Sale of public land Contents

1Performance against the land disposal targets

1.On the basis of a Report by the Comptroller and Auditor General, we took evidence from the Ministry of Housing, Communities and Local Government (the Department), the Cabinet Office, HM Treasury (the Treasury) and Homes England on the government’s approach to land disposals.

2.The government holds a significant amount of land. In 2016–17, central government owned land and property with an estimated value of £179 billion. The government has a policy to sell assets where it considers they no longer serve a public purpose, or that purpose can be more efficiently realised by transferring it to the private sector. There are benefits to selling public sector land, such as releasing land for new homes and reinvesting proceeds in the government estate, but there are also risks. If public land is sold without first understanding whether it is needed in the future, this can damage the delivery of important public services. Government owned land is a finite resource and needs careful management to ensure its sale delivers maximum benefit to the public.1

3.The government has two main targets for its land disposals:

(a) Land for new homes target: Under its Public Land for Housing Programme, the government aims to “increase housing supply by releasing surplus public sector land for at least 160,000 homes” in England between 2015 and 2020.

(b) Proceeds target: The government plans to “deliver £5 billion of receipts between 2015 and 2020 through the release of surplus public sector land and property across the UK”.2

4.Responsibilities for delivering public land disposals are spread across a wide range of bodies. The Cabinet Office, in addition to leading on progress toward the proceeds target, also sets the strategy for managing the entire government estate. The Ministry of Housing, Communities & Local Government (the Department) is responsible for the land for housing target while Homes England, a non-departmental public body of the Department, manages land sales on its behalf. The Treasury sets departmental budgets, taking into account proceeds from expected land sales. Each department is responsible for pursuing their own land and proceeds targets as well as ensuring individual sales represent value for money.3

5.This is the third time the Committee has reported on the Department’s Public Land for Housing Programme. In 2015 we reported on the first programme to release enough land for 100,000 estimated homes by 2015. We concluded that the Department could not demonstrate the success of the land disposal programme in addressing the housing shortage or achieving value for money. Furthermore, it did not collect basic information necessary to oversee the programme effectively and, where it did collect programme-level data, there were omissions and inconsistencies.4 In 2016 we recalled the Department to give further evidence on the progress of the second programme to dispose of enough land by 2020 for 16,000 estimated homes. We concluded that the Ministry for Housing, Communities and Local Government and individual departments had made progress, since our previous report, on managing the disposal programme, such as implementing guidance and monitoring arrangements. However, we found that progress towards the second target had been slow, with the future success of the programme requiring significantly accelerated land sales in the remaining years to 2020.5 This is the first time we have reported on the Cabinet Office’s progress against the £5 billion disposal target.

Land for new homes target

6.By the end of December 2018, departments had disposed of land with the capacity to build an estimated 38,000 homes, representing less than 25% of the overall programme commitment. The Department acknowledged that the target to dispose of enough land for 160,000 potential homes would not be achieved by the March 2020 deadline. Instead, it told us that there was a “high probability” of it releasing enough land for 69,000 homes by March 2020, although it acknowledged that this was an “upper estimate”.6 To deliver this latest estimate of 69,000 homes, departments would need to release additional land for 31,000 homes in the final 12 months of the programme. We were concerned that the Department’s expectations were not realistic, especially given the programmes previous track record of only releasing enough land for 38,000 homes in the first 33 months of the programme.7 The Department accepted that it had a lot to do in this financial year to achieve this.8 Even if this is achieved, by 2020 the programme will have failed to meet the original 160,000 target by a margin of 91,000 homes. The Department confirmed that it was still committed to the 160,000 homes target, but that this would now be delivered over a much longer timescale than originally envisaged. The Department forecasts land for a further 95,000 homes will be released by 2025.9

7.We questioned the Department on why the programme has performed poorly compared to its target. The Department told us that there were several barriers which prevent land being released. It said that in a third to a half of cases, the land was still being used for operational purposes by departments, for example, an old hospital site cannot be sold because the replacement hospital has not yet been completed.10 We were concerned that this highlighted that the Department had made little progress in addressing this barrier. In our 2016 report on progress with the disposal of public land for new homes, we found that 50% of sites identified for future disposal were ‘high risk’, with many still being used to deliver public services. We recommended that departments needed to identify capacity over their individual targets to create sufficient contingency to avoid feeling pressurised to include sites for disposal if they think it is unlikely that homes will be built on these sites.11 The Department told us that delays in the planning system and environment constraints also held up land disposals.12

