Coercive control and high-demand groups

How groups actually take hold of people — gradually, through ordinary social pressure, and almost never in the way films depict.

What it is

Coercive control is a pattern rather than an event. It is the cumulative narrowing of a person's choices, relationships, information and sense of their own judgement, achieved through ordinary social means applied consistently over a long time.

The word researchers increasingly prefer for the groups that do this is high-demand group, and the reason is practical. "Cult" is a term that ends analysis rather than beginning it: it implies a category of obviously deranged organisations that obviously sensible people would obviously avoid, which describes almost nothing that actually happens. High-demand describes an axis instead of a box, and axes can be measured.

That matters because demand by itself is not the problem. A Benedictine monastery makes enormous demands. So does medical training, military service, a serious meditation retreat, and any life organised around something larger than personal convenience. Demanding is not the same as harmful.

What distinguishes a harmful group is not how much it asks but what it does to a person's capacity to reconsider. Broadly:

Can you get accurate information, including criticism, without penalty?
Can you maintain relationships the group does not control?
Can you leave — practically, financially, socially?
When you raise a concern, is the concern examined, or are you?

Where it came from

The modern study begins with Robert Jay Lifton's interviews with Western civilians and Chinese intellectuals who had undergone "thought reform" in China in the early 1950s, published as Thought Reform and the Psychology of Totalism (1961). Margaret Singer added a large body of clinical case material through the 1970s and 80s.

Public attention arrived catastrophically with Jonestown in 1978, and the anti-cult movement that followed produced both genuine expertise and a good deal of panic — including the practice of forcible "deprogramming", which involved abduction and confinement, and which is now widely regarded as having been both unethical and counterproductive.

The most useful recent development came from a different field entirely. Evan Stark's Coercive Control (2007) reframed domestic abuse as a pattern of domination — isolation, surveillance, micro-regulation of daily life, control of money — rather than as a series of violent incidents. That shift has been legislatively influential: coercive control is now a criminal offence in England and Wales, in Scotland, and in several Australian jurisdictions including New South Wales and Queensland.

Stark's framework transfers to groups almost unchanged, and it has the advantage of describing mechanisms that are observable from outside rather than states of mind that are not.

How it is used

In assessment, the questions above are the working tools — the practical version is in evaluating a teacher or group.

In family support, the central finding is counter-intuitive and worth stating: confrontation does not work. Attacking the group attacks the person's judgement, which is precisely what the group has taught them to distrust, and it usually confirms the group's account of the outside world. Staying in contact, remaining someone they can call, and not making reconciliation conditional on their leaving is the approach with the better record.

What we can and cannot say

We can say the gradual mechanism is well documented. Nobody joins a group intending to hand over their savings and stop speaking to their family. It proceeds one reasonable-looking step at a time, and each step is small relative to the one before — which is what makes the written record recommended in the evaluation article useful, since it defeats the gradual recalibration of what counts as normal.

We can say recruitment is not random. It concentrates at transition points: bereavement, divorce, illness, migration, job loss, leaving home. Not because people at those points are foolish, but because they have fewer stabilising relationships and a live need for meaning and belonging, and because groups that recruit successfully have learned to notice.

We can say the strongest predictor is not credulity. Research on who joins consistently finds recruits to be roughly average or above in education and intelligence. Sincerity and idealism are more characteristic than gullibility — which is uncomfortable, because it removes the reassurance that this is something that happens to other kinds of people.

We cannot say the underlying models are uncontested. Whether groups exert something qualitatively different from ordinary social influence is the substance of a long disagreement between clinicians and sociologists of religion, treated separately in the article on the brainwashing debate. What both sides accept is that the harms — financial, sexual, medical, familial — are real and documented, whatever the mechanism turns out to be.

What it is

Where it came from

How it is used

What we can and cannot say

Further reading