An article at Chief Executive on How to “Rightsize” Your Board has attracted some attention on several of the Linkedin corporate governance groups I follow. Curiously, while the article opines that "boards must be the right size for their charge," it fails to propose an optimal number. Instead, it ticks of a checklist of 7 considerations to be taken into account in determining whether a specific board is the right size. It's all pretty vague IMHO.
In fairness, however, getting a handle on the optimal size for a board of directors is very hard. In my article Why a Board? Group Decision Making in Corporate Governance, I review the relevant legal, psychological, economic, and sociological literature to conclude that:
Corporate statutes historically required that boards consist of at least three members who had to be shareholders of the corporation and, under some statutes, residents of the state of incorporation. Today these requirements have largely disappeared. DGCL Section 141(b) authorizes boards to have one or more members and mandates no qualifications for board membership. MBCA Sections 8.02 and 8.03 are comparable. As a default rule, allowing single member boards probably makes some sense. It gives promoters maximum flexibility, while allowing the creation of multi-member boards at low cost. In light of the apparent advantages of group decision making, however, it is hardly surprising that multi-member boards are the norm for corporations of any significant size. To be sure, board sizes vary widely. A 1999 survey found that slightly less than half had 7 to 9 members, with the remaining boards scattered evenly on either side of that range.
Is there an optimal board size? It is mildly puzzling that the literature on group decision making has not paid more attention to questions of group size. Studies in which group size is an experimental variable are rare; worse yet, many studies of other variables fail even to hold group size constant. The principal exceptions are studies of optimal jury size. Unfortunately, those studies are inconclusive at best.
As for studies of board size in particular, one meta-analysis found a statistically significant correlation between increased board size and improved financial performance. Given the potential influence of moderating variables, however, it does not seem safe to draw firm conclusions from that survey. Other studies, moreover, are to the contrary.
Here then is an opportunity for further research. In theory, a number of factors favor large boards. Larger boards may facilitate the board’s resource-gathering function. More directors will usually translate into more interlocking relationships with other organizations that may be useful in providing resources, such as customers, clients, credit, and supplies. Board interlocks may be especially helpful with respect to formation of strategic alliances. Firms considering a joint venture need access to credible information about the competencies and reliability of prospective partners. Almost by definition, however, this information is asymmetrically held and subject to strategic behavior. Interlocks between prospective partners provide both access to such information and, by analogy to hostage taking, a credible bond of the information’s accuracy.
Larger boards with diverse interlocks are also likely to include a greater number of specialists—such as investment bankers or attorneys. This is relevant not only to the board’s resource gathering function, but also to its monitoring and service functions. Complex business decisions require knowledge in such areas as accounting, finance, management, and law. Providing access to such knowledge can be seen as part of the board’s resource gathering function. Board members may either possess such knowledge themselves or have access to credible external sources who do. This hypothesis is consistent with the new institutional economics view of specialists. In that model, specialization is a rational response to bounded rationality. The expert in a field makes the most of his or her limited capacity to absorb and master information by limiting the amount of information that must be processed through limiting the breadth of the field in which the expert specializes. As applied to the corporate context, larger, more diverse boards likely contain more specialists, and therefore should get the benefit of specialization. In addition, with reference to the debate over the best member hypothesis, specialization is a way for the group to identify the superior decision maker with respect to specific issues.
On the other hand, a number of considerations suggest that small boards may be preferable. Large boards will be contentious and fragmented, which reduces their ability to collectively monitor and discipline senior management. In such cases, the senior managers can affirmatively take advantage of the board through “coalition building, selective channeling of information, and dividing and conquering.”
The social loafing phenomena also suggests an upper limit on efficient group size. As group size grows, for example, the number of non-participants (loafers) likely increases. Conversely, larger boards may inhibit the formation of the sorts of close-knit relationships by which groups constrain agency costs.
 MBCA § 8.03(a) cmt.
 National Association of Corporate Directors, Public Company Governance Survey 1999-2000 7 (Oct. 2000) (44 percent between 7 and 9).
 Davis, supra note 98, at 16.
 Id. at 18-21 (summarizing studies).
 Dan R. Dalton, Number of Directors and Financial Performance: A Meta-Analysis, 42 Acad. Mgmt. J. 674, 676 (1999).
 See, e.g., Theodore Eisenberg et al., Larger Board Size and Decreasing Firm Value in Small Firms, 48 J. Fin. Econ. 35 (1998) (finding a significant negative correlation between board size and firm profitability in small and medium Finnish firms); see generally Sanjai Bhagat and Bernard Black, The Uncertain Relationship Between Board Composition and Firm Performance, 54 Bus. Law. 921, 941-42 (1999) (summarizing studies).
 See Ranjay Gulati and James D. Westphal, Cooperative or Controlling? The Effects of CEO-board Relations and the Content of Interlocks on the Formation of Joint Ventures, 44 Admin. Sci. Q. 473 (1999).
 Note that, because their decisions are publicly observable, board members have a strong incentive to defer to expert opinion. Because even a good decision maker is subject to the proverbial “act of God,” the market for reputation evaluates decision makers by looking at both the outcome and the action before forming a judgment. If a bad outcome occurs, but the action was consistent with approved expert opinion, the hit to the decision maker’s reputation is reduced. In effect, by deferring to specialists, a decision maker operating under conditions of bounded rationality is buying insurance against a bad outcome.
In a collegial, multi-actor setting, the potential for log-rolling further encourages deference. A specialist in a given field is far more likely to have strong feelings about the outcome of a particular case than a non-expert. By deferring to the specialist, the non-expert may win the specialist’s vote in other cases as to which the non-expert has a stronger stake. Such log-rolling need not be explicit, although it doubtless is at least sometimes, but rather can be a form of the tit-for-tat cooperative game. In board decision making, deference thus invokes a norm of reciprocation that allows the non-expert to count on the specialist’s vote on other matters. This prediction is supported by findings with respect to group polarization, in which the majority coalition makes small concessions so as to trigger the norm of reciprocity. See Kerr, supra note 140, at 92 (noting the use of such norms).
The normative payoff of this insight is at least two-fold. First, insofar as board decision making itself is concerned, directors should consciously ask whether deference to specialists is appropriate in a particular instance. Second, it validates state statutes relating to board reliance on expert opinion. Under Delaware code § 141(e) directors are “fully protected in relying in good faith” on reports or opinions of external experts. The statute requires that the director reasonably believe the matters in question are within the expert’s professional competence and that the expert have been chosen with reasonable care. Case law suggests that this standard requires at least some inquiry into the basis of the expert’s opinion. See, e.g., Smith v. Van Gorkom, 488 A.2d 858, 874-75 (Del. 1985) (interpreting DGCL § 141(e)).
 Jeffrey A. Alexander et al., Leadership Instability in Hospitals: The Influence of Board-CEO Relations and Organizational Growth and Decline, 38 Admin. Sci. Q. 74, 79 (1993). On the other hand, some commentators contend that large boards provide more opportunities to create insurgent coalitions that constrain agency costs with respect to senior management. William Ocasio, Political Dynamics and the Circulation of Power: CEO Succession in U.S. Industrial Corporations, 1960-1990, 39 Admin. Sci. Q. 257, 291 (1994).