Comparator Case Library¶
Case-precedent register — instantiates Bounded Discretion Governance
A searchable store of past decided cases — facts, outcomes, and reasons — so a new discretionary call can be checked against, and kept consistent with, how like cases went before.
A Comparator Case Library is a curated, searchable collection of previously decided cases — each with its facts, its outcome, and the reasoning behind it — that a decision-maker consults to see how comparable situations were handled, and that new deciders learn the standard from. Its distinguishing contribution is horizontal consistency and institutional memory: it makes "like cases treated alike" operable by turning it into a retrieval problem. It neither sets the limits of discretion nor records any single new decision; it remembers the ones already made and hands them back when they are relevant.
Example¶
A trust-and-safety reviewer faces a borderline post: satire that reads, to some, like targeted harassment. Before ruling, they query the library for decided look-alikes and get back ≈7 prior cases — say 5 left up, 2 removed — each with the reasoning attached. Read together, the distinguishing feature becomes visible: the removed ones named and aimed at a specific private individual; the allowed ones mocked a public figure or an idea. The reviewer aligns to that reasoning and cites the closest prior case in their own rationale. The same curated set is what new reviewers train on before they take live cases, so a hard edge case does not get opposite rulings depending on who happens to be on shift.
How it works¶
The library retrieves by case similarity — facts, tags, and outcome — rather than by rule, which is what lets it serve situations no rulebook anticipated. Each entry is a filed decision (a completed rationale form) enriched with tags and its result, so two uses fall out of one store: precedent lookup for a live case, and a training set of worked examples for onboarding. Its signature is treating consistency as memory retrieval: the question it answers is not "what does the rule say?" but "what did we actually do the last dozen times this came up, and why?"
Tuning parameters¶
- Similarity / indexing scheme — how cases are tagged and matched. Good indexing surfaces true analogs; bad indexing surfaces superficial ones and teaches the wrong lesson.
- Curation vs. completeness — every decision versus a curated canon. Completeness is unbiased but noisy; curation is clean but can quietly bake in whoever's judgment did the curating.
- Precedent weight — advisory versus presumptively binding. Stronger weight boosts consistency but ossifies practice and propagates old mistakes.
- Treatment of overturned cases — whether reversed or successfully appealed decisions are flagged and kept as counter-precedent or purged. Keeping them teaches the boundary; purging hides it.
- Refresh / retirement — how stale precedent is aged out as norms and rules change.
When it helps, and when it misleads¶
Its strength is equal treatment and fast onboarding: it gives a decider a defensible "we handled the last dozen like this" and gives a newcomer the standard in worked form. Its failure mode is anchoring[n1] on a case that merely looks similar — matching on surface features rather than the one that actually mattered — and its classic misuse is cherry-picking the single prior case that supports a predetermined outcome. A curated library can also launder yesterday's bias into today's "precedent." The disciplines that keep it honest are indexing on outcome-relevant features, keeping overturned cases visible rather than deleted, and periodically re-examining whether the canon still reflects current standards instead of frozen ones.
How it implements the components¶
The Comparator Case Library fills the memory-and-comparison components — the store of past decisions and the exemplars drawn from it:
comparability_and_precedent_memory— the searchable store and its retrieval, which let a new case be compared against decided ones.training_case_library— the same corpus used as worked examples to teach new deciders the operative standard.
It remembers and compares decisions but does not author any single record (the Case Rationale Form), set the limits cases must respect (the Discretion Matrix), or run the sessions where deciders reconcile their reading of the precedents (the Calibration Review Cycle).
Related¶
- Instantiates: Bounded Discretion Governance — the library supplies the precedent memory that keeps discretionary decisions consistent across people and time.
- Consumes: Case Rationale Form — its entries are filed rationale forms, enriched with tags and outcomes.
- Sibling mechanisms: Case Rationale Form · Calibration Review Cycle · Discretion Matrix · Appeal and Reconsideration Workflow · Discretion Audit Dashboard · Exception Review Board · Guideline-with-Reasons Manual · Structured Professional Judgment Tool · Waiver or Override Log · Peer Case Conference
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: A searchable store of past decided cases — facts, outcomes, and reasons — so a new discretionary call can be checked against, and kept consistent with, how like cases went before, making its operative form a durable record, ledger, register, or trace whose value depends on preserving actual state or history.
Independent corroboration: The frozen evidence defines Comparator Case Library as 'A searchable store of past decided cases — facts, outcomes, and reasons — so a new discretionary call can be checked against, and kept consistent with, how like cases went before', so its operative form is Record, Log & Register.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Law & Governance
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Case-law and administrative practice established decided-case repositories for consistent treatment of materially similar facts.
Related originating lineages:
- Library & Information Science — Indexing and retrieval practice supplies searchable classification by decision-relevant facts and reasons.
- Public Administration & Policy — Administrative case management independently institutionalized precedent repositories for consistent discretionary decisions.
Review resolution: Legal precedent is the primary lineage of treating like decided cases alike. Searchable indexing and administrative case repositories materially turn that doctrine into the source's operational library, so both information science and public administration are genuine alternates and the origin is synthetic.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The tendency to rely too heavily on an initial or salient reference point — here, a superficially similar past case — when making a judgment, described in Tversky and Kahneman's work on anchoring. Indexing on the features that actually drove past outcomes is the standard corrective. ↩