The buildings that appear in a neighbourhood are chosen from the set that planning rules permit and finance supports. Both filters are narrow, and their overlap is narrower still.
Rules define an envelope, not a design
A planning framework sets what may be built where, through use classes, height limits, distances from boundaries and how much of a plot may be covered.
Within that envelope a developer chooses a scheme, but the envelope usually determines the type of building before an architect is engaged.
This is why streets built under one set of rules look consistent. They are all solutions to the same geometric constraints.
Parking requirements consume the site
Many frameworks require a minimum number of parking spaces per home, and each space needs a substantial area once access is included.
On a small site this can make a scheme impossible, since the parking either fills the plot or forces expensive underground construction.
Where such minimums have been relaxed, the buildings that follow are typically denser and cheaper to build, which is why the requirement attracts so much attention.
Approval risk is priced into the land
A site is worth what can be built on it, so land is bought on an expectation of permission rather than a certainty.
The gap between the two is risk, and it is carried by whoever holds the land through a determination process that can take years.
Long or unpredictable processes therefore favour large developers who can afford to wait, and squeeze out the small builders who cannot.
Obligations attach to permission
Consent commonly comes with requirements to contribute towards infrastructure, affordable housing or public space.
These are negotiated against the scheme's economics, and a development that cannot bear them is either redesigned or abandoned.
The negotiation explains why two similar sites in one area can produce very different buildings, since each was tested against what its own numbers could support.
Existing residents have structured influence
Consultation gives neighbours a formal route to object, and their concerns tend to focus on traffic, overlooking and the character of a street.
Those who benefit from a new building are usually not yet living in the area and have no equivalent voice in the process.
That asymmetry is a recognised feature of most planning systems, and it pushes outcomes towards smaller, lower and slower schemes than the housing numbers alone would produce. Reforms in several countries have tried to correct it by fixing rules in advance so that individual schemes are judged against a standard rather than negotiated case by case.