Eight to Late

Sensemaking and Analytics for Organizations

Posts Tagged ‘Collaboration

Conditions over causes: towards an emergent approach to building high-performance teams

with 7 comments


Much of the work that goes on in organisations is done by groups of people who work together in order to achieve shared objectives. Given this, it is no surprise that researchers have expended a great deal of effort in building theories about how teams work. However, as Richard Hackman noted in this paper,  more than 70 years of research (of ever-increasing sophistication) has not resulted in a true understanding of the factors that give rise to high-performing teams.  The main reason for this failure is that:

“…groups are social systems. They redefine objective reality, they create new realities (both for their members and in their system contexts), and they evolve their own purposes and strategies for pursuing those purposes. Groups are not mere assemblies of multiple cause–effect relationships; instead, they exhibit emergent and dynamic properties that are not well captured by standard causal models.”

Hackman had a particular interest in leadership as a causal factor in team performance.  One of the things he established is that leadership matters a whole lot less than is believed…or, more correctly, it matters for reasons that are not immediately obvious. As he noted:

“…60 per cent of the difference in how well a group eventually does is determined by the quality of the condition-setting pre-work the leader does. 30 per cent is determined by how the initial launch of the group goes. And only 10 per cent is determined by what the leader does after the group is already underway with its work. This view stands in stark contrast to popular images of group leadership—the conductor waving a baton throughout a musical performance or an athletic coach shouting instructions from the sidelines during a game.”

Although the numbers quoted above can be contested, the fact is that as far as team performance is concerned, conditions matter more than the quality of leadership. In this post, I draw on Hackman’s paper as well as my work (done in collaboration with Paul Culmsee) to argue that the real work of leaders is not to lead (in the conventional sense of the word) but to create the conditions in which teams can thrive.

The fundamental attribution error

Poor performance of teams is often attributed to a failure of leadership. A common example of this is when the coach of a sporting team is fired after a below par season. On the flip side, CxOs can earn big-buck dollar bonuses when their companies make or exceed their financial targets because they are seen as being directly responsible for the result.

Attributing the blame or credit for the failure or success of a team to a specific individual is called the leadership attribution error. Hackman suggested that this error is a manifestation of a human tendency to assign greater causal priority to factors that are more visible than those that are not: leaders tend to be in the limelight more than their teams and are therefore seen as being responsible for their teams’ successes and failures.

This leader-as-hero (or villain!)  perspective has fueled major research efforts aimed at pinning down those elusive leadership skills and qualities that can magically transform teams into super-performing ensembles.  This has been accompanied by a burgeoning industry of executive training programmes to impart these “scientifically proven” skills to masses of managers. These programmes, often clothed in the doublespeak of organisation culture, are but subtle methods of control that serve to establish directive approaches to leadership. Such methods rarely (if ever) result in high-performing organisations or teams.

An alternate approach to understanding team performance

The failure to find direct causal relationships between such factors and team performance led Hackman to propose a perspective that focuses on structural conditions instead. The basic idea in this alternate approach is to focus on the organisational and social conditions that enable the team to perform well.

This notion of  conditions over causes is relevant in other related areas too. Here are a couple of examples:

  1. Innovation: Most attempts to foster innovation focus on exhorting people to be creative and/or instituting innovation training programmes (causal approach). Such approaches usually result in  innovation of an incremental kind at best.  Instead, establishing a low pressure environment that enables people to think for themselves and follow-up on their ideas without fear of failure generally meets with more success (structural approach).
  2. Collaboration: Organisations generally recognise the importance of collaboration. Yet, they attempt to foster in the worst possible way: via the establishment of cross-functional teams without clear mandates or goals and/or forced team-building exercises that have the opposite effect to the one intended (causal approach).  The alternate approach is to simplify reporting lines, encourage open communication across departments  and generally make it easy for people from different specialisations to work together in informal groups (structural approach). A particularly vexing intra-departmental separation that I have come across recently is the artificial division of responsibilities between information systems development and delivery. Such a separation results in reduced collaboration and increased finger pointing.

