How should corporate boards be developed and evaluated? Following is an older, but important and still timely “thesis” on how to structure the work of a board, but also how we manage the social system a board actually is.

As reported by HBR on September 2002 by Jeffrey A. Sonnenfeld.

What Makes Great Boards Great

In the wake of the meltdowns of such once great companies as Adelphia, Enron, Tyco, and WorldCom, enormous attention has been focused on the companies’ boards. Were the directors asleep at the wheel? In cahoots with corrupt management teams? Simply incompetent? It seems inconceivable that business disasters of such magnitude could happen without gross or even criminal negligence on the part of board members. And yet a close examination of those boards reveals no broad pattern of incompetence or corruption. In fact, the boards followed most of the accepted standards for board operations: Members showed up for meetings; they had lots of personal money invested in the company; audit committees, compensation committees, and codes of ethics were in place; the boards weren’t too small, too big, too old, or too young. Finally, while some companies have had problems with director independence because of the number of insiders on their boards, this was not true of all the failed boards, and board makeup was generally the same for companies with failed boards and those with well-managed ones.

In other words, they passed the tests that would normally be applied to ascertain whether a board of directors was likely to do a good job. And that’s precisely what’s so scary about these events. Viewing the breakdowns through the lens of my 25 years of experience studying board performance and CEO leadership leads me to one conclusion: It’s time for some fundamentally new thinking about how corporate boards should operate and be evaluated. We need to consider not only how we structure the work of a board but also how we manage the social system a board actually is. We’ll be fighting the wrong war if we simply tighten procedural rules for boards and ignore their more pressing need—to be strong, high-functioning work groups whose members trust and challenge one another and engage directly with senior managers on critical issues facing corporations.

The Inadequacy of Conventional Wisdom

Over time, good-governance advocates have developed no shortage of remedies for failures of governance. Most of these remedies are structural: They’re concerned with rules, procedures, composition of committees, and the like, and together they’re supposed to produce vigilant, involved boards. However, good and bad companies alike have already adopted most of those practices. Let’s take a look at some of the most common.

1. Regular Meeting Attendance

Regular meeting attendance is considered a hallmark of the conscientious director. It matters a lot and, still, as shareholder activist Nell Minow comments, “Some big names on the boards…barely show up due to other commitments, and when they show, they’re not prepared.” Indeed, some WorldCom directors were on more than ten boards, so how well prepared could they be? Fortune’s 2001 list of the most-admired U.S. companies reveals no difference in the attendance records of board members of the most- and least-admired companies. Data from the Corporate Library, a corporate governance Web site and database cofounded by Minow, show the same “acceptable” attendance records at both kinds of companies. Good attendance is important for individual board members, but it alone doesn’t seem to have much impact on whether companies are successful.

2. Equity Involvement

Board members are assumed to be more vigilant if they hold big chunks of the company’s stock—but data from the Corporate Library don’t suggest that this measure by itself separates good boards from bad, either. Several members of the board of GE, Fortune’s most-admired corporation in 2001, had less than $100,000 of equity, whereas all board members of the least-admired companies held substantial equity stakes. Not only did all but one of the Enron board members own impressive amounts of equity in the company, but some were still buying as the shares collapsed.

3. Board Member Skills

Patrick McGurn of Institutional Shareholder Services, like other expert observers, has frequently questioned the financial literacy of troubled companies’ audit committee members. It’s certainly true that many board members have their jobs because they’re famous, rich, well connected—anything but financially literate. But just as many board members have the training and smarts to detect problems and somehow fail to do their jobs anyway. At the time of their meltdowns, for example, Kmart had six current or recent Fortune 500 CEOs on its board, and Warnaco had several prominent financiers, a well-known retail analyst, and a top-tier CEO; all those excellent credentials made little difference. On this measure, again, we find that Fortune’s most- and least-admired companies alike had board members with the training and experience to analyze complex financial issues and to understand what kinds of risks a company is taking on.

Despite Enron’s disastrously complex financial schemes, no corporation could have had more appropriate financial competencies and experience on its board. The list includes a former Stanford dean who is an accounting professor, the former CEO of an insurance company, the former CEO of an international bank, a hedge fund manager, a prominent Asian financier, and an economist who is the former head of the U.S. government’s Commodity Futures Trading Commission. Yet members of this board have claimed to have been confused by Enron’s financial transactions.

