Corporate governance reform in Japan: A behavioral view

Japan is promoting corporate governance reform through amendments to its corporate law (“the Companies Act”) and stock exchange rules. One major change under the corporate law (§327-2) has been the introduction in 2014 of a requirement that at least one independent outside director (“independent director”) be present on the board, based on a comply-or-explain rule mandating explanations from firms which do not comply. The comply-or-explain concept was influenced by the OECD Principles of Corporate Governance of 1999, succeeded by the G20/OECD Principles of Corporate Governance of 2015 (Kanda, 2019). This legislation applied to large firms, defined as those, listed or unlisted, with at least 500 million yen (about $4.5 million) in paid-in capital or 20 billion yen (about $180 million) in debt. The amendment was followed in 2015 by the corporate governance code, a stricter requirement by the Tokyo Stock Exchange calling for listed firms to have at least two such directors. As of 2019, 93.4 percent of the firms listed in Section 1, its largest section, had complied (Tokyo Stock Exchange, 2019). A further amendment to the code in 2018 suggested a third of directors being independent when firms believe it necessary in view of their size and other relevant factors. In 2019 a stricter amendment was made to the corporate law as well, mandating all large firms, without exception, to have at least one independent director. As more institutional investors outside the country invested Abstract

in Japanese firms, Japan's corporate governance reform came to reflect a growing perception of the need for independent directors on the model of the U.S. and English law. The shareholdings of non-Japanese institutional investors in Japanese listed firms grew from 4.7 percent in 1990 to 29 Under Japanese corporate law, the traditional corporate format has been a unique one involving a company with statutory auditors. Unlike external auditors, statutory auditors have no qualification requirements and are elected at the shareholders meeting along with directors. While statutory auditors attend meetings of the board of directors, they have no voting rights. At least half of those for large firms are required to be outsiders, statutory auditors often being perceived as people who have missed being promoted to a directorship. The format dates back to 1890, when a modernizing Japan imported commercial laws from Germany, and underwent significant changes in 1950 by the introduction of the concepts behind the Illinois Business Corporation Act of 1933, which is also a foundation for the U.S. Model Business Corporation Act (West, 2001).
To accommodate more independent directors, an amendment to the corporate law in 2014 gave firms the option to abolish the unique system of statutory auditors and introduce in its place a board of directors having an audit and supervisory committee with independent directors as a majority of its members. While the authority of statutory auditors is limited to monitoring the legality of directors ' decisions, audit and supervisory committee members have the greater authority of monitoring whether such decisions are valid even when legal. The committee also expresses its views on board elections and remuneration at the shareholders meeting. The number of firms that have abolished statutory auditors and put in a board system with an audit and supervisory committee has increased to 26.8 percent in the four years since the provision's enactment in 2015 (Tokyo Stock Exchange, 2019). The corporate law has already offered a more advanced format since 2003, mandating nominating and compensation committees as well, each comprising a majority of independent directors (a "three-committee company"). Under the two corporate formats, only the board of directors, as a more diverse group including independent directors, will be answerable to shareholders (the unitary board system).
Reform of the corporate governance framework, as typified by the introduction of the new corporate format and the increase in independent directors, is often said to have the dual objectives of "offense" and "defense." Offense means realizing management efficiency and sustainable growth, while defense means preventing the kind of corporate misconduct recently seen in established Japanese firms. Unlike in past reforms, which focused on preventing wrongdoing, the emphasis on sound growth reflects the economy's two "lost decades," when the market capitalization of Japanese firms remained stagnant. On the other hand, despite the rapid adoption of unitary board systems requiring the involvement of independent directors, what is introduced as a rule or regulation often ends up as a mere formality, as if the placement of independent directors per se answers the needs of, for instance, institutional investors.
To examine the conditions needed for this change to produce its intended effects, it is worth looking at both the offensive and defensive aspects of board reform by applying the findings of psychological experiments while keeping a focus on group decision-making. This approach helps to build a foundation on which firms can make effective decisions, whether they adopt regulations either as a formality or in substance. In particular, it is helpful to consider whether adding independent directors is supported from the viewpoint of group dynamics, and how the findings of group dynamics suggest conditions required for an effective institutional architecture. Psychology tends to focus on how a change in a group's composition produces high-level collective intelligence by integrating diverse knowledge within a group (offense) while preventing corporate misconduct by introducing outsiders' viewpoints (defense). The objective of the analysis by this paper, and its contribution, lies in viewing corporate governance architecture from both offensive and defensive perspectives and in analyzing the functions of boards. For this purpose, the paper looks at boards through the behavioral lens, exploring the conditions needed for boards to produce collective intelligence and prevent corporate misconduct, and thereby contributing to the actualization of Japan's ongoing corporate governance reform. This paper is organized as follows. Section 2 reviews the existing literature. Section 3 describes the research methodology and conceptual framework by which the paper analyzes boards of directors. Sections 4 shows the results of analyzing offensive and defensive aspects based on the framework. Section 5 discusses our findings based on these results. Section 6 concludes the paper.

