Although the Greater London Authority (GLA) was established in 2000, it only assumed responsibility to administer, to allocate, and to deliver centrally funded affordable housing programmes in 2012 within a policy framework set by, and negotiated with, Whitehall.
The current mayor, Sadiq Khan, is responsible for two overlapping centrally funded housing programmes: the Shared Ownership and Affordable Homes Programme 2016 to 2023 (2016-23 SOAHP or 2016-23 programme) and the Affordable Homes Programme 2021 to 2026 (2021-26 AHP or 2021-26 programme).
The GLA was allocated £4.8 billion in 2016-23 SOAHP central government funding to deliver at least 116,000 affordable homes by March 2022, a deadline that due to the impact of Covid on delivery was subsequently extended to March 2023.
By April 2021, around 72,000 dwelling starts and 34,500 completions from that programme were recorded. Of the remaining 43,000, around 17,000, according to the mayor, are to be started in 2021-22 and 26,000 in 2022-23, accordant with targets previously agreed with Whitehall.
In November 2020 £4bn from the successor 2021-26 AHP was allocated to deliver 35,000 affordable homes across London by March 2026. London-wide allocations from the GLA’s first bidding round for that programme were announced on 31 August 2021, when £3.46bn was allocated to 53 housing providers to deliver 29,456 homes, none of which have started yet.
The mayor expects the 2021-26 AHP – running alongside the extended and outstanding preceding 2016-23 programme – to provide a combined total of c.79,000 affordable new home starts by March 2026.
The Housing Committee of the elected London Assembly at its 19 October meeting(the committee), coincident with the publication of its 2021 Affordable Housing Monitor (2021 Monitor), considered issues relating to the past and future delivery of these two programmes.
This post uses most current up-to-date information that the committee, other forums, and the official affordable and other housing statistics published in November 2021 provide(d) to illuminate London-wide trends and issues likely to impact on affordable housing delivery until 2029.
Associated policy development implications at both central and local level are identified and related, where applicable, to the core themes of A Social Democratic Future, including the real fiscal crisis of the state.
The core conclusion is that the driving principle of the mayor’s successful 2017 Affordable Housing and Viability Special Planning Guidance should be extended nationally, but in a way cognisant with regional, sub-regional, and local housing market characteristics and values.
Such a policy shift would align to the line of current political least resistance to make the Section 106 process simpler, with greater certainty and transparency, making paramount the principle that the planning gain attributable to granting of residential planning permission should be recycled to the maximum possible extent to the provision of affordable housing and supporting social infrastructure.
It should be combined with changes in the public accounting treatment of housing investment that reflected its productive income and wealth effects and its self-funding capacity over a 30-year period.
This should tap into a discernible overlapping political and technical consensus that the provision of affordable housing must be mainstreamed through measures that directly reduce the cost and price of housing into both public and private business models, and thus blur the distinction between market and social housing, so lessening the trade-off between maximising social rented and total affordable currently aggravated by constrained central housing capital budget allocations.
1 Defining and tracking affordable housing
What precisely counts as an affordable home is itself a complex and contestable concept that tends to be tautological rather than illuminating.
The government’s definition was published in the 2012 National Planning Policy Framework, last updated July 2021.
Broadly speaking, dwellings provided at least 20% below local market rents (including service charges where applicable) and 20% below local market values in the case of discounted home ownership, including shared ownership, are counted affordable under that definition, which must remain at an affordable price for future eligible households or include provision for any receipts to be recycled for alternative affordable housing provision is required when public grant is involved.
Defining a dwelling as affordable, of course, does not make it by itself affordable for households of varying circumstances subject to local housing market conditions.
The GLA definition of affordable housing attempts to address that consideration, whilst maintaining conformity with the government’s definition of affordability and other programme funding conditions, where applicable, by requiring it to be made available “at a cost low enough for eligible households to afford, determined with regard to local incomes and local house prices”.
For example, the GLA defines London Living Rent (LLR) as an: “intermediate affordable housing tenure with a London-specific rent introduced by the mayor that will help, through sub-market rents on time-limited tenancies, households save for a deposit to buy their own home. Rents are based on one-third of the estimated median gross household income for the local borough, varied by up to 20 per cent in line with ward-level house prices, and capped to reflect the maximum affordability for an eligible household. Providers of LLR homes funded through the Affordable Housing Programme (AHP) 2021-26 will be required to offer tenants the opportunity to buy the LLR home on a shared ownership basis during their tenancy and within ten years. Finally, the benchmark rents also vary based on the number of bedrooms within the home”.
In planning delivery terms, Delivering affordable housing (HP4) and Affordable housing tenure (HP6) policies of the 2021 London Plan, published in March, specified the following split of affordable products delivered (proportion of total affordable units provided expressed as 100%) expected to be provided in a development):
- a minimum of 30 per cent low-cost rented homes, as either London Affordable Rent or Social Rent, allocated according to need and for Londoners on low incomes (A1);
- a minimum of 30 per cent intermediate products which meet the definition of genuinely affordable housing (A2);
- the remaining 40 per cent to be determined by the borough as low-cost rented homes or intermediate products (as A1 and A2 define) based on identified need;
- Where the proportion of (total affordable) homes provided in in a development exceeds 35% (50% on public land or on industrial sites, and public sector landowners with agreements with the mayor across a defined portfolio of sites, or 60% in case of an appointed strategic development partner), their tenure (split) is flexible, provided that the homes are genuinely affordable, as A1 and A2
The presumption is that the 40 per cent to be decided by the boroughs under (4) will focus on social rent (SR) and London Affordable Rent (LAR). HP6 does, however, recognise that for some boroughs a broader mix of affordable housing tenures will be more appropriate, “either because of viability constraints or because they would deliver a more mixed and inclusive community”, in which case it should be determined through the local Development Plan process or through supplementary guidance.
Social housing is allocated with reference to statutorily defined need categories and is mainly allocated to those on low incomes or are otherwise medically or socially vulnerable.
