mug holders were asked to identify the minimum amount they would accept to sell the mug, while cash holders were asked to specify the maximum amount they would be willing to pay to purchase a mug. Subjects were told that a market-clearing price would be determined and trades executed between mug holders willing to accept that amount and cash holders willing to pay that price. It turned out that the price demanded by mug holders was about twice that cash holders were willing to pay, so that very few trades took place.

I do not deny that there is considerable empirical evidence for the endowment effect and the status quo bias. More generally, I also do not deny that behavioral economics is a potentially useful tool that any legal scholar should have in his toolkit. As the Economist explained:

[I]f the endowment effect is real, people's economic decisions are fundamentally different from what economists have assumed. The implications of this are profound. To take one example, the Coase theorem, which argues that initial allocations of wealth do not matter as long as markets allow people to trade their stakes—the rationale for government auctions of everything from radio spectrum to mobile-telephone licences—would no longer be valid. To take another, although economists have shown that you need only a few sharp traders for prices in financial (and other) markets to become efficient, the volume of trade with an endowment effect will be below what it might be without one.

But it is one that must be used cautiously; too many legal scholars are using it far too glibly. Consider, for example, the evidence that the endowment effect appears to vanish when people do not physically possess the commodity in question. Subjects trading tokens or vouchers demonstrate only a weak endowment effect. Because capital market transactions more closely resemble the token or vouchers context then experiments involving physical possession of a tangible commodity, for example, these results call into question the extent to which one can rely on the endowment effect as evidence of a capital market failure. [Update: Korobkin passed on one of his articles, which presents a more nuanced account of this evidence. The money quote is "Money itself does not create an endowment effect, but the effect does appear to exist for financial instruments that are valued only for the money they are worth (i.e., have no intrinsic value themselves) if the value of the instrument is uncertain. These results suggest that securities and other financial instruments can create an endowment effect even though they are held as stores of wealth rather than for their intrinsic or 'use' value." The Endowment Effect and Legal Analysis, 97 Northwestern University Law Review 1227, 1236-37 (2003). Is not clear to me, however, why the results Korobkin discusses are not better characterized as mere risk aversion as opposed to an endowment effect.]

More important, in light of the above update, as the Economist also reported, there is emerging evidence that market actors can learn their way out of biases like the endowment effect:

John List, an economist at the University of Maryland, recently tested the existence of the endowment effect in a new way. Instead of using callow students, he went to a real market with traders of varying degrees of experience: a sports-card exchange, one of many such, where Americans trade pictures of their favourite athletes. There, traders dealing in hundreds of cards mix with browsers who might buy only one.

List found:

[E]vidence for an endowment effect—but also that long experience as a card trader spilled over into his experimental mug-and-chocolate market. Only novices, like the students in earlier experiments, tended to be swayed by what they had been given. This implies that prospect theory can capture the behaviour of inexperienced people, of which the world has many in all sorts of markets. But experienced buyers or sellers in well-established markets get over their psychological “flaws”. They can even transfer their trading skills from one market to another. The neoclassicals, it seems, have scored a point.

I would add a different, but I think equally significant reservation: The claim that law can correct market failures caused by decisionmaking biases or cognitive errors treats regulators as exogenous to the system. Once the state is endogenized, however, regulators must be treated as actors with their own systematic decisionmaking biases. It thus becomes evident that behavioral economics loops back on itself as a justification for legal intervention. Public choice theory provides still another reason market failure is not a sufficient justification for government intervention. Again, the problem is one of treating an endogenous factor as exogenous. A welfare economics model that posits legal intervention as a solution to market failure ignores the fact that regulators are themselves actors with their own self-interested motivations.

In sum, there is a strong temptation to use behavioral economics too glibly. Advocates of government intervention are akways tempted to jump from positing the status quo bias, citing the coffee mug experiments, to an assertion that the government needs to shake up the status quo, without demonstrating that the bias is truly valid in the specific setting at hand. In addition, a certain degree of skepticism about the power of law to effect social change seems warranted. Indeed, behavioral economics itself offers additional reasons to doubt the capacity of law as agent for social change. Finally, one cannot justify government intervention without asking whether the case for it survives endogenizing the state.

