JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

The Predetermined Overhead Rate

In the last guide you traced direct materials and direct labor straight onto a job's cost sheet. But the rent, the supervisor's salary, the factory electricity — the overhead — refuses to be traced and refuses to wait. This guide shows the clever workaround accountants reach for: an estimated rate, set before the year even begins, that lets a job be costed the moment it leaves the floor.

The overhead you cannot trace, and cannot wait for

In the previous guide, [[job-order-costing|job-order costing]] worked beautifully for two of the three manufacturing costs. When a worker drew steel from the storeroom for Job 412, you wrote that steel straight onto the job's cost sheet; when she clocked six hours building it, those wages went on too. Direct materials and direct labor are [[direct-vs-indirect-costs|direct costs]] — you can watch them flow into a specific job and tag them as they go. The third bucket of [[manufacturing-costs|manufacturing costs]], factory overhead, is the troublemaker.

Overhead is everything the factory consumes that you *cannot* point at a single job: the rent on the building, the depreciation on the machines, the factory manager's salary, the electricity that lights every bench at once, the oil and rags and glue too cheap to bother tracking. None of it belongs to Job 412 in particular; all of it was needed for Job 412 to happen. So a fair share of overhead genuinely is part of what that job cost to make — yet there is no invoice you can staple to its cost sheet that says "Job 412's portion of the rent." That is the first half of the problem: overhead cannot be traced, only *allocated*.

Now the second half, and it is the one people miss. Even if you were willing to wait and split the rent fairly, *the actual overhead total is not known until the year is over*. The December electric bill, the unexpected machine repair in August, the bonus paid to the supervisor — you learn the true annual overhead figure only on the last day, with the books closed. But Job 412 ships in March, and the customer wants a price *now*. You cannot quote a price, bill a job, or value the inventory on the shelf if a third of every product's cost is a number you will not know for nine more months. Real time collides with year-end arithmetic.

Estimate first, divide once, apply all year

The accountant's escape from this trap is disarmingly simple: *stop using actual overhead for individual jobs, and use a rate set in advance instead.* Before the year begins, the company makes two budget estimates — total overhead it expects to incur, and total activity it expects to run, such as direct labor hours. Divide one by the other and you get the [[predetermined-overhead-rate|predetermined overhead rate]]: a dollars-of-overhead-per-unit-of-activity figure, frozen on day one and used unchanged for the whole year.

PREDETERMINED OVERHEAD RATE

             budgeted total overhead for the year
  rate  =  ----------------------------------------
            budgeted total activity for the year

  Example:
     budgeted overhead    = 600,000
     budgeted labor hours =  40,000 hrs
     rate = 600,000 / 40,000 = 15 per direct labor hour

  APPLYING IT to a job:
     Job 412 used 80 direct labor hours
     overhead applied = 80 hrs  x  15  =  1,200
Both numbers on top and bottom are budgets, set before the year starts. The rate, 15 per hour, is then applied to every job by multiplying it by the hours that job actually used.

Watch what this buys you. The moment Job 412 finishes and you read off its cost sheet that it consumed 80 direct labor hours, you multiply by the rate and add 1,200 of overhead — instantly, in March, with no waiting for the December bills. This step is called [[overhead-application|applying overhead]], and it is the whole point of the rate: it converts an untraceable, year-end-only number into something you can attach to a job in real time. The materials and labor were *traced*; the overhead is *applied*. Two different verbs for two different kinds of cost.

Choosing what the rate divides by

The denominator of the rate has a name worth taking seriously: the [[allocation-base|allocation base]]. It is the measure of activity by which you spread overhead across jobs, and the job that grabs more of the base grabs more of the overhead. In the example above the base was direct labor hours, so a job that soaked up 80 hours collected twice the overhead of one that used 40. The base is not a formality; it decides which products look expensive and which look cheap, so choosing it badly quietly distorts every cost you report.

The guiding rule is that the base should be a genuine [[cost-driver|cost driver]] — something that actually *causes* overhead to rise and fall, so that jobs which drive more overhead receive more of it. Historically, direct labor hours were the default, and for good reason: in a labor-heavy workshop, the more hands on a job, the more supervision, the more lit and heated floor space, the more payroll-clerk time — overhead really did march in step with labor. Where labor still dominates, it remains a sensible, easily measured choice.