Proceeds target

8.The government told us that it expected to meet its target of raising £5 billion in proceeds from land and property disposals by March 2020. Between April 2015 and March 2018, the government had generated £2.48 billion in proceeds.13 In February 2019, Network Rail’s sale of the railway arches contributed a further £1.46 billion, representing nearly 30% of the total target. This sale means that the government only needs to raise a further £1.06 billion to reach its £5 billion target. The sale of the railway arches was not originally earmarked to contribute towards the £5 billion commitment as it was not included in the Department for Transport’s proceeds target that was set out in Spending Review 2015. Without the inclusion of the railway arches, the government will need to raise an additional £2.52 billion before March 2020. The National Audit Office found that raising this additional money will be challenging because, at March 2018, nearly all departments were significantly underperforming against their individual proceeds targets. For example, the Department for Health and Social Care had generated proceeds of £752 million against a target of £1.95 billion.14

9.We challenged the Treasury on whether it would still have been able to meet its target for the programme if the unexpected sale of the railway arches had not taken place. The Treasury told us that the sale met the necessary criteria for inclusion against the target. The Cabinet Office asserted that a lot can happen in the six years since the target was agreed, and that while some planned sales had not happened, some things had happened that hadn’t been planned.15

10.The National Audit Office found that between April 2015 and March 2018, 176 sites were sold for £1 or less with the Department and Homes England being responsible for selling 160 of these. Homes England told us that there were typically three reasons for selling a site for £1. The first and most common reason related to land which had a negative value because a significant amount of remediation work was required, for example former industrial sites. The second relates to “small parcels of land” which do not have any development value and are passed to local authorities or local community trusts, with the third reason covering sites previously sold via a long-term conditional lease. These leases contractually require Homes England to transfer the freehold for £1 after a set time period has elapsed.16

Setting targets

11.In 2015, the National Audit Office found that the target to sell land for 100,000 homes by 2015 under the first programme was not supported by documentation or economic evidence on what could realistically be achieved.17 When challenged on how the 160,000 target was created, the Department told us that it was a “top-down” target set by the centre of government. The Department accepted that the target was set without a sufficient evidence base and was ultimately over-optimistic, but asserted that it had learned lessons about the need to ensure targets were better grounded in evidence and detailed analysis.18 The Department told us that while government performed some analysis centrally to assess how many homes could realistically be delivered, this analysis was not undertaken by individual departments.19

12.The Department and the Treasury told us that, unlike the land for homes target, detailed analysis supported the £5 billion proceeds target. The Treasury outlined that this work resulted in a recommended target of £4.5 billion, which was then increased by Ministers by around 10% to reach £5 billion. According to the Treasury, adding 10% “stretch” to a target was a “normal discipline” in the private sector. The Department told us that setting very stretching targets put more “energy through the system” and encouraged departments to deliver more. The Treasury accepted that this could become a problem if the degree of stretch was so unrealistic that departments have no confidence in it, but that this was not the case for the proceeds target. The Department told us that while the 160,000 target was not achievable, it was not clear that government could have achieved as much as it had through land sales to date if it had set a “more cautious” target. The Department added that Minsters made “no apology for setting very stretching targets”.20

Number of homes built

13.The government has committed to deliver an additional 300,000 homes a year by the mid-2020s.21 In our September 2015 report on the first land disposals programme, which ran from 2011 to 2015, we found that the Department had not intended to monitor the number of homes being built on land released under the programme. We found that the Department viewed the success of the programme purely in terms of whether enough land had been released to meet the target of 100,000 homes, without considering whether new homes were actually being built. We recommended that the Department should apply a broader test of value for money which includes the collection data on how many homes have been built as a result of the programme.22 The Department accepted this recommendation and in 2017 it commissioned the Ordnance Survey to start collecting this data.23 Government has released land with a capacity for 147,000 homes to date - 109,000 homes under the first programme, and 38,000 estimated homes under the second programme.24 The Department estimated that by March 2018, 40,500 homes had been “brought to market” on land released under both programmes, which is less than 30% of the total potential homes that could have been delivered.25

