{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","cell_id":"bounded_rivalry_governance__criminology_forensic","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_id":"brg_regional_net_harm_prevention_league_p4_v0","proposal_index":4,"version":0,"title":"Regional Net-Harm Prevention League","problem":"A city compares police districts on reductions in recorded crime while awarding a scarce pool of prevention-project continuation funds and analytical support. Districts can improve their standing without reducing regional harm by discouraging reports, changing incident classifications, concentrating visible enforcement on easily measured offenses, moving activity across district boundaries, or avoiding people and locations likely to worsen the metric. Neighboring districts and residents then absorb costs that the district leaderboard does not count.","actors":["Police district and community-organization partnerships","Residents, victims, and people seeking to report harm","Neighboring district teams","Municipal crime-data office","Independent audit and survey unit","Civilian public-safety oversight board","City budget authority","Emergency response and victim-service providers","People exposed to enforcement or displacement spillovers"],"observable_state":"The city can observe districts competing for a fixed pool of marginal prevention resources; standings based mainly on recorded incidents within administrative boundaries; classification corrections and report-access complaints; changes in calls or incidents near district borders; enforcement-contact volumes; differences between administrative records and independent resident reports; and prevention proposals that do not account for harms transferred to neighboring areas or affected populations.","consequence":"The contest can select districts best at managing recorded numbers or exporting costs rather than partnerships that reduce net regional harm. It can also burden reporting, intensify low-value enforcement, disadvantage districts that accurately record difficult conditions, and deprive decision-makers of evidence about which prevention approaches merit continuation.","affected_objective":"Use bounded comparison to identify prevention approaches associated with lower net serious harm while protecting truthful reporting, emergency response, civil rights, neighboring areas, and the noncontestable public-safety floor.","intervention":"Replace the raw district crime leaderboard for marginal prevention funding with a time-limited Regional Net-Harm Prevention League. Eligible district-community partnerships register a bounded prevention plan, affected area, permitted actions, resource envelope, expected mechanism, displacement risks, and stopping conditions before the period begins. Core emergency response, victim services, legally required duties, and base district budgets remain outside the contest. A frozen rulebook prohibits report suppression, opportunistic reclassification, quota-driven contacts, boundary dumping, retaliation, selective audit access, and unregistered substitution of enforcement for the proposed prevention method. Independent scorers use a preregistered composite of verified serious-harm indicators, reporting accessibility, classification integrity, boundary spillovers, resident-reported harm, enforcement burdens, cost, and implementation fidelity. Neighboring areas receive shared credit or penalty for measured cross-boundary changes. Teams operate under equal contest-resource caps, leading results are audited before awards, and a portion of each provisional award remains in an externality reserve until delayed spillovers are reviewed. Several complementary approaches may receive continuation support. A post-period review can change the metric, release or redirect the reserve, invalidate a result, or retire the league.","structural_mapping":[{"archetype_element":"Rivalry purpose statement","domain_realization":"Use district comparison to discover and continue bounded prevention approaches that reduce net serious harm, not to reward the lowest recorded incident count."},{"archetype_element":"Scarce prize or selection constraint","domain_realization":"A fixed pool of marginal prevention-project continuation funds, municipal analyst hours, and training support; core public-safety resources are not at stake."},{"archetype_element":"Competitor eligibility boundary","domain_realization":"A district may enter only through a documented partnership with affected community representatives, a registered prevention mechanism, data-integrity commitments, conflict disclosures, and an auditable resource plan."},{"archetype_element":"Contest arena boundary","domain_realization":"Permitted moves are the registered lawful prevention activities and service coordination. Report discouragement, record manipulation, retaliatory contact, boundary displacement, quota-driven enforcement, audit obstruction, and unregistered coercive substitution are fouls."},{"archetype_element":"Performance metric and scoring basis","domain_realization":"Scoring balances independently checked serious-harm indicators with reporting access, classification accuracy, resident-reported conditions, geographic spillovers, enforcement burdens, cost, and fidelity to the registered intervention."},{"archetype_element":"Fair process and due process layer","domain_realization":"Eligibility, weights, geographic attribution, audit sampling, tie-breaks, evidence access, conflicts, written score explanations, and a time-boxed appeal process are fixed before results are observed."},{"archetype_element":"Anti-sabotage and anti-collusion guardrail","domain_realization":"Common data definitions, immutable submission logs, random classification audits, protected reporting channels, and penalties constrain manipulation, obstruction, retaliation, or coordinated shifting of incidents between districts."},{"archetype_element":"Externality and spillover boundary","domain_realization":"Adjacent-area outcomes, reporting burdens, enforcement contacts, complaints, and delayed harms enter the score; a provisional-award reserve remains available for verified remediation rather than leaving spillovers with residents or neighboring districts."},{"archetype_element":"Escalation and arms-race damper","domain_realization":"Contest-funded spending, dedicated staff time, paid outreach, and analytical support are capped and audited, while enforcement volume provides no positive score by itself."