The Government has announced where it wants to build its twelve new towns. Here is my review.

Short version: they've done quite a good job. The key context is that England’s housing shortage is heavily concentrated in certain areas, as shown by Anna Powell-Smith's map of floorspace values below. We should generally focus housebuilding in these areas, because:

1. Doing so has far more effect on alleviating housing scarcity than building in places without shortages;

2. Because housing scarcity is normally driven by people’s desire to move towards employment centres, it has more economic impact;

3. Receipts from house sales are much higher; this means that the government can capture some of them through development levies and use it to fund superb infrastructure.

Over the last three years, housebuilding in London has collapsed. Molior recorded just 2,158 private starts in the first half of 2025, around 5% of London’s (low) targets, and still falling.

What is going on? I have posed this question to numerous specialists, most of whom cannot comment publicly for professional reasons. This thread is a summary of what I have gleaned.

The headline reason for the collapse of housebuilding is the establishment of a quango called the Building Safety Regulator, whose approval has been required since April 2024 (after a transitional period starting in 2023) for all tall buildings.

The BSR immediately became heavily backlogged, so many projects that would have started over the last year are stuck in the enormous queue of applications waiting for the BSR’s attention. Because a far higher share of new builds are apartments in tall buildings in London than elsewhere, this disproportionately affects London.

I originally thought this was the main cause of the current crisis, which made me relatively relaxed about it, since after all the Government can solve the backlog by simply increasing the BSR’s capacity. (It has indeed since done this.)

Unfortunately, things aren’t so simple.

Most of Inner London fell into poverty in the twentieth century, and expensive neighbourhoods like Notting Hill, Pimlico and Islington were seen as slums just decades ago. Here is a street in Notting Hill in the 1970s and today.

Why did this happen? Some thoughts.
First, let's remember when this happened. The decline starts around 1920. Population continues to fall until about 1990, although gentrification had already started in the 1970s and 80s. Since then, the population has risen steadily.

Japan is famed for its flourishing urban life, with peerless infrastructure and vibrant urban density. The bewildering image below is Tokyo's rail map, the world's most extensive network, and one of the true wonders of the world.

But it was not always so. Japan entered the twentieth century with large but impoverished cities, whose infrastructure was exceptionally poor. Most of these were then destroyed by war.

Japan developed perhaps the best cities of the modern world in spite of some of the worst starting conditions. An important part of the explanation lies in a policy called ‘land readjustment’. Tokugawa cities obviously did not have trains, trams, buses or cars. More surprisingly, they made little use of wheeled transport, relying on pack animals, canals and porters. This meant they only needed narrow and crooked roads.

The Shogunate divided up commoner areas with walls and gates to monitor and control the movement of the population. They often discouraged the building of bridges for the same reason.

The whole structure of Tokugawa cities was thus antithetical to modern transport needs.

In the nineteenth century, railways were the dominant form of mechanised transport. Vast networks were built in many countries. In 1846, the British Parliament mandated 9,500 miles of railways, of which two thirds were ultimately built. By contrast, HS2 will be 140 miles. This was an epoch-defining achievement. But when they reached existing cities, the Victorians faced a key constraint: they did not have the technology to bore tunnels deep under cities, and it was fantastically contentious and expensive to plough through them.

In the 1700s and 1800s, English and Irish developers laid out hundreds of garden squares, though there was no planning requirement that they do this, and though they never made any money on them. Why? The reason was that the garden squares raised the value of nearby houses, which were sold by the same developers. They raised it by so much that it outweighed the opportunity cost of not building houses on the squares themselves.

London is famous for its garden squares, laid out by private developers in the eighteenth and nineteenth centuries. Why did they do this? A short thread.

Most of the expansion of London took place on 'Great Estates', land owned by various church bodies, charities, guilds and aristocratic families. These are some of the survivors today, but originally there were far more.

Remarkably ambitious scheme for Waterloo Station. Waterloo is a vital piece of infrastructure and has some superb Edwardian architecture, but its surroundings are dismal and wayfinding is extraordinarily counterintuitive. Successful remodelling could be a huge improvement.
Proposals include a new concourse under the platforms running between new entrances on both sides, the pedestrianisation of Cab Road and Mepham Street, the opening up of viaduct arches for shops and restaurants, and the peninsularisation of the currently appalling 'Imax Plaza'.

From the C18 to the C20, Britain was extremely good at building transport infrastructure. In the Georgian period, 1100 private companies laid out and maintained 22,000 miles of tolled roads (‘turnpikes’), giving Britain the best road network in Europe.
Between 1830 and 1900, railway companies built some 20,000 miles of intercity railways, giving Britain the best railway system in the world. (For comparison, HS2 to Birmingham is 143 miles.)

