Why the fight over a single word leaves the mechanism that decides affordable housing untouched.
I have spent a good part of my career operating inside the contested regeneration zones of London, consulting with developers and local authorities, building and running mixed-use businesses in exactly the kind of neighbourhood where a development proposal and a protest tend to arrive together. I have sat on both sides of the table, with the residents who fear what is coming and with the people trying to build viable schemes. And across all of it, one word has done more to shape the public argument, and less to shape the actual outcome, than any other. The word is “gentrification”.
Let me start with the fact which is capable of ending half these disputes before they begin.
Gentrification is not a material planning consideration. In UK planning law a consideration only carries weight if it serves a planning purpose and relates to the use of land. Gentrification, as an abstract sociological label, does not qualify. For an objection framed that way to count for anything, it has to be translated into something the system recognises, such as loss of affordable housing, loss of employment floorspace, an equalities impact, or a daylight loss. The planning officer weighs the translated harm. The word itself is inadmissible. A resident can write gentrification on a hundred objection forms and the inspector is required, as a matter of law, to set the word aside and look for the planning reasons underneath it.
Everyone who does this professionally knows this. The developer’s consultants know it. The council’s officers know it. The result is a peculiar kind of theatre, where the loudest argument in public is the one with the least traction in law, and the arguments that actually decide the outcome happen somewhere quieter.
One word, four meanings, no traction
Here is where it helps to be precise, because gentrification is not one thing. The sociologist Ruth Glass coined the term in 1964, watching the middle classes move into the working-class quarters of North London, take over the modest houses, and change the whole character of the district as the original residents were, in her words, ‘displaced’. It was a sharp, specific observation about a particular process. Six decades later the word has swollen to carry almost any anxiety about neighbourhood change, and in swelling it has lost the original meaning which made it useful. What began as a description has become a container.
The research literature now uses it to mean at least four different processes, and objectors tend to mean all of them at once. There is direct displacement, people forced out by demolition or rising rent. There is commercial displacement, the trader or the market priced out to make room. There is demographic change without displacement, a neighbourhood’s character shifting through ordinary turnover. And there is the affordability of new homes priced beyond what local incomes can reach. These are different problems with different evidence and different remedies. Compressed into a single word, they become impossible to weigh, which means the word cannot guide a decision even when the worry behind it is real.
The worry is often real. Separating the meanings serves a purpose, as it reveals how much information the compression destroys. A specific, addressable grievance, “my trader relocation terms are inadequate,” or “the social rent replacement is a fraction of what stood here before,” disappears into an abstraction that the system is entitled to ignore. The resident reaches for the biggest word available because it feels commensurate with the fear. The system hears a word it cannot act on and has to dissect in an attempt to interpret it. Both sides then spend years fighting about it.
Meanwhile the mechanism that actually determines the outcome sits in a document almost nobody in the public argument reads. It is the viability assessment, and behind it, the cross-subsidy model. In the standard arrangement, the affordable homes in a scheme are paid for out of the profit on the market-rate homes. The affordable housing exists only because enough unaffordable housing was built alongside it to fund it. Which means the quantity of genuinely affordable housing is set by the margin on the market units and the assumptions in a spreadsheet, rather than by need or by the volume of objection.
The Heygate estate, in South London’s Elephant and Castle, is the case that shows what this produces. The regeneration that became Elephant Park carried a contractual affordable-housing obligation of 25 per cent, itself below Southwark’s own policy of 35 and well below the 50 that earlier guidance had floated. That is the headline most people argued about. The number underneath it is the one that matters. Campaigners calculate that of the delivered homes, roughly 4 per cent were social rent, against a documented net loss of around 919 social homes on the footprint of the estate that was demolished. About one in five of the original secure tenants remained in the postcode. The public fight was conducted in the language of gentrification. The homes were lost in the language of viability.
A word of honesty is owed here, because the nostalgia that often drives these campaigns deserves as little indulgence as the misdirection. The Heygate in its final years was not a lost idyll. It had become a byword for neglect, crime and the kind of concentrated deprivation that any serious government is right to want to end. Demolition was a defensible answer to a genuine failure. The criticism is not that the estate was replaced. It is that the replacement delivered a fraction of the social housing it removed, and that this outcome was determined by the funding model, not by the merits of regeneration itself. One can believe the old Heygate had to go and still notice that what replaced it quietly rehoused far fewer of the people it was meant to serve.
