Written evidence submitted by the New Economics Foundation [DSH 012]
Q1 In November 2019, the New Economics Foundation called for the establishment of a Public Land and Housing Corporation (PLHC) to use public and private land strategically to deliver the next generation of zero carbon social homes, delivering 50,000 homes a year itself.
a) Given the poor progress on delivering social and affordable housing on public land, do you believe it is still possible for surplus public land to deliver 50,000 homes a year?
Surplus public land has significant potential for affordable housebuilding. Across the two central government programmes 2011-2020, land sufficient for 178,000 homes is likely to be released. In addition to this, the larger capital receipts generated by local authority land and property sales, with at least 12,000 sites disposed of since 2014/15, suggest local authority land sales make a significant contribution towards land supply for housing. The programmes have produced very little (2.6%) social rented housing, as our research has shown. I outline the reasons for these failings in my answer to Q.4.
Our research showed that many of the UK’s major cities - which have high social housing deficits - have sold land with significant housing capacity, which could have been used to deliver social housing. 43% of the housing capacity from the public land sale to date has been in London and the South East, which together account for over a third of the national waiting list for social housing. Six of the top twenty local authorities with highest housing capacity on public land sold are also six of the top twenty local authorities in terms of housing waiting lists.
Public sector land can clearly make a large contribution to meeting more ambitious social housebuilding targets, although data availability is not sufficient to identify this exactly. In 2019, we estimated that, taking into account a discount for low hanging fruit and the amount of land already disposed of, while noting that the programme could be run more efficiently in the view of the Public Accounts Committee and National Audit Office, we could potentially forecast enough public land for 100,000 homes on central government land over five to ten years. Much less is known about local authority land sales, but given the significantly higher capital receipts, number of sites sold, and the fact that two-thirds of public sector land is owned by local government, we can estimate that 100-150,000 homes could be built on local authority land over the same timeframe.
Alongside its spatial potential, we have also shown that using no-cost public land in a social housing programme could deliver a 62% reduction in the grant levels needed for those homes built on former public land. Using surplus no or low-cost public land for building social housing reduces the need for government borrowing for many of the social homes we need, freeing up grant to be invested in further schemes.
b) Wouldn’t Homes England be best placed to fulfil this role, rather than a new body?
Our research outlines a body which would lead on developing surplus central government public sector land, as well as acquiring private sector land at more affordable prices for development (see Q. 3). This central government body would be responsible for assembling and developing land in the public interest, primarily for social housing, meeting around a third of our need for social housing (see Q. 6). It requires three elements: i) The creation of a strategic body, with regional implementation; ii) bringing together public land and private land, acquired under revised land acquisition powers; and iii) bringing significant amounts of housing delivery back in house, undertaken by central government. This body – the Public Land and Housing Corporation - would most appropriately sit within Homes England, as indicated in the report. Homes England already has a small direct delivery aspect, and has regional offices, as we suggest for the PLHC, and so as our report argues, scaling up this direct delivery element through a PHLC situated within Homes England would be appropriate.
Q2 Your report further estimates that the planning system will continue to deliver 10,000-20,000 homes a year through section 106 agreements.
What impact would you expect to this figure from the Government’s intention to include mandated percentages of First Homes within section 106 agreements?
As the government’s First Homes proposal stands, NEF shares the view of the National Housing Federation, Shelter, the G15, Unite, and a number of other sector bodies that the introduction of First Homes either as a percentage of section 106 affordable housing through developer contributions or as a percentage of all units delivered on suitable sites will come at the expense of building homes for those most in need; social rented housing.
While the reductions in government spending has reduced social house building over the past decade (a government spending reduction of 70% between 2010 and 2018 corresponded with a social house building reduction of 85%), there is also clear evidence that social housing delivery through the planning system has been affected by the introduction of new forms of affordable housing products. Since the introduction of new types of ‘affordable housing’, including Affordable Rents at 80% market value in 2011, social housing provided through Section 106 agreements has flatlined at an average of 3,180, whereas Affordable Rent homes have increased to 12,727.
The government’s First Homes consultation is ongoing, but our analysis is that First Homes, delivered either as a percentage of section 106 or as a percentage of all units delivered on suitable sites would squeeze out social housing. If First Homes were delivered as a percentage of section 106 affordable housing through developer contributions, it is likely they would squeeze out other types of affordable housing like social housing from the limited money available. But equally if First Homes were delivered as a percentage of all units delivered on suitable sites, it has the potential to impact the viability of schemes and might have negative consequences on other developer contributions, e.g. Section 106 affordable housing contributions and/or lead to developments on these sites being delayed. It is likely that delivering First Homes through either mechanism will significantly reduce the amount of social rented housing delivered through 106 agreements.
Q3 The new PHLC body you proposed would require changes to the Land Compensation Act 1961, so it could use expanded compulsory purchase powers to enable it to purchase land at existing use value. You noted examples of this in Germany and Denmark.
The Committee planned to visit Denmark but has been unable to do so due to the General Election and the ongoing COVID-19 pandemic. Could you explain how land purchase and the planning systems works in Denmark, contrasting this with the current system in England?
In the UK the Land Compensation Act 1961 gives landowners an entitlement to ‘hope value’ on their land, which essentially means that the value of a plot of land depends not on what it is currently used for, but what it could be used for (e.g. housing) with planning permission. A key finding of the Independent Review of Build Out undertaken by Sir Oliver Letwin was that land-awarded planning permission for housing is currently worth 275 times the agricultural value. When selling land in the UK the seller is entitled to this additional value.