Intermediate housing in the guise of LLR mainly caters for potential households in employment with moderate to average (middle) household incomes below £60,000; households with gross household incomes of up to £90,000 per annum can be eligible for some intermediate home ownership products, primarily London Shared Ownership (LSO).
As an indication of prevailing actual affordable intermediate costs, a proposed 100% intermediate affordable development in Ealing will offer one bedroomed dwellings attached with a market value of c£450,000; translating – assuming that purchased equity stake is 25% on mortgage with 10% deposit found paid – into a monthly cost c£1,239, including a £223 service charge; and two bedroomed units attached with markets value between c£550-600,000 translating into a monthly cost of c£1,500, including c£300 service charge. Properties will be initially offered to eligible local households demonstrating annual household incomes of £60,000 before higher income applicants are considered.
The official statistical sources
There are lies, damn lies, and statistics. Official housing statistics, of course, don’t propagate lies. They do, however, suffer from frequent definitional and consistency discontinuities, from subsequent revisions of previously published data, from gaps in coverage, and from reporting shortcomings (such as source data omissions and miscoding). All can impact on data accuracy, completeness, and transparency. Data owners can decide what to publish, in what format, over what period, and what is included and omitted.
Published data then gets subject to competing political spin narratives, selective in what data and what time periods are highlighted, presented to paint a picture of the past, present, and future that the narrator, accordant with their own organisational interests and preferences, want beholders to see.
All this makes the interpretation of housing data over time difficult, time-consuming, and an invariably less than definitive experience: confusion can consequently trump understanding.
It is important to understand what a particular statistical source in each case is describing and what it is not.
Tables 1 and 2 of GLA Affordable Housing Statistics reports affordable starts and completions that the GLA has funded or monitored as part of a wider regeneration programme. It can omit affordable housing activity undertaken by councils and other registered providers (mainly housing associations) not recorded by the GLA. It is thus a subset of total gross new affordable housing supply provided in London over any set period.
The affordable housing statistics series published by the Ministry of Levelling up, Housing and Communities (MLHC) in its Live Table 1011 is more complete, but it is only published annually in November. It “aims to provide a complete picture on affordable housing delivered, irrespective of funding mechanism”, including new build and acquisitions from the private sector, (although not losses through demolitions or sales), as well affordable completions reported in local authority as well as GLA statistical returns to MLCH, some section nil grant 106 units that the GLA in its 2019-23 returns may have omitted, as well as units funded through Right to Buy recycled receipts not counted in the GLA statistical series.
Although Table 1011S of that series reports total affordable starts only from 2015-16, Table 1011C provides a time-series for completions back to 1991-92, reproduced in Annex Table One.
The GLA and MLCH series are compared in Table 1 below.
The MCLG technical note to the latest published 2020-21 series explains the variance between the two series, thus: “local authorities are asked to only record (in their LAHS return) affordable housing that has not been reported by Homes England or the GLA. To assist them in doing so and minimise the risk of double counting, Homes England (outside London) or the GLA (within London) sends all local authorities a list of the new affordable housing recorded in their administrative systems. However, despite best efforts, double counting may still occur if local authorities misunderstand the instructions on the form or if, due to differing definitions of completion of housing, local authorities considered that a unit had been completed in a separate financial year”.
It appears higher for completions, where it can reach 40%, as it did, most recently, in 2019-20, reduced last year to 8%. Around 76% of the 2020-21 starts in the MLCH series were recorded as funded either by Homes England/GLA but only about a third of the same year completions.
Neither the GLA nor the MLHC affordable housing statistical series report net totals that take account of reductions in the affordable stock, including from right-to-buy (RTB) and from demolitions; both negative flows can be considerable.
A distinction on the margin can be made between losses due to RTB and demolitions: dwellings demolished are lost now and for ever as source of affordable housing and require the rehousing of previous occupants; yet when integral to a redevelopment or regeneration programme, they may be replaced in time.
Right-to-buy completions reduce the affordable stock and (all other things remaining the same) deflate the future flow of affordable housing opportunities (although receipts generated by the sale may enable partial replacement) but not the total dwelling stock: the dwelling sold can still be lived in by current and future occupants.
Also published annually each November, Table 118 Housing and Communities dwelling stock statistical series provides a net additional dwellings time series for London (E12000006) that is broken down into components, including new build completions, net conversions, and change in use, with demolitions netted off.
According to MLCH, it is the primary and most comprehensive measure of housing supply that provides the “only consistent data source for providing dwelling stock estimates and changes in net supply between census years in England (London), at least from 2006-7”.
That the series is adjusted to reflect the decennial Census result suggests the possible margin of error attached to it, however. And, like the GLA and MLCH affordable series it does not breakdown units provided by bedroom size (number of bedrooms).
It is not broken down according to tenure, and therefore it cannot be directly reconciled with the MLCH annual affordable housing series. It can only be inferred from both series, that gross affordable completions accounted about a third of total new build completions during 2020-21.
Programme phasing, starts and completions
The GLA affordable housing programme is subject to successive central government funding allocations, themselves subject to comprehensive spending review decisions. Each multi-year funding settlement then is allocated by the GLA to individual development partners following an application, competition, and moderation phase.
Development programmes then involve long lead-in times, requiring total scheme funding to be negotiated or identified, and the necessary scheme approvals to be secured and preparations undertaken, prior to any start on site.
Complex and multi-partner scheme developments take years to gestate and then complete. For example, the large-scale redevelopment of the Hackney Woodberry Down and Ealing South Acton council estates took or will take up to 30 years to complete.
The GLA’s Capital Funding Guide defines dwelling start as the date when the investment partner and contractor has signed and dated the building contract and the building contractor takes possession of the site or property and undertaken some necessary defined preliminary works, such as excavating for foundations or infrastructural drainage work.
The beginning of each programme cycle for such starts invariably is relatively fallow followed by a rapid pick-up that then is translated into a later bunching or concentration of completions towards the, or even beyond the end, of successive funding programme cycles.