The Government wants to build new towns. In a paper for @UKDayOne, @KaneEmerson and I propose a huge new town at the intersection of the East Coast Main Line and the planned East-West Rail. We believe this may be the best greenfield site in England. Here is why. @UKDayOne @KaneEmerson To succeed, new towns need to tie into existing cities with shortages of housing and laboratory space, as I argued in this recent BD piece. Doing so helps the new towns flourish and alleviates the shortages in existing centres.
bdonline.co.uk/opinion/what-m…

It looks likely that Britain is going to have some new towns. Many people think that all new towns are dysfunctional white elephants. This isn’t true: well located and well designed new towns have succeeded throughout history. The Romans founded planned new towns, like York, Chester and Gloucester. The Saxons founded them at places like Oxford and Wareham. Many planned towns were founded in the Middle Ages, of which Salisbury is famous.

It is generally assumed that making ornament is labour intensive. This leads many to think that the rising cost of labour in modern societies led to the decline of ornament, as it was outcompeted by smooth machine-made forms. Here, public buildings from the 1900s and 2000s. Like many myths, this is partly true. In premodern societies, making ornament generally did take a great deal of craftsmanship.

The glazing of this house is around 33% of its floor area. However, new regulations in Part O §1 allow maximum glazing of 11-18%. So *half to two thirds* of these windows would be blocked up. Do the British people really need their Government to protect them from this building? Now, §1 sets out only the ‘simplified method’. §2 allows builders to do ‘dynamic thermal modelling’, potentially allowing larger windows. But this is more complex and costly. There is a risk that large windows have effectively been banned in new homes for people on lower incomes.

Estate regeneration at le Plessis-Robinson, a ‘banlieue’ of Paris. Begun in 1989 and continuing today, replacing asbestos-heavy postwar blocks. Mixed use, mixed tenure, medium density. Here is the school, sports centre and market hall.

Notice how in Laon (L) the street forms an architectural 'frame' for the cathedral front, magnifying its importance and giving it a tremendous sense of scale (i.e. relative size). By comparison Notre Dame in Paris looks a little lost on its enormous plaza, cleared in the C19. Such framing is extremely common in older urbanism, creating a kind of artistic mutual dependence between the great public buildings and the fabric of the city around them. Here, Amiens.

St Edmundsbury Cathedral. The walls of the nave are C15, but most of today's cathedral dates to 1959-1970, and the tower was added in 2000-2005. One of the masterpieces of English Gothic, created in our time. The architect of the 1959-60 work, Stephen Dykes Bower, bequeathed everything he had to help fund the completion of the church, which was led by Warwick Pethers.

Hypothetical before-and-afters for Le Corbusier's Voisin Plan to demolish and rebuild central Paris. Of the street, in his characteristic striking style, LC wrote: 'There is neither the pride which results from order, nor the spirit of initiative which is engendered by wide spaces ... only pitying compassion born of the shock of encountering the faces of our fellows.'

When the Church of St Geneviève was turned into a secular mausoleum after the Revolution, its windows (L) were blocked up. Curiously enough this works fine internally, for which clerestory lighting turns out to be perfectly sufficient, and which remains remarkably beautiful. Externally however the change was rather a disaster, as may be seen by comparing the one facade that retains them (R) to the incoherent ones without. Suggested moral: windows are hugely important for enlivening and articulating facades, over and above their utilitarian value.

In the 1960s Paris adopted new regulations, inspired by London's postwar 'plot ratios', which incentivised tall setback slabs rather than mid-rise perimeter blocks. Amazingly it was envisaged that *2/3rds* of Paris's blocks would ultimately be demolished. Only a qualified version of the regulations applied in pre-1780 areas, but in arondissements XIII-XX the new rules had astonishingly rapid effects. Of course the mood shifted rapidly, and in 1974 Paris returned to a version of its ancient regulations.

Church of St Paul the Apostle in Salford. In 1968 the new priest, Canon David Wyatt, received a grant from the diocese to demolish and replace the semi-derelict building. He persuaded the bishop to let him restore it instead. Stephen Dykes Bower, perhaps the last great Gothic architect, volunteered to help pro bono. Over the early 1970s the church was restored to his remarkable designs, at once rich and restrained.

The facade of Reims Cathedral, lit with lasers to replicate the appearance of the original paintwork, albeit glowing surreally. Technically almost unbelievably impressive. Another illustration of the rich polychromy of medieval architecture. From frenchmoments.eu/reve-de-couleu… Here a similar experiment at Chartres. From: chartresenlumieres.com/en/chartres-li… and churchpop.com/2015/02/23/med…