In most organisations a myriad of decisions are made everyday. In established organisations most of these are delegated to management to be made within a framework of pre-established policy, procedure and decision criteria.
Decisions that do get to the board level are likely to have strategic significance and be ‘material’ within the scale of the business. Decisions are also elevated to the board because they address matters that are complex,uncertain and carry significant risk. The mere fact that a decision appears on the board’s agenda should signify that a lot is at stake.
Decision-making at the board level typically follows a common pattern. To illustrate that let’s say there is a long-established operation that no longer appears to be the best use of the resources involved. It is of enough significance, however, that it requires a board decision to close it down. Management’s concern may have been expressed to the board in a preliminary way, but the board has not given it any concentrated attention.
In the meantime, as it should, the management team has gone through an iterative evaluation process assessing a range of considerations. For the sake of this example, the team concluded that the operation should be closed down. The next step would be to put a paper with this recommendation to the board. In cases however, the management team’s conclusion will be suspect, having been badly affected by a range of cognitive biases.
These are described by Daniel Kahneman (of Thinking Fast and Slow fame) and two collaborators in a recent article. They point to problems like:
Excessive coherence. Getting this kind of issue to the board typically involves the progressive development by managementof an ‘impression’ or ‘mental model’ of what the board’s final decision should be. Unfortunately, this mental model is usually simpler and more coherent than the reality of the situation. Team members also bring a range of prior assumptions to bear on the information available. These assumptions influence the questions used to interrogate that information. As a consequence, it is increasingly likely that those prior assumptions will be confirmed.
A ‘quick and sticky’ quality. First impressions are likely to have a disproportionate impact on a final assessment. Mental models form rapidly - often on the basis of limited evidence - and are altered only slowly. In this example, management had doubts about this operation before the evaluation even started.
Biased weighting. These mental models often fail to give each pertinent factor the weight it deserves. Consequently, important pieces of information are discounted and factors that are less relevant are given more weight than they deserve.
Excessive coherence would have encouraged the management team to confirm an impression that was much less nuanced and less ambiguous than the reality. This bias would have had an even greater impact if, from the outset, the managers had worked together on developing a recommendation for the board rather than looking at the situation independently and forming their own views first.
As the team’s preference emerged the perceived benefits would likely have been exaggerated and its possible consequences underestimated. And, once an initial impression had taken hold – which typically happens quite fast- this would likely have been reinforced by managers asking leading questions that supported their preliminary view. It is likely team members would also have interpreted ambiguous facts in the light of their pre-existing attitudes (‘confirmation' bias).
Another bias (‘availability’ bias) means they would probably have also given extra weight to recent more obvious and readily available information. Associated is an over-emphasised, short-term and perhaps self-serving considerations of the kind often associated with financial incentives.
All of that is before we bring the board into the picture. The rapidly compounding deficiencies of the evaluative process previously described would be replicated, perhaps even more quickly, at the board level. It also has a head start by being ‘anchored’ by management’s recommendation.
It requires a considerable amount of compelling new evidence for people to back away from an initial judgement that is in error. The further down the track it is, the more difficult it is to retreat from. As this example shows, boards need an approach in which they can have more confidence.
Kahneman and co. contend that a more structured approach is less susceptible to this kind of bias and ‘noise’. What they call their mediating assessments protocol (MAP), this has the following features:
· the key attributes that are critical to the evaluation of alternative actions are defined in advance
· participants make evidence rather than opinion-based assessments of each attribute, evaluating and scoring one. attribute at a time
· they make these ‘mediating assessments’ independently so that they are not influenced by others who will. contribute to the final decision
· only once a complete profile of these individually developed assessments is available are the decision-makers. allowed to come together to discuss the composite picture and work towards a decision
This approach tempers the effects of bias by impeding the premature formation of a mental model of the ‘right decision’. It also increases the transparency of the process as all the assessments are presented to the decision-makers at the same time.
This approach has some important implications for the roles of the management team and the board in respect of decisions that must ultimately be made by the board (i.e. not ‘rubberstamped’). It suggests, for example, that management should not come to the board with a recommendation. Instead, what the management team should produce for the board is their initial assessment of each of the key attributes. Ideally, these attributes would have been agreed with the board before managers began their assessment, giving the board a better chance to shape the decision it must make. At the board meeting the board would debate with management - attribute by attribute – to reach finality on those attribute-specific assessments. Only then would the board look at the composite picture of the information now in front of it and come to a decision.
As Kahneman, Lovallo and Sibony state, important decisions are about the ‘distillation of complexity into a way forward’. Boards want and need to make good decisions. Therefore, when decisions are elevated to the board, they should be put through a thorough process, be as free of bias as possible, and be given the depth of attention they deserve.
Daniel Kahneman, Dan Lovallo, and Olivier Sibony. 'A Structured Approach to Strategic Decisions'. MIT Sloan Management Review, Spring 2019, pp.67-73