For example, in 2014-15 over 18,000 completions were recorded (see Annex Table One London 2029), more than three times the volume recorded the following year, reflecting the concentration of completions at the end of the 2011-15 Affordable Housing Programme.
The latest stipulated date for 2021-26 AHP funded dwellings to start is March 2026 and the end date for their completion is March 2029: a potential gap of eight years between programme inception and end, and three between the start and completion of individual dwellings.
The preceding 2016-23 programme is attached with no backstop completion date, meaning that completions funded by that programme could possibly trickle-in towards the end of this decade.
In that light, October’s Housing Committee (the committee) was reminded that a family can only live in a dwelling when it is completed, not when it is started; on the other hand, assessments of programme delivery need to relate to the applicable programme funding and associated target arrangements.
Using GLA Table 1 affordable starts within the capital funded or reported by the GLA since 2008 have averaged annually, overall, around 11,000: 12,000 during 2008-12; falling to less than 9,000 in 2012-16; rising again to approach 13,000 during 2016-21, as Table 2 of this post catalogues in the next section.
Much care is needed in interpreting such overall averages, however, as they are sensitive to the choice of period surveyed. And any average taken over a period can mask variations marking that period or an underlying trend within an inherently shifting story, compounded as it is by programme phasing and the gap between starts and completions identified above.
Over 30,000 dwellings under the 2016-23 programme were started during 2019-21 suggesting a total figure of c43,000 remaining programme starts between April 2021 and March 2023, nearly double the average annual of 11,120 starts during the preceding decade (excluding any that may also be funded under the 2021-26 AHP).
Such a level of starts, if achieved, will then feed into subsequent 2021-26 completion figures, pushing them beyond the previously record 2008-12 completion average, plus some.
On the other hand, the smaller 35,000 dwelling 2021-26 programme will subsequently come on stream from 2023, pushing down the average annual start and then the completion rate towards the end of the decade.
A total of c108,850 expected GLA dwelling starts between the decade April 2019 and March 2029 (c30,500 2019-21 starts plus 43,000 remaining 2016-23 programme starts plus 35,000 2021-26 programme starts) by itself, if achieved, would provide an overall annual average of just under 11,000 for that period, and an even lower completion average. That figure, however, excludes the unknown number of starts that can expect to result during 2027-29 from any successor programme to the 2021-26 programme, and from any other additional allocations.
All of this, of course, shorn of context is prone to presentational cherry-picking.
Improving statistical reporting
Neither the GLA nor the MLHC affordable housing series break down start and completion activity according to bedroom category. This is an important omission insofar that much new affordable housing in London, as noted earlier, is delivered through multi-decade redevelopment/regeneration schemes. These often involve the phased demolition and replacement of existing social rented and other housing tenures by a higher number of one to two bedroomed units.
In that light, the 2021 Housing Monitor recommended that bedroom numbers should be included in the GLA data, which should also capture all gains and losses to the affordable housing stock, and that starts and completion data should also differentiate between funding programmes.
This website has asked both the GLA and MLCH to make efforts to reconcile its respective affordable housing series, and the MLCH its affordable and net additions series.
And, as the London Plan highlights, the growing importance of Section 106 in providing additional affordable housing supply within London means that it is crucial that their implementation and review mechanisms are monitored to ensure that the additional homes contractually committed are delivered in the form promised.
Its monitoring of affordable housing (HP7) policy, in that light, requires boroughs to have clear monitoring processes to ensure that the affordable housing secured on or off site is delivered and recorded in line with the requisite Section 106 agreement, and to share that monitoring information with the GLA to incorporate within its annual monitoring process, and, to underpin the accuracy of their statistical returns to MLHC.
These returns, in the view of this website, should record all individual housing obligations contained in section 106 agreements, and to track them according to whether they are:
- delivered in accordance with the agreement; or,
- remain outstanding; or,
- were cancelled; or,
- differed in quantum and composition to the original agreement.
2 What the published available statistics tell us and what they don’t
The most telling message conveyed by Table 2 below concerns programme composition.
It shifted from social rent (accounting on average for 64% of starts 2008-12) to affordable rent (55% of starts, 2012-16), then followed by shifts in favour of intermediate housing provision (50% of starts) and of social rent (31% of starts), both away from affordable rent during 2016-21.
The total and proportionate share of social housing should increase significantly further in the future on current plans. An August 2021 mayoral press release advised that 57 per cent of the 2021-26 AHP programme will be for social rent, of which half will be delivered by councils; shared ownership or London Living Rent accounting for the rest.
That increased 57% proportion compares to the 46% share that social rent took of the total number of affordable dwellings started in 2020-21, and the 64%, 30%, and 17% of total affordable completions that on average social rent accounted for throughout the 2008-12, 2012-16 and 2016-21 mayoral terms, respectively.
The 2021 Monitor also broke down starts and completion according to borough. Four boroughs completed over 2,000 affordable dwellings in total between April 2016 and March 2021 (over 500, on average, each year): Tower Hamlets, Newham, Southwark, and Ealing. At the top, Tower Hamlets, 4,306 completions, while the other three ranged from 2,709 in Newham to 2,070 in Southwark.
Four contributed less than 400: Richmond, Harrow, Havering, and Hillingdon, with Richmond reporting the lowest: 210.
With respect to social rent/LAR completions, Tower Hamlets, Newham, and Ealing all reported over 500 completions over the same period (18% to 23% of their respective total completions). Sutton reported two; Kingston, three; Harrow, four; and, Richmond, 16.
These headline figures, however, do not provide a good guide as to the progress of each borough in meeting the 2021 London net additional dwelling requirements that Annex Table Two London 2029 reproduces.
Ealing, for example, has a 10-year net completion target ending March 2029 of 21,570 dwellings. An October Planning Appeal decision noted that the borough is delivering, at best, 40% of its objectively assessed need for affordable housing while the council accepts that it cannot demonstrate a five-year supply of deliverable housing sites (consistent with it delivering its London Plan targets).
