The board demography – firm performance relationship revisited: opening the ‘black box’ using Bayesian analysis

Abstract

Boards of directors have enjoyed an increased attention in recent management research. Because of scant and conflicting empirical evidence on the relationships between board demographic variables and firm performance, recent research has shown special interest for the antecedents of board performance in an effort to open the black box of the board of directors. We build on this research by testing a model that incorporates not only these board-level outcomes but also includes the preceding board demographics and subsequent firm performance. Moreover, building on behavioral theory, we argue that in order to study board behavior, focusing merely on actual board task performance is not enough. An integration of board task needs by studying how they differ from the actual board task performance is called for. We make use of a Bayesian estimation method that allows us to make firm specific estimates of the effect of board composition on firm performance, which we can then further analyze by using board behavior measures. Our findings indicate that the effects of the board demographic characteristics -board size and percentage of family directors- on firm performance, are negatively related to the size of the gap between the board task needs and board task performance for the service as well as the control task.