That said, let’s take a look at Hackman’s advice on how to create an environment conducive to teamwork.  Hackman identified the following five conditions that tend to correlate well with improved team performance:

  • The group must be a real team– i.e. it must have clear boundaries (clarity as to who is a member and who isn’t), interdependence (the performance of every individual in the team must in some way depend on others in the team) and stability (membership of the team should be stable over time).
  • Compelling direction– the team must have a goal that is clear and worth pursuing. Moreover, and this is important, the team must be allowed to determine how the goal is to be achieved – the end should be prescribed, not the means.
  • The structure must enable teamwork– The team should be structured in a way that allows members to work together. This consists of a couple of factors: 1) The team must be of the right size – as small and diverse as possible (large, homogenous teams are found to be ineffective), and 2) There must be clear norms of conduct. Note that Hackman lists these two as separate points in his paper.
  • Supportive organizational context– the team must have the organisational resources that enable it to carry out its work. For example, access to the information needed for the team to carry out its work and access to technical and subject matter experts.  In addition, there should be a transparent reward system that provides recognition for good work.
  • Coaching– the team must have access to a mentor or coach who understands and has the confidence of the team. Apart from helping team members tide over difficult situations, a good coach should be able to help them navigate organizational politics and identify emerging threats and opportunities that may not be obvious to them.

To reiterate, these are structural rather than causal factors in that they do not enhance team performance directly. Instead, when present, they tend to encourage behaviours that enhance team performance and suppress those that don’t. 

Another interesting point is that some of these factors are more important than others. For example, Ruth Wageman found that team design (the constitution and structure of the team) is about four times more important than coaching in affecting the team’s ability to manage itself and forty times as powerful in affecting team performance (see this paper for details). Although the numbers should not be taken at face value, Wageman’s claim reiterates the main theme of this article: that structural factors matter more than causal ones.

The notion of a holding environment

One of the things I noticed when I first read Hackman’s approach is that it has some similarities to the one that Paul and I advocated in our book, The Heretic’s Guide to Best Practices.

The Heretic’s Guide is largely about collaborative approaches to managing (as opposed to solving!) complex problems in organisations. Our claim is that the most intractable problems in organisations are consequences of social rather than technical issues. For example, the problem of determining the “right” strategy for an organisation cannot be settled on objective grounds because the individuals involved will have diverse opinions on what the organisation’s focus should be.  The process of arriving at a consensual strategy is, therefore, more a matter of dealing with this diversity than reaching an objectively right outcome.  In other words, it is largely about achieving a common view of what the strategy should be and then building a shared commitment to executing it.

The key point is that there is no set process for achieving a shared understanding of a problem. Rather, one needs to have the right environment (structure!) in which contentious issues can be discussed openly without fear.  In our book we used the term holding environment to describe a safe space in which such open dialogue can take place.

The theory of communicative rationality formulated by the German philosopher, Juergen Habermas, outlines the norms that operate within a holding environment. It would be too long a detour to discuss Habermas’ work in any detail – see this paper or chapter 7 of our book to find out more. What is important to note is that an ideal holding environment has the following norms:

  1. Inclusion
  2. Autonomy
  3. Empathy
  4. Power neutrality
  5. Transparency

Problem is, some of these are easier to achieve than others. Inclusionautonomy and power neutrality can be encouraged by putting in place appropriate organisational structures and rules. Empathy and transparency, however, are typically up to the individual. Nevertheless, conditions that enable the former will also encourage (though not guarantee) the latter.

In our book we discuss how such a holding environment can be approximated in multi-organisational settings such as large projects.  It would take me too far afield to get into specifics of the approach here. The point I wish to make, however, is that the notion of a holding environment is in line with Hackman’s thoughts on the importance of environmental or structural factors.

In closing

Some will argue that this article merely sets up and tears down a straw man, and that modern managers are well  aware of the pitfalls of a directive approach to leading teams. Granted, much has been written about the importance of setting the right conditions (such as autonomy)…and it is possible that many managers are aware of it too. The point I would make is that this awareness, if it exists at all, has not been translated into action often enough.  As a result, the gap between the rhetoric and reality of leadership remains as wide as ever – managers talk the talk of leadership, but do not walk it.

Perhaps this is because many (most?) managers are reluctant let go the reins of control when they know they will be held responsible if things were to go belly-up.  The few who manage to overcome their fears know that it requires the ability to trust others, as well as the courage and integrity to absorb the blame  when things go wrong (as they inevitably will from time to time). These all too rare qualities are essential for the approach described here to truly take root and flourish.  In conclusion, I think it is fair to say that the  biggest challenges associated with building high-performance teams are ethical rather than technical ones.

Further Reading

Don’t miss Paul Culmsee’s entertaining and informative posts on the conditions over causes approach in enterprise IT and project management.

Written by K

January 29, 2015 at 9:03 pm

%d bloggers like this: