Can the NHS Learn From Google About the Follies of Cutting Middle Management?

Much has been said about the role of middle managers in the NHS – some view them as a financial burden, whereas others see them as a key tool for unlocking the efficiency of the service. A new study on middle managers published in the Journal of Management Studies suggests that we should underestimate the role they play at our peril.

In 2002, Google launched Project Oxygen, which worked on the hypothesis that managers are at best a necessary evil, and at worst a useless layer of bureaucracy. The early work of Project Oxygen included a radical experiment – a move to a flat organisation without any managers.

The experiment was a disaster, lasting only a few months as the search giant found employees were left without direction and guidance on their most basic questions and needs. Nowadays, the company employs over 30,000 managers, although there have been recent talks of layoffs again.

As the NHS has struggled for efficiency in recent years, many have begun to ask similar questions about the usefulness of middle managers in the organisation. But are they really a good target for creating a more lean, efficient NHS, or would cutting their numbers result in ever-growing chaos?

We know that coordination matters. A study of over 250 companies found that teams are three times more likely to miss performance commitments because of coordination breakdowns with other teams than because of coordination issues within their own team. The main mechanisms for creating coordination described in academic literature are: (1) relying on the formal hierarchical structure, (2) having middle managers span the boundaries between teams to ‘bust silos,’ and (3) creating a shared understanding of common strategic goals.

So, which coordination mechanism should organisations choose to get interdependent teams to work together effectively?

The paper theorises that organisational structure can be a valuable coordination mechanism. In fact, in agreement with those that question the need for middle managers, it argues that when teams are part of the same division, this structural coordination is a sufficient means to foster collaboration between teams. Being part of the same division ensures both the sharing of local knowledge and perspectives, as well as shared understanding of local objectives. Here, middle manager boundary spanning and shared understanding of strategic goals do not add value.

However, there are many instances where teams do not share division membership and still very much need to collaborate. Then, the paper suggests, a shared understanding of strategic goals and boundary spanning by middle managers come into play because they can substitute for structural coordination.

In fact, in agreement with those that champion the need for middle managers, the paper argues that middle managers’ boundary spanning is a critical coordination mechanism for teams from different divisions. Here middle managers can ensure that both local and distal knowledge and perspectives are shared. Moreover, they can negotiate potentially conflicting interests and manage interactions across division boundaries in aid of collaboration between teams.

Furthermore, the shared understanding of strategic goals can also substitute for structural coordination because it offers common ground, facilitates communication and ensures that teams work together in line with the organisation’s strategy. Lastly, the paper argues that middle manager boundary spanning strengthens the positive influence of a shared understanding of strategic goals; the active engagement of middle managers in ‘busting the silos’ between teams can help translate an understanding of shared distal, strategic objectives into collaborative action.

To put their thinking to the test and gain a more thorough understanding of the topic, the authors of the paper collected data from a large service organisation from Western Europe. Using their organisational chart, they mapped out the formal hierarchical structure. In addition, they not only surveyed all middle managers about their boundary spanning activities, but also surveyed all their employees to determine the shared understanding of strategic goals, as well as all supervisors to determine how effectively different interdependent teams were working together.

The results confirmed their theorising. They show that –no– to ensure teams from the same division work together effectively, there is no added value from middle managers’ boundary spanning. However, the results also show that –yes– to ensure teams from different divisions work together effectively, middle managers’ boundary spanning is an important coordination mechanism that improves collaboration. Additionally, the results confirm that middle managers’ boundary spanning further strengthens the positive influence of shared understanding of strategic goals in aiding teams to work together effectively.

The implications of this research are relevant, first, to organisations and managers who want to ensure their teams work together effectively. With the knowledge that middle managers’ boundary spanning is most critical for teams from different divisions, organisations can ensure their boundary spanning training not only develops managers’ understanding of how to engage in boundary spanning, but also of when to do so. This is helpful to middle managers not only because boundary spanning is a taxing activity, but also because these managers have multiple strategic roles that require attention. To make such boundary spanning efforts even more effective, organisations would ideally also invest in building a shared understanding of strategic goals.

Second, this research is relevant to scholars and practitioners who question the need for middle managers, especially to those considering boundaryless organisational structures and eradicating middle manager ranks. The findings suggest that eradicating middle managers may be especially harmful to boundaryless organisations. Since boundaryless organisations offer very limited possibility for structural coordination, they greatly increase the need for both middle manager boundary spanning and shared strategic goals as substituting coordination mechanisms. Indeed, former boundaryless companies like Google and Github have (re-)introduced middle managers and Google now even identifies ‘collaboration with other teams’ as one of the critical skills for their managers.