I’ve mentioned one of my theories before:

I have a theory that whenever a single specialisation becomes the dominate or controlling specialisation it turns the coordination into a technocracy.  I might be using technocracy in the wrong sense but I’m using it to mean that a single specialisation (i.e. a single branch of technical knowledge) becomes dominate.

It is in this sense that I believe organisations are largely technocratic if they are run by technical specialists called ‘managers’.  In another example, I think this is part of the process that has occured to create the current ‘crisis’ in the financial services industry:

In this case, rather than financial knowledge acting as just one input into decision making, and as just one service available to individuals, financial knowledge became the single dominate knowledge used in decision making.  This in itself was a problem because it interfered with true and proper determination of ‘value’.  But it would also have the effect or corrupting the body of knowledge itself (perhaps not in the textbooks, but ‘in use’).

Eventually, the knowledge of finance not only supersedes all other knowledge, but the finance knowledge actually disappears as it is replaced with what is convenient or beneficial to those in finance.  This process could be a simple as people entering financial jobs even though they have no passion or knowledge of such things because that is where the money is (literally in this case, but it doesn’t have to be that way).

There are two sides to this coin that should be understood.  The fact that control by specialists is the dominant coordinating mechanism in a society (or an organisation) is bad for the society itself. This is basic technocracy.  On the other side of the coin – this control by a particular specialisation is bad for the specialisation.

It’s not that the people who are part of the elite and elevated specialisation don’t have any power.  They do have power, and they can have certain privledges.  However, the actual specialisation itself is hurt.  The knowledge in that specialisation is corrupted.

Not only is the knowledge in the specialisation corrupted, but the group of specialists is itself politicised.  I don’t just mean it becomes a ‘hot topic’ but that it becomes focused on power relationships.  Now I’m not saying anything new if I’m just say that management is political.  Everybody knows this.  But there are deeper, but related impacts of this politicisation process.

In addition to the corruption of knowledge, the politicisation of a group effects the way the group is entered or exited, the amount of diversity tolerated in the group, and ultimately the ability to make non-incremental impovements in performance.

If you understand that open systems (i.e. low political / legal barriers to entry and exit) are good over time, and that diversity is good (again, over time), and that operational innovation is good (if you can even tell when it’s occuring!) then you can see that politicisation is bad for the group and the organisation it supports.

Example from software testing

To use an example that not simply about general management – let’s look at testing.  I’ve spent some time over the years setting up testing capabilities – either for individual projects or across organisations.  I’ve also watched others do the same.  One of the mistakes that is often made during this process is to not hold the testing team to the same standards as developers.

As background, a testing capability provides an independant verification of the software that has been developed, often also ensuring it works with other software developed by other groups.  The purpose of this process is to find issues with the software, but more fundamentally to manage the completeness of the project.

Fundamentally, test management isn’t so-much about finding defects as it is about continuously asking ‘Are we finished?’ and then ‘If we don’t know if we’re finished how can we find out?’ and then ‘Are we finished finding out?’ and then ‘Are we finished?’…

In order control this process there are a number of rules the testing organisation needs to place on developers.  These rules ensure the software doesn’t change in an uncontrolled manner while it’s being tested.  For example:

  • developers should check their own software first, then we’ll check it
  • once the testing team is checking your code, you can’t change it
  • the testing team will tell you if you have to change your code, and we’ll say it was a defect
  • once you fix the defect, you have to check your code again before we check it

These are good rules.  And they work together to provide an element of control and governenace over the software development process.  They provide a gatekeeping mechanism at the end of a project that, if used effectively, will reduce issues with live production systems.  Though the process itself is very expensive (but that’s a different issue).

The problem occurs when the testing team doesn’t hold itself to the same rules.  Examples of this include the case where a defect is raised in error because the test being perform was itself incorrect. This is still a defect – in this case a defect in the test itself – but the testing team doesn’t like to see it that way.

A more subtle version of this is that the defect is never raised but rather the test is changed without a defect being raised.  In this more subtle example the testing team is not holding themselves to the same stardards they hold the development team because they are changing something without a defect.

The testing team sees these decision as ‘saving time’.  But the governance issue is that the testing team are not being ‘tested’.  These means that the testing team have no objective criteria for performance management. They have been elevated above the rules.

By elevating the testing team such that they have special privledges – i.e. that they don’t have to follow the rules – the group becomes politicised.  People who want the special prevlidges enter the group, the knowledge of testing becomes corrupted, and operational innovation is restricted as the group uses its power to focus on changing others and not themselves.

All of these ‘problems’ are not problems if you are in the group.  However, the testing group will be different group then had it not be politicised, the performance of the group may not be as it would have been, and the overall performance of the organisation that the group is a part of may also suffer.

This is melodramatic in many ways.  Because these things will only occur over time, and only if all others things are equal.  In reality other groups and specialisations will compete for power and the dance continues…