Legal Governance of Classification Thresholds
This manuscript isolates the operating point, the rule that converts a score, rank or similarity measure into governmental action, as a separate unit of public-law analysis. Documentary records from federal prison risk classification, New Jersey pretrial decision-making and English live facial recognition show that consequential boundaries can be adopted or revised without a record comparing alternatives or stating the resulting allocation of error. The analysis identifies where existing rulemaking, reason-giving and review doctrines already reach this act and where they stop. It then proposes a Threshold Decision Record that links the rule in its native form to an accountable owner, evidence about consequences, revision triggers and an available route for review.
Abstract
That the thresholds converting algorithmic scores into governmental action are policy choices is settled in scholarship. The act fixing those thresholds remains at most partially governed. This article isolates the operating point, the decision rule converting scores, ranks, similarities and constrained selections into official action, as a distinct unit of public-law analysis. Documentary records of PATTERN, New Jersey’s pretrial framework and English live facial recognition show operating points adopted and revised with, at best, authorship, publication and bounded reasons. Yet no located adoption record states the error allocation, compares alternatives or supports meaningful threshold-specific review. This article specifies when existing doctrine already reaches the threshold-setting act and where it runs out. It develops the Threshold Decision Record, a canonical record of adoption and material revision connecting the rule in its native technical form to an identified owner, comparative error-allocation reasons, upstream dependencies, an adopting legal vehicle and the review route.
How the inquiry is constructed
A documentary and doctrinal study examines public records, judgments, legislation and published guidance from the United States and England and Wales. A defined search method and coded treatment of absence claims support the documentary analysis. Decision theory is used narrowly to explain cost-sensitive thresholds and the conditions under which a threshold can imply a benchmark error-cost ratio. The institutional proposal is an eight-element Threshold Decision Record.
What the analysis establishes
- The operating point is a distinct public act because it converts an epistemic score into an official consequence.
- Published systems can identify an owner and a boundary while still lacking comparative reasons about the errors created by that boundary.
- Existing administrative law reaches some threshold-setting acts but does not provide continuous, symmetric or branch-wide governance.
- The Threshold Decision Record should document authority, score semantics, the native decision rule, alternatives, error allocation, human departures and revision conditions.
- A model revision, operating-point revision or changed deployment context can require a new record even when the displayed threshold number does not change.
Related research
SSRN manuscriptsPricing Contractual Vagueness
This manuscript examines what deliberately imprecise contingency language costs when the triggering event is genuinely unforeseeable. It separates uncertainty about whether a novel disruption occurs from uncertainty about how a court will interpret open-ended language. An alpha-maxmin bargaining model shows how the two parties' ambiguity premia combine into a wedge that reduces certainty-equivalent surplus. It also identifies conditions under which an interior degree of openness is optimal because openness limits exposure to the worst event outcome while creating an interpretive band. The analysis connects these results to Delaware material-adverse-effect doctrine and develops implications for indexed triggers, mandatory renegotiation and ambiguity-aware interpretation.
Moral Design of Algorithmic Governance
This manuscript argues that a government algorithmic enforcement system distributes the cost of official mistakes through its training loss function and deployment threshold. Together these settings form an error-allocation specification. The analysis treats that specification as a policy choice that may constitute a legislative rule under the Administrative Procedure Act and state analogues. It explains the persistence of inherited technical settings through choice-architecture defaults and institutional incentives. The proposed response is a penalty-default framework built around procurement-linked disclosure, legislative error budgets, public impact statements, periodic democratic review and a structured evidentiary presumption.