Constructive Desistance — Joanne Lockwood

By Joanne Lockwood (she/her) FIEDP FRSA FPSA – Written from lived experience, not from the sidelines.

Part two of two. In part one — Broad Door, High Bar — I read the bill on its own terms. This piece reads the world it lands in.

I write as an inclusion specialist working from lived experience, not as a lawyer. This is commentary on a draft law and the world it lands in, not legal advice. The bill applies to England and Wales; take your own advice for your own context

The draft bill as published 25th June 2026: Conversion practices draft bill - GOV.UK

When the draft Conversion Practices Bill was published, it did something the long, bitter argument that preceded it had rarely managed: it drew a clear picture of harm. In that picture there is a person who intends to change another person — to talk them, pressure them, coerce them out of who they are. There is intent. There is a victim. There is conduct, aimed and deliberate, and it is abusive, and it causes serious harm. It is a true picture. The harm it shows is real, and naming it in statute is a genuine step forward.

But every picture has a frame, and every frame has an edge. What I want to talk about is what sits just outside this one — because it is, I think, where most of the damage is actually being done, and it is precisely the part no conversion law will ever reach.

What the law was built to see

Read the bill’s structure and you can see the harm it was designed to catch. Three things must all be true before anyone has done anything criminal. Someone must intend to change or suppress another person’s sexual orientation or transgender identity. The conduct must be abusive — coercive, controlling, threatening, exploiting economic or psychological pressure. And it must cause serious harm: serious damage to health, or serious distress that substantially disrupts a person’s daily life.

That is a high bar, and deliberately so. It is what allows the law to exist at all without criminalising prayer, persuasion or honest conversation. I defended that design in my last piece and I stand by it. But notice the shape of what passes through those three gates. It is always the same shape: one identifiable actor, intending, doing something to one identifiable person, acutely, on an occasion you could point to in a courtroom. The bill is built on an individual-perpetrator model of harm. It sees the hand on the shoulder. It is constitutionally incapable of seeing the room.

The category error worth avoiding

Here is the question I keep being asked, in one form or another: do the Supreme Court’s ruling in For Women Scotland and the EHRC guidance that has followed it amount to conversion under this bill? Do they fail the test?

They do not fail the test. They never reach it — and the difference matters enormously.

To be a conversion practice, conduct has to clear the front gate before the high bar is even relevant: there must be a person, intending, directing conduct at an individual, to change who they are. A Supreme Court judgment has none of that. Statutory guidance has none of that. There is no perpetrator. There is no intent toward any particular person. There is no “conduct towards an individual.” These are general law and regulation — a different category of thing entirely. Holding them up against the serious-harm threshold is using the wrong instrument; they are excluded at the definitional gate long before the question of harm is ever asked.

So no — the post-For Women Scotland settlement is not a low-scoring conversion practice. It is not a conversion practice at all, and it never could be. And that, far from closing the matter, is exactly where the real story begins. Because the harm I am describing is not a failed prosecution. It is a harm the law has no concept for.

Naming the gap: constructive desistance

Employment law has a useful idea called constructive dismissal. Nobody sacks you. Instead, your conditions are made intolerable enough that you leave — and the law is sophisticated enough to treat the leaving as a dismissal, because it understands that an outcome can be engineered without anyone ever issuing the order.

We have no equivalent concept for identity, and we need one. So let me offer a name for it: constructive desistance.

Nobody coerces you out of being trans. No single actor intends it. Instead the environment around you is tightened, space by space, service by service, ruling by ruling, until living openly as yourself becomes a daily negotiation you grow too tired to keep having. You start mode-switching to get through the day. You stop correcting people. You shrink the parts of yourself that draw friction. Some people detransition not because anything inside them changed but because the cost of the outside became too high. The law has a concept for the coerced version of that — it is in the bill, with a five-year sentence attached. It has no concept whatsoever for the engineered-environment version, even when the end state is functionally identical: a person pressured out of living as who they are.

The conversion bill measures detriment one acute incident at a time. Constructive desistance is detriment delivered in a thousand pieces, each one individually too small to be “serious,” and that thinness is exactly what makes it invisible to an instrument designed to weigh single blows.

Hypothetical discomfort, actual detriment

I have written before about the asymmetry I think sits at the heart of this whole debate: on one side, hypothetical discomfort; on the other, actual detriment. The post-For Women Scotland framework, and the EHRC guidance tightening lawful single-sex provision around it, is justified very largely by the anticipated discomfort of others — discomfort that need not be evidenced, need not have occurred, and crucially carries no requirement of intent toward anyone. The actual detriment, meanwhile, lands on trans and non-binary people. But it lands distributed — spread so thinly across so many small frictions that no single instance rises to anything a law would recognise as a harm.