4. Board Member Age

According to one governance expert, “Enron melted down because it lacks independent directors and several are quite long in the tooth.” His remarks reflect a general belief that boards become less effective as the average age of their members rises. My research on executives over the past two decades has shown that, to the contrary, age is often an asset, and this general finding is supported by board data from the Corporate Library. Charles Schwab, Cisco, and Home Depot all have had several board members who are well into their sixties. Michael Dell (Dell Computer placed tenth on Fortune’s 2001 list of most-admired companies) told me that when he incorporated in 1987, as a 21-year-old college dropout, he found it invaluable to have then 70-year-old George Kozmetsky, Teledyne’s visionary founder and the former dean of the McCombs School of Business in Austin, Texas, serve on the board; Kozmetsky stayed for more than a decade.

5. The Past CEO’s Presence

The complicated reality is that sometimes a past CEO’s presence is helpful and sometimes it’s not. In the years I served on and even chaired commissions for the National Association of Corporate Directors (NACD), some commissioners regularly vilified the “old dragons” who haunted successors by serving on boards. In certain cases, this can be a problem; one can only imagine board meetings at Warnaco, where deposed CEO Linda Wachner voted her 9% of the company’s equity for several months after her November 2001 termination. Alternately, a retired CEO can play an invaluable internal role as a mentor, sounding board, and link to critical outside parties. It’s hard to imagine anyone arguing that Intel, Southwest Airlines, or Home Depot would be better off if their legendary retired CEOs Andy Grove, Herb Kelleher, or Bernie Marcus had just gone home to play golf.

6. Independence

Good-governance advocates and stock exchange heavyweights alike have argued that boards with too many insiders are less clean and less accountable. Some argue that Tyco’s confusing spiral of acquisitions and the apparent self-dealing of the CEO at Adelphia Communications might have been less likely if their boards hadn’t been dominated by insiders. Indeed, the New York Stock Exchange’s Corporate Accountability and Standards Committee recently proposed requiring that the majority of a NYSE-listed corporation’s directors be independent—this in response to the recent governance disasters. Governance reform proposals are also being developed by such business groups as the Conference Board and the Business Roundtable. Yet again, if you judge the most- and least-admired companies on Fortune’s 2001 list against this standard, no meaningful distinction emerges. Least-admired companies like LTV Steel, CKE Restaurants, Kmart, Warnaco, Trump Hotels and Casino Resorts, Federal-Mogul, and US Airways had only one or two inside directors on their boards; Enron had only two. By contrast, at various times in their histories, Home Depot had five insider directors on its 11-person board, Intel had three on a nine-person board, and Southwest Airlines had three on an eight-person board. Typically, half of Microsoft’s board are insiders. Currently, three of Warren Buffett’s seven Berkshire Hathaway board members have the Buffett name, and another is his long-term vice chairman.

United Parcel Service has ranked high on Fortune’s list of most-admired companies since the list was started, and half of the UPS management committee is on its board. Three outside board members have told me how well plugged-in they have felt over the years because the inside members are very candid and well informed. From what the outside directors have seen, none of the insiders has ever been afraid to debate a point with the boss, the CEO.

7. Board Size and Committees

A host of other issues that good-governance advocates propose turn out to be either not truly important or already in place at both good and bad companies. Take board size. Small’s considered good, big’s considered bad. But big boards exist at some great and admired companies—GE, Wal-Mart, and Schwab—along with some poorly performing companies like US Airways and AT&T. At the same time, small boards are part of the landscape at good companies like Berkshire Hathaway and Microsoft and some not-so-good companies like Trump.

Another area where good companies don’t necessarily conform to the advice of good-governance advocates: executive sessions, which give boards the chance to evaluate their CEOs without interference. Executive sessions are also sometimes coupled with a designated lead director. But GE, the most-admired company in the country in 2001, didn’t allow executive sessions in Jack Welch’s day. Said Ken Langone, who serves on the boards of both GE and Home Depot, “Jack will give you all the time in the world to raise any issue you want, but he wants to be there during the discussion.” GE’s not alone; many good boards never have meetings that exclude the CEO.

Another supposed safeguard of good governance—audit and compensation committees—turns out to be near universal. A 2001 survey by the NACD and Institutional Shareholder Services of 5,000 public company boards shows that 99% have audit committees, and 91% have compensation committees. Sunbeam, Enron, Cendant, McKessonHBOC, and Waste Management all had the requisite number of committees and guidelines, yet accounting scandals still penetrated this governance shield. Let’s not forget, either, that the audit committee at Enron was consulted about suspending the conflict-of-interest guidelines and willingly agreed to it.