Who pays when the proxy war runs its course
The uncomfortable part, and the part a developer, a councillor and a residents’ association all need to hear, is that fighting the word often does the opposite of protecting the neighbourhood.
Consider Wards Corner in Tottenham, North London, the Latin Village. After fifteen years of objection framed largely as resistance to gentrification, and a long chain of legal challenges, the developer withdrew in 2022, citing rising costs and the drawn-out nature of the fight. On paper, the objectors won. In practice, the market they were defending had already been closed since 2020, and the community-led alternative they hoped would replace the scheme remained undelivered when the dust settled. The victory was procedural, but the place was gone anyway. If popular expectation is that blocking a development freezes an area in amber, the reality is that it simply defers the investment, raises the cost of capital for whatever comes next, and triggers the viability reassessment that reliably returns with fewer affordable homes, not more.
Bishopsgate Goodsyard, on the edge of Shoreditch in East London, tells the same story in a different key. A scheme fought down from 1,356 homes to 500 over a five-year battle, with the absolute number of affordable homes barely moving despite the headline percentage roughly doubling. The objectors changed the ratio but they did not change the maths that produced the ratio.
None of this is an argument for waving schemes through, and it is emphatically not an argument that residents are the problem. Residents are the ones left holding the consequences either way. It is an argument that the fight has been misdirected, and that the misdirection is convenient for precisely the advocates of the cross-subsidy model.
And there are advocates, with a serious case, which is worth stating fairly before disagreeing with it. The cross-subsidy model, delivered through Section 106, is defended across most of the political spectrum for a reason that has nothing to do with ideology. It turns private development permissions into affordable homes without funding the full capital cost from general taxation. In a country reluctant to raise tax and reluctant to borrow for social housing at scale, it has been the path of least resistance for thirty years, and it has produced real homes that would not otherwise exist. Its defenders (in government, in local authorities, and among the housing associations that depend on the pipeline) are not acting in bad faith. They are choosing the instrument that delivers supply without a tax rise.
The difficulty is structural. Because the model funds affordable housing out of market-rate profit, its output rises and falls with market values, sales rates, interest costs and build costs. When those deteriorate, the viability negotiation reopens, and the first thing to come under pressure is the affordable quantum or the tenure mix. London’s threshold-and-viability system formalises exactly this trade-off. The result is an affordable-housing supply that is weakest precisely when housing need is greatest, because a downturn that raises need also erodes the profit the model relies on to meet it. It is worth adding that the cross-subsidy is rarely the whole story in any case. A parliamentary evidence submission found that nearly three quarters of Section 106 affordable homes also drew on Social Housing Grant. The private-profit engine has long run on a quiet public subsidy, which makes the pretence that these homes are delivered at no cost to the state doubly misleading.
I am not proposing more taxation, and I would not. The critique of this model, from housing economists at bodies like the Centre for Policy Studies and in the direction of travel the current planning reforms are taking, points somewhere more useful. It points toward hybrids that reduce the dependence on market-cycle profit rather than deepen it. Fixed affordable-housing tariffs set per unit or per square foot, agreed up front and not reopenable through viability, so the number stops moving every time the market wobbles. Land value capture that takes a defined share of the uplift that planning permission itself creates, rather than leaving it to be negotiated scheme by scheme. Public land released on the condition that affordable delivery is fixed and grant-backed from the outset. And a far greater role for genuinely mixed delivery, including build-to-rent and institutional capital, whose economics do not swing on for-sale margins in the same way. None of these is a fantasy of the state building everything. Each is a way of making affordable supply less hostage to the property cycle, which is the actual disease.
The way this connects to the wider pattern I write about in The Re-Alignment Era is simple. A loud, emotive narrative draws the eye while a quiet structural mechanism does the actual work, and the noise is tolerated, even encouraged, precisely because it changes nothing. Gentrification has become one of the most reliable forms of misdirection in British public life because its meaning is too large to act on, not because it is meaningless.
The word gentrification as the currency of these arguments should be retired and replaced with the questions the word is standing in for. How many genuinely affordable homes, at what tenure, fixed by what mechanism, protected against what happens when the market turns? Those questions have answers. They can be written into conditions, tariffs, land terms, and so on. They survive a downturn. The word cannot do any of that, and the longer we spend shouting it at each other, the longer the model that actually decides the outcome stays exactly where its advocates would like it, which is out of the room and beyond the argument.
