We have all heard that diversifying the skills on your board would bring you better performance in the long run. However, recent studies have shown that this is not necessarily the case. During the annual Center for Corporate Governance Academic Conference at Drexel University, leading financial academics were invited to present their current analytic and empirical research on corporate governance. One of the papers presented, titled “Director Skill Sets” and written by Adams, Akyol and Verwijmeren, takes a data analytics approach to examine how director skills cluster on and across various boards. Factor analysis (a process in which the values of observed data are expressed as functions of a number of possible causes) indicates that the key measurement by which boards vary is the variety of skills of their directors. The authors did not start with the premise that skill diversity matters. Instead, the research indicated that skill diversity arises naturally as an important characteristic from the factor analysis. A key finding of the report showed that firm performance was positively related to having more commonality in skill sets on the board and between inside and outside directors. In fact, it showed that there is actually a negative correlation between skill diversity and firm performance. A result contrary to the widespread adage that diversity improves performance.
The authors started with 1,481 firms in the RiskMetrics database. They eliminated firms headquartered overseas (57 firms) and removed 337 utilities and financial firms from the sample. They then collected descriptions of director skill sets following the 2009 amendment to regulation S-K from the 2010 proxy statements from SEC Edgar. They further excluded 181 firms that did not disclose director skills for all directors on their boards. Of the 906 remaining firms, the main performance measure was based on Tobin’s Q (the ratio of the total market value to the total asset value). The firms had an average market value of about 8.3 billion dollars and an average Tobin’s Q of 1.867, with a return on assets of 4.2%. The typical firm had 9 board members and 4 board committees. Seventy nine percent of the directors in a typical firm were classified as independent. Included in the study were 6,409 outside directors and 1,199 inside directors. In defining skills the authors tried to stay as consistent with the firms’ definitions as possible and settled on 20 skill sets.
The results showed that there is a negative correlation between skill diversity and firm performance. So, what drives the negative relationship between the skill diversity and performance? The differences! The authors suggested that skill diversity leads to less effective decision-making because directors have less common ground (as measured by skill overlap). It can be further argued that the depth of discussions and quality of decisions are potentially higher with more than one director knowledgeable on a subject. This is particularly evident in complex strategic decisions, like in M&A and technology acquisitions.
The authors also considered the overlap in skills between inside and outside directors and found that a high concentration of skills between inside and outside directors was, in fact, positively related to firm performance. This proposes that a lack of common ground between insiders and outsiders is a particularly important factor.
So, what are the implications for corporate governance?
First, the study found that boards whose directors have more commonality in skill sets have better company performance. This is not necessarily inconsistent with the adage that diversity is good. The differentiator is in the word “sets,” a collection of skills on the board that can lead to higher quality discussions and decisions. Second, new directors come with a set of skills. Nominating committees implicitly understand they face multi-dimensional constrained optimization problems in recruiting new directors. Third, a strong overlap in skills between outside and inside directors strengthens the outside directors’ interaction with the company. While some of the results appear intuitive, and others less so, the research is founded on a data analytics approach. This study is expected to lead to further research along several productive directions.