As the previous section noted, the GLA affordable housing statistics series only cover programme activity that the GLA monitors, thus excluding some local authority and housing association self-funded and nil grant section 106 activity. Nor does it breakdown starts and completions according to type of provision.
Table 3 provides that breakdown by utilising the more comprehensive and complete MLCH series
Strikingly, more than 50% of the total 10,800-odd affordable dwellings completed during the last two 2019-21 years were delivered through the nil grant Section 106 route – a progressive increase from the 13.7% recorded in 2014-15.
To put that secular trend into an even longer-term perspective, until 2013-14 (when they accounted for 12.7% of the total, a proportion that thereafter rose steadily year-on-year, as above) the share of the total affordable supply accounted for by nil grant Section 106 completions had had previously exceeded 10% in one year only: 2006-07 (see Annex Table One London 2029).
Table 3: Affordable completions, according to tenure type and section 106 status, 2014-21
In tune with that, GLA analysis of planning data, published in March 2021, reporting mayoral approvals of planning applications referred to the mayor during calendar year 2020 (largely because application concerned more than 150 dwellings), advised that 37% – the same proportion as 2019, a record – of the total 38,865 units approved (14,337), and 41% of habitable rooms (a better measure as it take account of dwelling size composition and mix) were affordable. That proportion had only reached 30% in 2017, after previously out-turning in the 18% to 25% range between 2011 and 2017.
Apparently discordant with that secular trend, however, MLCH Table 1011S indicates that only c16% of 2020-21 starts were expected to be funded through nil grant Section 106 route.
Table 4 shows that successive ten-year average completions have not diverged from the 30-year average by more than 8%.
Over successive five year periods more variation is discernible, but the table provides little evidence that the political complexion of the government in power proved a major factor; the two lowest five year averages reflected, first, a squeeze on public housing investment in affordable housing during the early New Labour years; and, second, the during 2016-21, the tail end of fiscal austerity acting disproportionately on housing capital investment presided over by David Cameron and George Osborne, tempered in outcome by increased nil grant section 106 supply.
Completions recovered to reach record levels during the later New Labour years; as the previous section noted, new record levels can be expected at some point during the next decade as 2019-23 start levels subsequently feed into completions.
On the whole continuity interrupted by macro-economic, public expenditure, electoral, and programme cycles, rather than any overarching ideologically-led policy intervention, consistent with post-Thatcherite governments recognising the need for additional affordable housing, if not the means.
The primary and overwhelming message is that the delivery of affordable supply has grossly undershot assessed requirements and aspirations, regardless of the political colour of the governing party.
Table 5 below uses the MLCH net additions series. It records a near-doubling of net supply from the levels recorded in the early noughties to reach c32,000 by 2009. The impact of the GFC then resulted in a drop to c21,000 in 2012-13, then a progressive increase (interrupted in 2017-18) took it to a record c41,000 dwellings in 2019-20 (including c37,000 new build completions). It then dropped back, probably due to Covid-related impacts, by 9% to c37,000 (including c33,000 new build completions) last 2020-21 year.
A sharp fall in demolitions was also recorded in 2020-21.
Table 6 below advises that new builds account for most net additions, although between 2015-17 changes in use from office and other uses to residential reached 22%, prompting wider concern about the space and other standards. It dropped back later in the decade.
The source statistical series (MLHC, Table 118) is not broken down according to tenure. The relative negative impact of demolitions on affordable net supply outcomes can be expected to be higher: large scale demolition of existing social housing estates has generally been more prevalent in recent decades than redevelopment of privately owned stock, more predominant in in the 60’s and 70’s.
In that light, a Southwark Monitoring Report covering the April 2004-19 period reported 15,237 planning affordable home planning approvals but a net increase in affordable supply of only 9,924.
The demolition of estates, such as the Heygate, close to the Elephant and Castle, subject to comprehensive and long-term redevelopment and replacement programmes can cause net affordable housing supply, taking account of demolitions, to subtract from gross affordable supply.
The committee was told that housing associations and other social landlords face ‘competing priorities’ between maintaining existing buildings and developing new homes, with the cost of remediating fire safety defects and decarbonising homes reducing the capacity to build new ones. A central headwind echoed in the last meeting of the London First convened Council-Led Housing Forum (the forum) held that same morning.
Richard Hill, vice chair of the G15 group of housing associations within London, advised the Housing Assembly committee that the ‘operating environment’ its members faced required them to re-focus on building safety and investment on their existing stock, causing their development numbers to go down, noting that housing associations face multi-million billion bill to fix safety issues over the next 15 years that came attached with an opportunity cost of around 72,000 new affordable homes foregone.
The London Plan, indeed, includes requirements for all developments of ten or more homes to be net zero-carbon and to incorporate sustainable urban green spaces as part of a wider drive for London to be a zero-carbon city by 2030.
The mayor in his August press release also confirmed that housing providers building homes funded by the new 2016-23 programme, on top of that core climate change sustainability requirement, will also have to meet new conditions on building safety and design, including:
- The installation of sprinklers or other fire suppression systems in new blocks of flats;
- A ban on combustible materials being used in external walls for all residential development, regardless of height;
- Minimum floor-to-ceiling heights and a requirement for private outdoor space;
- A ‘sunlight clause’ requiring all homes with three or more bedrooms to be dual aspect, any single aspect one- or two-bedroom homes to not be north-facing and at least one room to have direct sunlight for at least part of the day.
Switching to the crucial wider economic environment, deputy mayor for housing, Tom Copley, advised the committee that rising building costs along with shortages of materials and labour in the construction industry could “have a serious impact on scheme viability”, while Darren Levy, director of housing for the London Borough of Newham, starkly warned that: “We’re seeing materials, labour all increasing in cost. [There are] supply chain issues. There’s a general lack of stability. The labour market is becoming more challenging, there are fewer skilled operators available, and they have to a certain extent got a number of suitors in the construction industry in London… I do think, going into the next five to 10 years, it will have a significant impact into that future pipeline”.
