A lighting pole in the desert is not an asset

Plant a lighting pole in the middle of the desert. It is a real object. It has a cost, a supplier, a useful life and a replacement value, and every one of those can be entered into a register. What it does not have is a service. Nobody is lit. No journey is safer. There is no benefit to measure, and therefore nothing to manage it against.

Move the same pole to the edge of a carriageway and it acquires one instantly. It now contributes to a road being usable after dark, which is a service somebody receives and somebody funds.

That difference is not a philosophical point. It is the whole of the componentisation question, and getting it the wrong way round is why so many asset registers are simultaneously enormous and useless.

The exercise that has no end

The standard approach is to decompose. Take the estate, break it into systems, break the systems into equipment, break the equipment into parts, and record each part with its own identifier, cost and life.

On a small estate this terminates. On a large one it does not, and the reason is that nothing in the method says where to stop. A pump is a component. Its motor is a component. The bearing in the motor is a component. The bolts holding the bearing housing are components. Each step is defensible on the same grounds as the last, so the exercise recedes indefinitely and the register grows without becoming more useful.

Two failures follow reliably. The first is that spares get recorded as components, because a spare is also a discrete object with a cost and a part number, and nothing in a decomposition method distinguishes a thing you hold in a store from a thing installed in an asset. The second is that the exercise never finishes, so the register is permanently in a state of partial completion and never becomes the authoritative source it was built to be.

Both are symptoms of the same missing decision. Decomposition is a technique. It is not a criterion, and applying a technique without a criterion is how a programme spends three years tagging.

What the accounting standard actually says

Owners frequently believe the depth is prescribed by accounting requirements. It is not, and the standard is unusually direct about this.

There is a separate requirement about depreciation, and the two are routinely conflated. IPSAS 45 requires each part of an item that is significant in relation to the total cost of that item to be depreciated separately. That is a real obligation, and it does drive some componentisation. But it is triggered by significance and by differing consumption patterns, not by physical separability, and it says nothing at all about parts that are individually insignificant.

The standard also offers relief in the other direction. It permits an entity to aggregate individually insignificant items and apply the recognition criteria to the aggregate value, naming library books, computer peripherals and small items of equipment as its examples.

Put those together and the position is clear. Nothing in the accounting framework compels an entity to record a bolt. The framework hands the entity a judgement and expects it to be made. An organisation that has decomposed to the last fastener has not complied with something. It has declined to make a judgement and substituted exhaustiveness for it, which is the more expensive option and the less defensible one.

The road as the unit

Return to the pole. If the unit of account is the pole, the register will contain a very large number of poles, each with a life and a value, and no line in it will correspond to anything a citizen uses or a budget holder funds.

If the unit is the road, the register contains a road. The carriageway, the footway, the drainage, the lighting, the signage, the barriers, the planting and the rest of the right of way are its parts, and each is recorded at the depth its own life and significance justify. The lighting circuit has a different life from the pavement, so it is componentised. The individual pole does not have a different life from the other poles, so it is not.

What is the unit of account

Four parts are componentised because their lives differ and their cost is significant. Two are not. An even breakdown would not be more rigorous, only more expensive.

The test that does the work here is service. An asset is a thing from which service potential or economic benefit is expected to flow, and that test is applied to the road rather than to the pole because the road is the level at which the service exists. It is also the level at which a condition rating means something, at which a renewal decision is taken, and at which a budget is set.

Why this is not the low-value-items argument

There is a neighbouring argument about desks and monitors, which says the accounting standard permits aggregating individually insignificant items and that tagging them individually is effort spent on something too small to move the accounts. That argument is sound, and it does not apply here. Poles are not desks, and running the numbers is what shows why.

Neither of those is a rounding error. A class worth 3% of the estate sits above any materiality threshold an auditor would set, and a recurring charge four times the furniture's is not something to wave away. So a reader who has accepted the desks argument and expects it to carry across to street lighting should stop, because it does not.

The poles have to be recognised. The question was never whether, it is at what level, and the answer follows from service rather than from size. They are recognised inside the road, as a component of it, with the lighting depreciated over its own life because that life differs from the pavement's. What they are not is 400,000 separate lines in a register, each with an identifier, a cost, a condition rating and a depreciation schedule.

That distinction is worth stating precisely because it is where the two arguments part company. Materiality decides whether something is worth recording separately. The unit of account decides what the thing is. Conflating them produces the error in both directions: entities that tag a 200 monitor because it is an object, and entities that leave 600m of street lighting out of a register because individually a pole is cheap.

What the counting actually costs

The effort is the part that never appears in the business case for the exercise, and it is calculable.

Nineteen person-years buys a register line for every column. What it does not buy is a different answer to any question the authority actually faces, because the renewal decision is taken for a length of road, the budget is set for a network, and the condition that matters is the condition of the lighting circuit rather than of column 214,806.

What sets the depth: criticality

Once the unit is settled, depth becomes answerable, and the answer is not uniform across the estate. Some assets warrant decomposition to component level and most do not, and the discriminator is criticality: what happens to the service if this part fails.

