I recently came across a statement made by Warren Buffett in a 1965 letter to his Partnership where he mentioned group decision making: “My perhaps jaundiced view is that it is close to impossible for outstanding investment management to come from a group of any size with all parties really participating in decisions.” This made me reflect on the fact that the overwhelming majority of the literature and research around behaviour and decision making produced in recent decades has been about how individuals make choices, but in virtually every walk of life – politics, business, family and investing – decisions are frequently made by groups of people. Despite this little time appears to be spent seeking to understand how groups function. This is a severe oversight. If we thought individual choice was confusing and problematic, just wait until we start putting people together.

It is important to define what a group decision actually is. For me there are two key features:

1) People ostensibly working together to reach a decision.

2) More than one person has the ability to materially influence (implicitly or explicitly) the decision made.

A group decision does not have to be a situation where there are 11 people sitting on a committee and each holds a vote, in fact many collective decisions look like individual choices in that a single person makes the ultimate call, but are actually the outcome of group interaction and influence.

Why are most decisions made – in some form – by a group of people? There are two reasons. First is a desire to combine experience and expertise. In theory, better decisions should be made if there is an ability to access higher quality information and inputs. Second because there is a need to represent multiple stakeholders. Where a decision impacts different groups, there is often a requirement to provide an appropriate voice to those parties.  The makeup of most boards and committees should at least attempt to meet these two criteria.

While these both seem laudable aims, they can create profound decision-making problems if not handled considerately (and might still do even if they are). The investment management industry is overrun with implicit and explicit group decision making, but I have barely ever heard anyone talk coherently about how groups have been brought together. Yes, in recent years lip service has frequently been paid to the notion of cognitive diversity, but I have severe doubts that many people using the term have a clear idea of what it means and its implications.

Imagine involving a group of people in a decision all of whom have different personalities, incentives and beliefs, and expecting that to result in a coherent outcome. It seems entirely non-sensical, but this is exactly what occurs in most group decision making scenarios.

It is not that the construction of decision-making groups is random – there is generally some credible reason as to why people become involved in a decision – it is just that there is an inescapable complacency about how or why the interactions of the group will impact the type of choices being made. This almost inevitably leads to unintended consequences and undesirable outcomes.

There are several critical issues to consider when trying to avoid the major challenges of group decisions:

Goals and incentives:  The essential condition for effective group decision making is shared goals and aligned incentives. If the individuals with influence over a choice are not attempting to achieve the same objective then any decision it makes is likely to be greatly compromised. Organisations where group decisions take place are typically incredibly complacent about this, assuming that ‘of course’ all people involved have the same objectives. This is rarely the case, usually because the personal incentives of individuals are wildly misaligned.

The critical question to ask is – what is the primary incentive / aim of an individual involved in this decision? To take a heavily stylised example: Imagine there is a portfolio manager and a risk manager in a group involved in making an investment recommendation. The portfolio manager’s primary aim it to maximise return, the risk manager’s is to avoid disaster. These are both rational positions to take for them as individuals but it is very unlikely to lead to rational decisions at a group level – they have skin in different games. Now imagine it is not two people involved in an investment decision, but 17 all with somewhat contrasting motivations.

This is not to say that individuals with different specialisms and areas of focus cannot be involved in an effective group decision, it is just that for it to work they have to have their incentives aligned. Most groups involve a collection of people optimising for different things, resulting in sub-optimal results.

Philosophy and values: Consistent goals are integral to effective group decision making, but almost as important is shared philosophy and values. This does not mean that individuals within a collective have to think in the same way or attempt to tackle each problem in an identical fashion, but they must hold a set of common beliefs about the best way to reach their objectives. Group members can challenge each other without challenging their identity.

Decision making groups with a shared philosophy can discuss and debate the nuances of how best to apply that set of beliefs to the problem that they are trying to solve. Groups with philosophical differences will forever be trapped in unresolvable debates, often making little progress.

Accountability sinks: One of the primary risks of group decision making is the creation of what economist Dan Davies calls “accountability sinks”. This is where organisations remove individual accountability for decisions and instead place it into amorphous groups, policies and procedures. This creates a situation where nobody is to blame for anything – the system has responsibility. This can be quite attractive for people because being held to account is often unappealing, but it can lead to bad decisions, the complete removal of agency and chronic inflexibility.  The atrocious Horizon IT scandal that recently engulfed the British Post Office, is almost inevitably an extreme example of horrendous group / system-led decision making.  

Decisions can be categorised on a spectrum from those that feature individual accountability to those where there is no obvious accountability:

– Individual (Single accountability)
– Group (Shared accountability)
– Multiple Groups (Vague accountability)
– System (No accountability)

This is not to say that individual, sole responsibility decision making is always optimal, but we must be aware of the consequences of the shifting accountability structure that occurs as the number of people involved in a decision increases. Individual behaviour alters dramatically if we have 100% responsibility, compared to no ultimate responsibility.

Implicit group decisions: Another major accountability problem arises when decisions appear to have individual accountability but are actually heavily impacted or constrained by others. Similar to accountability sinks, this can be appealing for most group members because they can exert material influence without any consequence for their actions. It is not, however, appealing for the individual who is deemed to be the decision maker.  Davies, again, has a simple test for whether someone deserves to be accountable:  

“The fundamental law of accountability, the extent to which you are able to change a decision is precisely the extent to which you can be held accountability for it, and vice-versa”

It is quite common for individuals (or other groups) to hold the ability to materially alter a decision while seemingly being a great distance from any accountability for it. This typically occurs when people are able to apply constraints to a decision; here the accountable decision maker appears to have full responsibility for the ultimate course of action taken, but their set of options has already been greatly reduced by others.