The Importance of the Human Element

So if following good-governance regulatory recipes doesn’t produce good boards, what does? The key isn’t structural, it’s social. The most involved, diligent, value-adding boards may or may not follow every recommendation in the good-governance handbook. What distinguishes exemplary boards is that they are robust, effective social systems. Let’s see what that means.

1. A Virtuous Cycle of Respect, Trust, and Candor

It’s difficult to tease out the factors that make one group of people an effective team and another, equally talented group of people a dysfunctional one; well-functioning, successful teams usually have chemistry that can’t be quantified. They seem to get into a virtuous cycle in which one good quality builds on another. Team members develop mutual respect; because they respect one another, they develop trust; because they trust one another, they share difficult information; because they all have the same, reasonably complete information, they can challenge one another’s conclusions coherently; because a spirited give-and-take becomes the norm, they learn to adjust their own interpretations in response to intelligent questions.

The UPS board of directors has just that kind of chemistry, and as a result members have debated strategic decisions openly and constructively for years. The company’s 1991 move from Connecticut to Georgia was hotly debated within the management committee, for example, but once the plan to move was agreed upon, the board chose a new location unanimously and never looked back. In the mid-1980s, after forging partnerships with delivery businesses around the world, a revolutionary concept at the time, the company decided to reverse course and become truly global itself. In just two years, UPS was running operations in more countries than are members of the United Nations. This strategic reversal is generally considered a brilliant move, one that might never have happened had board members not respected and trusted one another enough to consider that a smart move could be trumped by an even smarter one. The board even tolerated an open debate in 1992, led by a former CEO, over the company’s widely recognized corporate color, brown—the hallmark of UPS’s current advertising campaign.

A virtuous cycle of respect, trust, and candor can be broken at any point. One of the most common breaks occurs when the CEO doesn’t trust the board enough to share information. What kind of CEO waits until the night before the board meeting to dump on the directors a phone-book-size report that includes, buried in a thicket of subclauses and footnotes, the news that earnings are off for the second consecutive quarter? Surely not a CEO who trusts his or her board. Yet this destructive, dangerous pattern happens all the time. Sometimes a CEO’s lack of trust takes even more dramatic forms. It’s stunning that Enron’s chairman and CEO never told the board that whistle-blower Sherron Watkins had raised major questions about financial irregularities. It is impossible for a board to monitor performance and oversee a company if complete, timely information isn’t available to the board.

It is, I should note, the responsibility of the board to insist that it receive adequate information. The degree to which this doesn’t happen is astonishing. Consider Tyco. In recent quarters, it’s suffered some of the worst strategic confusion I’ve ever witnessed: Seemingly every single public statement by the company’s senior management has been contradicted by subsequent statements. For example, in January 2002, then CEO Dennis Kozlowski announced a plan to split the company into four pieces, only to reverse that plan a few months later. On a single day, senior managers announced first that a financial unit would be IPO’ed, next that it would be sold to an investment house, and finally that neither would occur. Where was the board? Why didn’t directors demand a better accounting of the company’s direction and well-being? What brought down the CEO eventually was an apparently private financial matter—the board seemed content to keep him on indefinitely.

Another sign that trust is lacking is when board members begin to develop back channels to line managers within the company. This can occur because the CEO hasn’t provided sufficient, timely information, but it can also happen because board members are excessively political and are pursuing agendas they don’t want the CEO to know about. If a board is healthy, the CEO provides sufficient information on time and trusts the board not to meddle in day-to-day operations. He or she also gives board members free access to people who can answer their questions, obviating the need for back channels.

Another common point of breakdown occurs when political factions develop on the board. Sometimes this happens because the CEO sees the board as an obstacle to be managed and encourages factions to develop, then plays them against one another. Pan Am founder Juan Trippe was famous for doing this. As early as 1939, the board forced him out of the CEO role, but he found ways to sufficiently terrorize the senior managers at the company and one group of board members that he was returned to office. When he was fired again following huge cost overruns on the Boeing 747 the company underwrote, he coerced the directors into naming a successor who was terminally ill.

Most CEOs aren’t as manipulative as Trippe, and in fact, they’re often frustrated by divisive, seemingly intractable cliques that develop on boards. Failing to neutralize such factions can be fatal. Several members of Jim Robinson’s American Express board were willing to provide the advice, support, and linkage he needed—but the board was also riddled with complex political agendas. Eventually the visionary CEO was pushed out during a business downturn by a former chairman who wanted to reclaim the throne and a former top executive of another company who many felt simply missed the limelight.