LITERATURE REVIEW
The behavioral approach adopted in this paper derives from the discipline of law and economics, in which an economic framework is applied to the analysis of laws and institutions. Law and behavioral economics seek to reflect actual human behavior and psychology by applying behavioral findings to traditional law and economics issues. Generally speaking, as we see behavioral economics developing from economics, law and behavioral economics similarly develop from law and economics. Behavioral science, in particular, provides a powerful lens through which to observe board members' behavior, along with a framework to analyze group judgment and decision-making in that context.
An analytical approach to law and behavioral economics is already found in Conard (1972), but applications to modern corporate governance appear later, reflecting the development of behavioral economics: Sunstein (1997), who contrasts rational choice with actual human decisions; Jolls, Sunstein, and Thaler (1998), who expand the bounded rationality concept to government's regulatory design; Korobkin and Ulen (2000), who criticize assumptions of economic rationality in economics to explain the real social phenomenon; Baron and Wilkinson-Ryan (2018), who emphasize descriptive and prescriptive powers of behavioral findings; and Zamir (2019), who proposes behavioral foundations complementary with predictions of rationality-based economic analysis. This research critically compares traditional and behavioral economics and argues for the effectiveness of applying the latter to law.
Based on these foundations, more specific applications to boards of directors follow Bainbridge (2002), who looks at the functioning of the board as a decision-making group; Langevoort (2001Langevoort ( , 2012, who emphasizes the monitoring roles of independent directors; Morck (2008), who deals with the effect of bias on independent directors in their monitoring role; Marnet (2008), who points to insufficient monitoring arising from conflicts of interest between firms and independent directors; Van Ees, Gabrielsson, and Huse (2009), who look at political and negotiation processes among directors in their decisions on the board; Greenfield (2015), who extends psychology to the traditional contract theory describing board-shareholder relationships; and Zamir and Teichman (2018), who propose an integrative view of behavioral analysis of corporate law, securities law, and antitrust law.
Research that puts more emphasis on managerial decision processes has developed in parallel, such as Forbes and Milliken (1999), who construct decisions by boards of directors as a group decision-making model and connect it to board performance; Blair and Stout (2001), who deal with group productivity of boards and contrast it to the agency model of Jensen and Meckling (1976); Hambrick, von Werder, and Zajac (2008), who broaden the perspective to include comprehensive processes of decision-making and actions inside and outside firms; Powell, Lovallo, and Fox (2011), who examine strategy formulation as a managerial decision made by boards; and Westphal and Zajac (2013), who further position boards of directors in the social context in which firms and boards are situated.

RESEARCH METHODOLOGY
One of the early criticisms of law and economics is that discrepancies arise when trying to apply profit-maximizing, economically "rational" assumptions to laws when such assumptions are in natural contradiction to actual human behavior. As humans, we often make mistakes; we forget things, lose our temper and calm down again. Since corporate decisions are delegated to boards made up of humans, it is reasonable to assume that common psychological processes are going on. In seeking an optimal place between formality and substance in corporate governance reform, it is useful to analyze the psychological processes that arise in group decision-making by boards of directors involving independent directors from outside and to consider the conditions which must be present for the right judgments and decisions to be made and for boards to function as expected from the dual perspectives of offense and defense. Figure 1 shows the conceptual framework of the analysis from those perspectives.

Figure 1. Conceptual framework
Despite the theoretical effectiveness of the agency model developed by Jensen and Meckling (1976) for the manager-shareholder relationship, which formed a theoretical foundation for corporate governance, it was later pointed out by Jensen and Meckling themselves (1994) that a model is only a model, and that people do not actually behave as the agents described therein. Rather, an organization's members often behave as if they are its owner, emotionally attaching themselves to a firm, for example, and not necessarily pursuing self-interest at the expense of shareholders (Davis, Schoorman, & Donaldson, 1997). To fill the apparent discrepancies between theory and practice, it is worth looking at the psychological processes involved.
The combination of law and psychology is not new. Franks (1930) explores biases in judicial decisions by looking at judges and jurors not as mechanical interpreters of laws, but as human beings. These ideas have been developed over the years, leading to the combination of the two disciplines and application of the framework in the specific context of corporate law and governance, as seen in Langevoort (2012), Greenfield (2015),  Winter (2018), and Zamir and Teichman (2018), collectively offering an alternative framework to the traditional one proposed earlier by Easterbrook and Fischel (1991). Laws and institutions often become formalities if people treat them as such. There can be a discrepancy between written and unwritten rules depending on people's behavior in adopting or ignoring them. People's behavior even constitutes an informal institution in the form of their customs and practices (Aoki, 2001). Let us say that the first phase of corporate governance reform is when the overall framework takes shape, even as a mere façade, and the second phase is when the framework is more firmly established. An effective approach in achieving this evolution is not to add regulation after regulation, but to look at the behaviors of the people on the board -a group with human emotions and feelings. Within a behavioral (psychological) framework, rules and regulations do not automatically function per se. While corporate governance typically involves legal concepts, if we combine these with perspectives obtained by observing actual human behavior we can deepen our understanding of laws and institutions in the real world, thus helping to achieve their original intentions.