14.In our 2015 report we criticised the Department for adopting a very wide interpretation of what it could count towards achieving the land for housing target. We found that the Department was including against the target land that had been sold before the programme had even begun, in some cases land sold as far back as 1997. It also counted land which was not sold for housing, for example when British Waterways converted to charitable status.26 We were concerned that in reporting the number of homes built as a result of this programme, the Department had once again adopted a wide interpretation of what should count. The Department uses the term “brought to market” to calculate the number of homes that have been built. This figure includes 4,279 homes that pre-existed, rather than being newly built. 12,500 of the 40,500 homes which the Department reported as “brought to the market” also related to land which was sold before the first programme started in 2011. We were concerned that the inclusion of pre-existing houses artificially inflated the number of homes that had been delivered by the programme. The Department told us that the 4,279 were homes that the public sector previously had use for, but no longer needed and therefore had been released into private sale. It told us that these homes were still adding to the available housing stock for ordinary private buyers and renters across the country.27

Affordable homes

15.The Department recognised that the private sector was not incentivised to build affordable homes without significant investment from central and local government.28 We received written evidence from the National Housing Federation which estimated that in order to address the current housing crisis, 42% of all new homes built in England should be affordable, but that in 2017 only 23% of new housing stock was deemed affordable. It also estimated that 20% of the homes due to be built on public sector land are likely to be affordable and as few as 6% of new homes were likely to be social rented housing.29 We asked the Department about whether the land for housing programme could be used to prioritise the delivery of affordable homes. The Department told us that it does not to seek to determine the type of housing to be built on land sold, as this was the responsibility of the local authorities. It told us that, instead, the Department aimed to support the delivery of affordable homes at a national level through interventions such as removing local authority borrowing restrictions. The Department said that it took this approach to avoid adding complexity to the local authority planning system or attempting to centrally determine what was needed locally.30

16.The Department told us that there were exceptions to its decision not to intervene in the type or mix of homes that should be provided by new developments. For example, it told us that the NHS has committed to prioritising up to 3,000 new homes for its staff and was piloting this at five sites. Homes England also told us that it had several tools available which could allow greater influence over the type of housing being delivered, including extensively using building leases as opposed to straight freehold disposals.31 Under a building lease, Homes England retains the freehold during the length of the development and can attach conditions to the development such as the number of affordable homes that should be built. Once the homes are completed, the freehold then reverts to the homeowners.32 Homes England told us that it was probably one of the larger users of building leases across government with some local authorities also using this method.33 We asked the Department why it did not set a specific target for the number of affordable homes to be delivered on land released under the programme. The Department told us that while this was something that it could do, this would be a policy decision and could create a risk that affordable homes are concentrated in specific locations which might not create “great communities”.34

Transparency and data collection

17.In our 2016 report on the progress of the land disposal programme, we found that the Department had not yet published an annual report setting out its progress towards the 160,000 potential homes target.35 The Department published its first annual report on the programme in February 2017, with a commitment to produce these reports each year.36 The Department failed in this commitment, only publishing a second report more than two years later in May 2019.37 The Department apologised for the time lag between publications and said that this failure was due to the high turnover of Housing Ministers. It said each new Housing Minister wanted to understand the programme and discuss performance with individual departments before reporting to Parliament.38 We challenged the Department on whether it would now commit to publishing an annual report. The Department told us that it would now regularly publish data on programme performance, including the number of sites sold and homes built. This would take the form of a statistical release which the Department will aim to publish every six months. It could not confirm whether an annual report will be published alongside the data, although Ministers had previously made the commitment to do so.39

18.The Cabinet Office publishes annual Transparency Reports which outline progress against the £5 billion proceeds target. These reports provide details on the proceeds generated, sale location, method of disposal, size of the land sold and whether there are any arrangements in place that allow the government to benefit from future increases in the value for the land sold.40 However, the Transparency Reports do not include information on the identity of the buyers, nor does the Cabinet Office perform any data analysis at an aggregate level. For example, the Cabinet Office does not look to see the types of buyers attracted to certain government properties, whether departments are selling land to the same people or companies, or what the future use of the land will be.41

19.In our 2015 report, we concluded that the Department had adopted a very wide interpretation of what it counted towards achieving its target which overstated the performance of the programme. For example, we found that the Department had counted 15,000 homes from land sold before the programme started and 10,000 homes from land which had simply moved outside the public sector with the privatisation of Royal Mail.42 We were concerned that the Department continued to use loose definitions which risked undermining the transparency of the programme’s performance. For example, in its definition of what constitutes a “home” the Department included “essential local worker accommodation” such as homes for health service staff and “communal accommodation” such as care homes and sheltered housing.43 We questioned the Department about whether the creation of a single room, which could be the case for health service staff accommodation, would count towards the number of homes built. The Department confirmed that for any type of housing to count towards the programme it needs to be a “dwelling”, which is determined by specific rules about the number of bedrooms in relation to share cooking and bathroom facilities, and that a single room would not count towards this.44 We similarly asked the Department whether care homes should be included under the definition of a new home, given that they will not benefit families requiring new housing. The Department recognised that this would be a concern only if every new home built was a care homes. The Department believed that care homes are only likely to represent a minority of the homes built. The Department told us that it does not have data on the number of homes built by category, such as care homes.45

