The cost entity is the part of the Programme and Project Data Standard (PPDS) most closely tied to government finance. Its 21 attributes track spend across three time horizons (monthly, financial year and whole life) and split costs into resource and capital spending.
For organisations whose finance data already follows government finance data standards, many of these attributes should map to existing fields. For others, the cost entity is likely to be one of the larger pieces of work on the road to compliance.
This post covers the three things teams most often miss, how the attributes fit together and where the common gaps are.
The departmental baseline and HMT approved estimate record the most recent approved cost profile. If no approval exists yet, the standard asks you to return nil.
A nil tells anyone reading the data that there is no approved figure. A zero could be read as an approved cost of nothing, which distorts any analysis that combines figures across projects.
The standard records four cost types, and each one has a separate resource (RDEL) and capital (CDEL) attribute. Many teams track a single total for each cost type. That works for internal monitoring, but cross-government analysis needs the split.
The departmental baseline year records the financial year that counts as Year Zero (Y0) for the baseline, with dates beginning 1 April. Without it, nobody can tell which year the baseline figures relate to.
The standard also asks for both the departmental baseline and the HMT approved estimate in nominal terms. That means the cash amounts in the years they fall, not adjusted for inflation.
The standard monitors costs at three levels: monthly, by financial year and over the whole life of the project. Alongside these sit the approved reference points, stage funding and the cost types. All values are entered in £m.
| Group | Attributes | What they show |
|---|---|---|
| Approved reference points | Departmental baseline, HMT approved estimate, departmental baseline year | The approved cost profiles that actual and forecast figures are compared against |
| Whole life | Actual cost to date, forecast outturn, variance | The project's full financial position from start to finish |
| Financial year | Actual cost to date, forecast outturn, current budget, variance | The project's position within the current financial year |
| Month | Actual cost | Spend in the given month, giving early warning before variances build up |
| Stage funding | Agreed funding for current stage, agreed funding for next stage | Funding agreed for each stage of the project |
| Cost types (RDEL and CDEL) | One-off new costs, recurring new costs, recurring old costs, non-government spend | What the money is spent on, split by resource and capital |
Both variance attributes compare what has been funded with what has been spent: over the project so far for whole life, and within the year for financial year.
Eight attributes are split into four RDEL and CDEL pairs. Getting the split wrong weakens any analysis that combines cost data across projects, so it is worth being clear on the difference.
RDEL (Resource Departmental Expenditure Limit) covers day-to-day spending that is used up rather than creating a lasting asset. In a project, this is typically staff, contractors, consultancy and running costs.
CDEL (Capital Departmental Expenditure Limit) covers investment that creates or buys assets with long-term value. In a project, this is typically the cost of building, buying or developing what the project delivers, such as infrastructure, systems or equipment.
The four cost types are:
One-off new costs: the investment in change, meaning the one-off project costs incurred until project closure
Recurring new costs: ongoing spending needed to maintain or service the asset the project delivers
Recurring old costs: spending that was ongoing before the project began, recorded only if included in the business case
Non-government spend: spending met from sources other than HMT public finances, such as supplier contributions
Not every project will have all four types, and not every type will have both an RDEL and a CDEL element. Record only what applies.
The departmental baseline and the HMT approved estimate are the reference points for every actual and forecast figure in the entity.
Departmental baseline: the cost and benefit profile approved by your department nearest to the current reporting period
HMT approved estimate: the most recent cost and benefit profile approved by HM Treasury nearest to the current reporting period
Both are reported in nominal terms, and both return nil if no approval exists. In practice, the departmental baseline will usually come from an approved business case or an internal investment board decision.
The two stage funding attributes record the funding agreed for the current stage and the next stage of the project. They are most useful for projects funded in stages, where each approval covers part of the work rather than the full cost upfront.
The standard says cost data should be created with reference to existing finance data standards. Six cost attributes list related Common Chart of Accounts (CCoA) and finance data standard IDs, which are held in the NOVA Data Dictionary.
These related IDs are project identifiers, such as project ID, project name and parent and child IDs. So for organisations whose finance systems already follow these standards, the main work is likely to be linking finance data to the correct project ID in the format the standard requires.
For organisations whose finance systems do not yet follow them, the cost entity gives a clear reference for what project-level finance data should contain.
When we review project cost data against the standard, the same three gaps come up most often:
No RDEL/CDEL split: costs are tracked as a single total, which works internally but not for cross-government analysis
Monthly actuals not separated: teams track cumulative or financial year actuals but not the month's figure on its own
No baseline year: the baseline figure exists, but the year it relates to is not recorded, so historical comparison is impossible
If any of these gaps apply to your cost data, start there. Split your cost types into RDEL and CDEL first, because most other cost reporting depends on it. Then record the baseline year, check that missing approvals show as nil and separate out monthly actuals.
The standard is in a 12-month internal government trial. Version 2 is due in early 2027, and organisations are expected to comply before 2030. That gives teams time to plan, but these gaps are worth addressing now.
For the full list of cost attributes, with definitions, formats and related data standard IDs, see the full standard overview.
Next in this series, we cover the final three entities: benefits (15 attributes), people (8) and project features (11). This includes the Green Book benefit classifications and the GMPP status field.