The units of production method charges depreciation according to what an asset produced during the period rather than according to how much of the period elapsed. Each unit of output carries the same slice of the asset’s cost, so a machine that ran hard in one year takes a larger charge that year and a machine that sat idle takes almost none.
Every other common method measures time. This one measures use, which is why it is reached for on assets whose wear tracks output rather than the calendar.
The Formula
rate per unit = (cost less salvage value) / total units expected over the asset's life
period charge = rate per unit x units produced during the periodThe numerator is the same amount every method has to recover. Deductions for depreciation may not exceed such amounts as may be necessary to recover the unrecovered cost or other basis less salvage during the remaining useful life of the property, whatever method is used (26 CFR 1.167(b)-0(a)-0)). What varies between methods is the denominator, and here it is a count rather than a span. The amount subtracted at the top is set out in salvage value, and the starting figure it is subtracted from is in fixed assets.
The denominator is an estimate of total lifetime output, so the method carries an estimate the time-based methods do not. Revising it revises every remaining period’s rate.
Why It Counts as Depreciation at All
The cost of a productive facility is one of the costs of the services it renders during its useful economic life, and generally accepted accounting principles require that this cost be spread over the expected useful life of the facility in such a way as to allocate it as equitably as possible to the periods during which services are obtained from the use of the facility. That procedure is depreciation accounting, described as a system of accounting which aims to distribute the cost or other basic value of tangible capital assets, less salvage if any, over the estimated useful life of the unit, which may be a group of assets, in a systematic and rational manner. It is a process of allocation, not of valuation (ARB No. 43, Chapter 9, Section C, paragraph 5).
Read against that sentence, an output-based charge is not an exception to depreciation accounting. It is one way of being systematic and rational, and on an asset that gives up its service potential by running rather than by ageing it is the closest allocation to the services actually obtained. Where the charge lands on the statements is the same as for any other method, and that placement is set out in accumulated depreciation.
The reason the method appeals is the same reason it is fragile. Matching output to charge is exactly what the matching principle asks for, and the match is only as good as the lifetime-output estimate underneath it.
Where the Method Sits in the Tax Rules
The method is often described as simply not permitted for tax. The statute is narrower than that, and the difference decides whether a client can use it.
Section 168, the cost recovery system, does not apply to any property where the taxpayer elects to exclude that property from the section, and, for the first taxable year for which a depreciation deduction would be allowable for the property in the taxpayer’s hands, the property is properly depreciated under the unit-of-production method or any method of depreciation not expressed in a term of years, other than the retirement-replacement-betterment method or similar method (26 U.S.C. 168(f)(1)).
Both conditions have to hold. An election alone does not get the property out of section 168, and using the method without electing does not either (26 U.S.C. 168(f)(1)). What the exclusion leaves behind is depreciation under section 167 instead, where any reasonable and consistently applied method of computing depreciation may be used or continued in use, and where the methods that continue to be acceptable are named as the straight line method, the declining balance method with the rate limited to 150 percent of the applicable straight line rate, and, under appropriate circumstances, the unit of production method (26 CFR 1.167(b)-0(a) and (b)-0)).
Three words in that regulation do the work. Under appropriate circumstances is a condition, not a courtesy, and the burden sits with the filer: it is the responsibility of the taxpayer to establish the reasonableness of the deduction for depreciation claimed (26 CFR 1.167(b)-0(a)-0)).
So the accurate statement is narrower than the common one. The method is not available inside MACRS, which assigns recovery periods in years, and it is available to property properly outside it under the two-limb test above. On a client that never makes the election, the books can still run units of production while the return runs MACRS, and the two figures are reconciled rather than merged.
What It Is Not
Depletion. The arithmetic is nearly identical, basis divided by units, but the subject differs. Depletion recovers the cost of a natural deposit as it is extracted and has a second, rate-based form the depreciation rules have no equivalent for, both of which are covered in depletion.
Straight line depreciation. Straight line divides by periods and produces the same charge each year regardless of activity, which is worked through in straight line depreciation. Units of production produces a charge that can be zero in a year the asset never ran.
An accelerated method. It is neither accelerated nor decelerated by design. The pattern is whatever the output pattern is, so the same method front loads on one asset and back loads on another.
Common Questions
Cost less salvage value, divided by the total units the asset is expected to produce over its life, gives a rate per unit. That rate multiplied by the units produced in a period gives the period’s charge. The cap on the total is statutory rather than conventional: the deductions may not exceed what is necessary to recover the unrecovered cost or other basis less salvage over the remaining useful life (26 CFR 1.167(b)-0(a)-0)).
Whatever the asset’s service is measured in, provided the measure can be estimated for the whole life and counted each period. Miles, machine hours, tons, impressions and cycles are all used. The test is not which unit is most natural but which one the client can actually evidence at period end, because the charge is an assertion about a count.
Where lifetime output cannot be estimated with any confidence, the denominator is a guess and the charge inherits the guess. Where output is roughly steady year over year the method produces nearly the same answer as straight line at more cost to maintain. And where the asset deteriorates with age rather than use, a storage tank or a building, output is the wrong driver whatever the machine’s counter says.
As any depreciation is: a charge against income and an increase in the contra account that sits against the asset. Nothing about the method changes the entry, only the amount, so a firm switching an asset onto it changes a schedule rather than a chart of accounts.
Accountably places trained offshore accountants and tax preparers inside US CPA and EA firms, and a firm can run a real block of its own fixed-asset and depreciation schedule work through a trial task before anything client facing depends on it.
