What Do Boards Really Do? Evidence from Minutes of Board Meetings

In our paper, What Do Boards Really Do? Evidence from Minutes of Board Meetings, which was recently made available on SSRN, we analyze a unique database from a sample of real-world boardrooms – minutes of board meetings and board-committee meetings of eleven business companies for which the Israeli government holds a substantial equity interest. We use these data to evaluate the underlying assumptions and predictions of models of boards of directors. In recent years, more than a dozen economic models have attempted to examine what boards actually do. However, because board meetings are generally a “black box” to which scholars have very limited access, these models proceed from wildly different underlying assumptions, and accordingly, make very different predictions. These models generally fall into two categories:

• (1) “Managerial models” – assume boards play a direct role in managing the firm, and that they make the actual decisions pertaining to the business of the firm.
• (2) “Supervisory models” – assume that the board only observes the CEO’s actions, but does not make any business decisions. Based on the boards’ observations, it evaluates/reevaluates the CEO, and decides whether to retain or fire the CEO.

We evaluate the extent to which these models depict a reasonable reality and realistic board dynamics: we analyze a unique database from a sample of real-world boardrooms – minutes of board meetings and board-committee meetings of eleven business companies for which the Israeli government holds a substantial equity interest (GBCs).  This approach allows us to examine the day-to-day functions of boards.

Consistent with the supervisory models, our minutes-based data suggest that monitoring is the most common and typical work of boards. In our sample, 67% of the issues the boards discussed were of a supervisory nature, boards were presented with only a single option in 99% of the issues discussed, and at the voting phase they disagreed with the CEO only 3.3% of the time. Nevertheless, we find that managerial models describe a reality that also exists at times. Boards requested to receive further information or an update for 8% of the issues discussed, and they took an initiative with respect to 8.1% of them. Furthermore, in 63% of the meetings, boards took at least one of these actions or did not vote in line with the CEO, indicating that in most meetings boards are active, and take at least one action.

We also examine the extent of dissension among directors. In most economic models boards are modeled as a monolithic entity. We find support for this approach: in only 3.3% of the cases in the sample, boards did not vote unanimously. The minutes data allows us to draw some inferences that are impossible to make using publicly available data. For example, our sample suggests that prior work understates the fraction of CEO departures that are “forced”. While our sample is too small to draw reliable estimates of the understatement, there are at least two cases in our sample  (from a total of four departing CEOs) where the CEO was clearly coerced to leave by the board, yet there would be no way to know about this coercion using only publicly available data. The existence of these cases suggest that estimates of the fraction of CEO turnovers that are forced using publicly-traded data will underestimate perhaps substantially, the fraction of turnovers that are initiated by the board.

Last, we find evidence that larger boards are more active – in our sample larger boards are more likely to request further information or an update. However, larger boards also seem to have more coordination problems, as reflected by their voting patterns: they were more likely not to reach a consensus by the voting phase – i.e. not to vote unanimously. In addition, the CEO was also less likely to receive the boards’ consent to his proposal in larger boards, perhaps, indicating that CEOs with larger boards are less able to align the board to their proposal prior to the board meeting.

A potential concern is the extent to which the boards of our sample of Israeli government-controlled companies reflect other companies. It is impossible to know exactly how different our firms’ governance is from that of privately held companies in both in Israel and the rest of the world. Nevertheless, like directors of privately-held companies, these directors still have a fiduciary responsibility to maximize their firm’s profits, and our reading of the minutes suggests that they take this responsibility seriously. Furthermore, as we specify throughout the paper, the board dynamics we document are similar to those reported in interview-based studies, which are most often based on publicly-traded U.S. companies. For these reasons, we believe that the relationship between a CEO and a GBC board, and among the directors of GBCs, is likely to be similar to the corresponding relationships in other boardrooms.

To understand the role of boards of directors, we believe it is necessary to observe to the extent possible how they actually function. This paper is the first to document in any systematic fashion what boards actually do. It is our hope that future research will be able to perform similar analyses for other samples of companies.

The full paper is available for download here.