Corporate governance reform in Japan
In introducing independent directors, firms and lawmakers expect that a diverse group will produce collective intelligence for decision-making by offering a wider range of expertise and viewpoints. Nevertheless, biases such as groupthink, escalation of commitment and polarization of decisions are also widely observed. Likewise, shareholders and regulators cannot expect boards to check fraudulent activities promptly and appropriately simply by increasing the number of their independent directors. It is unclear if a compensation design for directors, for example, takes such defensive elements into proper consideration. It is not inconceivable that a corporate governance framework introduced as a result of either a mandatory obligation or a comply-or-explain rule will become a mere formality that gives the impression of fulfilling a minimum requirement. Furthermore, businesses want requirements to be more concrete and specific in order to fulfill them. A paradox surfaces here, with firms that supposedly want to be left alone by regulators turning to those same regulators for yet more intricate regulations (Winter, 2018). While corporate governance typically involves legal concepts, if we combine these with perspectives obtained by observing actual human behavior we can deepen our understanding of laws and institutions in the real world, thus helping to achieve their original intentions. The framework of behavioral law and economics is useful for this purpose, as we will show in the following sections.

Collective intelligence
A key offensive aspect of corporate governance is the pursuit of group dynamics that produces collective intelligence by integrating diverse expertise within the group. Japanese boards of directors traditionally have made a key investment and financing decisions, reflecting requirements placed on them by Japanese corporate law, while statutory auditors have played an oversight role. Even while the unitary board system permits the delegation of such managerial decisions to the top management team for companies with an audit and supervisory committee and a three-committee company, the board of directors has still functioned as a top management team. As independent directors are typically expected to approve or advise on strategy from a different viewpoint to insiders, the question is whether such a combination actually produces collective intelligence.
As people have limits on their cognition, to the extent that they often miss things that are very obvious when asked to watch other things in their field of sight (Simons & Chabris, 1999), group discussion may be an effective way of sharing information useful for decision-making and noticing points which have been missed. It is an illusion, however, that group action enhances productivity per se (Nijstad, Stroebe, & Lodewijkx, 2006 In comparing group and individual performance, Sniezek and Henry (1989) show that the performance of a group tends to exceed the average performance of its individual members and even that of the best-performing individual. Larrick and Soll (2006) show theoretically that errors by a group's individuals balance each other out, resulting in better performance by the group as a whole. Smith et al. (2009) show that because group deliberations gradually exclude wrong answers and approach the right ones, a group can arrive at a correct answer even when no individual member knows what it is.
These experiments point to the existence of collective intelligence, but their results do not generally hold true. Group performance does not always produce collective intelligence and exceeds individual performance (Hill, 1982), and, as is often pointed out, generalized experiments in a laboratory do not necessarily reflect conditions at a real firm (DellaVigna, 2009). However, the results of many experiments testing collective intelligence suggest general tendencies for human behavior. Tetlock and Gardner (2015), for instance, compare the accuracy of the group and individual performance in forecasting future political and economic events. They show that groups exceed individuals, particularly when behaviors such as listening to one another's views are observed, even if participants do not meet in person. Blinder and Morgan (2005) model a decision process for financial policy by central banks to show that groups perform better than individuals in deliberations.
These results point to the importance of conditions to a group's production of collective intelligence. Regarding this point, Woolley et al. (2010) list smooth interaction within the group, a process for giving equal consideration to all participants, and, interestingly, participation by women, as such conditions. Quantitatively, they measure the social perceptiveness of group participants, which tends to rise when women participate. Van Der Vegt and Bunderson (2000) similarly cite close communications and interactions among dissimilar individuals, and their equal participation in discussions, as conditions for the diverse expertise of a group to have positive effects. As for the attributes of participants in a group, Jolls and Sunstein (2006) show that discussions among similar individuals produce more errors than those among dissimilar ones, and Paulus and Brown (2007) point to the importance of a diversity of knowledge as opposed to the social attributes of participants as a condition for success in brainstorming.
If we look at the common elements of these findings, we realize the importance of the sharing of views and information among participants and a supportive environment for doing so. But despite the importance of these elements, participants in group decision-making tend to discuss information they have already shared, rather than contributing new, previously unshared information to the group, even when aware that such new information is critical to their decision (Stasser & Titus, 1985;Sunstein, 2005;Sunstein & Hastie, 2015). When board members in a strategic planning session recognize weak points in a strategy, they will often refrain from sharing their concerns out of a failure to see that other members recognize the weakness as well. Change to the strategy is thereby inhibited or delayed. Further, despite a general requirement for diversity, people tend to share their concerns only with groups of independent directors with similar backgrounds (Westphal & Bednar, 2005; Westphal & Zajac, 2013). These findings mean that a group does not guarantee smooth interaction, and that introducing independent directors, while presumably adding to the diversity of expertise and viewpoints, does not automatically lead to better decisions.

