Cost-Benefit Analysis in Accounting: Relevant Costs, Incremental Analysis, and Decision Framework

What Makes a Cost Relevant

A cost is relevant to a decision if — and only if — it meets both of these conditions: (1) it is a future cost (historical/sunk costs are never relevant), and (2) it differs between the alternatives being evaluated (costs that are identical regardless of which option is chosen do not affect the decision). Apply these two tests before including any cost in a decision analysis.

This sounds simple but is surprisingly easy to get wrong in practice. Allocated overhead rates, historical purchase prices, past research expenditures, and depreciation on existing equipment all frequently appear in accounting reports alongside truly relevant costs. The decision-maker's job is to strip them out. The make-or-buy decision guide shows this in detail for one of the most common decision types.

Sunk Costs: Always Irrelevant

Sunk costs are amounts already committed and impossible to recover regardless of the decision. The classic examples: the original purchase price of equipment being replaced (already spent, unchanged by whether you keep or replace), past research and development expenditures (the money is gone whether you launch the product or cancel it), and setup costs already incurred for a project. None of these affect future cash flows, so none are relevant.

The sunk cost fallacy — allowing past expenditures to drive future decisions — is one of the most persistent cognitive errors in business. Continuing a failing project to "justify" money already spent makes the situation worse, not better. The correct framing is always: given where we are now, what are the future costs and benefits of continuing versus stopping?

Opportunity Costs: Always Include Them

Opportunity costs are the benefits foregone by choosing one alternative over another. They are not recorded in accounting systems — they exist only in decision analysis — but they are just as real as out-of-pocket costs. Ignoring them systematically biases decisions toward options that consume scarce resources, because the cost of those resources is understated.

Classic examples: a business owner's time (if they manage the business themselves, the foregone salary from alternative employment is an opportunity cost); use of owned floor space (the foregone rental income is an opportunity cost); and production capacity constraints (when operating at full capacity, making one product forecloses making another — the contribution margin of the foregone product is the opportunity cost).

Fixed Costs: When They Are and Are Not Relevant

Fixed costs are relevant when they are avoidable — when they will actually be eliminated if a particular alternative is chosen. A supervisor's salary that disappears if a product line is dropped is an avoidable fixed cost and is relevant. Factory rent that continues regardless of production decisions is a non-avoidable fixed cost and is irrelevant to the make-or-buy decision.

Allocated fixed overhead — overhead assigned to products or departments through cost allocation — is almost never relevant, because the total fixed overhead remains the same; allocation just redistributes it. Removing a product from the portfolio does not eliminate the overhead unless specific, identifiable fixed costs genuinely go away.

Application 1: Special Order Decisions

A special order is a one-time request from a customer at a price below normal selling price, often for a large volume. The relevant question: does the special order generate a positive contribution margin, accounting for any incremental fixed costs required to fulfil it?

The special order is profitable at $4.50 per unit. The $6.00 allocated fixed overhead is irrelevant — it will be incurred whether or not the order is accepted. A naive total cost comparison ($22 price vs $22.50 total cost per unit) would incorrectly recommend rejection. Capacity must also be confirmed: if the factory is at full capacity, the opportunity cost of displaced regular-price production changes the analysis entirely.

Application 2: Keep or Drop a Product Line

When a product line appears to be unprofitable, the relevant question is not "does it show a loss?" but "will total company profit increase or decrease if we drop it?" The answer depends on which costs are truly avoidable.

A product line showing a $50,000 accounting loss may have a $120,000 contribution margin. If only $60,000 of fixed costs are avoidable (the rest stay regardless), dropping the line eliminates $60,000 of fixed costs but loses $120,000 of contribution margin — a net $60,000 decline in profit. Keep the line. The contribution margin framework from the contribution margin guide is essential here.

Application 3: Capital Investment Decisions

Capital investment cost-benefit analysis compares the present value of future incremental cash flows to the upfront investment required. Relevant cash flows: incremental revenues, avoidable costs, salvage values, and tax effects. Irrelevant: sunk costs already spent on research or feasibility studies, and allocated costs that will not change.

The time value of money — discounting future cash flows back to present value — is critical here. A $500,000 benefit in Year 5 is worth less than a $500,000 benefit in Year 1. The net present value (NPV) method handles this correctly. See the complete framework in the NPV guide.

The Decision Framework

1. Define the alternatives clearly — including the status quo
2. Identify every cost and benefit that differs between alternatives (relevant)
3. Exclude sunk costs — already spent, irretrievable
4. Exclude non-avoidable allocated fixed costs — they stay regardless
5. Include opportunity costs — forgone benefits of the rejected alternative
6. Quantify incremental cash flows and, if long-term, discount to present value
7. Identify qualitative factors that could override the quantitative conclusion
8. Recommend the alternative with the highest net relevant benefit