You can read an even more extended version of my argument, with application to the longstanding debate over mandatory disclosure in securities regulation, in my article Mandatory Disclosure: A Behavioral Analysis, 68 University of Cincinnati Law Review 1023 (2000).

Can a Board Meet Online?

The Wall Street Journal (subscription required) has a good article in Monday's technology section on how corporations are using dedicated board of directors websites to let directors communicate with management and each other. But here's the part that caught my eye:

The main functions of a directors' Web site are simple enough: easy online access to briefing papers and internal company data, so directors can do their homework at home or on the road a few days before board meetings and show up fully informed; the ability to fine-tune a document online, so directors can hold a brief but urgent committee meeting online from different locations; and an exclusive e-mail system.

Can they really meet online? It depends on whether they use Internet telephony or a real-time text-based mechanism (like a chat room). Delaware General Corporation Law sec. 141(i) provides that directors may conduct a meeting:

by means of conference telephone or other communications equipment by means of which all persons participating in the meeting can hear each other....

Web-based supplements to a telephone conference call thus are fine, but real-time text based messaging is OUT. To conduct a meeting exclusively online, in a Delaware corporation at least, you must use Internet telephony. I wonder if any Delaware corporations have held invalid text-based board meetings online? I emailed this post to the reporter who wrote the story and asked him. I'll report back if he answers. In the meanwhile, I once wrote an article motivated by DGCL section 141(i)'s requirement that directors be able to hear one another. Why a Board? Group Decision Making in Corporate Governance, 55 VANDERBILT LAW REVIEW 1 (2002). To be clear, the article's not about online meetings -- its a theoretical inquiry (neoinstitutional and behavioral economics with a dose of social norms theory) into why corporations are run by a board rather than an individual autocrat (of course, in practice, many CEOs are individual autocrats). But section 141(i) was what got me thinking about the theory issues.

The requirement that members be able to “hear” one another seems quaint in an era of electronic mail, instant messaging, and internet chat capabilities. Yet, when Delaware recently amended its corporation statute to permit much greater use of electronic forms of communication, it retained the requirement that board meetings be conducted in such a way that all members may hear one another. As it turns out, this appears to have been the right choice. Research on decision making has found that groups linked by computer make fewer remarks and take longer to reach decisions than do groups meeting face to face. Kiesler and Sproul, for example, not only found that meetings conducted through computers result in greater delays, but also that the decisions made in such meetings were more likely to exhibit the risky shift phenomenon. Sara Kiesler and Lee Sproul, Group Decision Making and Communication Technology, 52 Org. Beh. & Human Decision Processes 96 (1992). They also found that time-constrained groups exchanged much less information when meeting electronically than when meeting face-to-face.

Electronic communication takes place mostly through text-based mediums. For many people reading and typing are slower and require greater effort than verbal communication. Text-based communication also deprives participants of social cues, such as body language and tone of voice, that may be important signals. Social norms constraining behavior apparently function less well in text-based communication, as illustrated by the flame wars that plague Usenet newsgroups.

As with other aspects of the rules governing board meetings, accordingly, there seems to be a legitimate basis for otherwise formalistic rules. Even such housekeeping rules as notice requirements prove to be consistent with the research on group decision making. Unless the articles of incorporation require otherwise, no notice of regularly scheduled board meetings is required. Special meetings require at least two days notice. As a matter of statutory law, the requisite notice need not announce the purpose of the meeting. Because the directors’ duty of care requires them to make an informed decision, however, it is advisable whenever possible to provide directors of advance notice of the reason for calling a meeting and any relevant documentation. As with other requirements relating to board meetings, the notice rules are intended to ensure that the board functions as a collegial body, all of whose members participate and get the benefit of the participation by all other members. MBCA § 8.23 cmt.

That notice requirements effectively carry out that function is suggested by research on group performance. Michaelsen et al. conducted a study in which individuals were pre-tested and then re-tested as members of a group. Under those conditions, groups outperformed individuals. Michaelsen et al. analogize this testing order to organizational decision-making processes in which “group members prepare a position paper and circulate it to other group members prior to problem-solving discussions.” Larry K. Michaelsen et al., A Realistic Test of Individual Versus Group Consensus Decision Making, 74 J. App. Psych. 834, 834 (1989). A board meeting conducted after meaningful notice likewise replicates this testing order.