On top of that, NIMBY-related opposition to developments that might meet London Plan and local plan requirements but are perceived as detrimental to the local environment and/or future house price prospects, will also need to be navigated.
One of the appealing characteristics of many parts of London is that high quality and high value residential areas are often cheek-by-jowl to social housing or mixed tenure developments. But residents of the former tend to possess more effective voice power in a local planning decision context to oppose and sometimes block high density new affordable housing developments.
The role of councils
On a positive note, a ‘renaissance of council housing’ was celebrated, which the forum was advised involved the delivery since 2018 of more than 8,000 affordable dwellings by London councils; a figure expected to rise to c.10,000 by March 2022, notwithstanding the recent impact of Covid.
As was noted earlier, councils have been allocated more than 50% of the 2021-26 programme for social rented accommodation, which is expected to translate into an additional c12,000 such units funded with GLA support over the next five years.
Southwark, for instance, has a longstanding commitment to provide an additional 11,000 council homes through ‘various means’ by 2043, as part of its efforts to maintain a net increase in the council stock; an aim helped in recent years by rising values that makes the right-to-buy increasingly unaffordable to existing tenants and by higher social housing re-provision levels after a period when losses of social rented stock resulting from estate -regeneration -linked demolitions and RTB’s exceeded gross new affordable provision..
Barking and Dagenham established in 2017 a wholly owned company, Be First, that with 400 hectares of development land, proclaims plans to “provide 50,000 high quality new homes and 20,000 new jobs by 2037”. More modestly, its 2020-25 business plan that the local cabinet approved in April 2020, envisaged “the delivery of 116 new homes in 2020/21 from four development projects and the commencement of seven new development projects to deliver a further 938 homes up to 2024-25, with an average of 73% affordable housing”.
Moving west, Ealing’s November 2020 Cabinet approved a £390million business plan for BLRP, a subsidiary of its wholly owned housing development company, Broadway Living. Along with the £99m grant that the council secured from the Greater London Authority in 2018, the investment will be used to build at least 1,300 affordable homes within the next six years. In total, the plan expects to see BLRP build 1,513 homes, with the sale and let of some of the homes subsidising the development of more homes for affordable let. The business plan projected that BLRP to pay for itself over the course of half a century by “recouping the initial investment through sales and rents”.
A lot can happen over such timeframes, however, and such local development company initiatives come attached with uncertain and elongated risk profiles embracing, but not exclusively, execution, development linked to market conditions, interest rate, and funding source, risk.
The experience of Croydon’s locally owned development, Brick by Brick, created in 2016 demonstrates their potential impact provides a salutary warning. It received £200m in development loans from the council, but by 2021, after its development programme stalled, had not produced any dividends or returns, contributing to its parent council and owner becoming in effect insolvent and unable to make a balanced budget.
Mr Hayes at the forum related a shift from the previous reliance on housing associations to develop affordable or acquire housing from developers to a reaction to the growth of ‘mega’ associations with confused commercial and social ends, drawing on his previous experience as Ealing’s regeneration-lead to note the actions of two associations in embarking on a bidding war for sites contrary to the local public and social interest.
Councils with their greater democratic accountability and their ability to co-ordinate planning with provision activity and to overview local housing and other markets are relatively well-placed to step-in and step-up provision to fill the gap left by HA’s hamstrung by post-Grenfell pressures to spend on their own stock.
Councils, however, compared to private housebuilders, face higher development costs, including across their material supply chains; in that light, moves to aggregate purchasing across organisations and to secure greater vertical integration of their development activity (such as setting up in-house construction arms) can be expected.
On a different but related note, drawing on his current role at Be First, he also noted the rising value of industrial land, reaching £6m per acre, as a constraint on direct provision, although he felt that values would plateau into the medium-term.
The forum was also told that some London authorities are also approaching their Housing Revenue Account (HRA) borrowing cap and that, overall, the scope for councils to expand further direct provision beyond current plans is likely to be limited.
Taken in the round, the conclusion must be that council provision of social housing will remain a sub-theme rather than a transformative one in London housing the next decade.
The funding environment
Pat Hayes, the current managing director of Be First, also highlighted at the forum possible policy risk, cautioning that the current mercurial ‘right-wing populist’ Johnson-led government, gyrating between radical market- and state-led interventions, could simply decide to move the policy goalposts concerning council provision.
The biggest risk, as it seems to this website, is the replacement of Johnson by a successor more wedded to Rishi Sunak’s proclivity for a smaller state and fiscal conservatism – a far from unlikely scenario, even though one still likely to be tempered by the wider political imperative to retain the recently won Conservative – in the absence of a more precise description – ‘working class’ and ‘red wall’ vote.
Certainly, central government cannot be relied upon to provide increased funding support. Although the 2021 Spending Review (SR21) did announce an additional £1.8 billion in total for housing supply, on top of the existing £11.5 billion investment through the Affordable Homes Programme (2021-26), of which £7.5 billion is over the SR21 period, 65% of the funding will be for homes outside London.
The government’s political driver to target resources to the North and the Midlands – the home of its ‘red wall’ seats – to further its Levelling Up agenda combined with the chancellor’s commitment to bring down the share of national income taken by public debt within three years, means that any further increase in centrally supported housing investment for London is unlikely.
The £4.9bn that Tom Copley advised the committee that is needed ‘every year for 10 years’ to provide affordable housing on the scale needed will remain ‘for the birds’ in the absence of imaginative and radical systemic change to the housing finance system, requiring at least some measure of cross-party support.
The concluding section will return to that.
Changing programme composition
The funding and other sources of affordable housing are multiple subject to frequent change. The average amount of public grant or other support required to deliver affordable housing varies according to tenure, location, and bedroom mix, the prevailing policy environment, as well as wider macro-economic and housing market conditions.
