Beyond Anecdotes, CEOs Really Do Matter

Tuck professor Constance Helfat uses a statistical approach to prove the impact of CEO behavior.

So it makes sense that CEOs should also have a large impact on the strategy and success (or failure) of firms. For decades, that’s what many people have thought. Boards consider it one of their most important jobs to find the right corporate leader. Industry observers follow CEO searches closely, thinking that the future of a firm turns to a substantial degree on the skills and nature of the person on top. This belief—call it the Great Man Theory of Business—has been confirmed by academic research, but evidence of CEOs’ role in strategic behavior of firms has been scant.

With a new working paper, Tuck professor Constance Helfat changes that. In “Disentangling the Microfoundations of Dynamic Capabilities: Evidence from Microdata on Acquisitions,” Helfat, the J. Brian Quinn Professor in Technology and Strategy and the associate dean for research innovation, and colleagues Philipp Meyer-Doyle of INSEAD and Sunkee Lee of the Tepper School of Business, show that the wide variance in the number and quality of corporate acquisitions within and across industries is due in part to differences in the ability of CEOs to sense and seize opportunities, and then to reconfigure the business to capitalize on them—otherwise known as dynamic capabilities.

Helfat and her co-authors don’t study CEOs directly. Instead, they infer their importance through the lens of mergers and acquisitions—corporate activities that every CEO is closely involved with, and that can bring large changes to a firm.  “Acquisitions are major strategic decisions, so they are intrinsically relevant,” Helfat says, “and CEOs are going to have a lot of input.”

Some firms complete many acquisitions, others do it infrequently or never at all. Some of these acquisitions are very successful—the new company is integrated smoothly into the parent firm, and the business grows (think ExxonMobil)—while others are akin to a misguided marriage that just doesn’t work out (DaimlerChrysler).

Helfat and her colleagues take advantage of this heterogeneity, using a type of statistical analysis called “variance decomposition.” In simple terms, the analysis takes the variance in behavior and performance and parcels it out into different factors, including CEOs. They then compare the relative impact of those factors on acquisition behavior and performance. In this case, the measures that the researchers use enable them to make inferences about the dynamic capabilities of CEOs and firms.

They find that CEOs make a larger contribution to sensing and seizing behavior in the context of acquisitions than other factors associated with firms as a whole, and that CEOs play an important role in the quality of those acquisitions. This helps explain the variance in the number and quality of acquisitions across firms, and their performance after the acquisition.

“We started looking at the big things CEOs could affect,” Helfat says, “and it turns out they do matter for explaining why firms undertake fewer or greater numbers of acquisitions, and why there’s variance in the quality of acquisitions.”