Home›Blog›Cost breakdown structure: what it is, and what the WBS does not carry
FundamentalsBeginner
Published September 13, 2026 · By Vincent KENNEL
Cost breakdown structure: what it is, and what the WBS does not carry
No general project management standard defines the cost breakdown structure, and three incompatible versions of it circulate. This article establishes which one an organisation is actually practising, and what decides how deep the structure has to go.
In brief
A cost breakdown structure is a hierarchy of cost categories covering all of a project's costs, and a framework shared by its suppliers (ECSS-M-ST-60C, 2008). It answers what kind of expense a cost is; the work breakdown structure answers what that cost is spent on. Some costs have a kind but no work package, and their volume decides how deep the structure goes.
On a large contract programme, a cost structure is often obtained by copying the work breakdown structure and hanging a budget on each work package. That holds until the first cost report, when project management, quality assurance, insurance and corporate charges turn out to have no cell anywhere. A work breakdown structure carries only what progress can be measured on, and those costs cannot be (AACE International, Skills & Knowledge of Cost Engineering, 5th revised edition, 2007). The rule that prevents the problem was written down in 1963.
What a cost breakdown structure is
A cost breakdown structure is a hierarchy of cost categories used to break down the whole cost of a project. ECSS-M-ST-60C (2008) states it in one clause: a set of cost categories covering all the costs of the project, a common framework for every supplier on the same project, and the total cost planned for each work package broken down per category. The European glossary of the time, ECSS-P-001B (Revision B, 2004), had the short form, a hierarchical structure which depicts elements of cost.
Several names cover the same ground and are not interchangeable. The PMBOK Guide (1996 edition) had a chart of accounts, any numbering system used to monitor project costs by category, distinct from its code of accounts, which numbers the elements of the work breakdown structure. American public practice says cost estimating structure, and the NASA Cost Estimating Handbook (NASA CEH v.4.0, 2015) treats that term and ours as the same object. AACE International says code of accounts, a bridge Asbjørn Rolstadås had already made explicit (Project Management Techniques Applied to Industrial Manufacturing, 1995).
No general standard defines it
No general project management standard defines the object. The PMBOK Guide carries the phrase in none of its editions from 1996 to 2021, with one exception, the 3rd edition (2004), where it appears once in a clause on risk simulation, undefined and never picked up. PRINCE2 7 (2023) has none. ISO 21502:2020 has none either, and the absence runs deeper than a word count: its cost management clause estimates, budgets and controls, and attaches cost to elements of work without ever giving cost an axis of its own. ISO 21511:2018 names a cost breakdown structure once, in a list, and leaves it undefined. The definitions come from ECSS, from the APM Glossary (consulted on 19 August 2026), from the IPMA (IPMA Individual Competence Baseline, ICB 4.0, 2015) and from AACE International.
One criterion, and three forms
One criterion decides which version an organisation is actually practising, and it comes from AACE's Skills & Knowledge of Cost Engineering (5th revised edition, 2007): not all costs relate to a specific activity, so overhead and general and administrative costs appear in the cost breakdown and not in the work breakdown, and where costs and actions coincide a control account is created. How much of that a project carries decides which of three forms the term takes.
In the first, the cost structure is the cost side of the work breakdown structure, the hierarchical breakdown of costs when allocated to the work packages of a WBS (Albert Lester, Project Planning and Control, 4th edition, 2003). In the second, it is an axis cutting across the work and applied to each work package, which is ECSS-M-ST-60C (2008) and Rolstadås in 1995, and which USAF PERT Volume III, PERT COST System Description Manual (Advance Copy for AFSC Implementation, December 1963) already required. In the third, it is the larger set that contains the work breakdown structure, obtained by adding every account that carries a budget without being used to measure progress (AACE International, 2007).
None of the three is the definition.
What the work breakdown structure does not answer
A project asks two questions about its money. The work breakdown structure answers what the money is spent on, and so does the control account that sits at the intersection of the work breakdown structure and the organizational breakdown structure (ISO 21511:2018, figure B.4). The cost structure answers what kind of expense that money is. Control accounts are the subject of their own article; here it is enough that they answer the first question and not the second.
Two axes, then, and not two levels. USAF PERT Volume III (December 1963) asked that cost categories be related to work packages so that no distortion of the work breakdown structure is required to segregate the data, and that once established they remain as originally defined for the life of the programme, and it gave the cost-category axis its own input form.
Figure 1: the costs that have a kind and no work package
Some costs have a kind but no work package. ECSS-M-ST-60C (2008) makes one of them a mandatory category in its own right, non-production expenses, alongside direct manpower, internal special facilities, supplies and other direct costs, and subcontracts.
How much sits in that band decides how deep the cost structure needs to go, and the only sizing rule in the literature runs against intuition: when an extensive work breakdown structure exists, there is reduced need for a fully detailed code of accounts (AACE International Recommended Practice No. 20R-98, Revision of 27 January 2003). A work breakdown structure split by discipline makes discipline cost codes redundant. The same document describes the practice without tidying it up: some organisations use the work breakdown structure in place of a code of accounts, others call their code of accounts a work breakdown structure, and many run both. On the 28 real codes AACE gathered and dissected, the work breakdown structure came out as a tertiary characteristic, present in a quarter to a half of the sample, a figure that belongs to AACE and to a sample the document declares confidential in its detail.
Read this way, the three forms stop competing. They are three sizes of the same band. Empty band, and the cost structure collapses onto the work breakdown structure. Band held to one side, and it is a transverse axis. Band counted inside the total budget, and it contains the work breakdown structure. That reading is ours; no source states it in those terms.