Errors and biases
While groups have the potential to produce collective intelligence, they are also rife with errors and biases. Chief among them is groupthink (Asch, 1956;Janis, 1972). This is a process where a group creates pressure to conform, suppressing potential dissent to a point where participants accept an obviously erroneous answer -even coming to believe it is the correct one -and collectively reach an erroneous decision. Groups are also conducive to polarization, a phenomenon in which opinions made as a group become more polarized than the original ones held by individuals (Myers & Lamm, 1976;Burnstein & Vinokur, 1977;Sunstein, 1999;Zhu, 2013), and to risky shift, the tendency for a group to make riskier decisions than those made by individuals (Davis, 1992). Conforming to others also appears in the form of cascading, where participants follow the first opinion expressed, even when they know it is wrong, causing errors contained in the first decision to be amplified in subsequent ones; this is especially apparent in situations where conforming is rewarded (Hung & Plott, 2001).
Further, there exists an escalation of commitment, where the failure of an initial decision is followed by risky decisions in an attempt to make up for losses initially incurred (Staw, 1981;Bazerman, Giuliano, & Appelman, 1984). This kind of escalation is amplified in a group context (Whyte, 1993). Groups also show the diffusion of responsibility (Darley & Latané, 1968;Latané & Darley, 1969) and social loafing (Williams, Harkins, & Latané, 1981), which arise particularly when the responsibilities of each participant in a group are unclear. Also, groups are more susceptible than individuals to the kind of problem-framing that leads to biases when making decisions (Kerr, MacCoun, & Kramer, 1996). These findings suggest that while the group is a social unit with the potential to produce collective intelligence, it is also prone to errors and biases. They also show that adding independent directors to boards does not automatically lead to an increase in collective intelligence.

Women's participation
When we consider these effects in the context of corporate governance reform, what is most notable is the effect of female presence on the board. In 2017, women made up 5.3 percent of all directors on Japanese boards, far below the figures of 24.5 percent for the U.K., 21.7 percent for the U.S., 31.9 percent for Germany, and 43.4 percent for France (OECD, 2019). The appointment of independent directors may reflect a recent effort to put more women on boards, but in reality, women make up a mere 11.6 percent of independent directors in Japan, compared to 27 percent in the U.S. in 2019 (Mishra, 2019). The State of California mandates that there be at least three women on boards of six or more directors by 2021. While a meta-analysis of research on corporate performance in terms of gender suggests there is no business case either for or against appointing women to boards (Klein, 2017), if women's participation is behaviorally effective in pursuing collective intelligence, it has an importance beyond the general need to appoint female directors as a matter of social fairness and justice.
Even when appointed, women typically sit on boards as a minority group, raising concern that their presence may be a mere formality, or tokenism (Elstad & Ladegard, 2012). A minority group feels pressure to provide opinions that do not diverge from those of the majority if it is to maintain its reputation with other members (Sunstein & Hastie, 2015). Under this pressure, it is likely that female directors, as a minority group, will hesitate to oppose the male members in the majority. The situation thus fails to meet the conditions for collective intelligence. Regarding this balance, it is shown empirically that three female directors achieve a critical mass in the group setting of a board of directors (Konrad, Kramer, & Erkut, 2008 (2000) show, an environment allowing participants to share opinions and information is a condition for collective intelligence, and, as Stasser and Titus (1985) and Westphal and Zajac (2013) argue, unshared information tends not to surface. Further, a minority female group is subject to the pressure just described. From these findings, we see that appointing female directors does not, per se, satisfy the conditions for collective intelligence as intended. An additional measure is necessary to create such an environment, and, in line with the Californian regulation, achieving a critical mass of female directors is suggested as an avenue to mitigate the pressure to conform to the group that arises from their minority status. These findings provide the behavioral foundation for an argument for more female directors on boards -even aside from the social justice issue -as a condition for boards to avoid sliding into tokenism when pursuing collective intelligence.

Incentives and compensation
The roles of independent directors comprise a spectrum of levels ranging from passive ratification to advice and networking to making proactive proposals, but they all aim at achieving a higher quality of group decision-making that leads to sustainable growth and value creation. One problem, however, is where independent directors get the incentives to perform such complex tasks; in other words, whether there is clear support in the form of incentives for such directors to function as designed. Crystallizing this problem is the question of how independent directors can ever act effectively without high pecuniary compensation. At Japanese firms, where the compensation of even top managers is relatively low (METI, 2018), independent directors are also compensated modestly, even when stock-related compensation is included in the package. With their other full-time assignments and the need to attend board meetings several times a year, it appears that too much may be expected of them.
To consider this question, it is worth first exploring the relationship between compensation and performance from the behavioral perspective. One notable finding is that pecuniary compensation is a better fit for simple tasks that are easily quantifiable than for complex and creative ones Further, pecuniary compensation is accompanied by the endowment effect, where workers' performance worsens if their compensation does not increase because they take for granted an amount of compensation once received (Bareket-Bojmel, Hochman, & Ariely, 2014). The endowment effect makes it hard for firms to lower compensation once a certain amount is established, and forms the psychological background for a firm's continuing to ratchet it up to chronically excessive levels. Pecuniary compensation is simply a form of status for an industry's executives to compare among themselves (Winter, 2018) or within their firm (Statman, 2004). Top managers are wealthy in general, and the marginal effect of additional compensation is obscure, aside from its effect on status. If the disutility of a lower stock price, with its connection to stock-related compensation, is greater than the utility of a higher stock price -often the case in a value function based on the prospect theory (Kahneman & Tversky, 1984) -it might even cause managers to try to prevent the stock price from falling, a passive behavior, rather than working to raise it.
To motivate people, non-pecuniary compensation such as internal recognition, the meaning given to work, and the flexibility of work are more effective than pecuniary payment (Thibault-Landry, Schweyer, & Whillans, 2017). And it is on non-pecuniary elements such as status and honor that independent directors place a premium, given that they are typically accomplished individuals (Winter, 2018). The relatively modest level of compensation in Japanese firms derives from a culture that does not place high regard on pecuniary compensation in general -one senior executive at a top Japanese firm, Nippon Steel, has commented publicly that he does not work for pecuniary reward, nor does his compensation serve as an incentive (FSA, 2019) -though the country is gradually adopting global market rates to attract and retain talent. These findings of psychology warn that higher pecuniary compensation is not always effective, as its effectiveness is suppressed by the endowment effect to which it leads. It also has the psychological effect of whetting the appetite for non-pecuniary, higher socio-industry status, paradoxically in the form of escalating pecuniary compensation.

