The Middle Territory

The Middle Territory
Why the best available explanation for your retention problem still leaves you with nothing to do
You have a pattern, and you can describe it precisely.
Capable people leave after about two years. Not the weak ones. The exit interviews are thoughtful, and the reasons are varied enough that no single cause presents itself.
Concerns that used to reach you now reach you late, or in a softened form, or not at all. The same complaints recur in the same department under three successive managers.
Meetings are agreeable, and the real conversation happens afterwards, in twos.
None of this shows up as a failure. The numbers are adequate. Nobody is behaving badly in a way you could name.
So you go looking for an explanation, and the field offers you two.
The first explanation, which is still there, but mostly gone
It used to be about the people. They weren’t resilient enough, weren’t a cultural fit, didn’t manage upward well.
That account has lost most of its credibility over the last twenty years, and rightfully so.
It fails the simplest test: when the same seat produces the same outcome under four different occupants, the explanation cannot be the occupants.
The second explanation, which is where most thinking now rests
The cause is the environment.
This arrives in three traditions, which disagree about almost everything and end in the same place.
The physiological.
Threat response, nervous system regulation, psychological safety as a felt state. Rock’s SCARF model – status, certainty, autonomy, relatedness, fairness – is the default vocabulary, with the advantage of sounding like science. Change triggers threat; threatened people cannot think well; leaders should manage the five dimensions to lower the response.
The identity.
The question is not what needs to change but who we need to become. Transformation as collective identity work: what parts of our current self are worth carrying forward, what story the organisation tells about itself, how people belong.
The interior.
Leadership as the capacity to shift the inner place from which one operates. The success of an intervention depends less on what is done than on the interior condition of the person doing it.
Different disciplines, different authorities, and they would argue vigorously in a room together.
And this is genuine progress.
Moving the cause outside the individual matters: a person who struggled in a particular organisation no longer carries it as evidence about themselves. The observations are usually accurate too. People in difficult organisations do become vigilant, guarded and slower to speak.
So the second explanation is not wrong. It is incomplete in one specific way, and the incompleteness is easy to miss.
A test that takes a minute
Here is how to tell whether an explanation will change anything.
Name what would have to change. Name who would have to change it. Name what it would cost them.
Three answers and you can act. A vague first answer, “everyone” or “leadership” for the second, and silence on the third, and you have a description rather than a cause.
Run it.
The culture is toxic.
Set aside that cultures can not be toxic – toxicity is a property of behaviour. What the architecture does is enable it, reward it, or fail to punish it. But let’s take the claim anyway. Who made it so? Not the visibly bad actor: remove them and another grows in the same seat.
Which is the clue. If the behaviour survives the person, something else produces it. And usually it is that the organisation measures results and does not measure conduct. Hit your numbers, and it notices. Treat people badly, and it has no instrument that would.
That is not weather. Somebody built a performance system that counts one and not the other, and somebody could build a different one.
People are in threat response.
Triggered by what? Uncertainty, usually. And who decides how much uncertainty people live with, and for how long? There is an answer. The framework does not require it.
The inner condition determines the outcome.
Whose? Everyone’s. Which distributes the work across every individual and leaves nobody responsible for beginning it.
Each of the three has the shape of a cause and the properties of weather.
The same two years, described twice
Take one of the people who left.
The middle-territory account.
The environment was chronically unsafe. Their temperament made them more exposed than colleagues. They eventually burned out.
Possibly all true.
The structural account.
They were accountable for a delivery date set by a commitment they were not consulted on. They could not change the scope, resourcing, or date. They raised the gap twice, in writing. The first time it was noted. The second time, the conversation turned to whether they were managing their team’s capacity well.
Everyone watching learned what raising it produces.
Both describe the same two years. Only the second one tells you that a commitment needs a capacity check before a date is given, that whoever makes the commitment should carry the delivery risk, and that the second conversation was the moment the organisation taught forty people something it did not intend to teach.
The first account is about what happened to a person. The second is about an arrangement that is still in place and will do it again.
Why the better explanation hands the problem back
The middle territory is worse than incomplete, for one reason. Its remedies are addressed to the people experiencing the problem.
Regulate your nervous system.
Examine your inner condition.
Build your resilience.
Do the identity work.
Even the kindest version: the organisation should create psychological safety, resolves in practice into leaders behaving differently through sustained personal effort.
So an explanation that correctly moved the cause off the individual gives the remedy back to them. The person is no longer blamed for the problem and remains responsible for solving it.
Which is why the programs work and then stop working. Nothing in them changed what the next difficult conversation costs the person having it.
What to do instead
Three things, none of them a program.
Find the gap between accountability and authority.
Take the three people who left and ask, for each, what they were accountable for and what they could actually decide without asking. The gap is the finding. In most organisations nobody has ever written it down, because accountability is documented – objectives, reviews, pay – and authority lives in custom and precedent.
Trace what happened to the last people who raised something.
Not whether a policy exists. What happened afterwards. Did they stay in post, move sideways, go quiet? What did their next review say? That information sits in systems you already have, and almost nobody assembles it. A leadership team that saw it once would not need telling what its reporting rate means.
And ask what the arrangement teaches, rather than what it intends.
Every structure instructs. The approval that sits two levels above the knowledge. The commitment made without a capacity check. The conversation that turned to the person raising it. None of these was designed to teach anything, and all of them did.
Each has a named owner and an attached cost, which distinguishes them from the remedies in the second explanation.
One thing to be clear about. An architecture is not arbitrary: it is the structural form a culture takes, and culture is the organisation’s working answer to what pays off here. Which is why a redesign holds where the two already agree, and reverts where someone has to keep insisting on it.
Why the comfortable explanation wins
Why is it so much more available than the costly one?
The character explanation implicates the person who left. The structural explanation implicates whoever designed the arrangement: usually in the room, often sponsoring the work, sometimes paying for the advice.
The middle explanation implicates a condition. Conditions have no author, no address and no budget line.
So you can run a program on nervous system regulation, on collective identity, on self-observation under pressure. Each produces a workshop, a vocabulary, a set of practices and a measurable improvement in how people report feeling.
You cannot run a program on decision rights, because it is not a program. It is a redesign; you have to do it, and it takes something away from somebody.
The people doing this work are generally sincere and frequently excellent at it. That is the point rather than an exception to it: an explanation that costs nobody anything outcompetes one that does, regardless of which is more accurate.
A cause nobody authored is a cause nobody can remove. An environment is not a thing: it is the accumulated effect of arrangements somebody made, somebody could change, and that would cost somebody something to change.
Naming those is harder, less comfortable and much less saleable than describing their effects. It is also the only version that tells you what to do on Monday.