The CEO, the chairman, and other board members can take steps to create a climate of respect, trust, and candor. First and most important, CEOs can build trust by distributing reports on time and sharing difficult information openly. In addition, they can break down factions by splitting up political allies when assigning members to activities such as site visits, external meetings, and research projects. It’s also useful to poll individual board members occasionally: An anonymous survey can uncover whether factions are forming or if members are uncomfortable with an autocratic CEO or chairman. Other revelations may include board members’ distrust of outside auditors, internal company reports, or management’s competence. These polls can be administered by outside consultants, the lead director, or professional staff from the company.

2. A Culture of Open Dissent

Perhaps the most important link in the virtuous cycle is the capacity to challenge one another’s assumptions and beliefs. Respect and trust do not imply endless affability or absence of disagreement. Rather, they imply bonds among board members that are strong enough to withstand clashing viewpoints and challenging questions.

I’m always amazed at how common groupthink is in corporate boardrooms. Directors are, almost without exception, intelligent, accomplished, and comfortable with power. But if you put them into a group that discourages dissent, they nearly always start to conform. The ones that don’t often self-select out. Financier Ken Langone tells the story of a widely admired CEO who was invited to join the board of a famous corporation that is suffering great distress today. He was told that, as a matter of custom, new directors were expected to say nothing for the first 12 months. The candidate said, “Fine, I’ll see you in a year,” and of course never got the appointment. Langone explained that directors generally feel that they are under pressure to fit in so they’ll be renominated. As he put it, “Almost no one wants to be a skunk at a lawn party.”

Even a single dissenter can make a huge difference on a board. Bill George, a former CEO and chairman of the board of Medtronic, reported that a lone dissenter had forced his company to reconsider near unanimous decisions on several occasions. One pharmaceutical director held out in opposition to Medtronic’s acquisition of Alza, a maker of drug delivery systems, saying it would take Medtronic into an area it knew nothing about. He was so convincing that the acquisition was abandoned, and in retrospect, that was the right decision. Another dissenter convinced George and the board to reverse themselves and not to get out of the angioplasty business—and, indeed, to intensify those services—and that shift has paid off handsomely.

Frequently, executive recruiters looking for leads during board candidate searches will ask, “Is this fellow a team player?” which is code for “Is this person compliant, or does he make trouble?” If a board member challenges major decisions, a company sometimes goes to great lengths to discredit the person. Consider Walter Hewlett—an academic; the cofounder’s son, who controlled 18% of Hewlett-Packard stock; and someone with a deep understanding of the computer business—who had the temerity to question HP’s proposed merger with Compaq in the fall of 2001. Despite the fact that technology mergers rarely work, his point of view was summarily dismissed internally. When he was forced to go public with his objections, he was ridiculed publicly in a smear campaign.

CEOs who don’t welcome dissent try to pack the court, and the danger of that action is particularly clear right now. Recall that Enron board members Rebecca Mark and Clifford Baxter resigned reportedly because they were uncomfortable with paths the company had taken. And one can imagine a happier ending at Arthur Andersen had somebody said, “Wait a minute,” when the document shredding began, or at Tyco when the board learned of millions in undisclosed loans to the CEO and didn’t question them.

The CEO, the chairman, the lead director, and the board in general need to demonstrate through their actions that they understand the difference between dissent and disloyalty. This distinction cannot be legislated through nominating committee rules and guidelines for director résumés; it has to be something that leaders believe in and model. Home Depot chairman Bernie Marcus notes that, for one simple reason, he’d never serve on a board where dissent was discouraged: When he serves on a board, his reputation and his fortune are on the line. A lost reputation can’t be regained, and director’s insurance won’t necessarily protect anyone’s fortune, because there are always exemption clauses. Marcus has remarked, “I often say, ‘I don’t think you want me on your board. Because I am contentious. I ask a lot of questions and if I don’t get the answers, I won’t sit down.’ That’s the kind of board member that I want on my board…because our company needs help. We think we’re bright, but we’re not the smartest people in the world.” Ken Langone corroborates this view of the Home Depot board. Both he and Marcus describe times when the board disagreed with management about strategic questions—when reformulating the small-store concept, for example, and when revisiting expansion into Latin America. The upshot wasn’t that the board won and management lost, but rather that, after passionate disagreements had been voiced, together they arrived at new conclusions.

According to data complied by Kathleen Eisenhardt and L.J. Bourgeois, the highest-performing companies have extremely contentious boards that regard dissent as an obligation and that treat no subject as undiscussable. Directors at these companies scoff at some of the devices more timid companies use to encourage dissent, such as outside directors asking management to leave while they discuss company performance. What’s the point of criticizing management, they ask, if management isn’t there to answer the criticism? It should be noted that skepticism and dissent don’t constitute disagreement for its own sake but rather are the by-products of a constantly evolving view of the business and of the world.