Setting of management goals
Relatedly, management goals, which are linked to compensation to make managers commit to achieving them, are also changing under corporate governance reform. Japanese firms have regarded it as a virtue to emphasize long-term, rather than short-term, goals, something which has enabled managers to run their firms from a long-term perspective and produce unique products and processes. But despite this long-term emphasis, management compensation is, in fact, linked mainly to single-year earnings rather than mid-term earnings and stock price (METI, 2018). This suits the recent preference of managers, however: as more firms mature and their stock prices stagnate, managers whose pay is based on single-year earnings are well compensated. Therefore, the government suggests that firms increase stockrelated forms of compensation such as stock options and restricted stock, once avoided as shortterm goals, in order to switch their incentives from raking in a stable cash flow toward pursuing future growth reflected in rising stock prices. However, it is important to recognize that, as we have found in our analysis, stock rewards do not work as simply as is generally assumed and may even produce shorttermism in place of the cherished focus on longterm value.
When linked to compensation, management goals are by no means immune to biases. In an environment that prioritizes established goals, people pay little attention to other issues. Rather than pursue several goals simultaneously, they focus on a limited number, particularly ones which are short-term and easily quantifiable. They also tend to take excessive risks. Further, goals undermine internal motivation, negatively affect happiness when not achieved, hinder learning, make people competitive rather than cooperative, and promote unethical behavior (Ordóñez, Schweizer, Galinsky, & Bazerman, 2009). When compensation is linked to the achievement of goals, managers are criticized for lying twice: first by setting lower goals from the start, and secondly by taking whatever measures necessary to achieve them, to the neglect of other ongoing issues (Jensen, 2003). Despite the doubtful effectiveness of pecuniary compensation as an incentive, the setting of goals and their linkage to pecuniary compensation form the backbone of corporate management systems. However, it is often unclear how these factors relate to the way people respond, especially regarding productive behaviors such as enhancing collective intelligence within a group, while it is known that the standard system of compensation and goal-setting will often produce adverse effects.

Obedience to authority
To survive in the marketplace, it is important to prevent the corporate value from falling as well as taking proactive action to enhance it. This section deals with corporate misconduct and subsequent loss of reputation and corporate value. Corporate governance reforms often appear in response to corporate wrongdoing. In the U.S., the revelation of accounting fraud by Enron and WorldCom led in 2002 to the enactment of the Sarbanes-Oxley (SOX) Act, under which an audit committee consisting only of independent directors became mandatory for listed firms. Establishing a checks-and-balances mechanism is important not only to prevent groupthink and other corporate errors in regular times but also to avert damage to a firm's reputation and the losses accompanying misconduct that occur irregularly.
From the behavioral perspective, independent directors can become a dysfunctional formality in the absence of conditions supporting their defensive roles as well. Most fundamentally, this is because it is hard for independent directors to object to the views of managers who are essentially their sponsors, with de facto authority to elect and pay them subject to the typically passive approval of shareholders. This difficulty with objecting is linked in part to the tendency to obey authority (Morck, 2008), but obedience to authority and pressure to conform also occur in situations that do not involve compensation (Asch, 1956;Milgram, 1963). In the context of board monitoring, moreover, it is known that people receiving compensation will make decisions favorable to their sponsors, and, importantly, that they will make such favorable decisions unconsciously, with no awareness that they are influenced by the fact that they are being paid (Harvey, Kirk, Denfield, & Montague, 2010). It is not surprising that independent directors are sometimes mocked as "passive stooges who always champion the incumbent management" (Brealey, Myers, & Allen, 2019). A case in point is Toshiba, which was fined by the Japanese Securities and Exchange Surveillance Commission over accounting irregularities in 2015. It had already had four independent directors out of sixteen directors on the board, even before the governance code was introduced. But the scandal, which led to the resignation of eight internal directors, illuminates the danger of a passive, dysfunctional oversight role by independent directors. Also, it is worth noting that Toshiba had only one woman, an independent director, on its board and none on its top management team of 37 people.

