஽ Academy of Management Executive, 2004, Vol. 18, No. 1 ........................................................................................................................................................................ Good governance and the misleading myths of bad metrics Jeffrey Sonnenfeld In the aftermath of the well-publicized corruption governance metrics ratings services, the best and malfeasance in several large public corpora- known of which are Institutional Shareholder Ser- tions, especially at the executive and board levels, vices (ISS) and Governance Metrics International investors and analysts are searching for manage- (GMI). While firms such as Moody’s and The Cor- ment tools to measure the vulnerability of firms to porate Library use a wide mix of criteria to evalu- dishonesty, fraud, and corruption. While this effort ate companies, including their openly qualitative to improve governance through uniform guidelines judgment, ISS and GMI rely more on crisp numer- is understandable, at times boards and companies ical scoring systems. Some even believe that it is are reaching out for any life preserver that comes dangerous for firms to challenge the influential ISS along. Some firms are capitalizing on this desper- and GMI ratings services, given the attention paid ation by setting themselves up as corporate gover- to them by credit analysts, institutional portfolio nance experts. In 1999, when William Donaldson managers, and liability underwriters.3 was chairman of Aetna, he said prophetically, “I ISS and GMI look at public records to score firms fear that there is a growing cottage industry of on their governance effectiveness by using sim- superficial thought about corporate governance.”1 plistic checklists of standards or metrics based The swelling number of governance consultants heavily upon cliche´s and myths, rather than on has made Donaldson’s statement truer than ever. genuine research. They also may cross the line The vogue in the consulting world, in fact, is gov- from being independent raters to becoming active ernance—supplanting business process reengi- consultants for the firms they study in ways which neering, the “new economy,” transformational lead to questions about their objective credibility. leadership, diversity training, right-sizing, total Finally and most importantly, their methods do not quality management, and the like. work; reliable, accurate governance ratings are not Some of what is being sold by the close to 100 really produced despite all the charts and lists governance training programs offered by consult- published. These three aspects of corporate gover- ing firms and universities is truly disturbing be- nance ratings services—using evaluation stan- cause it is often anchored more in cliche´s and dards based on Wall Street superstitions rather myths than in careful research. In a recent review than research, potential conflicts of interest, and of academic studies on governance, the Financial providing ratings that don’t work—are discussed Times suggested that many of the supposedly pre- in the following sections. ventive practices advocated are not truly related to better performance and concluded, “Perhaps it is time the corporate governance activists came un- Governance Expertise: Mixing Fact and Fiction in der the sort of scrutiny to which they subject listed Measurement companies.”2 Certainly the ratings services examine such worth- while factors as financial disclosure, shareholder rights, related-party transactions, and executive The Metrics Rating Services compensation. These are sensible, research-sup- The problematic nature of what is often being sold ported dimensions to include in measures of the by commercial governance consultants is epito- effectiveness of corporate governance.4 But ISS mized by the offerings of the powerful and feared and GMI blend these dimensions with supersti- 108 2004 Sonnenfeld 109 tious ones to create checklists of highly stringent of outside/independent directors without manage- standards, regardless of the genuine research ment present; or (c) substantial adherence to the foundation to support them. They cite the col- well-known General Motors guidelines for corpo- lapsed firms of Enron and Worldcom as examples rate governance. This was thus not a study of the of poor governance without demonstrating how structural attributes of boards.6 A recent study by well these firms met many prominent structural Paul Gompers, Joy Ishii, and Andrew Metrick found dimensions of supposed good governance. that companies with strong shareholder rights had They perpetuate unfounded myths and cliche´s higher annual returns, profits, and sales growth by downgrading firms for such reasons as failing than companies with weak rights. But again, to have a retirement age for directors and failing to though sometimes offered as substantiation for the separate the chairman and CEO roles. They claim need to reform board structure, this was not a study that the downfall of many corporations has re- of structure.7 sulted from a lack of financial expertise on the Finance studies by Sunil Wahal and Michael board. Other reasons for poor ratings are failing to Smith suggest that even when shareholder activ- require that managers and directors have a for- ists have been able to change firm governance mally set amount of equity holdings; prior history structures, the changes have not translated into of service on boards of firms suffering financial improvements in operating performance.8 Simi- distress; failure to have a formal retirement age, larly, in research I have been doing with Sanjay board size, and code of conduct; allowing a former Bhagat of Colorado and Dick Wittink of Yale on CEO to serve on the firm’s board; failing to have a 1500 public companies, we are finding no support separate chairman and CEO; and failing to have a for a relationship between structural dimensions supermajority of outside, independent directors. In of board governance and company performance. sum, the ratings services evaluate the corporate governance of firms by mixing together empiri- cally based standards and the myths and cliche´s The Age Myth of “the Street.” Let us examine some of these myths and superstitions on which many corporations are There is no research suggesting that increased measured, to see how wrong they can be. director age leads to impaired judgment. In fact, experience is often found to be advantageous in decision-making. Cognitive and developmental The ratings services evaluate the psychologists have mapped a strong correspon- corporate governance of firms by mixing dence between age, wisdom, and judgment on and together empirically based standards off the job.9 In particular, these studies have indi- and the myths and cliche´s of “the cated an age-related strength in competency in the Street.” face of uncertainty and in perceiving others’ inten- tions, as well as stronger communications skills. Term limits and age limits for board members are commonly discussed, but age-biased policies for The Structure Myth board turnover lack genuine validation. One problem is that certain studies not actually showing a relationship between board structure and performance are often cited as justification for The Split CEO/Board Chairman Myth structural reform, while true structural studies do not find relationships that matter between struc- The Conference Board recommended either split- ture and performance. While a frequently cited ting the CEO and chairman roles or using lead McKinsey study suggests that investors were will- directors or presiding directors.10 The metrics ser- ing to pay an 18 per cent premium for a well- vices also favor firms that divide these functions. governed firm, such “good governance” was not And yet, many if not most of the highest-profile defined in terms of any explicit board structure scandals in the US and Europe, (e.g., Enron, World- requirements.5 com, Vivendi, Adecco, Royal Ahold, ABB, Manes- Millstein and MacAvoy studied the relationship mann, Deutsche Telecom) involved firms that had between board independence and corporate per- separated the CEO and chairman roles, but the formance to suggest that an active board made a split hardly prevented subsequent scandals. Ac- difference. A board was deemed active if it met any cordingly, there is no research that has estab- one of the following criteria: (a) a non-executive lished a link between the split leadership roles chairman or lead director; (b) scheduled meetings and firm performance. 110 Academy of Management Executive February The Financial Expertise Myth their results showed a significant correlation be- tween the amount of stock owned by individual A recent advertisement I received suggested a outside directors and firm performance as well as higher level of financial literacy as the solution to an increased likelihood that CEOs would be termi- governance crises such as those experienced at nated in poor-performing firms. Since this was not Enron, Worldcom, the New York Stock Exchange, a longitudinal study, however, the findings are and Freddie Mac, despite the fact that these and suggestive but do not prove causality. many other struggling and collapsed firms had Moreover, how much stock is enough, and does it boards dominated by wide-ranging financial wiz- matter if the policy is observed but not codified in ards including Ph.D. academicians in finance. In- a formal written mandate? Equity holdings by di- sufficient financial expertise has rarely been the rectors in firms