Public grant support was squeezed down practically to nil during the 2014-19 period when public housing investment became a key victim of fiscal austerity. A social rent unit will require more than twice as much public support than an intermediate unit.
As way of example, the GLA’s Building Council Homes for Londoners fund offered £100,000 per Social Rent/London Affordable Rent compared, compared between £28,000 and £38,000 for London Living Rent, London Shared Ownership (LSO) or other genuinely affordable intermediate unit started between April 2018 and March 2022.
Programmes comprising 90% social housing and discounted home ownership, respectively, will generate quite different numbers of additional housing opportunities for households of different social-economic characteristic.
Politics abhors a vacuum, and Tom Copley at October’s committee highlighted that the mayor had successfully argued (or, put another way, the government for its own reasons had been willing to accept the principle) for the majority of the 35,000 homes to be provided (started) by 2026 to be social rent dwellings, the “most affordable and most needed” homes, noting that “the government has come full circle from when it originally wouldn’t allow social rented homes to be created”.
He further pointed out that the 2016-23 programme of 116,000 dwellings was attached with £4.9bn grant support compared to £4bn for the 35,000 dwellings currently expected to be delivered through the 2021-26 programme.
The average unit public grant cost of £42,000 to provide each of the 116,000 affordable dwellings under the 2016-23 programme will rise (mainly due to the change in programme composition towards the more grant-intensive social rent, but also due to build cost inflation and the cost of design, quality, sustainability, safety and equality, diversity, and inclusion benefits) to a projected c£114,000 for each of the expected 35,000 dwellings of the 2021-2026 programme dwellings.
The reversion back to social rent should help to contain the central government housing benefit bill to a lower figure than would be the case if the same units were to be provided at the ‘affordable rent’ tenure (higher rents require more benefit where tenants are eligible, which the majority are) and at the margin should enhance work incentives; crucially, too, the shift from intermediate to social rent should more directly assist authorities to get a greater number of homeless families out of expensive and unsuitable temporary accommodation.
The downside, however, is the much higher subsidy requirement involved in providing a social unit and the resulting trade-off that exists at any given total level of grant availability between, on the one hand, maximising social housing and, on the other, total affordable numbers, largely explains why the total number of affordable dwellings that the 2021-26 programme will deliver (start) will deliver less than a third than its predecessor (expected delivery according to current conditions and plans).
The cross subsidy model
Providers, mainly the mega housing associations, ‘stretched’ receding levels of public grant over the last decade, buttressed by reserves providing a capital risk buffer, and by building units for market sale to cross-subsidise the provision of affordable homes, between and within schemes, for example, when large council estates are progressively demolished and replaced with a higher number and density of predominately private dwellings, which when sold to finance succeeding scheme phases.
But, as noted above, post-Grenfell and other spending pressures bearing down on such providers will curb its future scope.
Across recent decades the predominate development model has been cross subsidy, certainly within London and the main urban conurbations. The process has helped to maintain gross affordable supply during a period of constricted public grant support availability.
But the cross-subsidy model will not by itself provide the affordable housing on the scale required; certainly, in the absence of a scale increase in public grant that the current unreformed public expenditure and wider policy environment will not provide.
It suffers also from other shortcomings that can militate against or even undermine its purported purpose. Cross-subsidy is intrinsically sensitive to wider market conditions, which – as in the wake of the 2009 Global Financial Crisis (GFC) – can interrupt, delay, or even forestall development progress.
The model encourages – and even at the margin is dependent – on providers to maximise market sale proceeds. This incentivises them to market and sell units, for example, to foreign investors or wealthy parents of foreign students, pushing-up prices way beyond the reach of local people on moderate to above average local incomes.
Even where developers can only maximise sale proceeds by targeting local buyers, many of whom are likely to have far from high household incomes, who then, in effect, cross subsidise the affordable units, the numbers of which, in turn, in the absence of additional public grants support, can be crimped when market sales proceeds are limited by local purchasing power.
More fundamentally, but related to above, the cross-subsidy model is predicated on, and perpetuates, a housing system that become increasingly riven and driven by multiple market failures, concentrated in the more economically buoyant sub-regions concentrated south of the Humber and Trent and east of the Tamar.
Use of nil grant Section 106
London in recent years have, as Table 3 demonstrated, increasingly has come to rely upon nil grant Section 106 planning agreements (obligations) to deliver affordable housing. It is, in effect, a mechanism to partially recycle or ‘tax’ the uplift in value secured through residential planning permission; as such it is integral to the cross- subsidy model: a palliative to the imperfections of the broken wider housing market on which it relies.
In the past, information and expertise imbalances between local authorities and developers allowed the latter to game the Section 106 system to their own advantage, increasing developer profit at the cost the number of affordable units provided: St. Mary’s Residential Case study provides a high profile Southwark example
The progressive step-increase in the proportion of total new gross affordable supply taken by nil grant Section 106 since 2013 provides some evidence of greater local authority focus and effectiveness in negotiating and then securing delivery of higher levels of affordable housing within regeneration and other schemes using the cross-subsidy model (as well as to the sterling and painstaking work done by organisations, such as the Southwark-based 35% Campaign.
A key catalyst of that step-increase was the 2017 Affordable Housing and Viability Supplementary Planning Guidance (2017 SPG), now formally enshrined in the 2021 London Plan, supporting its polices including HP5 and HP6 (discussed below), are to be implemented, which must be taken into account as a material consideration in planning decisions.
A streamlined Fast Track Route was introduced for schemes that include at least 35% cent affordable housing, and 50% on public or redeveloped industrial land that enables them to progress without the need to submit detailed viability information and without late viability review mechanisms which re-assess viability at an advanced stage of the development process.
This, according to the mayor, “provided greater certainty to the market, sped up the planning process, and has helped to increase the level of affordable housing secured in new developments”. The consequent greater consistency and certainty has substantially improved outcomes.