UPDATE: The WSJ reporter (Geroeg Anders) have given me permission to blog his response: "As far as I can tell, boards (either deliberately or inadvertently) are staying within the bounds of the Delaware code as currently written. Even when computer-to-computer interactions are a meaningful part of a remote board/committee meeting, there usually are parallel phone lines that are open, so that people can ask questions the old-fashioned way."

SEC injunction against Hollinger

The SEC's successful pursuit of an injunction against changes in composition or mission of the Hollinger special committee has gotten less attention than it deserves. At base, the Hollinger scandal is about allegedly unauthorized payments by Hollinger to Conrad Black and some of his cronies. This is classic corporate governance stuff - governed by Delaware General Corporation Law § 144's provisions on conflict of interest transactions. Granted, there are related federal fraud issues arising out of the nondisclosure of the challenged transactions. The federal claims, however, are essentially seconday to - even parasitic on - the state law claims. For the SEC to invoke federal authority therefore allows the tail to wag the dog. When that federal authority is used to impede core corporate governance powers - selecting the membership of a board and its committees - we have a dramatic federal preemption of state law. I think the creeping federalization of corporation law is a very bad idea. I think federalizing corporate law without even having a debate on the subject is an even worse idea.

Sarbanes-Oxley Winners and Losers

Dwight Klingenberg, a business journalist, writes:

A little over a year ago, Congress passed the most sweeping legislation affecting businesses since the creation of the Securities and Exchange Commission in the '30s. Although the Sarbanes-Oxley Act - we'll call it Sarbox from here on - itself is a scant 66 pages, the subsequent implementation rules and regulations will fill numerous hard drives over the next few years. At this stage, who are the winners and losers?

66 pages is scant? Anyway, go read the whole thing. To whet your appetite, here's a summary of winners and losers:

Winners
Treadway Commission
The SEC
Law firms
Software vendors
Outsourcers

Losers
Research analysts
The AICPA
Private companies and agencies

Too close to call
In-house counsel
Public companies [This was the one I was prepared to question, leaning towards putting public companies into the loser category for the reasons stated in thefirst sentence that follows, but I am semi-persuaded by the second sentence. I'll need more data before making up my mind on that one.]

Public firms are now strapped with significant new costs: whistleblower hotlines, expanded internal audit departments, extensive control documentation, higher director compensation, shocking increases in directors' and officers' insurance and bigger fees from their auditors. However, most CFOs we speak to will admit privately that the documentation effort resulted in finding control issues and shortcomings they are fixing.

CPA firms
Directors

New Article on Sarbanes-Oxley Section 307

The Social Science Research Network is a site for legal, economic, management, and accounting scholars to post working papers before they are formally published in a professional journal. All of my articles since 1998 are collected there, along with most of the ones published before that date. SSRN just released to its data base my newest article, co-written with UCLA grad Christina Johnson, entitled "Managerialism, Legal Ethics, and Sarbanes-Oxley Section 307. This article was prepared for a conference on the Sarbanes-Oxley Act (a.k.a. the 'Public Company Accounting Reform and Investor Protection Act' of 2002), this Article focuses on the professional responsibility rules promulgated by the Securities and Exchange Commission under Section 307 of the Act. According to the theoretical model of corporate governance espoused by all business corporation statutes, a corporation is to be run by its board of directors for the benefit of its shareholders. In practice, however, corporations frequently are run by their top managers for the benefit of those managers.

A number of recent trends have empowered boards of directors vis-a-vis management. As this Article's review of the statutory text and its legislative history demonstrates, Congress intended the Sarbanes-Oxley Act to further that trend. We further demonstrate that Section 307 should be understood as part of the Act's overall anti-managerialist intent. Congress sought to enlist legal counsel in strengthening the board. Specifically, Congress directed the SEC to create an up the ladder reporting requirement pursuant to which a firm's legal counsel would report evidence of misconduct to the board of directors, thereby redressing one of the information asymmetries between boards and managers.