This is not an unusual position. The national asset management manual issued by Saudi Arabia's Expenditure and Projects Efficiency Authority assigns the public entity responsibility for managing the process of establishing the asset hierarchy, and states that this includes asset criticality. Hierarchy and criticality are given as one task rather than two, which is the correct construction: the shape of the hierarchy is a consequence of what matters, not a neutral description arrived at first.

A criticality-led hierarchy is uneven by design, and that unevenness is what makes it usable. The switchgear that takes out a hospital wing is decomposed. The identical switchgear in an unoccupied store is not. Anybody who has tried to defend an uneven register to an auditor knows the objection this attracts, and the answer to it is that a uniform register is not more rigorous, only more expensive.

What sets the depth: the payment mechanism

The second discriminator is sharper, and it is the one most likely to be underestimated during a conversion exercise.

The same manual states that performance-based contracting is the preferred contracting option wherever possible, and describes bundled contracts built on service output specifications, with performance measurement used to calculate the incentives and penalties embedded in the payment mechanism.

That preference has a consequence nobody costs at the time it is adopted. A payment mechanism that pays against measured performance can only measure what has been identified. If lighting availability is a payment line, then the lighting has to exist in the register as something whose availability can be established. If it does not, either the payment line cannot be operated or it will be operated against a number the parties do not agree on, which is the same thing arriving later and with lawyers.

This gives the cleanest available rule for depth, and it is a rule an engineer and an accountant can both apply.

The ten-year supplier

The payment argument gets stronger the longer the contract runs, and long is where this class of contract is heading.

A supplier engaged for a decade against performance measures has to price a risk it cannot inspect its way out of. It will be paid or penalised according to the condition and availability of assets it did not build, using a register it did not compile. If that register is silent on which components exist and what state they are in, the supplier has two rational options. It can price the uncertainty, which the owner pays for in the tender, or it can price optimistically and recover the difference through claims, which the owner pays for later and with more friction.

Neither of those is bad faith. They are the only two responses available to a bidder asked to accept a measured obligation over an unmeasured asset base. The register is what converts that from a gamble into a priced position, and its depth needs to reach exactly as far as the measures do.

A note on the relationship

There is a governance condition attached to all of this, and it is worth naming because the technical work does not survive without it.

A procurement culture built around repeated purchase of small, well-understood works carries an assumption that the supplier's interest runs opposite to the buyer's. Where the works are simple and the supplier is easily replaced, that assumption is cheap and roughly harmless. It does not survive the move to a ten-year performance contract, where the owner is dependent on a party it cannot quickly change and the information asymmetry runs in both directions. A supplier who expects to be treated as an adversary prices for it, and the price is paid whether or not the expectation was correct.

The criticism here is not ours to originate, because the same manual that prefers performance-based contracting also requires the governance framework to declare an engagement policy, and names integrity, transparency, fairness and working in partnership as its examples. An organisation adopting the contracting model while retaining the adversarial posture has taken half the document.

Three questions

What is the asset? Not what are the objects. What is the thing that delivers a service and against which a condition, a budget and a renewal decision make sense. If the answer is a pole, the hierarchy is wrong before any componentisation has begun.

Which components will be measured? By the payment mechanism, by the criticality assessment, or by a statutory obligation. That set is the scope. It is usually far smaller than the set of components that physically exist, and saying so out loud is what ends the exercise.

Where do spares live? If the answer is the asset register, the register has already lost the distinction between what is installed and what is held, and no amount of further tagging will recover it.

Sources. IPSAS 45, Property, Plant, and Equipment, effective 1 January 2025, replacing IPSAS 17: paragraph 8, which states that the standard does not prescribe the unit of measure for recognition and that judgement is required in applying the recognition criteria to an entity's specific circumstances, and which permits aggregation of individually insignificant items, naming library books, computer peripherals and small items of equipment; and paragraph 41, requiring each part of an item that is significant in relation to the total cost of that item to be depreciated separately. Expenditure and Projects Efficiency Authority, Kingdom of Saudi Arabia, National Manual for Assets and Facilities Management: volume 2 chapter 2, Asset Register, document EOM-ZA0-PR-000004-AR, revision 000, 29 March 2020, for the public entity's responsibility for managing the process of establishing the hierarchy including asset criticality; and volume 9 chapter 2, Asset and Facilities Management Contracting Strategies Procedures, document EOM-KD0-PR-000001-AR, revision 000, 11 March 2020, for performance-based contracting as the preferred option wherever possible, for bundled contracts built on service output specifications with performance measurement calculating the incentives and penalties in the payment mechanism, and for the requirement that the governance framework declare an engagement policy naming integrity, transparency, fairness and working in partnership. The cited manual volumes are published in Arabic. English quotations are translated by Projects Advisors; the Arabic original governs. The service test applied to the road rather than the pole follows the IPSAS definition of an asset as a resource from which future economic benefits or service potential are expected to flow to the entity.