20.The Department does not outline what type of housing is included in its definition of “new affordable homes”.46 The Department told us that this was a broad definition and included different types of tenure such as social rent and shared ownership. The Department committed to elaborating on this definition in the next version of the Programme Handbook, but could not guarantee that future annual reports would include a breakdown of the types of affordable homes built.47

21.In our 2015 and 2016 reports on progress with the land disposal programme, we recommended that the Department should collect data on the number of affordable homes and key worker homes that the programme delivered.48 The Department has still not collected and published this data. The Department told us that it had now let a contract with Ordnance Survey in spring 2017 that introduced a system for counting the number of new homes built under this and the previous programmes. It explained that the process matched postcodes to all the sites that have been sold. The Department said that this processed was hampered by the poor departmental record-keeping under the previous programme. This system provides data on whether the site is complete and whether it was sold under freehold or leasehold arrangement. It does not differentiate between affordable and key worker homes. The Department told us that to gather affordable homes data it had to go back into the planning system for each site and manually review individual planning applications.49 It told us that it was collecting this information and that it hoped to publish data on affordable homes with the next data release.50 This data collection methodology will only be able to provide data on the “planned” affordable homes because it extracts data from planning applications. Homes England told us that that a common problem with planning applications is that what was expected to be built and what was actually built can be quite different.51

22.The Department told us that one of the reasons why it had been so difficult to collect data is because the necessary data collection processes were not put in place at the outset of the programme.52 Given that the purpose of the programme was to lead to new homes being built, we asked the Cabinet Office whether a system for counting new homes should have been considered at the start. The Cabinet Office told us that it would be nice to have all the systems in place at the beginning, but the government “struggles with this all over”.53

1 C&AG’s Report, para 1

2 C&AG’s Report, para 2

3 C&AG’s Report, para 2

4 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

5 HC Committee of Public Accounts, Progress with the disposal of public land for new homes, Twenty-Second Report of Session 2016–17, HC 634, November 2016

6 Qq 71, 105

7 C&AG’s Report, para 2.5

8 Q 71

9 Qq 67, 105

10 Q 66

11 HC Committee of Public Accounts, Progress with the disposal of public land for new homes, Twenty-Second Report of Session 2016–17, HC 634, November 2016

12 Q 66

13 C&AG’s Report, Investigation into the government’s land disposal strategy and programmes, Session 2017–18, HC 2138, 2 May 2019

14 Q 116, C&AG’s Report, Figure 10

15 Qq 115, 116

16 Qq 118–119, C&AG’s Report, para 3.5

17 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

18 Qq 61–63, 69

19 Q 66

20 Qq 63, 72, 106, 110–111

21 Q 135, C&AG’s Report, 2.1

22 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

24 Q 71, HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

25 Q 78

26 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

27 Qq 77–78

28 Q 129

30 Qq 81–82

31 Qq 81, 83

32 Qq 83, 86, 88

33 Q 90

34 Q 128

35 C&AG’s Report, Disposal of public land for new homes: a progress report, Session 2016–17, HC 510, 12 July 2016

36 Ministry of Housing, Communities and Local Government, Public Land for Housing Programme 2015–20, Annual Report, February 2017

37 Ministry of Housing, Communities and Local Government, Public Land for Housing Programme 2015–20, Progress Report, May 2019

38 Qq 93–94

39 Qq 93, 148

41 C&AG’s Report, para 3.9

42 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015

44 Qq 137–138

45 Qq 136–139

47 Qq 140–144

48 HC Committee of Public Accounts, Disposal of public land for new homes, Second Report of Session 2015–16, HC 289, September 2015, HC Committee of Public Accounts, Progress with the disposal of public land for new homes, Twenty-Second Report of Session 2016–17, HC 634, November 2016

49 Q 97

50 Q 95

51 Q 100

52 Q 98

53 Q 104

Published: 24 July 2019