Symmetry Based Fairness

Treat equivalent cases equivalently unless a relevant asymmetry justifies different treatment.

Solution archetype #

1045

Problem family

Exclusion, Inequality & Distributional Harm

Problem subfamily

Distributive Allocation & Equal-Treatment Harm

The Diagnostic Story

Symptom: Comparable cases are being handled differently, but no one can clearly state which differences justify the different treatment. Exception records are sparse, invented after outcomes are chosen, or simply absent. Audits reveal unexplained variation after controlling for stated criteria. Policy language says cases are treated equally, but operational practice has developed hidden asymmetries that look like discretion to insiders and look like bias or arbitrary preference to everyone else.

Pivot: Define which cases are equivalent for the purpose at hand, constrain decision-makers to treat those cases the same, and require an explicit, reviewable rationale when treatment diverges—so that deviation from equal treatment is justified differentiation rather than undocumented discretion.

Resolution: Treatment of comparable cases becomes more consistent, arbitrary variation across reviewers, units, and time periods decreases, and the distinction between justified differentiation and unjustified inequality becomes auditable. The system gains legitimacy because deviations from equivalent treatment are reasoned and attached to the outcome record.

Reach for this when you hear…

[loan underwriting] “Two applications with the same financials got different rates last quarter—I need a documented reason for that difference or we have a fair-lending exposure.”

[academic review] “We approved that research protocol but rejected an almost identical one from a different lab, and the file contains no explanation of what the relevant difference was.”

[benefits administration] “The same exception has been granted to some employees and denied to others with identical circumstances, and nobody can point to a rule that explains it.”

When This Archetype Applies

No catalog groundingNone of the structural conditions is currently represented by an accepted prime or domain-specific abstraction.

Diagnostic problem

A decision system treats cases differently without showing which differences are relevant to the decision, making unequal treatment look like discretion, noise, bias, or hidden preference rather than justified differentiation.

What this problem means

The structural problem is unexplained asymmetry. A system claims to use a rule, but treatment shifts across cases that appear equivalent. Sometimes the variation comes from hidden bias, habit, convenience, reviewer preference, historical accident, or local interpretation. Sometimes the variation is justified, but the justification is not recorded.

The result is a legitimacy gap. People cannot tell whether different treatment reflects a relevant difference or an arbitrary one. Decision-makers may also lose track of their own standards, creating drift over time.

Show the applicability expression

Applicability expression4 distinct conditions

Unequal similar-case treatmentandUnjustified differential treatmentandUnexplained rule variationandOutcome-irrelevant labels

Algebraic1∧2∧3∧4

groundedpartly groundedopen

4 conditions, all required.

4Required in every casenumbered 1–4

These hold no matter which pattern applies.

1

Unequal similar-case treatment · open

Similar cases receive different outcomes, burdens, access, penalties, rankings, or benefits. more

and — also required

2

Unjustified differential treatment · open

Decision-makers cannot state which differences justify different treatment. more

and — also required

3

Unexplained rule variation · open

Rules vary across units, reviewers, time periods, or categories without an explicit rationale. more

and — also required

4

Outcome-irrelevant labels · open

A classification or allocation system depends on labels that may be irrelevant to the outcome being decided. more

Other requirements and context (1)

Why these sit outside the expression

Supporting context — it may accompany or help interpret the situation, but it is not a load-bearing condition in a sufficient diagnostic set.

• Supporting contextStakeholders suspect favoritism, discrimination, arbitrary exceptions, or inconsistent precedent.

Coverage

0 of 4 conditions grounded · 4 open.

Read the methodology · Download the trigger-logic data

Mechanisms / Implementations

• Anti-Discrimination Check: Holds a case fixed and flips only a protected characteristic — race, sex, religion, disability, age — to see whether treatment moves; a targeted symmetry test for the markers the law and ethics forbid from counting.
• Classification Fairness Review: Measures how a classification's errors — false inclusions and false exclusions — distribute across affected groups, exposing the burden and invisibility a formally neutral boundary can still produce.
• Consistency Audit: Measures decisions already made across reviewers, units, and time to surface unexplained variation — treatment that shifted when only the decider or the date changed, not the case.
• Equal-Treatment Checklist: A fixed set of questions a reviewer works through on every case — what treatment must stay invariant, what difference would justify departing, and what to record — so the same discipline lands on each decision as it is made.
• Exception Register: A living ledger of every approved waiver — with owner, rationale, compensating control, and expiry — so deviations stay visible and time-bound instead of quietly becoming the norm.
• Policy Symmetry Test: Interrogates a rule as written or as coded — would it hand equivalent cases different treatment when only an irrelevant feature changes? — to catch asymmetry and smuggled values in the policy itself, before a single case is decided.
• Precedent Analysis: Lines a current case up against specific prior decided cases and asks whether it is relevantly like them — so a decision rests on 'we handled the last one this way and nothing relevant has changed,' or an explicit, defensible distinction.

Related Abstractions

Abstractions this archetype builds on — directly (a source ingredient) or as a related pattern. Links follow the typed catalog namespace.

Built directly on (3)

• Equivalence Relation: Groups elements into equivalence classes.
• Procedural Fairness (Due Process): Due process.
• Symmetry: Invariance under transformation.

Also references 7 related abstractions

• Accountability: Responsibility for actions.
• Adjudication (Dispute Resolution): Dispute resolution.
• Constraint: Limits possibilities to guide outcomes.
• Equity: Context-sensitive fairness.
• Invariance: Properties unchanged under transformation.
• Reciprocity: Mutual exchange.
• Trade-offs: Balancing competing priorities.

Variants

Narrower or domain-specific specializations that share this archetype's core structure. Recognized variants are established; candidate variants are provisional.

Policy Symmetry Review · governance variant · recognized

Reviews whether a policy treats equivalent cases the same across units, populations, time periods, or rule interpretations.

Precedent Consistency Review · governance variant · recognized

Compares a current case to prior analogous cases so departures from precedent require an explicit relevant difference.

Anti-Discrimination Symmetry Check · risk or failure variant · candidate

Tests whether identity, status, location, or group markers are changing treatment when they are irrelevant to the decision rule.

Reciprocal Burden Symmetry · governance variant · candidate

Checks whether burdens, duties, or obligations are distributed symmetrically among parties in reciprocal positions.

Editorial Notes

Problem Classification

Classification: Exclusion, Inequality & Distributional Harm → Distributive Allocation & Equal-Treatment Harm

Problem kernel: equivalent cases receive unequal treatment without relevant justification

Rationale: Equivalent cases receive different treatment without an explicit showing that the differences are decision-relevant, creating unjustified distribution through discretion, noise, bias, or hidden preference. Fair hearing concerns whether affected parties receive evidence, voice, impartial judgment, reasons, review, and remedy; unequal treatment can exist even when no hearing or dispute process is implicated.

Boundary considered: Authority, Accountability, Legitimacy & Fair-Process Failure → Fair Hearing, Dispute & Remedy Failure

Why this classification prevailed: Equal-treatment harm concerns the substantive relevance of differences used to allocate outcomes; fair hearing concerns the procedure for evidence, voice, judgment, review, and remedy.

Review outcome: Adjudicated after independent review; high confidence.