Errors and biases
The comparison of groups and individuals for their offensive aspects helps analyze defensive aspects as well. Concerning the role of monitoring, for instance, it is arguable that in comparison with individuals, a group is more likely to produce collective intelligence and expand the limits of memory and cognition to which individual directors are bound, while at the same time being susceptible to such negative biases as groupthink, diffusion of responsibility, and escalation of commitments among directors. Unless they conduct their own research, independent directors generally depend on their information on internal reports by the firms they monitor; hence the homogeneity of the information that each independent director receives. As a result, they base their approval of management proposals on a portion of the information already held by internal directors. They are thus more susceptible to groupthink and diffusion of responsibility, in the sense that they can claim ignorance and that the other directors were in the same position.
The legal liability of independent directors is known to be light, as indicated by the negligible number of lawsuits filed against them compared to the number filed against internal directors (Black, Cheffins, & Klausner, 2006;Fairfax, 2011). Given that the compensation of independent directors is lower than that of internal directors due to their different levels of their responsibility and risk, and that they are typically insured against litigation under directors and officers (D&O) liability insurance, it is hard to expect too much of independent directors on this front.
Under the Japanese corporate governance code that requires listed firms to have two or more independent directors based on the comply-orexplain rule, on average, these firms have 2.2 independent directors on their boards, which average 8.3 directors. 48.9 percent of them have two and 44.6 percent three or more (Tokyo Stock Exchange, 2019). If we compare a board having multiple independent directors with a hypothetical board having only one, for the larger number of independent directors to overcome the cognitive biases related to group decision-making and perform more effectively than the individual, there must be conditions similar to those required to produce collective intelligence in offensive terms, including the critical mass condition. Achieving such a situation incurs costs, such as the need to create an environment enabling the collection of previously unshared knowledge and perspectives. If a mechanism to monitor misconduct is left as a dysfunctional formality despite its plausibly appearing to comply with the rules, the costs of overcoming biases and establishing the mechanism will exceed the benefits of producing collective intelligence from a group (Asaoka, 2018). In a similar sense to a group's not always performing better than an individual, more independent directors do not always raise the quality of monitoring unless conscious measures are taken to overcome group biases.
In addition to these errors and biases, it is worth noting some conflicts of interest. For instance, there is no binding regulation that prohibits independent directors from sitting on multiple boards in the same industry and getting paid accordingly. The corporate governance code requires disclosure in such situations. However, while the disclosure is used widely in conflicted situations, it has the adverse effect of giving disclosers a moral excuse for their conflicted actions, even emboldening them toward further escalation. Recipients of such disclosure, meanwhile, underestimate the extent of possible harm to themselves and continue to trust the disclosers to an excessive degree (Cain, Loewenstein, & Moore, 2005. Therefore, one can argue that disclosure has a limited, or even exacerbating, effect on conflicts of interest.

Internal reporting
Despite the errors and biases affecting independent directors, they do have an effective role to play in making a credible declaration of, and precommitment to, their independence of the boards on which they serve. Through internal reporting, including whistleblowing, on corporate misconduct, they are part of the firm's checks and balances.
While not yet adopted by all boards, it is possible that they, as opposed to top managers, stand at the top of a firm's internal hierarchy for reporting misconduct. As a corollary, they collect information on misdeeds by deploying the internal audit division or outside experts. Independent directors are, by design, remote from the execution of management decisions while being well-positioned to monitor them, and, as we have described, the information they receive is usually filtered to fit their limited time. But the key to this reporting hierarchy is the knowledge within the firm that outside directors are well placed to receive information on misconductinformation, including whistleblowing, that goes beyond regular reports and is independent of the management hierarchy.
Given their outsider position, this role is important because the sense that one might be seen and heard by outsiders tends to influence human psychology and behavior, functioning as a check on management and operators, particularly when the "outsider" status is emphasized (Bateson, Nettle, & Roberts, 2006). This is more effective than setting up an internal line of reporting ending with top management. Even if the information is shared concurrently with independent directors, the latter essentially involves insiders who may well be concerned about internal retaliation. Typically, the heads of administrative and audit divisions are direct subordinates of top managers, their responsibilities often assigned by rotation. It is hard for them to detect or prevent misconduct when their monitoring activities are monitored by their managers. As already described, independent directors are biased by obedience to authority and tend to make decisions favorable to their sponsors, but clearly to a lesser degree than insiders. By declaring and pre-committing to the role they play, they gain greater credibility and mitigate such biases. Groupthink can be controlled if there is even one person present who objects (Asch, 1956), and the stated commitment of independent outside directors has the effect of characterizing them more or less continuously as possible objectors.