},{"archetype_element":"Prize decomposition or multiple-winner design","domain_realization":"Continuation awards support a bounded portfolio of nonredundant prevention mechanisms rather than crowning one district or method as universally superior."},{"archetype_element":"Winner power and lock-in review","domain_realization":"Awards expire, confer no control over future scoring, and require a fresh eligibility and audit review before renewal."},{"archetype_element":"Learning and recalibration loop","domain_realization":"The oversight board compares measured outcomes, spillovers, reporting integrity, burdens, and metric-gaming evidence after each period and revises or ends the contest accordingly."}],"mechanism_mapping":[{"mechanism_slug":"contest_rulebook","role":"Freezes eligibility, protected public-safety functions, allowed prevention actions, fouls, scoring, geographic attribution, audits, tie-breaks, and appeals before the performance period.","counterfactual_removal":"Without a binding rulebook, officials could redefine success after seeing district results, and teams would not have a stable boundary separating legitimate prevention from reporting or enforcement manipulation."},{"mechanism_slug":"ranked_leaderboard_with_audit","role":"Produces confidential criterion-level standings and independently audits leading districts' classifications, source records, resource use, reporting access, and claimed cross-boundary outcomes before any continuation award.","counterfactual_removal":"Without audit, the leaderboard could reward suppressed reports, coding changes, selective data submission, or unsupported attribution rather than verified net-harm performance."},{"mechanism_slug":"spending_cap_or_resource_cap","role":"Caps contest-funded money, dedicated staff hours, purchased analytics, and promotional spending while auditing transfers into uncapped categories.","counterfactual_removal":"Wealthier or administratively favored districts could win through additional personnel, surveillance, or publicity, provoking an escalation unrelated to prevention quality."},{"mechanism_slug":"externality_bond_or_liability_rule","role":"Holds a defined portion of each provisional continuation award in reserve until boundary displacement, reporting barriers, enforcement burdens, and delayed remediation needs are reviewed.","counterfactual_removal":"A district could receive the full prize after improving its internal score while residents, neighboring districts, or later budget periods bear the exported costs."},{"mechanism_slug":"multiple_award_or_portfolio_selection","role":"Allocates continuation support to several approaches selected for complementary mechanisms, settings, and risk profiles rather than only to the highest aggregate district score.","counterfactual_removal":"A single winner could concentrate future resources, encourage imitation of one possibly fragile metric strategy, and suppress useful but nonredundant approaches."},{"mechanism_slug":"sabotage_or_foul_penalty_schedule","role":"Publishes graduated consequences for report suppression, reclassification manipulation, retaliation, audit obstruction, unauthorized data access, boundary dumping, and undisclosed resource circumvention.","counterfactual_removal":"Off-arena conduct could remain a profitable route to winning, while ad hoc discipline could be applied selectively against disfavored districts."},{"mechanism_slug":"post_contest_impact_review","role":"Examines reporting integrity, serious-harm measures, geographic displacement, enforcement burdens, distributional effects, resource concentration, and delayed outcomes before renewal.","counterfactual_removal":"The city could repeat a contest whose visible winners merely shifted harm, weakened reporting, or learned to optimize a proxy disconnected from the stated purpose."}],"causal_chain":["Districts compete for a scarce pool of marginal prevention funding and analytical support.","A raw recorded-crime ranking permits advantage through report suppression, classification changes, selective enforcement, or geographic displacement.","The league limits the rivalry to registered prevention plans while placing core public obligations outside the contest.","Independent measures and classification audits make accurate reporting part of winning rather than a source of competitive disadvantage.","Boundary-adjusted scoring and a provisional-award reserve attach exported costs to the district result instead of leaving them outside the arena.","Resource caps and a zero reward for enforcement volume dampen staffing, surveillance, and contact-count arms races.","Portfolio awards preserve rivalry while avoiding a single district or method becoming the mandatory model for the city.","Post-period review tests whether standings reflect net serious-harm reduction without unacceptable reporting, rights, or spillover costs.","If the link between winning and the stated contribution breaks, the board revises the rules, withholds the affected award, or retires the league."],"baseline":"Continue publishing district comparisons based primarily on recorded incidents or percentage change and use managerial judgment to allocate marginal prevention funds. Data-quality reviews, complaint processes, regional coordination, and ordinary budget oversight remain separate. This baseline is administratively simple but does not systematically incorporate reporting access, cross-boundary displacement, enforcement burdens, equal contest resources, provisional awards, or post-contest validation into the rivalry itself.","nearest_rivals":["A conventional recorded-crime or clearance-rate leaderboard, which is easy to communicate but leaves reporting, classification, enforcement, and geographic gaming routes open.","Formula-based funding by population, need, or workload, which can protect equity and stability but does not use rivalry to compare bounded prevention approaches.","Collaborative regional prevention planning without rankings, which can support information sharing and joint action but provides no scarce comparative selection mechanism for continuation funds.","Independent district audits without a contest, which can detect data problems and rights concerns but do not redirect the route to winning a scarce prevention award.","Random selection among eligible prevention pilots, which limits favoritism and proposal-polish advantages but does not use comparative evidence to choose which completed approaches receive continuation."],"remaining_contrastive_claim":"The proposal's bounded claim is that, where districts already compete for marginal prevention resources, incorporating