Problem diagnosis and decision boundaries · Lesson 5

Problem brief and decision criteria

5 min read

Lesson objectives

  • Synthesize a bounded problem statement, define decision criteria and evidence gaps, and make a conditional recommendation that can be challenged.
Reading tools

After diagnosis, a decision maker needs a concise brief, not a pile of observations. Open with an observed fact that states denominator and period: 'Among applications completed in the specified quarter, a defined proportion exceeded the published service time.' Then state a relevant public consequence, such as delayed entry by eligible firms or unequal treatment of comparable applicants. Do not smuggle the answer into the problem statement. 'We need a new rule' excludes alternatives before appraisal. Distinguish sample findings from unresolved issues, and say who can take which kind of decision. A service manager may adjust guidance within delegated authority, while changing a general duty follows another route.

Choose comparison criteria before deciding which option the team likes. For permits these may include protection of the public purpose, applicant journey time, administrative burden, distribution of cost among small firms, feasibility and legality. Do not bury them in a single number without explaining weights and judgment. Legality may be a condition for eligibility rather than a score that faster processing can offset. For each criterion define a measure or question. Is time measured from first submission or after completeness? How will rare protective harms be detected without claiming that absence of reports means absence of harm? What cost falls on people without digital access? Some criteria are quantitative and others descriptive; keep that visible.

Prepare an initial decision matrix with four rows: observed problem, leading explanation and its limit, options to examine, and evidence that could reverse the recommendation. The last row makes advice testable. If complete files wait mainly at an interagency handoff, better coordination may precede changing a document requirement. If the document does not detect material risk and authority to amend exists, reviewing the requirement becomes stronger. Do not force every possible finding back to one favourite option. A material change in case facts should alter the choice, or the brief should explain why it does not. Set a date and owner for obtaining decisive evidence so that the recommendation can be acted upon.

Separate conditional advice from a final decision. A conditional memo might say: 'Trial clearer guidance within established authority and measure before and after; meanwhile request an authorized interpretation of whether the document is mandatory.' A final choice requires verified authority and appraisal of options. Even then, the decision maker may reasonably weigh public values differently from the analyst. Show the trade-off: a faster route might increase protective risk, while a precautionary route may impose disproportionate burden on a particular group. Impact appraisal makes alternatives, baselines, costs and benefits visible; it does not turn analytical estimates into binding legal rulings or an automatic vote. State confidence, error risk and a short review plan.

Lesson summary

Inputs: 20 of 80 files are returned, or 25%, and processing after completeness takes four days. Model: 'A quarter of applications are returned once because the image is unclear, while processing after completeness takes four days. We do not know whet