Incentives and compensation
A common issue involving both offensive and defensive aspects is the incentives of independent directors. As noted, their compensation is lower than that of internal directors, and they are unlikely to be sued by shareholders should misconduct occur. It is not even unusual for monitoring and finding misconduct to be exempted from their scope of responsibility (Tricker, 2015). This being so, what is the foundation for appointing them in terms of their incentive to properly monitor firms? A key point is, again, non-pecuniary, psychological compensation, which in this case is the maintenance of their established reputation.
Typical independent directors are accomplished in their original careers and willing to take on a board position for relatively low compensation, meaning they are more interested in promoting and protecting their non-pecuniary reputation as an honorable person than getting well paid. In accepting the position, they are essentially putting their reputation on the line. This reputational stake incentivizes them to avoid exposing themselves to blame for failing to prevent or foresee misconduct. The significance of strictly defining and interpreting the independence of outside directors, over and above their general role of protecting shareholders independently of management, lies in their having declared their autonomy from the firm and its insiders. They have put their reputation at stake by doing so and provided themselves with non-pecuniary incentives to defend it. Because their rewards are intangible, this is consistent with the low probability of their getting sued for losses caused by internal misconduct.
On the other hand, the system of independent directors is inherently incomplete from the behavioral perspective, as independent directors are paid by the firms they serve. Despite the title, independent directors are dependent on firms for their compensation, and, compared with, say, regulators, are not free from conflicts of interest. But despite the dependency of independent directors on firms, the answer to the monitoring problem is not to replace them with larger numbers of regulators who, of course, are not paid by firms. The effectiveness of regulations and the cost of executing them must be balanced against their benefits (Fairfax, 2011). A greater regulation can lead to irresponsibility when a firm only adheres to the letter of the rules, giving the appearance of fulfilling minimum requirements while bringing on further regulation in a vicious cycle (Winter, 2018). Also, in the absence of relevant regulations, regulators might switch roles with managers at any time, making them susceptible to sponsor bias and conflicts of interest.
While the system is incomplete, independent directors can function if their incentives are properly addressed.

Fairness
It is also helpful to look at the psychology of the monitored insiders who may initiate misconduct, but not through the simple lens of crime and punishment. As observed in experiments for the ultimatum game, people tend to value fairness even when they know a decision will incur losses in terms of economic rationality. This holds true even for one-time decisions which have no effect on subsequent decisions, as opposed to repeated decisions where prior behavior and creditworthiness are known to others (Güth, Schmittberger, & Schwarze, 1982;Fehr & Fischbacher, 2004). To manage and prevent misconduct, it is worth exploring and addressing this sense of fairness. For instance, merely making people believe that their firm has a code of ethics is effective in curbing misconduct, even when such a code does not exist (Mazar & Ariely, 2006), suggesting that codes of conduct should be placed within the cognitive domain of corporate workers.
Also, people tend to justify their misconduct in an effort to maintain their self-image as honest individuals (Mazar, Amir, & Ariely, 2008). Because they value their self-image as honest citizens, people do not commit as much fraud as they could, even when they know they will never be exposed and will stop at a point where they can psychologically justify their conduct. As a corollary, people tend to escalate unethical behavior when told that it benefits not them but others, as they now have an even stronger moral excuse (Gino, Ayal, & Ariely, 2013). If corporate miscreants excuse themselves by maintaining that their misdeeds are for the benefit of the firm -a very plausible excuse in Japan, considering the collectivistic culture in which dedication to one's company is highly valuedintroducing independent directors with a mandate to stand at the top of the internal line of reporting serves to delegitimize, and therefore reduce, the psychological excuses for wrongdoing. Also, in a development suggesting a sea change in the culture, internal reporting already exceeds internal auditing as a source of discovering corporate misconduct in Japanese firms, with the former comprising 58.8 percent of discoveries and the latter 37.6 percent (CAA, 2017). Arguably, therefore, reinforcing the mechanism through the introduction of independent directors is an effective way to accommodate such change.

Pursuing both ends
The previous section showed that, from the offensive viewpoint, boards are expected to produce collective intelligence (while groups inherently bear biases that hinder that effect), and that, from the defensive viewpoint, the introduction of an outsider's view is expected to prevent corporate misdeeds (while ineffective design can make this a dysfunctional formality). To establish corporate governance architecture it is necessary to deal with these biases, and here is where the findings of psychological experiments can help.
On the offensive side, the analyses reveal that, when seen from the perspective of human interaction in the management of organizations, conditions exist which support the production of collective intelligence. Specifically, we can deduce the importance of processes enabling equal consideration of the views of all participants and the sharing of information beneficial to making decisions, as well as an emphasis on communication, recognition, cooperation, and fairness. Also notable is the effectiveness of women's participation for these purposes, pointing to the need for greater diversity in group decision-making. Achieving a critical mass is proposed in this regard, over and beyond the social fairness and justice perspectives. The fact that the 2018 amendment to the corporate governance code suggests a third of directors be independent points to the importance of the relative power of minority groups on the board to bringing it about.
Further, based on the paradox that people are reluctant to share information that has not been already disclosed, as opposed to adding information relevant to that which has already been shared, there is an important role for a devil's advocate -someone who, as Sunstein (2005) argues, puts forward points that have been missed and suggests directions contrary to those implicitly shared and predicted. The impact of the devil's advocate is limited, however, if those assuming that role make objections merely as a formality on the assumption that their objections will not affect the ultimate decision anyway (Nemeth, Brown, & Rogers, 2001; Schulz-Hardt, Jochims, & Frey, 2002). The existence of true dissenters promotes the collective intelligence of a group in the sense that they not only offer opposing opinions to consider but, by doing so, they prompt other participants to share the information which is beneficial for decisionmaking but has not been shared among the group On the defensive side, imposing rule after the rule is a questionable method of preventing corporate misdeeds. Doing so risks condemning independent directors to an uncertain existence, with limited access to information and negligible legal liability, depending on how they are positioned in the corporate governance architecture. In such a situation, corporate governance architecture can well become a formality that fails to produce its intended results. Their status as independent outsiders thus makes outside directors effective candidates to head the communication line for reporting misconduct. Independent of management, they influence workers' psychology in a way that reduces misconduct. Workers perceive themselves as being seen and heard, mitigating fears of internal retaliation should they report a misdeed. This perception becomes more credible when outside directors have announced and committed to their independence. This role is consistent with their compensation, in that reputation and honor are more at stake than monetary payment.

