Timothy J. Dixon – Emeritus Professor at the University of Reading and Visiting Fellow/Research Associate, Kellogg College/GCHU
In the run up to the UK Budget on 26th November, this blog explores the data around the Government’s housing target and asks, do the figures add up in terms of land use?
There has been a lot of scepticism about the UK Labour Government’s plan to build 1.5 million new homes in England by 2029, including some critical expert commentary in a recent BBC Panorama programme. Much of the focus so far has been on ‘blockers’ to housing growth, such as capacity and skills levels in the construction industry, the lack of progress in local plans and planning decision delays. However, there has been little discussion of the way in which different types of land use for development may or may not help achieve this highly ambitious target. In this short blog, I want to dive deeper into the data to ask three linked questions:
- How realistic is the overall target?
- What types of land could be developed to help meet the target?
- What are the implications of the evidence on house building to date?
Firstly, the 1.5 million target means that at least 300,000 net additional dwellings will need to be built each year over the period 2025 to 2029 (the maximum lifetime of the current parliament). The government has now also increased the mandatory annual target to 370,000 homes per year for local authorities, although the overall target remains unchanged. The net additional dwellings (NAD) figure for England includes homes that are newly built; homes that are created by change of use, for example a shop into a house; conversions, for example a house into flats; demolitions; and other changes to alternative dwellings such as caravans or houseboats. Data for ‘homes that are newly built’ are used here, rather than ‘permanent dwellings completed’, because they are more comprehensive, and focus on a wider set of data than just building control data.
Figure 1 shows the NAD trend from 2007 to 2025. During this period NAD has never exceeded 250,000 per annum. Indeed, in the current year (2024-25) recent data show that NAD is highly unlikely to exceed much more than 200,000 per annum. This would then have knock-on effects on the remainder of the period 2026-2029. Although the government is taking action to speed up the planning system through a new Planning and Infrastructure Bill and, controversially, is relaxing planning regulations and affordability criteria in the quest for growth, all the evidence, based on past performance, suggests that achieving the overall target will be very challenging. Recent research suggests that no government since records began in 1931 has managed to deliver 300,000 net additional dwellings. It’s also unclear how the new initiatives, such as the twelve new towns announced by government (each with at least 10,000 new homes), new housing on former railway brownfield land developed by the company Platform4 (40,000 homes), or the recently announced ‘forest towns’ in the Oxford-Cambridge arc, will complement the existing target.
Figure 1 New build completions and NAD for England (2007 to 2024) (Source: based on MHCLG data[i])
Turning to the second question, this is where it gets a little more complicated. The government has emphasised the need for brownfield land development (or what is known as previously developed land) to be prioritised through the National Planning Policy Framework (NPPF) 2024 (para 125), which focuses on a presumption in favour of sustainable development on brownfield sites. Brownfield land can include industrial sites, aerodromes, former retail and car parks, and some of this land may also be contaminated from previous industrial uses making development even more costly and complex. The main data source on brownfields in England today is Brownfield Land Registers (BLR) which are maintained by local authorities, and which may in some instances not be accurate or up to date. Nonetheless, based on this data, recent work by the CPRE suggests that there is sufficient brownfield land available for 1.41 million homes, although only just over half of these sites (enough for 770,000 homes) have planning permission. It’s also worth noting that, according to available data, the largest number of brownfield completions in England was in 2008, when some 78% of new dwelling completions (excluding extensions) was on brownfield land (with the remaining 22% on greenfield sites).
In its quest to ‘get Britain building’, the Government has therefore focused on what it terms ‘grey belt’ land (para 143 of the NPPF). These are: areas of land in the green belt that were (i) either previously developed; and/or (ii) contribute little to either the prevention of urban sprawl, keeping settlements separate, or preserving the character of historic towns, but may be considered suitable for development. Examples could include wasteland, petrol stations or carparks, and other types of previously developed land. However, it is not clear how much grey belt land there is in England and how many houses it could support. Recent evidence suggests the figure may be anywhere between 50,000 and 4 million! Work by Paul Cheshire at LSE, and other recent research, has also shown how existing railway stations could be used as hubs for housing development on grey belt (or greenfield) sites.
Finally, England’s green belt makes up what is a small but important part of overall greenfield land (or ‘undeveloped land’). Green belt areas were established some 70 years ago in England to restrict urban sprawl and comprise about 13% of England’s land area. With an emphasis on ‘build, baby, build’ it seems likely that green belt and other greenfield land will also need to be an important part of the mix if the targets are to be fulfilled.
