The Puzzle of Independent Directors: New Learning

In this symposium paper, I discuss and critique some new empirical learning on independent directors. The independent director has always offered a sort of magic bullet for corporate governance, representing the idealized monitor of executives’ behavior. Yet we corporate law scholars also harbor some ambivalence about the magic of this bullet. As much as we want to trust in the promise of independent directors, no solid empirical evidence exists to suggest that independent directors add value. Moreover, we have seen spectacular failures in the face of independent boards. How do we account for this disconnect between our intuitions and best intentions, on the one hand, and the stubborn refusal of the empirical evidence to confirm our faith in independent directors? Several possibilities come to mind. First, existing definitions of independence may be too lax. Independence requirements may fail to screen out important conflicts. A second possibility (not exclusive of the first) is that firms may be heterogeneous, such that optimal board composition may vary across firms. Independent boards may add value at some firms but not others. An emerging theoretical literature argues, for example, that firms’ information environments matter for independent directors’ efficacy. Monitors’ incentives and information are central to constraining agency costs. Happily, two recent empirical studies tackle these important issues. On incentives, one study investigates whether the existence of common backgrounds between CEOs and their nominally independent directors may affect directors’ monitoring. Another recent study focuses on directors’ costs of acquiring information about their firms and the effects of these costs on independent directors’ effectiveness. These studies suggest that the independent director’s place in corporate governance may be more complicated than we thought. We may need to refine our trust in independent directors and tailor our expectations about their utility