03Brighton – a high-demand, low-supply city
Cities consistently build very few houses in their existing suburbs, but the amount they build in their city centres and outskirts varies. There are large parts of the suburbs in every city where no or almost no new construction is taking place, regardless of the level of demand for new housing.
For instance, consider where homes have been built in Brighton in Figure 2. The colour of the neighbourhood shows how many new homes have, on average, been built in that neighbourhood every year from 2011 to 2019. The 3D map available online shows the same colours and total construction, with the highest column in Shoreham supplying 241 homes over this period, or just over 30 a year.
Brighton is an expensive city, where the average house costs 14 times the average income in 2019. As a result, the city faces a severe housing shortage, and building more homes is a critical priority for the city’s economy and social inclusion. But despite the city’s unaffordability, new housing supply has increased by just 3 per cent since 2011, even though the city’s population increased by 6 per cent over that period.
The city centre of Brighton has seen housing supply increase by 7 per cent, less than the national average for city centres of 16 per cent. Shoreham town centre to the west of the map has seen a 28 per cent increase in the number of homes. But as it sits in the neighbouring local authority of Adur, it faces a distinct housing target from Brighton local authority.
In both Brighton and Hove and Adur local authorities though, there is little new housing on the outskirts of the city. Partly this is because Brighton faces severe topographical constraints in the South Downs to the north of the city and with the coast to the south.
But Brighton does not respond to these constraints by building more homes in the existing suburbs. The supply of homes in the suburbs has since 2011 only grown by 2 per cent, less than the city as a whole. Since 2011, 59 per cent of neighbourhoods across all of the suburbs built less than one house each year, and 17 per cent built none at all. This does not appear to have any connection with existing transport infrastructure, as little development takes places around a number of railway stations in Brighton.
In more unaffordable places such as Brighton, high land values mean redevelopment should be financially viable for developers. As the price of land reflects the demand from households and firms for that location, cities with strong economies and a growing population will see their land values rise. By sharing the cost of increasingly expensive land between more households, such redevelopment by the private or public sector helps keep housing affordable in these cities and towns. But this process does not appear to occur in Brighton.
Alongside this low growth in housing stock in the existing suburbs, Brighton does not have much concentrated new supply in its city centre and little on its outskirts, so total housing supply is low, despite high demand and a very unaffordable housing situation. Similar unaffordable places with patterns like this include Southend, Luton, Worthing and Ipswich.
Box 2: Data and method
Two major sources of data were used:
- The domestic Energy Performance Certificate register, for the supply of new homes from the beginning of 2011 to the end of May 2019.3
- Census records on dwellings by neighbourhood, or Lower Super Output Area (LSOA) in On average, each of these neighbourhoods contained about 700 houses in 2011.
The domestic EPC register was created as a database for research into building energy efficiency, but it has recently been made available for research on other areas of housing policy. It is updated more frequently and at smaller geographies than other sources of housing data.
EPCs can be granted to properties for a variety of reasons which are identified in the register, but they must be granted to new buildings whenever they are constructed, sold, let, or converted from commercial to residential use. As a result, while they are not a complete record of all the buildings in the UK, they do record all the new homes which have been built since the late 2000s.
All the new dwellings which cities have built since 2011 were identified in the EPC register, and assigned on the basis of their postcode to every LSOA within cities. These changes in housing supply, relative to the complete Census 2011 records of housing stock are then displayed in the maps, averaged out across eight years to produce a measure of new homes per year by neighbourhood by city in England and Wales across the 2010s.
One limitation of the data is that it is not able to record when homes are removed from cities’ housing supply, for instance due to demolition and redevelopment, or conversion into another use such as Airbnb lettings. It therefore cannot be used to investigate net change in dwellings at the local LSOA level, even if it can serve as an accurate measure of new builds. ONS data does indicate that demolition rates are low (7-10,000 a year nationally, compared to 240,000 net new dwellings a year), and short-term Airbnb holiday lets appear to be concentrated in city centres.4
The EPC data can also be used to calculate the average floorspace each resident of a city has, and how this has changed since 2011. These relationships are investigated in more detail in the Centre for Cities briefing Making Room.5