Changes in organizational structure are an important lever for CEOs of large companies who are thinking about how management can make better decisions.
Its potential is still not fully exhausted.
But decision-making processes can be organized in such a way that we reduce wrong decisions.
It is important to distinguish between two types of mistakes: the so-called error of commission leads to bad suggestions being accepted, while the so-called error of omission prevents the implementation of good ideas.
Banks, for example, are organized across multiple levels of management, so lower-level decisions must be approved by a higher level.
This reduces the risk of a commission error, but at the same time increases the risk of an omission error.
So how should companies organize themselves to create an adequate balance between both types of error and thus between caution and risk?
My research team analyzed how bank managers can assess loan applications and infer how a loan's risk profile correlates with the likelihood of accepting that loan.
This assignment is called the screening function.
Based on this research, we developed a process to design organizational structures in such a way that decisions strike the desired balance between risk and prudence.
Majority voting and its limitations
Exactly the opposite is the case, for example, with the financing of innovative business models.
The primary goal here is to implement good suggestions, i.e. to reduce the error rate when good projects are rejected.
In this case, the decision levels should be reduced.
However, this is accompanied by an increasing number of risky decisions.
Removing too many layers is bad for business, adding too few is bad for business as well.
So how do you find the right balance?
Our research can provide the right answer for every practical application.
It is possible to reduce both types of errors at the same time, as illustrated by the following example: Delegate a decision to two managers.
If both agree, respect their decision - otherwise let a third decide.
My colleague Michael Christensen used this technique to make better software development decisions.
This resulted in faster procedures and fewer programming errors.
The decision rule described mimics majority decisions in a group of three people.
Majority voting, while efficient, has significant limitations.
As a result, they cannot strike the necessary balance between omission and commission errors.
When prudence prevails, walking errors need to be reduced by requiring more yes votes than a simple majority.
However, decisions often take place in several consecutive steps.
Influence of Hierarchy
How can such a decision be influenced?
The logic is to delegate a decision to two managers in parallel and let a third decide if the first two do not reach a unanimous conclusion.
In our research we have been able to work out general principles and based on this we have developed a procedure that can determine the optimal decision-making structure for each specific case.
The behavior of the employees is influenced by how hierarchically a company is shaped.
We have found that people adapt their behavior in centralized structures, i.e. when corporate management makes central decisions about corporate policy, processes and directions of action and also controls them very closely.
Things are different in decentralized structures, in which the individual business units act and make decisions independently.
This is because centralized structures encourage a sense of interdependence.
Colleagues see how their decisions affect others.
In a hierarchy, it is advantageous to leave difficult decisions to the next higher level.
A decentralized structure does not offer this possibility, so it reduces dependency and awareness of other colleagues.
This insight points to new ways in which organizational design can contribute to better decisions.
Thorbjørn Knudsen is Professor of Strategic Organization at the Frankfurt School of Finance & Management.