This Article argues that, as a normative matter, Sarbanes-Oxley Section 307 was well-intentioned. As a practical matter, however, Section 307 seems unlikely to effect significant changes in corporate governance. In our view, the nature of legal practice, the largely unchanged relationship between lawyers and managers, and the problematic approach taken by the SEC to implementing Section 307 suggest that the new legal regime is unlikely to result in significantly better information flows within the corporate hierarchy.

You can download the full paper HERE.

Why the Problem at the NYSE is the same as the Problem at Freddie and Fannie

The Wall Street Journal editorializes today (subscription required) on Treasury Sec'y Snow's testimony yesterday on Freddie Mac and Fannie Mae:

Arguing that publicly traded companies should have only private directors, he suggested the time had come for Fan and Fred to eliminate the directors on their boards who are now appointed by the President. Good corporate governance, he crisply noted, requires that the people who are running companies serve the stockholders.

Apropos of that point, a reader writes:

My opinion is that Freddie Mac's board [mild expletive deleted] its shareholders by ordering an investigative law firm to try to pin blame on senior management (and therefore not on the Board) for an accounting scandal. The result may have covered the board's posterior, but it damaged the company both internally and externally. Is there any solution for this in terms of corporate governance? My guess is that shareholder lawsuits are a lousy tool for creating Board accountability, but is there a better tool that would not have the side effect of giving board such a strong incentive to cover itself that it hires a hit squad investigator to go after management and trash the entire company?

How does this tie into the on-going Grasso pay imbroglio? The trouble with all three entities is that they are neither fish nor fowl. Their government connections insulate them from discipline by markets and investors. The NYSE is part of a trading market oligopoly with very high barriers to entry, some of which are attributable to SEC rules. (The SEC, for example, long let the NYSE get away with listing standards making it almost impossible for a firm to de-list.) In the case of Freddie and Fannie, though, its even worse: the markets believe (with some justification) that the federal government has (implicitly) guaranteed their debts, which allows them to avoid by market discipline by borrowing at below-market rates. Moreover, because their board includes political appointees, both are further insulated from investor discipline. The solution for all three is privatization. Let them run as for-profit corporations in competitive markets, with full disclosure (none of them are very transparent as to either governance or finances).

UPDATE: Broc on Grasso:

And don't get me started about the NYSE's governance structure. True, its not a public company but it does have all the earmarkings of a poster child for bad governance. A CEO who handpicks the board - and the compensation committee. A compensation committee who approves CEO compensation arrangements that it doesn't understand. A compensation committee comprised of CEOs from companies that have inherent conflicts of interest with the CEO by virtue of him being their regulator.

Check.

UPDATE 2: Gregg Easterbrook blogs:

Turns out the compensation board that approved Grasso's grotesque bonus contained executives from big companies that had a keen interest in insuring that the NYSE did not act against stock-market manipulation. Citigroup, Morgan Stanley and Merrill Lynch--the key offenders who admitted in the recent $1.4 billion Securities and Exchange Commission settlement that they deliberately misled investors--all had seats on the committee setting Grasso's pay. So did AOL Time Warner which, during the period Grasso was being assigned the grotesque bonuses, was engaged in the most determined campaign to wipe out shareholder value in American business history. These tainted companies were in effect offering Grasso huge amounts of money if he played along and made sure the NYSE did nothing to expose them--and Grasso, a faithful water-carrier, made sure the NYSE did absolutely nothing. There's a word for all this, and the word is "corruption."

Check again.

Sarbanes-Oxley Ethics Codes

Kansas city corporate attorney Arthur Chaykin sends along the following observations:

I had been hopeful that the Sarbanes-Oxley Act would have one saving virtue: I thought, perhaps, that CEO's would do what they could to improve business decision making in their companies if only to avoid immediate disclosure through anonymous "hotline" directly to the Board. I reasoned that such calls could, at the very least, become an annoyance for CEO's and, at worst, could give the Board the impression that the CEO did not know how to run the company. Therefore, I thought it was at least possible that CEO's would "pump up" corporate ethics codes and programs in an earnest attempt to reduce the number of "bad acts" within the company.