In Denmark, and other countries like Germany, public bodies are able to purchase land at closer to ‘existing use value’. So a field is sold at (or closer to) its value as a field. This makes purchasing land for state-led development more affordable, and would reduce the cost of acquiring land for affordable housebuilding. In both of these countries, this valuation is backed by compulsory purchase powers. There is evidence suggesting that these powers are rarely needed, as the valuation framework with the risk of compulsory purchase results in transactions which are at fairer values. Letwin’s report recommends a reduction of land prices from 275 times to around ten times. Bringing the land compensation act in line with the Danish approach could help achieve this, and enable social homes to be built for a reduced investment. Such powers were possessed and used by the new town development corporations, which undertook significant amounts of postwar development in the UK.
In addition, Denmark uses novel forms of land value capture, to share the financial benefits of new infrastructure and development to deliver social goods. In building Ørestad, a high-density satellite new town at the edge of Copenhagen, a system of land value capture was used to fund Copenhagen’s first metro line. Uplift in land prices created by the new development and infrastructure was used to fund the metro. A study by Savills and KPMG into the jubilee line extension found that the new line created an uplift in values in the surrounding areas of 53%, and so the potential is significant. Following the Danish approach to taxing land value increases could provide vital funding for infrastructure such as social housing.
Q4 In February 2020, NEF analysed the amounts of affordable and social housing that is being built on public land that is being sold and found that: the Government has sold enough public land for developers to build 131,000 homes -2.6% of those homes will be for social rent. 15% of homes built on public land will be classified as ‘affordable housing’.
What did your research uncover about the reasons for these low percentages?
There are two foremost reasons for the low percentages of affordable housing on public land sites. Firstly, the disposal programmes have no requirements or targets for affordable housing. The land disposal programmes have a general housing target and a financial target, but no social housing target. The Greater London Authority has a target - through the London Plan - of 50% affordable housing on sold off public land. A similar approach could be followed nationally.
Secondly, the reliance on private developers to develop public sites reduces the potential for social housebuilding. Our investigation into the public land sale, and conversations with officials, has demonstrated that the programme’s financial target encourages an approach to surplus public land in which land is sold to the highest bidder, and not the one who could deliver a higher proportion of social housing, or other social policy objectives. A high price of land stymies attempts to build social housing by private developers. We have found that between 2011/12 and 2014/15, the top 10% of local authorities in England in terms of land prices experienced a 70% drop in the numbers of new affordable/social rent homes built. This compares to a drop of 20% in the rest of England.
How much homes can be sold for, and the value of costs (eg s106 contributions) required feeds into the price of land. If a high price is paid for land by a developer, the houses need to be sold for a higher price and/or costs avoided to deliver their costs plus a profit. As land price is such a significant proportion of house prices - up to 70% - high prices significantly reduces their ability to make financial or in kind social housing contributions.
One way that this plays out in practice is through viability negotiations on affordable housing contributions. Our analysis of planning documents has found that on public land, developers have used the viability system to avoid building affordable housing on former public land sites, even when their planning permission requires it. This ‘loophole’ can inflate the price paid for land, and the viability framework should be closed over time. Social housing planning targets should be mandatory, and feed into land pricing. In cases where the developer will not proceed with the development under these rules, there could be a role for a central government body, like the PLHC described above, to acquire the land at a fair price and develop the site.
NEF’s research has recommended that over time, public land should be developed in-house, to avoid having to rely on private developers who have an incentive to avoid delivering affordable housing. We recommend that central government public land is developed by a Public Land and Housing Corporation, or by Housing Associations through public land leases, and that local government public land is developed by councils or again leased to housing associations.
Q5 Why is available data on public land so poor, especially for local government-owned land, and could improving land data transparency help the delivery of social and affordable rented homes?
Until recently, the availability of data on what happens to central government owned land when it is sold was very poor. Despite requests from the Committee of Public Accounts, the National Audit Office, and ourselves at the New Economics Foundation, MHCLG did not collect or publish data on the number of affordable homes built for the first decade of the combined public land sale programmes. The only “annual report” to ever have been published was in 2017. When NEF requested the second annual report in 2018 using a Freedom of Information Request, the information was denied on the grounds that it was due to be published. But in fact, no other annual report has ever been released. A ‘progress report’ was publsihed in May 2019, with associated data, however this release did not monitor the use of public land for affordable housing.
To plug the information gap during these years, NEF has undertaken our own research into the national public land sale. We have analysed planning documents to find out how many affordable and social homes were planned on sites sold under the government programme. MHCLG have now commissioned Ordnance Survey to undertake this research for them. In February 2020, ten years after the programme began, MHCLG published data on affordable and social housing built on public land sites for the first time. Our analysis of this data showed that just 2.6% of the homes built through the programme were for social rent.
Data on land sold at the local authority level is even harder to locate, because there is no public centrally held database of local government-sold land.
Improved data availability is an important step towards greater transparency for the disposal of public land. However, this alone will not deliver more affordable housing within a model designed around i) sale for the maximum price obtainable and ii) sale to private developers primarily. Social housing targets, and greater in-house and Housing Association development, will achieve this.
Q6 In August 2019, you wrote an article for the Guardian in which you imaginelooking back from a notional housing position in 2029where 200,000 social homes per year had been builtfor the last 10 years.
The latest figure for the number of social homes build is that there were 53,044 new build affordable homes completed in 2018-19, the third highest value since 1991-92 and the highest since 2014-15. The figure of 53,044includes all affordable housing and not just social rent, which is only…
What needs to be done to change this current figure into a figure of 200,000 social homes a year?
There is agreement among major housing bodies – Shelter, Crisis, NHF, RTPI and so on – that millions of social rented homes are needed to solve the housing crisis. Two of the three major parties at the last election – although not the winning party – had building 100,000+ social homes/year in their manifestos. To rapidly scale up delivery, the following four policies are crucial: