You just told your operations manager her department "lost money" last quarter — and she pushed back, correctly, that she sets no prices, approves no discounts, and never sees a dollar of revenue. She controls a budget, not a business. The moment your company grows past the point where you review every expense yourself, this mismatch shows up everywhere: people judged on numbers they cannot move, and nobody clearly accountable for the numbers that matter. Responsibility accounting is the fix. It divides your company into centers, hands each manager exactly one financial target that matches their actual authority, and measures them on that and nothing else.
Responsibility Accounting in One Paragraph
A responsibility center is any part of your business whose manager is accountable for costs, revenues, or invested capital — a department, a store location, a product line, a regional division. Responsibility accounting is the system of measuring each center's results against its budget or target and holding its manager to the outcome.
The whole idea rests on one principle: authority must match accountability. Judge a manager on costs they approve, and you get careful spending. Judge them on corporate overhead allocated from headquarters that they never approved, and you get resentment plus gaming. Textbooks describe segments and responsibility centers as two sides of the same coin — segments organize who does the work, responsibility centers organize who answers for the money — but for a small or mid-size business, the practical version is simpler: every dollar of cost and revenue should roll up to exactly one person who can actually influence it.
Centers come in three main kinds, in order of how much authority the manager holds: cost centers, profit centers, and investment centers. (You will also see two minor types mentioned in textbooks: revenue centers, judged on sales alone, like a reservations team; and discretionary cost centers, such as HR or accounting, where spending has no clean relationship to output. Everything below applies to them with small adjustments.)
The Three Centers, Side by Side
| Cost center | Profit center | Investment center | |
|---|---|---|---|
| Manager controls | Costs only | Revenues and costs | Revenues, costs, and invested capital |
| Judged on | Budget vs. actual spending | Segment profit and margin | ROI and residual income |
| Typical examples | Maintenance, support desk, warehouse crew | A store location, a product line, a regional sales office | A subsidiary, a division with its own equipment budget |
| The question it answers | Did we spend wisely? | Did we make money? | Did we earn enough on the money tied up? |
Picking the wrong type is the most common mistake in the whole system, and every section below shows what "wrong" looks like in practice.
Cost Centers: Judge the Spending, Not the Sales
A cost center is a unit whose manager is responsible only for costs — the maintenance crew, the customer support desk, the warehouse team. Nobody expects these units to generate revenue, so their scorecard is a simple budget-vs.-actual report: each expense line shows what was budgeted, what was actually incurred, and the variance.
Suppose your maintenance team budgeted $2,000 for supplies in December but incurred $2,900. The variance alone is not a verdict; it is the start of a conversation. Maybe a storm burst two pipes and the overtime was unavoidable — in which case the manager did the right thing and the budget was simply wrong about the weather. Or maybe supplies are being reordered without checking the stockroom. A cost-center report does not answer that question by itself. It tells you exactly where to ask it.
The trap to avoid is loading a cost center's report with costs its manager never approved. Corporate rent allocations, headquarters salaries, depreciation on a building chosen years ago — these are real costs, but putting them on a supervisor's scorecard teaches them that the numbers are rigged. The standard solution is to split every report into controllable and non-controllable lines, and to evaluate the manager only on the controllable ones. If you remember one rule from this article, make it this: never hold someone accountable for a line they cannot change.
Profit Centers: Revenue Minus Cost, One Bottom Line
A profit center manager controls both sides of the ledger — what comes in and what goes out — and answers for the difference. The classic examples are a second store location, a product line with its own pricing, or a regional office that quotes its own jobs. The report is a mini income statement: revenue, minus the center's direct costs, equals segment profit, often with a margin percentage so you can compare centers of different sizes.
The key word is "controls." A store manager who cannot set prices, choose suppliers, or run local promotions is not running a profit center; they are running a cost center wearing a costume, and judging them on profit will produce exactly the argument from this article's opening paragraph. Before you label a unit a profit center, check that its manager genuinely decides pricing or sales mix on the revenue side and staffing or purchasing on the cost side. If headquarters makes those calls, keep the unit a cost center and keep the profit target for yourself.
One more thing profit centers force you to decide: transfer pricing. When your manufacturing unit "sells" parts to your installation unit, the internal price moves profit from one scorecard to the other. Any consistent method works — market price, full cost, cost plus a markup — but pick one deliberately and write it down, because whichever you choose, one manager will eventually argue it is unfair.
Investment Centers: When the Manager Also Controls the Capital
An investment center goes one step further: the manager controls revenues, costs, and the capital invested in the unit — equipment purchases, inventory levels, facility expansions. Think of a subsidiary president, a division head with a capital budget, or the partner running a branch office that buys its own trucks.
Profit alone cannot evaluate these managers, because profit ignores how much capital produced it. A division earning $200,000 on $500,000 of equipment is a star; the same $200,000 earned on $4 million of equipment is a warning. Investment centers need return-based measures, and there are two that matter.
Return on investment (ROI) is segment profit divided by the investment base — profit per dollar tied up. It can be split into two drivers: margin (profit divided by sales) times turnover (sales divided by operating assets). That split is the diagnostic: it tells you whether a weak ROI comes from thin margins or from assets sitting idle.
