Frame constrained choices and bounded economic claims

Harbor Components' controller has one uninterrupted review block before a lender-reporting cutoff. A supplied reporting-priority scale assigns 95 points to acquisition modeling, 90 to a covenant review, 75 to a vendor control exception, and 40 to board-deck formatting. The 95-point task cannot be completed because required evidence has not arrived.

Here, a covenant review checks compliance with a lender agreement; a vendor control exception is a documented break from the purchasing process. You do not need accounting expertise to choose. Treat all four as competing uses of the same block and focus first on which uses are actually feasible.

Before reading further, choose the task you would perform and write one sentence explaining what your choice costs. Keep that sentence. The lesson will return to it after feasibility, opportunity cost, and claim boundaries are visible.

Build the feasible set before choosing

Scarcity is the condition that prevents Harbor from completing every use of the review block at once. It does not imply that Harbor is poor or that a product is unavailable in a market. The constrained resource is specific: one controller's uninterrupted time before a stated cutoff.

A reviewable choice frame records:

1. Decision maker: the controller, subject to Harbor's governance and escalation rules.
2. Feasible alternatives: tasks whose evidence, systems, time, and authority make completion possible before the cutoff.
3. Binding constraint: one review block.
4. Objective: the supplied reporting-priority criterion, not an unstated preference.
5. Information date: the moment the queue is evaluated.
6. Revision trigger: new evidence, a changed deadline, delegation, or an escalation that changes the feasible set.

An acquisition model whose required diligence file is unavailable is not a current alternative merely because its score is attractive. It may become feasible later, or management may change the constraint by obtaining evidence or assigning resources. Until then, including it in the current ranking would compare unlike sets.

Complete the scarcity and choice checks before moving on. If your answer relies on Harbor being wealthy or poor, or ranks an option before establishing feasibility, reopen the frame.

Name the trade-off and opportunity cost

A trade-off names the outcome sacrificed when the scarce resource is moved from one use to another. It is conditional on the current boundary. Automation, delegation, or a later deadline may allow more of both outcomes, but a possible future improvement does not erase today's constraint.

Opportunity cost narrows the sacrifice to the value of the next-best feasible alternative. In Harbor's verified queue, the 90-point covenant review is selected. The unavailable 95-point acquisition model is removed. The 75-point vendor control exception becomes the next-best feasible use of the block, so the opportunity cost is 75 points on the supplied scale.

Do not add every rejected score. The controller could not have completed all other tasks with the same block. Do not turn the points into dollars, either: the fictional scale ranks tasks for one supplied objective and does not claim a monetary valuation.

Now revisit your opening sentence. A complete revision identifies the chosen task, the next-best feasible task displaced, the constrained resource, and the value basis. If it names the unavailable model or sums all rejected tasks, use the opportunity-cost remediation before proceeding.

Compare a change at the margin

A second decision asks whether Northstar Services should add one review to an existing control process. The baseline process already has benefits and costs. The question is not whether the entire process was worthwhile or how much it has cost since inception. It is what changes with one more review.

The verified example raises total expected benefit from $45,000 to $55,000 and total expected cost from $18,000 to $24,500. The increment therefore contributes $10,000 of expected benefit and $6,500 of expected cost, leaving $3,500 of net marginal benefit under the supplied one-year assumptions.

Northstar previously paid $20,000 to design the workflow. Because that amount is nonrefundable and unchanged under both current alternatives, it contributes zero to the incremental difference. It does not disappear from accounting, impairment analysis, governance, or process review. The question determines its decision relevance; it does not rewrite history.

Before adopting the positive net result, challenge the expected-loss method, double counting, reviewer capacity, supplier effects, legal constraints, and uncertainty. Marginal analysis is a disciplined comparison, not a guarantee that its inputs are correct or a license to ignore nonfinancial duties.

Complete the marginal-analysis, marginal-benefit, marginal-cost, and sunk-cost items. A correct number without an explanation of the baseline and unchanged amounts is incomplete evidence of the objective.

Treat incentives as mechanisms to investigate

Alder Software pays a bonus when quarterly revenue exceeds a target, and the supplied schedule shows more contracts signed near quarter end. The bonus changes the payoff from reaching the target. That is a plausible incentive channel.

It is not proof that managers manipulated recognition, caused the timing pattern, or acted with improper intent. Seasonality, customer procurement, contract renewal cycles, legitimate selling effort, and other changes remain rival explanations. Recognition also depends on contract facts and applicable accounting criteria, not signing date alone.

Use the incentive to design procedures: inspect contract terms, approvals, override activity, communications, recognition evidence, comparison periods, customer mix, and control operation. Then preserve contrary evidence. The incentive item tests whether you can move from mechanism to investigation without turning risk assessment into a verdict.

Audit the model before reporting its output

An economic model deliberately simplifies. Cedar Transit's capacity model uses demand, staffing, and response-time assumptions to compare feasible plans. Its reproducible output is a model implication. Reproducibility shows what those inputs and rules produce; it does not prove that the assumptions describe the world or that one plan caused an observed result.

Audit seven elements:

• the question and unit of analysis;
• actors, choices, and constraints;
• variables included and omitted;
• the proposed mechanism;
• assumptions and calibration;
• comparative prediction and sensitivity; and
• evidence and domain that would support or limit use.

The right response is not to dismiss every simplified model. A model can isolate a mechanism or organize a decision. The discipline is to label its output and retain the assumptions that give the output meaning.

Separate positive evidence from normative judgment

Positive economics concerns what is, was, or would occur under stated conditions. A positive claim can be descriptive, causal, or predictive; those forms require different evidence. “Positive” does not mean favorable or true.

Normative economics evaluates an outcome against a value, right, objective, or distributional criterion. “The fee should be eliminated because access matters more than revenue” names a criterion but leaves affected parties, constraints, consequences, alternatives, and authority to be developed.

A responsible recommendation exposes both layers. Evidence can test whether a fee affects access and revenue. It cannot, by itself, decide how those outcomes should be weighted. A decision record should state the value bridge, identify rights and non-negotiable constraints, compare feasible alternatives, name who has authority, and preserve uncertainty.

Cedar's example shows a conditional recommendation: the model ranks one plan highest on the board's declared criterion under stated assumptions. The board may use that result conditionally while requiring sensitivity and pilot evidence. The recommendation is structurally reviewable even though no realized causal effect has been observed.

Exit check

Return to the task and cost you chose before the lesson. Copy that opening prediction, revise it if feasibility or opportunity cost changed your answer, and then write a six-sentence decision note about that packet:

1. identify the constrained resource and feasible set;
2. name the selected use and opportunity cost;
3. state one marginal comparison or incentive mechanism;
4. label one observation or model implication correctly;
5. expose the normative criterion and decision authority; and
6. request the next evidence that would change the conclusion.

Each sentence must stop at its own evidence boundary. Do not append a general recommendation unless the alternatives, criterion, constraints, consequences, and authority are all stated.