Schemes that conversely do not provide the threshold level of affordable housing or meet other relevant policy criteria, or that provide off-site or cash in lieu contributions, must continue to follow the Viability Tested Route and are subject to viability scrutiny and late, as well as early, review mechanisms.
Tom Copley was asked at the committee whether more affordable housing could be squeezed from private developers through the Section 106 route.
He was not hopeful, however, no doubt taking his cue from his peer, the Deputy Mayor for Planning, Regeneration, and Skills, former Hackney mayor, Jules Pipe, who a month earlier advised the September London Assembly Planning and Regeneration meeting in a Question and Answer session that continuing to rely on developer section 106 contributions to provide affordable homes, is “not a viable way of delivering the amount that we need as a capital city even combined with housing grants. Neither is sufficient….it is still obviously a very valuable tool and we would be loath to lose it. If anything, it needs to be enhanced, but quite how … when you are still requiring all these other things from a development, it will only pay for so much. Going back about ten years, it would have been standard for developers to expect about a 20% return. I think we are getting them down to about 12.5% on many of the developments that we see. Much below that, the real estate community will say, “Well, actually, we do not get enough return in London, we will go and develop elsewhere, thank you very much.”
That last point is addressed below.
4 What needs to be done
The 2021 London Plan (plan), based on a Strategic Housing Market Assessment (SHMA) undertaken in 2017, publicised that London needs to deliver each year c66,000 net additional completions across all housing tenures for the next ten years, two thirds (c43,500 or 65%) of which should be in social or affordable tenures.
Yet the Planning Inspectorate report that reviewed it concluded that over the next 20-25 years capacity only existed to deliver only 52,000 net new homes per year. Consistent with that, the total ten-year target for new dwelling additions across London contained in the London Plan (reproduced in Annex Table Two) only totals 522,870 dwellings. Little evidence exists to provide confidence that authorities are on track to meet their targets with some, such as Ealing, reported earlier, unable to demonstrate sufficient progress.
The stark fact is that the past, the current, and the planned future provision of both total and affordable homes within London, as demonstrated in this post’s data tables, will continue to fall well short of the plan’s target and capacity.
The improved levels of annual net additions achieved in recent years, reported in Table 5, running at c40,000 still are less than two-thirds of the SHMA requirement, while the proportion of that total taken by affordable dwellings seldom exceeds a third.
Much therefore must change if future delivery is to approach declared needed levels.
On a positive note, a general cross-party – national and local – political acceptance has developed that new dwelling supply, and the share of affordable housing within that total, should, indeed, increase.
That acceptance, albeit sometimes masked by tactical discordant rhetoric – whether emanating from Whitehall or City Hall – has, however, been largely limited to the payment of lip service to a recognition that the ‘housing market’ does not work, save for some measure of government pragmatism in terms of policy change at the margin, such as lifting of the HRA borrowing cap.
But the radical and concerted changes to the strategic policy framework that real progress requires has stubbornly remained backstage.
Why land and multiple housing market failures bedevil new housing opportunity
The slippery slope began with the progressive divorce of the twin-principles enshrined in the 1947 Town and Planning Act that both development control and value vested with state. During the 50’s and 60’s the reduction and abolition of development (betterment) and changes in compulsory purchase legislation, in combination provided scope for developers to realise future gains from value uplifts attributable to the granting or even hope of residential planning permission.
That process and its resulting consequences has been clearly charted by the work Derek Bentley has done for Civitas, most specifically in The Land Question, which by piecing together various sources, noted that between the mid-1950’s and 1990’s, the average price of land attached with residential planning permission, at 2016 prices, spiralled from £150,000 per hectare to £1.3m, surging again to £5m by 2007, reaching £6.2m prior to the onset of the GFC. Values have risen subsequently. Land in agricultural use without prospect or hope of development is attached with an average value of £22,000.
This website in The Measurement of Land Prices used similar sources to report 1994-2010 valuation trends of land vested with planning permission, while noting problems inherent in the compilation of land price indices (largely ceased since 2010) which try to provide weighted averages of an enormous array of transactions that vary according to site size, location and connectivity, other site circumstances, on the density of its development, and on the planning conditions imposed, as well, of course, on actual and expected future market conditions,
More recently, indicative land values of a hectare of land attached with residential planning permission across London ranged from £6m to more than £35m.
Technical detail should not, however distract from the clear and instructive story that Figure 1 below paints of the upward step-change in land prices that commenced in mid-fifties accelerating across cyclical oscillations, subsequently.
The financial liberalisation of the eighties, lubricated by the secular downward trend in interest rates, easier access to mortgage finance, and growth of dual income households, increased monetary demand for housing.
It was not accommodated by supply, constricted, as it was, by the cessation of mass council building programmes and by a tendency towards market concentration in the private housebuilding sector.
The prospect of higher market sale proceeds, crucially, bids up the future cost of development value of land, propelling prices further upwards, sometimes in a frenzy to be followed by a subsequent bust that then compresses supply even further. Higher and rising proportions of residential development value, up to 70%, are now accounted for by land.
The housebuilding industry has been transformed from one marked by dispersed multiple small and medium-sized suppliers in often local markers competing on both quality and price to an oligopolistic structure dominated by a few large, mainly national, builders. The three largest Persimmon, Wimpey and Barratt, account now for around a quarter of the market, compared to 1988, when Daily Telegraph columnist, Liam Halligan, in his recent book, Home Truths, advised that 40% of completions were provided by 12,200 companies.
The end-result, now widely noted and recognised across the political and ideological spectrum, is that the these now dominant companies have become price-givers driven by an incentive to restrict supply to maximise their short-term profit rather than output, allowing them to make supra-profits – as much as £60,000 or more per house built, and to pay chief executive and other salaries and bonuses in the millions, while engaging in restrictive practices, such as inserting covenants that involve regular doubling of ground rents on properties that could be sold freehold.
