Why people stay

Herman’s third condition is captivity, which forecloses exit and keeps the person within reach of the threat. Two things hold a person in place: whether they can leave, and whether leaving would change anything. The first translates her captivity into organisational terms. The second has no equivalent in her contexts.

Two kinds of captivity

Financial captivity is income dependency, accumulated pension entitlements, mortgage commitments set to current earnings, and the absence of comparable alternatives in the labour market. These are Siegrist’s (1996) dependency conditions, the reason workers tolerate sustained high effort and low reward.

Investment captivity is what a person has put into one place and cannot take out of it: years spent climbing a particular hierarchy, expertise in navigating this culture, a professional identity built around the position, a reputation made inside this community, a network built from decades of working relationships. None of it transfers. A better-paid job answers the mortgage; nothing answers a reputation that exists only where it was made. The longer a person stays, the more of who they are is held in the position, and leaving costs that as well as the income.

Neither should be read as a trap someone walked into carelessly. Robert names what held him, and none of it is an economic calculation. He loved the work and the people he worked with, and loathed the culture and the people he worked for, and it stayed bearable because he was part of a small crew who were tight. Looking back, he says he was trapped by his own enthusiasm and commitment and by the bonds he had built. He had also decided he would not be driven out by the people running the place: he was going to outlast them.

Hobfoll’s Conservation of Resources theory (1989) provides the mechanism by which investment captivity deepens over time. His loss spirals, in which initial resource depletion leaves individuals more vulnerable to further loss, predict that those who have already invested heavily are least able to leave, because leaving would be another loss on top of what the organisation has already cost them. Captivity is not a stable condition. It intensifies.

Robert describes the risk of leaving as feeling larger every year he stayed. A job search takes confidence and energy, and those were the two things the situation was draining. He sets the commitments beside it — mortgage, bills, children, parents — which grow with age and are heaviest at the point the work is hardest to carry.

Whether leaving would change anything

Most organisations are hierarchies running HDM. A hierarchy could work the other way, with the reasoning behind a binding choice given and answered before the work starts. Few do.

Some employers are far better to work in, but what differs is the practice rather than the architecture. Nor is a flatter organisation an escape: where directive authority is distributed across roles instead of concentrated by position, the same three components are present in a different arrangement (Verweij, 2026c).

So changing employer may do no more than relocate the person, leaving the conditions unchanged. How much of working life this covers has never been counted, because nothing is required, so nothing happens that anyone would record. What can be said is what the person finds by doing it: the next job usually resembles the last in the respect that produces the harm, however different it is in every other respect.

Robert made the move twice. He left the telecommunications company for a smaller one in the same industry, was successful there, and was pushed out once the person who had hired him left. He went on to a broking firm in the City of London, ran two programmes successfully, and left after eighteen months. He was pushed out of one and walked out of another. He was already damaged when he made the first move, and says he was not willing to admit it. He took the third knowing it would not last, because the money was good and it was a chance to repair the financial damage — not, in his words, the real damage. He calls it three strikes and out.

What makes this hard to act on in advance is what has to be compared: whether an organisation obliges anyone to explain a binding choice and answer questions about it before the work starts. That appears in no policy, no interview and no incident record. What a person can observe is the practice — how the place talks about itself, what the manager is like, whether people seem to be listened to — and that is precisely the part that varies. A person can therefore move in good faith, on accurate observation, and still land somewhere the requirement is missing. The only reliable way to find out is to go.

This is where the workplace case parts from Herman’s, and it parts in both directions at once. Her captivity contexts have an outside. The home, the camp and the cell are bounded, and the world beyond them runs on different conditions; escape is difficult, and it is escape to somewhere. Here there is no captor, no locked door, and no outside either. What the person faces is less acute and more extensive. Leaving remains possible and is often worth doing. What it does not do is end the exposure.

What follows from that is duration. In her contexts escape ends the exposure, so it has an end a person can reach. Here it does not, and what bounds it instead is the length of a working life. Leymann’s own account of what produces the injury is the frequency and the duration, not the severity of any one instance (Leymann, 1990). DFS reads the same variable as operative here, without an actor: lower intensity with no end is not a lesser case.

Defeat and entrapment

Gilbert and Allan’s defeat-entrapment model (1998) provides the closest clinical vocabulary for being unable to prevail and unable to leave. Defeat is a sense of failed social struggle, loss of rank, inability to prevail, and it maps to having expertise structurally overridden by unsubstantiated authority. Entrapment, the desire to escape from an unbearable situation combined with the perception that escape is blocked, maps to investment captivity. Their concept of arrested flight, the human equivalent of an animal that wants to flee but cannot, names the same shape DFS arrives at from its own starting point. The animal in the trap is not irrational. It has correctly assessed that escape is impossible. It suffers not from cognitive distortion but from an accurate perception of its structural situation.

Taylor, Gooding, Wood, and Tarrier’s systematic review (2011) of fifty-one studies found strong convergent evidence linking defeat and entrapment to depression and PTSD. Siddaway, Taylor, Wood, and Schulz’s meta-analysis (2015) of forty studies found these constructs predicted symptoms across four distinct psychiatric disorders, suggesting they may be mechanisms whose effects operate across diagnostic categories.

What DFS adds is the institutional specification. Defeat in Gilbert’s framework arises from interpersonal competition; DFS reads defeat as arising from an architecture that overrides expertise regardless of any interpersonal dynamic. Entrapment there is a matter of what the person perceives: that escape is blocked. DFS specifies the structural conditions that produce the blocking. And DFS brings the individual attribution of failure, which is what completes Herman’s conditions in the organisational case.

These mechanisms are not individual psychological failures. They are what the conditions predictably produce. What leaving costs comes from the employment relationship rather than from the architecture, and it deepens with tenure. What the architecture adds is that the exit those costs are weighed against leads into the same absence. Together they make escape “difficult or impossible” in precisely the language the ICD-11 requires, and no one has to be physically prevented from leaving for that to be so.