This type of scenario often precedes the formation of accountability sinks – as when the accountable individual realises that they hold little agency they baulk at this arrangement, and prefer something more anonymous.

When assessing a group decision structure, it is vital to look at everyone with involvement and ask which individuals (or groups) have the power to significantly shape the ultimate choices being made.   

Accountability shields: A less common structure for a group decision is an accountability shield. In this situation, one dominant and influential party has an overwhelming power to dictate decision making. They are, however, reticent to bear responsibility for potentially negative outcomes and therefore create what appears like a group / committee / board structure to insulate them from full accountability. This creates an attractive asymmetry for the individual who holds power. The groups involved are functionally pointless, bar providing some protection in the event of disappointing results.

Groupthink: Although most of us never seriously consider the implications of group decision making, there is one topic that is always mentioned if the subject is raised – groupthink. This is a situation where a collective makes a poor decision because there is not enough challenge or critique from within the group. While this can be a valid concern, the issue is nuanced. Certain elements which may be classed as groupthink – such as shared ethos, values and principles – are an integral part of a cohesive group structure. It is, however, crucial that a broadly shared philosophy is allied with psychological safety, freedom of expression and a diverse set of skills. For harmonious groups to be effective they must be able to allow new ideas to permeate and accept that they can be wrong.

Groupthink is undoubtedly an issue worthy of serious consideration, but attempting to mitigate it by putting together a collection of ideologically opposed individuals with conflicting incentives is a bad idea.

The behavioural melting pot: Individual decision making is chaotic and idiosyncratic. Group decision making takes all of our personal foibles, puts them into a pot and stirs. It is impossible to understand how the choices made by groups will react to this confluence. As much as we might try to refine and reshape a group to improve its decision-making capabilities, a significant element will remain the wholly unpredictable noise that emerges from a convergence of personal behavioural biases.

Power dynamics: Power is perhaps only second to incentives in understanding what drives human behaviour. If we have a clear understanding of individual incentives and the power dynamic at play in any given situation, we can get a long way to understanding the choices that are likely to be made. Talking and thinking about how power impacts the workings of a group is unfortunately something few people are comfortable discussing, probably because we are reticent to upset the people that wield the power. In large organisations power doesn’t have to be held by individuals, it can be held by the system – a structure of policies, procedures and groups can wield an overwhelming but largely silent power.

Group size: When attention is given to group decision making, a common question pertains to the appropriate size of the group – what is the most effective number of participants? There is no right or simple answer here – it depends on the type of decision being made. One overlooked consideration is seeking to understand exactly how large a decision-making group is, which in many cases we probably don’t really know. Returning to my original criteria – we need to identify the people who both have input and can influence a decision. That is our real group size, which could be very different to the number of people sitting on a formal committee or board.

A large group size does have benefits as it should bring with it a greater diversity of skill and provide a voice to multiple stakeholders (where relevant). This notion only holds, however, if some deliberate thought is given to its construction. The main issue with sizeable groups is that most of the problems discussed in this piece become more likely and more pernicious the larger they get.

Group size and composition have a huge bearing on the type and quality of decisions made, but we pay barely any meaningful attention to them.

What defines a group that makes high quality decisions?

The sheer complexity of group decision making may make it seem like a problem that is not even worth confronting, but given how influential these dynamics are to most of the choices we make it is not something that we can continue to ignore. As always, the best way to deal with something that appears unfathomably complicated is to apply some simplicity to it. To my mind, there are three characteristics that a group needs to possess to have a chance of making smart choices:

1) Aligned goals and incentives:  A high-quality decision-making group must have shared goals – they have to be trying to achieve the same thing.  Companies tend to get this spectacularly wrong by creating vague purpose statements that will have no meaningful impact on anyone’s behaviour. Shared group goals are derived from shared individual incentives. Do good and bad outcomes look the same for each member of the group?

2) Shared philosophy and values: A consistent set of values held by the group members is essential for high quality decisions. This doesn’t mean that they need to agree on every aspect of a task or problem, but rather there are no major philosophical gaps between individuals. If there are then the group is likely to fall at the first hurdle. Shared principles should be the foundation of any effective group.  

3) Complementary skills: The key reason for having a group decide is that it grants input and influence to individuals with distinct and complementary skillsets. The more complex the task, the more useful this can be. The key challenge here is that people fall into what football fans might think of as the ‘Paris Saint-Germain trap’ (or maybe the English national team) where we believe that an effective group is simply a collection of the most ‘talented’ individuals. This approach almost always ends badly as it disregards the vital notion that a strong team combines different characters, skills and expertise to meet a particular goal. Haphazardly combining some good individual decision makers will not create an effective decision-making group.

—

I often find myself frustrated at the lack of thought given to individual decision making, but this pales into comparison with the neglect of group decisions. The majority of the choices that we make (particularly in a professional context) are made by some form of collective, but we spend incredibly little time considering how groups function and the type of choices they are liable to make. It is undoubtedly a messy and intricate topic, but most of us don’t even try to get the basics right. Instead groups risk being blighted by murky accountability, misaligned incentives, opacity and compromise – not the ideal setup for sound decision making.

–

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).