What Characteristics Make a Good CEO?

I gather that David Brooks only read the abstract of the paper that he used as a springboard for today’s column:

Should C.E.O.’s read novels?

The question seems to answer itself. After all, C.E.O.’s work with people all day. Novel-reading should give them greater psychological insight, a feel for human relationships, a greater sensitivity toward their own emotional chords.

Sadly, though, most of the recent research suggests that these are not the most important talents for a person who is trying to run a company. Steven Kaplan, Mark Klebanov and Morten Sorensen recently completed a study called “Which C.E.O. Characteristics and Abilities Matter?”

They relied on detailed personality assessments of 316 C.E.O.’s and measured their companies’ performances. They found that strong people skills correlate loosely or not at all with being a good C.E.O. Traits like being a good listener, a good team builder, an enthusiastic colleague, a great communicator do not seem to be very important when it comes to leading successful companies.

Here’e the abstract:

We study the characteristics and abilities of CEO candidates for companies involved in buyout (LBO) and venture capital (VC) transactions and relate them to hiring decisions, investment decisions, and company performance. Candidates are assessed on more than thirty individual abilities. The abilities are highly correlated; a factor analysis suggests there are two primary factors with intuitive characterizations — one for general ability and one that contrasts team-related, interpersonal skills with execution skills. Both LBO and VC firms are more likely to hire and invest in CEOs with greater general abilities, both execution- and team-related. Success, however, is more strongly related to execution skills than to team-related skills. Success is, at best, only marginally related to incumbency, holding observable talent and ability constant.

As I suspected you get a somewhat different perspective when you read the digest which adds the following to the abstract:

The authors suggest that their results have several implications. First, it is possible to measure individual CEO talents and skills over and above the usual observable variables like age, industry, and college SAT scores. Second, CEO talents and skills appear to matter because they are consistently correlated with hiring, investment, and performance. Third, success and performance are more strongly correlated with execution-type skills than with interpersonal and team-related skills. In other words, CEOs like Jack Welch appear to be more successful than CEOs like Jeff Immelt. This is consistent with other researchers’ results (in non-CEO contexts) showing that steadfastness — and traits such as unwavering resolve, fanatical drive, and workmanlike diligence — is more important than being a good listener. The authors’ findings are also consistent with results in the psychology literature that suggest that “conscientiousness” is the best predictor of performance, but they do not support previous findings that successful CEOs exhibit compelling modesty, build strong teams, give credit to others, and take blame on themselves.

Finally, the authors point out that their results reflect buyout and VC-funded companies only. While these are two quite different groups, these types of companies may have specific needs and, therefore, the results may not generalize to all companies. Second, the performance data are coarse and potentially “noisy.” But, that said, the authors’ results correlate strongly with Peter Drucker’s description of the effective executive. The attributes Drucker describes are largely execution-related and appear to correspond well to the “efficient,” “persistent,” “proactive,” “commitment,” and “analytical” skills in this study.

The emphasis is mine and those were my precise reactions on reading the abstract. Are the needs of buyout companies and venture capital funded companies different from those of companies, generally? I honestly have no idea. And notice how small the sample size is. Someday I’m going to have to write a post on legitimate small sample statistics. It’s actually quite an interesting subject but I find small samples incredibly misused and abused.

Over the years I’ve met (as in had lunch or dinner one-to-one) with the CEO’s of a number of large as in Fortune 500 companies. I’m not sure I could make an honest generalization other than it was clear that they enjoyed being the CEO’s of Fortune 500 companies and I wouldn’t want to be in a dark alley with them if I had something they wanted.

I think the qualities one might look for in a Fortune 500 CEO are probably different from those that make a good team leader. I also wouldn’t be surprised if the qualities needed to run technology companies successfully are different from those required to run, say, a manufacturing or services firm successfully. Can one really make a true generalization about what makes a good CEO? I’m not so sure.