Making dumb groups smarter
Harvard Business Review, December 2014
Boards may make better decisions if directors have some understanding of the psychology of decision-making. In an article in the Harvard Business Review, ‘Making dumb groups smarter’, Sunstein and Hastie explore the idea of “groupthink”, which they describe as the tendency of groups to go astray and to fail to live up to their potential.
The authors note that groups err for two main reasons: (i) incorrect informational signals from other group members; and (ii) reputational pressures, which lead individuals to silence themselves or change their views in order to avoid a penalty (often being the disapproval of others). As a result, groups run into four problems, which may lead to poor decisions:
- amplification of many individual biases and mental short-cuts (e.g. overconfidence, the sunk-cost fallacy etc.);
- the “cascade effect” or “herding”, where group members follow the statements and actions of those who spoke or acted first;
- polarisation, where members take more extreme positions than those held before deliberations; and
- “the common knowledge effect”, where information held by all group members has a greater influence on group judgments than information held by one or a few.
The authors conclude that a central goal in group decision-making should be to ensure that groups aggregate the information their members actually have and do not let faulty informational signals and reputational pressures get in the way. They describe several ways to achieve that goal, including:
- leaders and high-status group members can encourage others to express their own views by refusing to take a firm position at the outset, and indicating a willingness and desire to hear uniquely held information;
- informing the group that each member has a different and relevant role, or distinctive information to contribute;
- encouraging authentic dissent or, potentially, appointing a “devil’s advocate” or “red team” (for example, to construct a case against a proposal or plan);
- voting anonymously to insulate group members from reputational pressures and reduce the problem of self-silencing; and
- structuring incentives to reward group success, thereby encouraging information disclosure.
The application of behavioural research to the issue of group performance has important implications for boards. In particular, it demonstrates that board decision-making may be improved by:
- educating board members on potential individual biases and mental short-cuts, as well as the nature of and reasons for group errors; and
- adopting certain practical safeguards and correctives to help facilitate effective decision-making. The chair will play an important role in this regard.