However, on further reflection, I am concerned that the Sarbanes-Oxley Act may have the exact opposite impact. As you know, the Act requires an ethics code, at least for financial officers, or an explanation as to why no ethics code is presented (and no company will want to explain that). Furthermore, the Act requires a company to explain any "deviation" from its ethics code. As I read it, that means that anytime a company decides to make a reasonable exception to a conflict of interest policy, it has to do a public filing. In order to avoid that, I am sure companies will try to (a) obfuscate their codes so that it will be hard to know whether a violation has occurred or not (thus defeating the entire purpose) or (b) attempt to reserve discretion in some body within the corporation so that they do not have to report the exception as a deviation (but it is not clear that will work). So, I expect a lot of companies to select (a). I have had one general counsel indicate that she had just written the new code and made sure that it only did the bare minimum because she did not want to incur additional disclosure obligations.

Can you believe it?

Yes, I can believe it. Arthur is exactly right that no company in its right mind will want to explain the absence of an ethics code (in the trade we call this therapeutic disclosure, about which I will be posting soon). He's also exactly right that no company will want to be put in the position of disclosing deviations from the code. Hence, we're going to get bare bones ethics codes, which is exactly what seems to be happening.

A Review of "The Company"

The Company: A Short History of a Revolutionary Idea. John Micklethwait & Adrian Wooldridge. New York, The Modern Library, 2003. Although I recommend this text to generalist readers seeking a (remarkably) concise introduction to history of the business corporation, personally I came away somewhat disappointed. There is little doubt Micklethwait and Wooldridge are correct in their claim that the corporation is now the key economic institution in Western nations. Yet, it did not have to turn out that way. As Micklethwait and Wooldridge usefully remind us, two centuries ago, leading business and economic thinkers (including the great Adam Smith) derided the joint stock company. What explains the relatively rapid development in the mid-19th century of a recognizably modern corporation and, in turn, that entity's emergence as the dominant form of economic organization?

Micklethwait and Wooldridge offer a fairly conventional answer to that question, based largely on new technologies -- especially the railroad -- requiring vast amounts of capital, the advantages such large firms derived from economies of scale, the emergence of limited liability that made it practicable to raise large sums from numerous passive investors, and the rise of professional management. Readers familiar with the work of business historian Alfred Chandler will find relatively little new in this part of the story, although Micklethwait and Woolridge's treatment has the advantage for generalist readers of being considerably more accessible than is most of Chandler's work. Instead of offering any novel historical analysis, Micklethwait and Wooldridge's principal potential contribution (albeit one they failed adequately to realize) is the normative thesis to be derived from the historical account.

In their introduction, Micklethwait and Wooldridge lay out a claim that will be familiar to readers of Michael Novak's work (surprisingly, however, they seem unaware of his seminal work). Like Novak, Micklethwait and Wooldridge argue not only that the corporation is one of the West's great competitive advantages, but also that the number of private-sector corporations a country boasts is a relatively good guide to the degree of political freedom it provides its citizens. Unfortunately, this insight goes nowhere.

The normative claim is entirely plausible. The rise of modern corporations did more than just expand the economic pie. The legal system that facilitated their rise necessarily allowed individuals freedom to pursue the accumulation of wealth. Economic liberty, in turn, proved a necessary concomitant of personal liberty -- the two have almost always marched hand in hand. In turn, the modern public corporation has turned out to be a powerful engine for focusing the efforts of individuals to maintain the requisite sphere of economic liberty. Those whose livelihood depends on corporate enterprise cannot be neutral about political systems. Only democratic capitalist societies permit voluntary formation of private corporations and allot them a sphere of economic liberty within which to function, which gives those who value such enterprises a powerful incentive to resist both statism and socialism. As Michael Novak has observed, private property and freedom of contract were indispensable if private business corporations were to come into existence. In turn, the corporation gave liberty economic substance over and against the state. Regrettably, after laying it out, Micklethwait and Wooldridge fail to pursue this thesis. Instead, their book lapses into mere narrative history.

Having said that, however, it is exceptional narrative history. As journalists for the Economist, which I regard as the best-written magazine around, they write clearly yet powerfully. There are numerous insights, cleverly turned phrases, and interesting anecdotes. All of which makes for a compelling read, if not a compelling normative argument.