But factories changed. In an automated plant, robots and computer-controlled machines do the heavy lifting, and the overhead — machine depreciation, power, maintenance, programming — now rises and falls with how long the *machines* run, not how long *people* do. There, machine hours are the honest driver, and clinging to direct labor would charge a hand-finished job far too much overhead while letting a machine-intensive one off far too lightly. The lesson is not that one base is right; it is that the base must mirror what genuinely drives the overhead in *this* factory. When no single measure fits well, a plant may even run several rates — one per department — rather than force everything through one number.

When the estimate misses: under- and over-applied

Here is the honest catch that follows from using a forecast. Because the rate came from a budget, the overhead you *apply* to jobs all year is built on estimates, while the overhead you *actually incur* is the real bills as they arrive. There is no reason on earth these two should match to the dollar — your forecast of overhead was a guess, your forecast of labor hours was a guess, and reality rarely honours either. At year-end you compare the two totals, and the gap has a name: [[under-and-over-applied-overhead|under- or over-applied overhead]].

The two words are easy to mix up, so anchor them to direction. If you *applied less* overhead to jobs than the factory actually rang up, you under-applied — the jobs were charged too little, and some real cost is still sitting unassigned. If you *applied more* than was actually incurred, you over-applied — the jobs were charged too much. A quick numeric feel: suppose the rate sent 590,000 of overhead onto jobs over the year, but the actual bills totalled 600,000. You applied 10,000 too little; overhead is under-applied by 10,000. Flip it — actual bills of 580,000 against 590,000 applied — and you over-applied by 10,000.

Cleaning it up at year-end

That leftover gap cannot just be ignored, because the financial statements must eventually reflect the *real* cost the factory incurred, not the estimate. So at year-end the company makes one adjusting step to dispose of the under- or over-applied balance. The common, simplest treatment is to clear the whole gap into cost of goods sold. Under-applied overhead means costs were undercharged during the year, so the fix *raises* cost of goods sold (lowering reported profit); over-applied means costs were overcharged, so the fix *lowers* cost of goods sold (raising profit). After this entry, the overhead account is back to zero and the income statement carries the actual cost.

  1. Before the year: estimate budgeted overhead and budgeted activity; divide to set the predetermined rate.
  2. During the year: as each job uses the base, apply overhead = rate x base used, and add it to the job's cost.
  3. Alongside it: record actual overhead bills as they arrive — these will not equal what you applied.
  4. At year-end: compare applied with actual to find the under- or over-applied gap, then clear it (usually into cost of goods sold).

One honest qualification keeps you from over-trusting the result. Dumping the entire gap into cost of goods sold is an approximation, justified only when the gap is small relative to the year's costs — which it usually is, and the [[under-and-over-applied-overhead|materiality]] judgment lets you take the shortcut. When the gap is large, simply shoving it into cost of goods sold would misstate profit, so the firm instead spreads it proportionally across the jobs still in inventory and those already sold, putting each portion where the over- or under-costing actually landed. Same goal either way: end the year with reported product costs that reflect what the factory truly spent.

Why this scaffolding matters

Step back and notice what the predetermined rate really is: a deliberate, disciplined estimate that buys you *timeliness* at a known, small cost in precision — a cost you then mop up honestly at year-end. It exists because two of the things you already believe pull in opposite directions. The [[accounting-period-assumption|accounting-period assumption]] insists you chop the endless life of a business into reportable years, yet a single job lives and dies inside that year and demands its cost *immediately*. The predetermined rate is the bridge between the impatient job and the patient calendar.

Carry one caution onward. The rate is only ever as good as the two budgets behind it and the base you chose, and it spreads overhead by an *average* — so it can never tell you which product is truly the costly one inside a single pooled rate. A job that is light on labor hours but ravenous for an expensive, automated machine will be undercharged by a labor-based rate, and you may price it too low without ever noticing. That blind spot is exactly the door through which [[overhead-application|activity-based costing]] walks in the next guide, slicing overhead by many drivers instead of one. For now, hold the core move firmly: estimate the rate before the year, apply it in real time, and reconcile the inevitable miss at the end.