Cultural shift
Corporate governance design differs according to national culture. Independent outside directors are, literally, outside a firm's internal hierarchy, in which the position of director is seen as the top career goal under lifetime employment. For Japanese firms, which have been characterized as "tightly-knit hierarchies" (Nakane, 1967) and "collectivistic corporate communities" (Hashimoto, 1991) based on their "collectivistic culture" (Licht, 2018), appointing outsiders as directors charged with making managerial decisions may seem odd. Their increasing adoption by boards appears to be a turning point, not only for the country's corporate governance design but also for its corporate culture. Japanese firms are often said to be homogeneous and to place great value on consensus. But the precipitate shift that began in 2015 with the abolishing of traditional statutory auditors and introduction of audit and supervisory committees, including independent directors, may lead to a convergence of corporate governance formats led by globalized firms and those with a high level of shareholding by institutional investors outside the country. The increase in the number of independent directors is a sign that mediated by their role, Japanese firms are becoming more open to the outside. The architecture of corporate governance is changing continually, as is the design of boards, from those comprising mostly internal directors to those which involve more outside directors and are more open to shareholders.
Further, although a significant part of the power of top managers derives from their authority to select and promote successors and lower-tier managers within their organization, a new corporate landscape is on the horizon. In this new model, the nomination and compensation of top managers are subject to approval by committees with a majority number of independent directors. The adoption of the model, while voluntary, has been steadily increasing, particularly since both the nomination and compensation committees became subject to the comply-or-explain rule under the corporate governance code amended in 2018. If managers delegate part of their power to these independent directors, despite their inherent biases of obedience and favoritism, the resulting format might change the way that managers exert power and direct firms. Greater openness and increased attention to institutional investors may work as a force that sets the direction of the organization, in contrast to the times when more attention was paid to "main banks" which valued stability rather than growth. The number of listed firms adopting anti-takeover measures such as poison pills has been decreasing constantly, from a peak of 569 in 2008 to 386 in 2018 (Nishiyama, 2018).
Japan, which imported its modern corporate law from Germany and later from the United States, has a long tradition of absorbing ideas from outside the islands and digesting and improving them to create its unique version. The modern corporate governance architecture of the country is in a state of ongoing change. It may well digest more ideas and evolve into a form with its characteristics. There is no endpoint when establishing corporate governance architecture, but in pursuing collective intelligence the key is to recognize the central role of humans in the process. We see this in the conditions needed to produce it (corporate law stipulates that directors must be natural persons) and the psychological processes that surround it. Effective corporate governance architecture is the result of conscious efforts to overcome errors and biases while pursuing collective intelligence in both its offensive and defensive aspects.

CONCLUSION
This paper analyzed some behavioral aspects of corporate governance reform, as exemplified by the introduction of independent directors to boards under the corporate governance code, which are enhancing the quality of group decision-making by producing collective intelligence and preventing corporate misconduct by introducing the viewpoints of outsiders. Our analysis was undertaken from the offensive and defensive perspectives.
For the offensive aspects, we positioned the board of directors as a group decision-making body pursuing collective intelligence by opening itself to outside knowledge and networks rather than focusing on close-knit groups of internal directors. It is not immune to psychological biases such as groupthink and escalation of commitment, but these can be mitigated by ensuring equal consideration of all participants' viewpoints. Notably, we suggested the possibility of women's participation as a condition that promotes collective intelligence, a growing need in society in general. For the defensive aspects, we emphasized the importance of an outsider's viewpoint in preventing internal misconduct. It is affected by other biases, such as obedience to authority and diffusion of responsibility, but establishing an internal system for reporting misconduct, with outside directors at the top, can be effective if the outsiders' position is perceived as credible.
The development of corporate governance in Japan is a move toward the development of group dynamics and networks that transcend the boundaries of an internally constrained, tightly-knit organization, which previously had consisted largely of internally promoted directors enjoying lifetime employment. We looked at the increasing adoption of independent directors as a signal of a cultural shift toward the opening of organizations to outside influences and resources.
The limitations of this paper and the future research agenda are multi-faceted. First, while the paper views the board as a cornerstone of corporate governance architecture, analyzing the interactions between the board and top management team will also be important as the increase in independent outside directors causes the two groups to become more distinct, creating a new aspect of group dynamics. Second, the outsider's viewpoint is not limited to independent directors. Japanese firms seem more attentive to the voices of shareholders than in the past when they looked mostly at their main banks. Such a change in dynamics will further affect the decision processes of boards. Since corporate decisions are made by groups, a change in group composition creates new chemistry, pointing to a rich field for future research.
As all phases of corporate activity must involve humans in order to continuously deliver creative work, an understanding of psychological processes is essential. We hope this paper will contribute to the development of such an understanding and the ongoing evolution of corporate governance architecture.