Have planning agreements increased the supply of affordable housing?

An increasing proportion of affordable housing is now being delivered through Section 106 (S106) planning agreements (where planning permission for private developments is conditional on a contribution to affordable housing). Research at the Universities of Cambridge and Sheffield looked at whether increasing use of S106 was at the expense of other forms of provision. The findings are presented below.

• Government Housing Investment Programme (HIP) data for the period 2000/01 to 2002/03 HIP data (which includes renovations and acquisitions in addition to completions) shows that the number of new affordable homes fell by over a third, from around 45,000 to 29,000.
• The mix of provision has been changing rapidly – in 1999, S106 provision accounted for just under a third of the total, in 2003 it was almost half.
• Planning permissions including S106 agreements are now rising rapidly: even if the amount of housing funded through other mechanisms continues to fall, overall the number of new affordable homes should increase over the next few years.
• It is becoming increasingly difficult for social landlords to acquire sites to develop exclusively for social housing. S106 is thus ever more critical to the overall programme.
• Much of the land that is available for non-S106 sites comes from the public sector. Such sites bought on the open market tend to be small, brownfield sites which are problematic to develop.
• Public subsidy is still very important for S106 provision – with over 75 per cent of all completions requiring some Social Housing Grant (SHG). Grant costs per home appear similar across the two approaches, with contributions from private developers covering the higher costs of land in S106 locations.
• Negotiation and partnership arrangements remain expensive in both time and money but help to make the system more acceptable to communities and to meet government policy on mixed communities. Stakeholders do not expect the proposed introduction of an optional charge to reduce these costs significantly. They are also concerned about possible negative effects on land supply for affordable housing and on mixed communities.

Background

There are currently two main mechanisms for delivering new affordable housing:

• The use of S106 agreements, which deliver affordable homes as part of a planning agreement for a private development (referred to here as 'S106 provision');
• The acquisition of sites by housing associations, which build affordable homes funded through public subsidy (Social Housing Grant - SHG) – the older of the two systems (referred to here as 'non-S106 provision').

The government has been placing increasing emphasis on the planning route. The most recent guidance places the planning system in a pivotal position for meeting the need for affordable housing.

The consultation paper, 'Contributing to sustainable communities: a new approach to planning obligations' (November 2003) proposed a number of changes to the planning system. These included introducing an optional planning charge which applicants may pay instead of the current negotiated route to agreeing planning obligations.

The government has now issued a draft revised circular for consultation with a view to putting the new arrangements in place later in 2005. These will entail a revised Planning Policy Guidance Note 3: Housing (PPG3) and a good practice guide. It has deferred any introduction of optional charges until a decision has been made on the recommendations, from the Barker Review, for a Planning Gain Supplement.

This project asked:

• Has the increased use of S106 agreements been at the expense of a decline in non-S106 sites?
• If so, does this mean that, overall, the number of new affordable homes is going down?

The study examined three specific elements: whether the supply of land for non-S106 provision is declining; whether lack of SHG is constraining development; and whether the process of negotiation and partnership is limiting the capacity of housing associations and others to build more homes. It also looked at how well the current system is operating, what improvements stakeholders feel might be made and, with government policy developing during the project, what their attitudes to the new proposals are.

The overall supply of affordable homes

In the period 2000/01 to 2002/03 HIP data (which includes renovations and acquisitions in addition to completions) shows that the number of new affordable homes fell by over a third, from around 45,000 to 29,000.

Combining Housing Corporation and HIP data to record new build completions only (and exclude renovations and acquisitions) shows a small decline in numbers of around 3 per cent between 2001/02 and 2002/03.

Data on completions from the Office of the Deputy Prime Minister (ODPM) suggests a fall of 19 per cent (2001/02-2002/03), and of 26 per cent since 1999. So although the data sources provide different figures because of different definitions, they all point to a decline in the quantity of new affordable homes in the period prior to 2002/03.

Homes built through S106 agreements are a rapidly increasing proportion of all affordable completions. Between 2000/01 and 2002/03 the proportion built with a planning agreement increased from 30 to 47 per cent of all affordable completions. However, of these the proportion that did not require public subsidy increased from 78 per cent to 82 per cent. These schemes were likely to include a developer subsidy to bring the scheme costs to a level within funding limits and meant that full funding was required.

The completions of non-S106 homes decreased from 21,451 in 2000/01 to 13,949 in 2002/03. Thus both the proportion and the absolute number of homes produced through the older, non-S106 route are declining.

The evidence on affordable homes with planning permission looks more promising. For example, in 2002/03 although only 12,600 affordable homes were completed, 23,727 were granted planning permission.

Land availability and affordable housing

The study included a survey of housing associations undertaking development work. Only 38 per cent stated that S106 sites were directly replacing 'traditional' sites. However, respondents commented that such sites were increasingly difficult and expensive to obtain and to develop.

Survey responses outlined the typical non-S106 site. The most common were small brownfield infill sites and former local authority sites. The case studies also showed that non-S106 sites were sometimes really part of a larger S106 site where a specific area had been identified for social housing. This land was also often owned by the local authority. This suggests that much of the declining quantity of non-S106 housing is being built on surplus public land, often within existing council estates.

