What is the Headline Social Housing Cost CPU?
The Headline Social Housing Cost per Unit (CPU) is a standardised metric used by local authorities and housing associations to express the average total cost of delivering a new socialrent property. It aggregates land acquisition, construction, professional fees, financing, and postcompletion expenses into a single figure, allowing comparison across projects, regions and time periods.
CPU is expressed in pounds sterling per square metre (or per square foot in some jurisdictions) and is usually presented as an allinone headline number that already includes contingency allowances and inflation adjustments.
Why It Matters
Understanding the headline CPU is essential for several reasons:
- Budget Planning: Local councils allocate capital based on expected CPU values, ensuring that housing programmes remain financially viable.
- Affordability Targets: The CPU directly influences the rent level that can be set while still meeting socialhousing affordability criteria.
- Investment Decisions: Private and public investors use CPU to assess risk and potential returns on mixeduse projects.
- Policy Evaluation: Shifts in CPU over time signal the impact of regulation, market conditions, and construction technology on housing delivery.
Key Cost Components
Although the headline figure is presented as a single number, it comprises several distinct elements.
| Component | Description | Typical Share of Total CPU |
|---|---|---|
| Land Acquisition | Purchase price of the site, including any compulsory purchase costs. | 2030% |
| Construction Cost | Materials, labour, plant, and subcontractor fees. | 4050% |
| Professional Fees | Architects, engineers, quantity surveyors, and project management. | 1012% |
| Financing | Interest on borrowings, fees, and lenders risk premium. | 58% |
| Contingency & Inflation | Allowance for unforeseen events and price escalation. | 57% |
| PostCompletion Costs | Commissioning, handover, and initial warranty reserves. | 24% |
These percentages are indicative; actual splits vary according to site constraints, design ambition, and local market dynamics.
How to Calculate the Headline CPU
Below is a stepbystep guide that housing providers commonly follow.
- Determine the Net Internal Floor Area (NIFA): Measure the usable floor area of the dwelling, excluding walls, shafts and service spaces.
- Estimate Individual Cost Elements: Use costbook data (e.g., BCIS, RICS) to assign unit rates to each component and multiply by the relevant quantities.
- Apply Financing Charges: Calculate the commercial loan interest over the construction period and any arrangement fees.
- Add Contingency: Apply a percentage (usually 510%) to the subtotal to cover unexpected cost overruns.
- Inflation Adjustment: If the project timetable extends beyond a year, index the construction cost element by the appropriate construction inflation rate.
- Total Cost: Sum all adjusted elements.
- Headline CPU: Divide the total cost by the NIFA. The result is the cost per square metre (or foot).
For example, a 75m twobedroom flat with a total estimated cost of 150,000 would have a headline CPU of 2,000/m.
Policy Implications and Recent Trends
Recent government guidance has highlighted the importance of reducing the headline CPU to increase the volume of affordable homes. Key policy levers include:
- Land Supply Initiatives: Accelerating brownfield site releases cuts landcost percentages.
- Modular Construction: Prefabricated elements can reduce construction cost by up to 15% and shorten build time, affecting financing charges.
- Standardised Design Frameworks: Reusing approved plans limits architect fees and planning delays.
- Funding Reform: Lower interest rates on socialhousing loans directly shrink the financing component of CPU.
Data from the UKs Ministry of Housing, Communities & Local Government show that the average headline CPU fell from 2,400/m in 2020 to 2,150/m in early 2024, driven largely by modular pilots and targeted landvalue capture.
Nevertheless, regional disparities remain pronounced. Londons CPU still exceeds 3,200/m, whereas many northern authority areas report figures under 1,800/m. Policy responses therefore need to be regionally tailored.