3. Fluid Portfolio of Roles

When board members don’t challenge one another, individual directors’ roles—the ruthless cost cutter, the damn-the-details big-picture guy, the split-the-differences peacemaker—can become stereotyped or rigid. Effective boards require their members to play a variety of roles, in some cases dipping deep into the details of a particular business, in others playing the devil’s advocate, in still others serving as the project manager. Playing different roles gives directors a wider view of the business and of the alternatives available to it.

Occasionally board members can so thoroughly transcend their normal roles that they’re able to change their minds about something they once built their lives around. This happened at PepsiCo in 1997 when the board decided to sell the various components of its well-run restaurant group. CEO Roger Enrico had previously turned around the unit— which had been the brainchild of two of Enrico’s predecessors, Don Kendall and Wayne Calloway—and must have felt great pride of ownership. Yet he eventually convinced all that the restaurant unit should be sold so that it could flourish freely beyond the controls of the parent company. It’s proved to be a brilliant decision.

4. Individual Accountability

Board accountability is a tricky problem for CEOs, as a 2002 survey by the Yale School of Management and the Gallup Organization underscores. In that survey, fully 25% of CEOs claim that their board members do not appreciate the complexity of the businesses they oversee. In addition, we’ve all seen instances when individual responsibility dissolved in large groups. This certainly appears to have happened at Enron: Practically everyone involved has pointed the finger of blame at others or proclaimed his or her ignorance as a badge of honor. The fact that many board members were financially sophisticated seemed to have encouraged the other board members to defer to their expertise.

There are various methods for enforcing accountability. Home Depot’s board members are expected to visit at least eight stores outside their home state between board meetings; GE’s board members dine with the company’s largest suppliers and distributors the night before the annual meeting. Perhaps the most effective enforcement mechanism, though, is old-fashioned peer pressure. Directors who take their duties seriously, and let their fellow directors know they’re expected to do the same, are the best insurance against a board whose first question, upon receipt of the quarterly earnings report, is, “When’s lunch?”

5. Performance Evaluation

I can’t think of a single work group whose performance gets assessed less rigorously than corporate boards. In 2001, the NACD surveyed 200 CEOs serving as outside directors of public firms. Sixty-three percent said those boards had never been subjected to a performance evaluation. Forty-two percent acknowledged that their own companies had never done a board evaluation. A 2001 Korn/Ferry study of board directors found that only 42% regularly assess board performance, and only 67% regularly evaluate the CEO.

This lack of feedback is self-destructive. Behavioral psychologists and organizational learning experts agree that people and organizations cannot learn without feedback. No matter how good a board is, it’s bound to get better if it’s reviewed intelligently.

A performance review can include a full board evaluation, individual directors’ self-assessments, and directors’ peer reviews of one another. Most often, the nominating or governance committee drives these evaluations. A full board review can include an evaluation of such dimensions as its understanding and development of strategy, its composition, its access to information, and its levels of candor and energy. In individual self-assessments, board members can review the use of their time, the appropriate use of their skills, their knowledge of the company and its industry, their awareness of key personnel, and their general level of preparation.

The peer review can consider the constructive and less constructive roles individual directors play in discussions, the value and use of various board members’ skill sets, interpersonal styles, individuals’ preparedness and availability, and directors’ initiative and links to critical stakeholders. This process is often best driven by a board committee such as a nominating or governance committee, which is assigned the execution and follow-through responsibilities for this process.

Annual evaluations led PepsiCo and Target to change their processes for reviewing strategy with their boards. Instead of the mind-numbing, back-to-back, business-unit dog and pony shows that boards often suffer, each company decided to spend a full day of each board meeting looking in depth at the strategic challenges of a single business unit.

We all owe the shareholder activists, accountants, lawyers, and analysts who study corporate governance a debt: In the 1980s and 1990s, they alerted us to the importance of independent directors, audit committees, ethical guidelines, and other structural elements that can help ensure that a corporate board does its job. Without a doubt, these good-governance guidelines have helped companies avoid problems, big and small. But they’re not the whole story or even the longest chapter in the story. If a board is to truly fulfill its mission—to monitor performance, advise the CEO, and provide connections with a broader world—it must become a robust team—one whose members know how to ferret out the truth, challenge one another, and even have a good fight now and then.