Overall, housing associations increasingly see themselves as only able to gain access to land in the areas of greatest housing pressure. Many housing associations stressed that land supply was their main constraint and that, for them, the prime rationale of S106 is that it provides land.

Costs and funding

The evidence from the HIP data confirms that the amount of SHG used for non-S106 developments has fallen quite rapidly and evidence shows that funding has been diverted away to make S106 sites viable.

In 2002/03 just over 2,260 affordable homes were completed through the S106 policy without any public subsidy, 9 per cent of the total. The remainder rely on SHG and subsidy from other sources including the now abolished Local Authority SHG (LASHG), Single Regeneration Budget and Safer Communities Grant. The level of funding depends on tenure and the cost of development. Land costs are a significant element in the total cost of S106 provision and contributions from private developers are important in reducing these costs and bringing total development costs within levels that are within the limits imposed by the Housing Corporation funding regulations.

Nearly 70 per cent of respondents to the housing association survey believed that the growth of S106 meant they were developing in more expensive areas (in terms of land costs) and a similar proportion (68 per cent) stated they were able to develop in areas not normally associated with affordable housing. This can be seen as contributing to the government's policy on promoting mixed communities.

The statistics also show that an increasing proportion of the homes being delivered are shared ownership. For instance in the South East, shared ownership as a proportion of the overall number of affordable dwellings rose from 9 per cent in 1999/2000 to 29 per cent in 2002/2003. This could be evidence, in part, of: the need to limit grant per dwelling in expensive areas; an agreed way of meeting need or giving developers additional comfort with respect to market sales; and a response to government initiatives for key workers.

There are only minor differences in the amount of SHG required on S106 sites compared with other sites because contributions from private developers bring S106 site costs down to funding limits in line with non-S106 sites. S106 sites funded through the use of public subsidy will thus produce a similar number of homes as the same level of funding on non-S106 sites, despite being located in areas of generally higher land costs. Were the rapid increases in planning permissions to translate into development it seems likely that limitations on SHG would become more binding.

Stakeholder views

The current system

All parties now appear more comfortable with the S106 approach than in 2000 when the same team undertook earlier work. In particular, there is a sense that, now the framework is in place, it is possible to negotiate 'proper' prices for land. Interviews with developers, housing associations and local authorities confirmed that the S106 policy was working 'reasonably well'. However, problems remain over the length of negotiations, with the start of the process to occupancy taking up to four years for some S106 provision. Evidence of the extreme variability across individual sites makes it difficult to see how the time taken can be reduced in all cases.

The majority of S106 homes are delivered on the same site as those homes available on the open market. This integration of social and market housing contributes directly to the mixed communities policy of the government. However, there is still work to do in persuading developers and local residents that social housing tenants do not directly equate to 'problem tenants'.

There were some examples of very good practice. This included having an extremely clear policy; involving all parties, especially housing associations, in the negotiations at an early stage; and replacing site-specific S106 agreements with a single standard version to increase clarity and transparency.

Housing associations stated that considerable resources were required find non-S106 sites. They have to ensure that they are well known in the land market, so that when a site is available land agents and others know that they are likely to be interested. In addition the growing emphasis on funding through SHG and the abolition of LASHG makes it less clear that funds will actually be forthcoming to enable the housing association to go ahead with the purchase of the site. On the other hand, buying land after grant has been allocated, rather than on the open market, can often be significantly more expensive.

Government proposals

Opinions on the government's proposals were generally negative. Housing associations and local authorities felt that introducing an optional charge would make it even more difficult to gain access to land. In particular, it was expected to reduce the extent to which affordable and market homes were built on the same site and so increase the need to purchase land elsewhere. Respondents felt that the overall cost would be higher and that it would also undermine policy to encourage mixed communities.

Developers were broadly in favour of introducing an optional charge, but not overwhelmingly so. Some were concerned that it would alter their capacity to negotiate an appropriate planning brief. All parties said that such a system would not remove the need for negotiations, since they take place over all types of planning gain as well as about the layout and design of the site.

Conclusions

The researchers conclude that S106 policy is becoming increasingly important in the delivery of affordable homes. Homes delivered in this way are becoming ever more important because housing associations are finding it increasingly difficult to secure sites to provide affordable homes through the methods previously used. Developers of affordable housing depend more and more on getting associated S106 provision for affordable homes through increased market provision.

Land availability remains the key issue. SHG has not so far been a major problem in most areas outside London, although it is likely to become more of a constraint if current levels of planning permissions translate into development.

Stakeholders are largely happy with the operation of the policy and expressed concerns over the changes proposed by the government. The changing planning context - with greater certainty in local development plans and greater consistency between regional policies and local policies - is likely to make the existing negotiating system work better. Paradoxically, the introduction of an optional charge could actually frustrate this by requiring far too much detail before the event.

About the project

The study was based on: an analysis of Housing Investment Programme and Housing Corporation data; a questionnaire survey of 380 housing associations undertaking development (response rate of 68 per cent); site-specific case studies involving face-to-face interviews with local authority and housing association representatives; and telephone interviews with developers and other stakeholders.

The researchers were Professor Christine Whitehead, Sarah Monk, Diane Lister and Christina Short (University of Cambridge) and Professor Tony Crook and Steven Rowley (University of Sheffield).