It is a ‘market’ in name for a product – vital to individual and community well-being – distinctive in that it has risen in price so much in both absolute and relative terms in tandem with a quality decline, where consumers can exercise so little real choice. It is truly a market broken by the imperfections and distortions outlined above.
Two pathways – given the prevailing and expected future political and financial environment – appear to provide the best and most feasible routes to future London and national housing affordability.
1 Public accounting for productive housing investment
Across the short-to-medium term, there is little or no prospect of any further step-increase in central government housing investment funding arriving to push up that total substantially to lessen that trade off; the downside risk, rather, is that fiscal conservatism will have re-exerted itself within the Conservative government by the time of the next spending review.
Even in the unlikely event that a Labour or hung parliament was elected in 2023-24, a scale increase in housing investment would require it, not only to set fiscal rules that lift the current 3% of GDP ceiling on investment expenditures, but to juggle and prioritise competing public investment demands in such a way that both maintains financial market confidence and avoids construction industry material and labour shortages.
Recent speeches by the Labour leader and his shadow chancellor have emphasised that Labour next in office will be ‘fiscally responsible’ and will not ‘throw public money’ at problems for the ‘sake of it’.
Quite so, but also redolent of New Labour whose early headline adherence to previous Conservative spending plans to demonstrate its fiscal probity resulted in a squeeze on affordable housing investment during the early noughties, which had deleterious economic and social consequences.
A symptom of the real fiscal crisis of the state marked by a mismatch between the public expenditure requirements of the UK and the political and electoral willingness for them to be met through salient and efficient forms of taxation, or borrowing where economically and financially sustainable and appropriate.
Given that, it is both vital for, and incumbent on, the housing sector to make as strongly and clearly, as possible, both the economic and financial, as well as social, case for increased affordable housing investment, and to integrate that with supporting supply side reform to the construction industry.
Affordable housing investment due to rising rents and values, can be self-financing over a 30 year or plus time span. An increased supply of social housing should reduce and deflate the future costs of keeping priority households in expensive and unsuitable temporary accommodation.
A strong invest-to-save case, therefore, can and should be made, alongside, and integrated, where possible, with innovative reforms to the financing of affordable housing and its associated public accounting treatment consistent with the above.
2 A root remedy to engage with an overlapping political consensus: deflating future land prices directly by embedding affordability requirements
The 2017 SPG adopted the Existing Use Value Plus (EUV+) approach (current use value of a site plus an appropriate site premium), to determine the benchmark land value. It serves to embed known and increased affordable housing requirements, crucially, into the generation of shared future public and developer expectations on expected sale receipts and thus acceptable planning gain surpluses, then feeding into future land value and price expectations. That is getting to the root of the matter (addressing a core housing market failure: escalating land prices based on speculative rent.
Housing Policy Five (HP5) incorporated into the 2021 London Plan (para 4.5.4) confirmed that the approach seeks to embed affordable housing requirements into land values and create consistency and certainty across London, noting earlier where there had been a relaxation in affordable housing and other planning requirements, it typically led to higher land values, not an increase in housing delivery. It further advised that the 35 per cent affordability threshold level will be monitored and reviewed in 2021 to determine whether it should be increased, with any changes consulted on as part of an updated Affordable Housing and Viability SPG or through a focused review of the London Plan.
Jules Pipe did shed doubt, however, in September, as the preceding section noted, whether much more affordable housing can be squeezed out of the Section 106 route at the London-level, at least without wider, supporting, and concerted changes to the national policy framework.
A point, indeed, that this website’s March 2018 Response to London Plan consultation advised that it : “fully supports the aim and content of the Mayor’s AHVSPG (2017 SPG) and believes that it sets a needed template for national housing policy reform as part of a needed package to repair Britain’s broken housing market. This requires the effective direct deflating of land values and a reduction in developer profit margins. But by the same token it is concerned that its intentions to be achieved requires similar reinforcing strategic policy reform at the national level, without which there is a possible danger of a developer ‘strike’ occurring at London-level as the largest housebuilders/developers migrate to locations where they can continue to enjoy excess super-profits. Cross party support for a sustainable reform – based on a credible and sustainable tailored incremental approach – is again a pre-requisite for the effective implementation of the needed package of reforms”.
The driving principle of 2017 SPG to close the gap between existing use and developed land value across all planning authorities should now be extended nationally, but in a way cognisant with regional, sub-regional, and local housing market characteristics and values, where necessary. In some lower value areas an EUV+ of, say, 30%, could, unlike London, render specific affordable housing thresholds almost superfluous.
In short, greater universality and consistency in ends and purpose at a national policy level, which allows a locally tailored treatment, rather than a uniform ‘one size fits all’ prescription, is required.
Such an extension and deepening of the principles of the 2017 SPG would directly reduce the cost of new housing: the root of the matter.
It would align along the line of current political least resistance to make the Section 106 process simpler, with greater certainty and transparency, making paramount the principle that the planning gain attributable to granting of residential planning permission should be recycled to the maximum possible extent to the provision of affordable housing and of supporting social infrastructure.
To take one example of that, the 2017 Conservative Manifesto (p41) committed a future Conservative government to ‘work with private and public sector house builders to capture the increase in land value created when they build to reinvest in local infrastructure, essential services and further housing, making it both easier and more certain that public sector landowners, and communities themselves benefit from the increase in land value from urban regeneration and development’, consistent with an overlapping political and technical consensus that the provision of affordable housing must be mainstreamed through measures that directly reduce the cost and price of housing and that blur the distinction between market and social housing, into both public and private business models.
The corollary is, of course, is that London political stakeholders should gravitate towards such a broad-based cross- party campaign given that London’s housing outcomes largely depend upon the national policy framework.
Whether any government, especially a Conservative one with strong links with the housebuilding industry with its now entrenched vested and lucrative private interests, would possess the will, focus, and strength of purpose to secure such a change, remains open to doubt. After all, not much has come from the 2017 manifesto pledge.
Perhaps, before long, political necessity will breed boldness to do what is right. The time for helping that moment along is now.