That is the trap, and it is a beautifully engineered one, whether or not anyone engineered it deliberately. Hypothetical discomfort gets to be general, pre-emptive and intent-free. Actual detriment is forced to be specific, evidenced and acute before it counts. One side of the scale is allowed to weigh feelings that have not happened yet. The other is told its real, accumulating harm does not register because no single ounce of it is heavy enough.

The honest counter-argument

I owe you the strongest version of the other side, because this is contested territory and pretending otherwise would be dishonest.

The good-faith position runs like this: none of these measures touches who anyone is. Gender reassignment remains a protected characteristic; nobody’s identity is being changed or suppressed by law. What has changed is the scope of access to certain sex-segregated services and spaces — where you may go, in specific and genuinely contested contexts, not who you are permitted to be. On that view, “conversion” is a category error in the other direction: a constraint on expression in particular places is simply not the same as a suppression of identity, and to call it so cheapens a serious word.

That argument deserves to be met, not dismissed — and meeting it is where my own claim becomes properly contestable. My bridge across the gap is this: that a sufficiently sustained, sufficiently total constraint on expression becomes a suppression of identity over time. That you cannot wall off “where you can be yourself” from “who you are” indefinitely, because a self that can never be expressed in public life is being slowly extinguished in private. That is the load-bearing claim, and I want to be clear that it is contestable — not on its logic, which I think holds, but on its evidence. Whether the cumulative effect actually reaches “untenable,” and for whom, and how often, is an empirical question, not a rhetorical one. It is the fight worth having honestly, and it is not won by assertion.

Thanks for reading! This post is public so feel free to share it - please support my work

Why this is an institutional question

Here is the turn, and it is the part I most want leaders to sit with.

The criminal law has now drawn a tidy line around the rare, acute, intended harm and walked away from everything else. That is what criminal law does; it is not a failing. But it means that legality has gone quiet precisely where the harm is loudest — in the ordinary, lawful, day-to-day operation of institutions. And a great many organisations have quietly concluded that quiet means clearance. If it is lawful, if no one could be prosecuted, then the matter is closed.

It is not closed. It has barely been opened.

Because the harm of constructive desistance is not done by lone bad actors the law can catch. It is done — assembled, really — by institutions operating entirely within the law: the employer whose facilities policy makes a trans colleague calculate their hydration around which building they are in; the service that is now lawfully entitled to exclude and reaches for that entitlement as a first resort rather than a last; the public body that updates its guidance to the letter and never once asks what the letter feels like to live under. No single one of these is a crime. No single one of these intends a person’s desistance. And their aggregate is doing more population-level damage than every prosecutable conversion practice in the country combined.

That is an institutional responsibility, and it does not end where criminal liability ends. This is the whole point. Leadership is the thing that operates in the space the law leaves empty. If your only question is “could we be prosecuted,” you have outsourced your ethics to the Crown Prosecution Service and you will sail straight through every harm the bill cannot see — which is most of them.

The bill itself, almost poignantly, hands us the better question. For the acute harm, it imposes a duty on those who govern organisations to take all reasonable steps to prevent it. We could choose — voluntarily, as a matter of culture rather than compulsion — to adopt exactly that posture toward the chronic harm the law will never touch. To ask not “are we running a conversion practice” (you are not) but “what is the cumulative human effect of the environment we are lawfully operating, and are we taking reasonable steps to prevent that.” To measure the friction, not just the incidents. To treat dignity as the operating standard that sits above legality, not the thing you are grudgingly entitled to withhold once a court has said you may.

The takeaway

The draft Conversion Practices Bill criminalises the hand on the shoulder. It is right to. But the harm being done to trans and non-binary people in this country is mostly not a hand on a shoulder — it is the slow, lawful, intended-by-no-one tightening of the room until staying in it as yourself becomes more than you can manage. The law calls the first thing a crime and the second thing nothing at all, and that silence is not neutrality. It is a vacancy. And vacancies in the law are filled by leadership or they are filled by harm.

Institutions cannot make the cumulative untenable a criminal matter, and they should not try. But they can refuse the comfort of “it’s lawful, so it’s settled.” They can own the outcomes they did not intend. They can take reasonable steps to prevent a harm no one will ever be prosecuted for — because it is the right thing to do, and because the people living inside the room they built are watching to see whether legality was ever the same thing as decency.

It never was. We get to decide what we do about that.

Hold the rope.