A useful answer names a choice, explains a mechanism, and compares feasible alternatives. It also identifies a cost of its own proposal and something that would change the recommendation. Economic terminology helps only when it does that work.
The two examples below concern an unnamed fictional printing business, separate from the book’s recurring cases. They illustrate reasoning, not findings about a real firm or solutions to the chapter questions.
Early example: diagnose before changing the reward
Read after Chapter 1. Suppose a printing business rewards operators for completed orders. Its order count and complaints about unusable prints both rise. The owner proposes eliminating the completion bonus.
A possible response
I would first compare the share of orders requiring correction, using the same follow-up period. More complaints could simply reflect more orders. If the correction rate has risen, rushing is one explanation, but a change toward harder print jobs could produce the same pattern. I would review a small sample of corrected and uncorrected orders, comparing job complexity, the submitted file, and any recorded print check. This costs review time and may reveal that the existing records are inadequate. Removing the bonus immediately could reduce attention to timely completion without solving a file-quality problem. The current report justifies investigation; it does not establish that operators became careless or that the bonus caused the complaints.
Why this works. The response checks whether performance actually deteriorated before explaining why. It gives competing mechanisms and asks for evidence that could distinguish them. The proposed review has a cost, and the recommendation does not treat “remove the incentive” as a costless cure. A small sample may guide further investigation without proving causation.
Later example: recommend a package, not a slogan
Read after Chapter 12. Now consider a separate stipulated problem at the printing business. Counter staff can compare a finished print with the customer’s approved file, but every free reprint requires the owner’s approval. The owner is often unavailable. Staff are rewarded for closing complaints quickly, and the business does not distinguish justified reprints from courtesy replacements.
A possible response
I would test local approval for a defined category of routine printing errors, with a total reprint-cost limit per order and a recorded comparison against the approved file. Requests outside that category would go to the owner or a named backup; escalation would not authorize waiving the policy. Before the trial, the owner would set the category and limit and retain authority to amend them. Evaluation would consider justified resolution, resource use, and recurrence rather than closure speed alone. The owner would review a sample of local decisions, and another designated reviewer would check disputed owner-approved cases. This requires records and review time. I would narrow delegation if ordinary cases could not be classified reliably or staff routinely reclassified courtesy requests as printing errors.
Why this works. Authority, evaluation, and rewards change together. The response names who sets the boundary, who handles escalations, and who checks decisions. It identifies a foreseeable response to the new rule and a reason to revise the proposal. It is a trial recommendation, not proof that delegation is superior. Before implementation, the business would need actual limits, review capacity, and a comparison of resolution time, repeat problems, and resource use on comparable orders. A later improvement alone would not establish that the trial caused it.
Check your own response
Could a reader identify the decision, the person whose choices change, the mechanism, and the cost of your proposal? Have you separated what the facts establish from what you still need to learn? Name a condition that would reverse or materially narrow your recommendation. Different recommendations can be defensible when their assumptions are explicit.