One caution before you compute anything: define your investment base once and stick with it. Gross book value, net book value, average operating assets, even replacement cost — accountants defend all of them, and switching definitions mid-year makes every comparison meaningless. Write your choice into the policy and use it consistently.
Why ROI Alone Makes Managers Reject Good Projects
ROI has a well-known dark side: it makes managers with a high current ROI reject projects that would help the company. Walk through the math once and you will never forget it.
Imagine your Eastern division earns $260,000 on $1,000,000 of operating assets — a 26% ROI. A new delivery van costs $100,000 and would add $15,000 a year in profit, a 15% return. Your company's cost of capital is 10%, so this project clearly creates value: 15% beats 10%.
But look at it from the division manager's chair. Accept the van and the division earns $275,000 on $1,100,000 of assets — ROI falls to 25%. If her bonus depends on ROI, she just paid for the company's gain with her own scorecard. So she rejects a profitable project, and she is behaving rationally given the rules you wrote. Every ROI-only system contains this incentive; the higher a division's ROI climbs, the more good projects it turns down.
The mirror-image problem hits struggling divisions: a manager sitting on a 4% ROI will accept almost any project above 4%, even one below the company's cost of capital, because anything lifts the average. ROI congratulates both decisions. Neither serves the company.
Residual Income Fixes the Incentive
Residual income (RI) solves this by measuring profit in dollars above a required return, instead of as a percentage. The formula:
RI = operating income − (required rate of return × operating assets)
The required rate is usually the company's cost of capital or a hurdle rate set by ownership — the same rate for every center, which is what aligns each manager with the whole company.
Run the delivery-van example again with a 10% required return. Before the van, the Eastern division's residual income is $260,000 − (10% × $1,000,000) = $160,000. After accepting the van: $275,000 − (10% × $1,100,000) = $165,000. Residual income goes up by $5,000, so the manager now wants the project — exactly matching the company's interest. And the struggling division's below-cost-of-capital project would show negative residual income, so its manager rejects it. One measure, both incentives fixed.
Residual income has its own weakness: because it is a dollar amount, big divisions almost always show bigger numbers than small ones, so it compares poorly across different-sized units. The standard answer is to use both — ROI for comparing efficiency across centers, residual income for project decisions and bonuses. (If you encounter the term "economic value added" or EVA, it is residual income dressed up with accounting adjustments; the idea is identical.)
Setting This Up in Your Own Books
You do not need enterprise software to run responsibility accounting. You need a chart of accounts that tags every transaction with its center, and a monthly one-page report per center showing budget, actual, and variance on the lines that center's manager controls.
In a plain-text ledger, cost-center tagging falls out of the account hierarchy naturally. Give each center its own branch under expenses, and the center's report is just the balance of its subtree:
2026-01-05 * "Staples" "Printer paper - support desk"
Expenses:Support-Desk:Office-Supplies 42.18 USD
Assets:Checking 42.18 USD
2026-01-12 * "Power Co" "January electricity - warehouse"
Expenses:Warehouse:Utilities 318.44 USD
Assets:Checking 318.44 USDRevenue accounts get the same treatment for profit centers (Income:Portland-Store:Sales), and a balance-sheet tag or separate inventory and equipment accounts per division give an investment center its asset base. A dashboard that groups balances by account prefix then renders each center's report with no extra bookkeeping — if you can see every subtree balance at a glance, the monthly review takes minutes. The Fava dashboard does exactly this kind of hierarchical breakdown, and the setup is covered in the docs.
Three practical rules make the system stick. First, assign every account to exactly one center — orphans and double-counting destroy trust in the reports faster than anything else. Second, keep allocated headquarters costs on a separate, clearly labeled line that everyone knows is informational, not evaluative. Third, freeze the budget before the period starts; a target that gets renegotiated every time it bites is not a target.
Mistakes That Draw the Wrong Conclusions
Watch for these five when you review your first few rounds of center reports:
- Judging a cost center on profit. If the manager touches no revenue lever, a profit target is fiction. Score the budget variance, plus service quality if you can measure it.
- Acting on allocated costs as if they were controllable. Allocations are fine for full-cost visibility, but the moment a manager's rating moves because headquarters hired more lawyers, the system loses credibility.
- Redefining the investment base to flatter a result. Net book value makes old, depreciated divisions look brilliant and punishes new investment. Whatever base you choose, hold it steady across periods and centers.
- Moving the required rate mid-stream. Changing the residual-income hurdle after projects were accepted rewrites history. Set it annually, alongside the budget.
- Promoting every team to a profit center. Autonomy you do not actually grant — over pricing, suppliers, hiring — turns the "profit" label into a morale problem. Match the label to the authority, not to the org chart's aspirations.
Start With One Cost Center Report This Month
You do not need to reorganize the company to get value from responsibility accounting. Pick the one department whose spending you question most often, build its budget-vs.-actual report with controllable lines only, and sit down with its manager to walk through the variances. That single conversation — about numbers they can actually move — usually pays for the whole exercise. Add the next center when the first one runs without drama, promote units to profit or investment status only as you hand over real authority, and let ROI and residual income split the job of judging capital decisions fairly.