So, what does the available data tell us? To provide an answer requires some broad-brush assumptions based on existing evidence:
- An annual target of at least 300,000 NAD is maintained to meet the overall target of 1.5 million.
- The current overall brownfield stock is sufficient for 770,000 homes initially (based on CPRE data).
- The overall number of completions on brownfield grows in parallel with overall new build completions from 2026 onwards, as more sites with permission come on stream.
- The long-term average trend of 60% of new completions on brownfield (2007-22) will increase slightly to 61% from 2025-29 because of additional brownfield sites coming onstream[ii].
- There is enough grey belt land for 100,000 houses over the 5 years (20,000 pa) (based on Knight Frank data).
- The long-term trend of 90% of NAD as new build completions continues from 2025 onwards.
- The NAD figure for 2025 is estimated to be 199,300 (based on Housing Forum analysis).
Using this data reveals the outcome in Figure 2. Here we can see that even with grey belt provision (albeit a modest figure of 9% completions) and with greenfield land utilisation at 30% of completions, annual total brownfield completions would need to exceed the high point of 2008 completions (156,000) from 2027 onwards. Also note the dramatic increase in annual housebuilding required to meet the target of 1.5m new homes over the period 2026-2029, which represents a highly optimistic compound annual growth rate (CAGR) of 21% per annum in both NAD and new build completions, as brownfield completions increase at the same annual rate over the same period.
Figure 2 Annual trends and projections for brownfield and greenfield completions, new build completions and NAD for England (2007-2029) (Source: based on data and assumptions above)
Figure 3 Total and average annual projected brownfield, grey belt and greenfield completions in England needed to meet 1.5 million target (2026-2029) (source: based on data and assumptions above)
Figure 3 shows the projected overall split between the three types of land use for the whole period (2026-2029) and the important and crucial role that brownfield would play in the overall NAD completions ‘mix’, alongside grey belt and greenfield sites (in terms of both overall completions during the period, and the annual average completions figure). In short, at least 325,000 NAD would be required annually on average over the period 2026-29 to meet the remaining total completions of 1.3 million homes required.
All of this is a very big ‘ask’ for local authorities, housebuilders, government and other stakeholders to deliver the NAD required. It also raises important questions for environmental sustainability and the overall balance with ‘growth’. On the one hand, it might seem sensible to prioritise brownfield sites, and the government is aiming to fast track development on these sites with ‘brownfield passports’, but we also know that many are home to diverse natural habitats and species. Again, building on greenfield sites (whether grey belt, or otherwise) will also have far-reaching consequences for the environment, as Tom Oliver (University of Reading) and Eleanor Milner-Gulland (University of Oxford) have pointed out.
Ultimately the devolution of decision-making and local housing plans has put the onus on local authorities to help achieve the national priority of building more homes, with the added emphasis on affordability. But, given the historic data we’ve seen, is this really feasible? As the political balance tips towards ‘Yimbyism’, this surely comes at a price for the environment, whatever type of land is to be developed. We might even challenge the notion that additional supply will create lower prices, as recent analysis suggests this won’t necessarily be the case. Additionally, there are some 700,000 empty homes in England, which is an underutilised element of housing stock, but is often overlooked or treated as too difficult to renovate or retrofit.
Surely, then, we need a new comprehensive national housing strategy[iii] that focuses on affordability, and (i) properly joins up housing, planning, land use, and environmental policies; (ii) prioritises brownfield land development more strongly in the NPPF; (iii) really gets to grips with defining and measuring grey belt in more detail; and (iv) improves data collection and analysis (for example through BLRs) to enable better informed decision-making locally, regionally and nationally.
Greater clarity in the budget this November will be crucial if the government is to offer more certainty on its housing ambitions and on how the new town, railway land and forest town developments fit into the target. Although progress has been made on a social housing strategy, a comprehensive strategy for future housing provision more broadly, with a detailed roadmap/timeline, would go a long way to resolving the current uncertainty over achievability and the resultant environmental impacts.
For more information on Tim Dixon’s previous research work on brownfields visit: https://www.timothyjdixon.com/
[i] Data sourced from https://www.gov.uk/government/collections/net-supply-of-housing
[ii] This makes some assumptions over the data on brownfield completions, and additionally, the official basis for such land use change was modified during 2011-12. But the available data have been adjusted to exclude extensions, and applying the average rate to new build completions gives a reasonably consistent estimate over the period involved.
[iii] Interestingly, work by Lucy Smith (and colleagues) for the Heywood Foundation at University of Oxford also points to the need for a long term comprehensive UK ‘National Strategy’ to encompass a range of themes (see https://www.bsg.ox.ac.uk/people/lucy-smith)