What it produces on a contract project
Under contract, the structure is not a matter of internal preference. ECSS-M-ST-60C (2008) puts it on the customer's side: the supplier works from the cost breakdown structure defined by the customer, any extension has to be agreed by that customer, and the supplier identifies its own indirect cost elements on the customer's structure. For the two neighbouring structures the standard defines, it is the supplier who submits and the customer who approves. Of its four project structures, the cost breakdown structure is the only one that comes down from the customer.
What it produces is comparability: one framework shared by every supplier on the project, so costs can be read across work packages and across companies. The standard fixes five main categories, direct manpower, internal special facilities, supplies and other direct costs, subcontracts, and non-production expenses. Margin and profit are not among them, on the ground that they do not reflect an expense borne by the supplier to execute the project.
Two constraints come from AACE rather than from the norm. A code of accounts serves many users but should have one master, and arbitrary changes by an individual user should not be permitted. And it should stay small: large contractors in the sample held between 10 and 15 primary categories for direct costs and as many for indirect, and the highest level of summarisation below total project level should fit on one page. Coding that grows too complex produces short cuts, and short cuts produce inaccuracy.
Where it comes from
Figure 2: the object precedes its name by about seventy-five years
The object is a great deal older than the term.
The oldest complete example we have read is from 1906. George A. Damon published, in The Iron Age of 22 November 1906, the code used on a railway shops contract, already built on two explicitly crossed axes, sections of classification against parts of the work, with a detail classification down to the individual item. It was not a cost tool alone: the code is described as the key placed in each drawing, specification, data sheet, report or letter produced as the work progressed.
Three years later, Halbert P. Gillette and Richard T. Dana published the plate the literature recognises as the first published cost breakdown, in Cost Keeping and Management Engineering (First edition, 1909), and the same book states the separation that still causes the most expensive mistakes: cost keeping was evolved by managing engineers and bookkeeping by merchants, and one of the most common blunders arises from attempts to graft a cost keeping system upon a bookkeeping system. Patrick Weaver (The Origins and History of Work Breakdown Structures, Enlarged edition, 2020) credits that plate and notes that Damon precedes it. Neither is an invention or a first appearance: they are the oldest examples the collection holds. What is striking is the shape of the tree, since the 1909 skeleton and the normative structure of ECSS-M-ST-60C are the same tree ninety-nine years apart.
The doctrine of a separate cost axis was written down in 1963. The term arrived much later: the oldest occurrence we have found is from 1981, in a Booz Allen & Hamilton paper by Anthony C. Leggitt (Cost Data Base Development. A Twelve-Year Perspective, 1981), describing models built in the late 1960s. Roughly seventy-five years separate the object from its name. The European glossary then took the term in and let it go again: ECSS-P-001B (Revision B, 2004) marked it a new definition, ECSS-S-ST-00-01C (Version C, 2012) listed it among the terms deleted, and the glossary in force, ECSS-S-ST-00-01C Rev.1 (2023), has no occurrence of it. Since 2012, the standard that requires a cost breakdown structure has pointed at a glossary that no longer defines it.
Two confusions worth heading off. The first is with the accounting chart of accounts, and the two cannot be merged for a structural reason: project coding tends to be multidimensional, because cost has to be analysed in many different ways, and an accounting chart of accounts is not (AACE International Recommended Practice No. 20R-98, Revision of 27 January 2003). Organisations that extend the general ledger to project use, or the reverse, leave neither side with a good solution. Compatible is not identical: the code of accounts must be homogeneous and permanent over time, and compatible with the cost accounting code, which is technically far less detailed (Dictionnaire de management de projet, AFNOR-AFITEP, 3e édition, 1996).
The second is with an acronym. The abbreviation CBS, which this article has not used, also stands for contract breakdown structure, the hierarchy of contractors, subcontractors and suppliers that shows the supply chain (Mosaic Projects, Breakdown Structures Revisited, online, undated, consulted on 3 September 2026).
In short
The move is not to pick one of the three definitions. It is to list, first, the costs that no work package on the project will ever carry: project management, quality assurance, insurance, corporate charges, everything the programme spends money on that progress cannot be measured against. That list is the band, and its size settles the rest. Close to empty, and the work breakdown structure is already serving as the cost structure. Substantial, and the second axis has to exist, with its depth set by what the work breakdown structure does not already carry rather than by how finely the categories could be cut. The first question is never how to code the work. It is what the coding of the work leaves out.
Stop guessing. See the real impact.
Frequently asked questions
Q.Is a cost breakdown structure the same thing as a control account?
No. A control account sits at the intersection of the work breakdown structure and the organizational breakdown structure, and answers what the money is spent on. The cost structure is an axis of expense categories. The two are often aligned, which is where the confusion starts.
Q.Is there a standard cost breakdown structure to reuse?
There is no general template, and the published codes of accounts are all sector specific. Under an ECSS contract the question does not arise, since the structure comes down from the customer with five mandatory categories.
Q.Where does cost estimating come in?
Afterwards. The cost structure settles which categories an expense falls into, not how much it weighs. Under contract the structure comes first and serves as the frame the estimate is then built on.
References
AACE International - Skills & Knowledge of Cost Engineering - 5th revised edition, with new appendices, 2007
AACE International - AACE International Recommended Practice No. 20R-98 - Project Code of Accounts - Revision of 27 January 2003
AFNOR, AFITEP - Dictionnaire de management de projet français-anglais-espagnol - 3e édition, 1996
Albert Lester - Project Planning and Control - 4th edition, 2003
