Brian Tayan is a researcher with the Corporate Governance Research Initiative and David F. Larcker is the James Irvin Miller Professor of Accounting, Emeritus, at Stanford Graduate School of Business. This post is based on their recent piece. Related research from the Program on Corporate Governance includes What Matters in Corporate Governance? (discussed on the Forum here) by Lucian Bebchuk, Alma Cohen and Allen Ferrell.
We recently published a paper on SSRN (“Seven Gaping Holes in our Knowledge of Corporate Governance”) that reexamines foundational assumptions within corporate governance.
Nine decades after Berle and Means proposed a theory of corporate governance, our knowledge of its “best practices” remains woefully incomplete. Corporate governance is a social science, which means that while the factors that determine its effectiveness are complex, they are at their core subject to theory, measurement, and analysis. From the conversation today, however, one would hardly recognize this fact. Instead, the dialogue about corporate governance is dominated by rhetoric, assertions, and opinions that—while strongly held—are not necessarily supported by either applicable theory or empirical evidence. Having to choose between the results of the scientific record and their gut, many “experts” prefer their gut.
While some of the blame for this state of affairs lies with these experts, the academic and institutional research literature itself is not above reproach. Although many aspects of governance have been the subject of empirical study, our knowledge of its central characteristics is incomplete. Organizations are complex entities, and the ability of social scientists to distill their effectiveness to prescriptive best practices is limited. Many studies involve large samples of data. Large samples enable a researcher to identify patterns across many companies, but generally do not tell us how corporate governance choices would impact a specific company. Case studies or field studies can help answer firm-specific questions, but the results tend to be highly contextual and difficult to generalize. Most observational social science studies suffer from the challenges of measuring variables and demonstrating causality based on data. Empirical tests can identify associations and correlations between variables, but it is exceedingly difficult to prove that a variable caused an outcome. And in the case of corporate governance, many important variables are not publicly observable to outside researchers—forcing them to develop proxies to estimate the variable they want to measure. It is extremely difficult to produce high-quality, fundamental insights into corporate governance because of these limitations.
