2.16 In March 2008, the Department published an exercise benchmarking the capital cost of new build PFI projects to grant-funded housing association projects. The analysis was used to develop an acceptable cost range for projects and those PFI projects falling above the range were examined further. As a result, some projects were dropped from the programme. The Department concluded that the exercise demonstrated the cost of PFI projects could be acceptable.
2.17 We extended this benchmarking exercise to account for the different sizes of housing. Using data from over 1,000 housing association projects from 2008-09, we established the average capital cost for different types of new build housing (Figure 8). Owing to the method of data collection we had to make some broad assumptions, for example, about the floor area of properties with different numbers of bedrooms. We found that there was huge variability in the cost of both PFI and housing association projects which makes it difficult to reach firm conclusions. The Department told us that this variability is inherent because of geographical location, site conditions and different specifications.
2.18 The Department now plans to develop its benchmarking further which will be useful in identifying outlying projects for further scrutiny. Around a third of the housing units across all the rounds are new build and the percentage has increased over time. In round six about seventy per cent of the units will be new build, these have been included in our analysis.
2.19 At a programme level the Department has not compared the cost-effectiveness of PFI relative to the other investment options available for bringing homes up to the Decent Homes standard. Although the Decent Homes Programme is coming to an end, PFI projects are likely to continue to include some refurbishment so there is still a need to benchmark these costs. The Department now collects planned cost data for refurbishment work and is assessing the level of PFI grant per unit.
2.20 Some local authorities and private sector providers felt that the cost of procuring PFI housing projects was too high relative to other procurement routes. We have not been able to collect any data on procurement costs for non-PFI so we have been unable to make direct comparisons.
Figure 8 | ||||||
| Average cost of housing association projects | Level above which PFI projects should be scrutinized further1 | Minimum cost of PFI projects | Maximum cost of PFI projects | Percentage of PFI projects requiring further scrutiny | Percentage of PFI homes requiring further scrutiny |
| £000s | £000s | £000s | £000s | % | % |
General needs housing |
|
|
|
|
|
|
Flat |
|
|
|
|
|
|
1 bed | 109 | 149 | 67 | 155 | 10 | 2 |
2 bed | 112 | 151 | 87 | 188 | 25 | 11 |
3 bed | 175 | 233 | 124 | 221 | 0 | 0 |
House |
|
|
|
|
|
|
2 bed | 113 | 143 | 101 | 206 | 54 | 32 |
3 bed | 126 | 164 | 116 | 236 | 38 | 30 |
4 bed | 151 | 205 | 148 | 276 | 42 | 46 |
Extra care housing |
|
|
|
|
|
|
Flat |
|
|
|
|
|
|
1 bed | 118 | 155 | 86 | 199 | 14 | 12 |
2 bed | 139 | 183 | 100 | 223 | 10 | 2 |
|
|
|
|
|
|
|
NOTE 1 This is defined as the housing association average plus one standard deviation. Standard deviation is a statistical measure of spread. Source: National Audit Office analysis of data from the Department for Communities and Local Government | ||||||