reporting integrity, cross-boundary spillovers, enforcement burdens, independent audit, and delayed liability into the contest can couple winning more closely to net regional harm reduction than a recorded-crime ranking. It does not claim that policing duties, victim services, civil rights, or base public-safety funding should themselves be competitive.","authority_safety":{"decision_authority":"The civilian public-safety oversight board and city budget authority may authorize only the shadow evaluation and, after separate approval, allocation of the designated marginal prevention fund. Statutory officials retain emergency, investigative, employment, disclosure, records, and enforcement authority; courts and established review bodies retain authority over legal violations and individual remedies.","authorized_first_step":"Approve a retrospective, nonpublic shadow scoring of one completed reporting period using already-authorized aggregate or de-identified data. Pre-register the alternative metric, audit sample, boundary rules, burden measures, scorer conflicts, and stop conditions. Do not change rankings, budgets, personnel evaluations, public communications, or operational practices from the shadow result.","excluded_actions":["Reducing core emergency response, victim services, legally required duties, or base district budgets because of contest performance","Using standings to direct stops, searches, arrests, surveillance, charging, or attention toward any person or group","Setting arrest, citation, contact, search, or incident-reduction quotas","Discouraging, delaying, refusing, or recoding a report to improve a score","Publishing district rankings or identifiable resident-level information during the shadow phase","Using shadow scores for promotion, discipline, termination, or collective-bargaining decisions","Treating survey disagreement or anomalous data patterns as proof of misconduct","Changing an individual case, record, or legal outcome through the contest","Allowing competitors to control the audit sample, scorer selection, or neighboring districts' data"],"halt_rollback":"Stop scoring upon re-identification risk, unequal data access, unverifiable geographic attribution, scorer conflict, evidence of retaliatory use, pressure to change a record, or an attempt to operationalize the shadow ranking. Quarantine affected data, preserve an access log, notify the independent privacy and oversight reviewers, suppress all standings, and invalidate comparisons that cannot be reconstructed. Because the first step is retrospective and carries no award or operational authority, rollback leaves existing budgets, public reports, employment decisions, and district practices unchanged."},"negative_tests":{"strongest_counterevidence":"The composite score remains gameable, produces unstable rankings under reasonable weighting or boundary choices, primarily reflects neighborhood resources outside district control, or rewards concealed coercion. Independent reporting measures may be too noisy or delayed, shared boundary penalties may obscure responsibility, and portfolio awards may distribute funds without meaningful selection.","problem_falsifier":"The inferred problem is unsupported if no scarce marginal resource or standing depends on district comparisons, recorded-incident measures do not influence strategic behavior, reporting and classification remain equally accessible and accurate across districts, and independent review finds no plausible cross-boundary or enforcement spillover omitted from current decisions.","intervention_falsifier":"The intervention fails if shadow standings are not more consistent than the baseline with independently audited serious harm, reporting access, and rights-burden measures; if rankings are highly sensitive to defensible scoring choices; or if participation encourages suppression, displacement, coercive substitution, audit manipulation, administrative overload, or withdrawal from regional cooperation.","risks":["Composite weights may encode contestable value judgments while presenting a misleadingly precise rank.","Resident-reporting measures may differ because of trust, access, sampling, or willingness to disclose rather than underlying harm alone.","Districts with stronger community organizations or administrative capacity may retain structural advantages.","Teams may move gaming into uncapped budgets, contractors, classifications, or neighboring agencies.","Shared boundary attribution may penalize a district for changes it could not control.","The externality reserve may be too small, released too early, or treated as a routine cost.","Confidential standings may still leak and stigmatize neighborhoods or personnel.","Competition may weaken cooperation or timely cross-district information sharing.","A registered prevention plan may conceal unrecorded coercive activity.","Portfolio selection may become an entitlement system instead of a genuine comparative down-select." ]},"next_evidence_step":"Run the preregistered shadow evaluation on one completed period without awards or publication. Calculate the baseline recorded-crime standing and the proposed net-harm standing from the same cutoff data. Use an independently selected sample to check incident classification, report accessibility, boundary attribution, enforcement-contact burden, resource accounting, and implementation fidelity. Conduct sensitivity analysis across disclosed defensible weights and boundary assumptions, and record ranking stability, missingness, scorer agreement, appeals that would have been triggered, and any re-identification or retaliation risk. Proceed no further if the score cannot be reproduced, changes materially under modest defensible assumptions, or requires person-level data that cannot be safely governed.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Proposal 1 governed rivalry among forensic laboratory teams by rewarding controlled error discovery and corrective learning. Proposal 4 instead addresses geographic district-community partnerships competing for marginal prevention resources, where recorded-crime metrics can omit reporting integrity, enforcement burdens, and displaced harm. It neither tests forensic analysts nor rewards defect detection; its causal path uses boundary-adjusted outcomes, protected reporting, and an externality reserve to internalize regional spillovers. Proposal 2 governed commercial digital-forensics vendors competing for time-limited deployment leases through sealed tool benchmarks, portable evidence formats, performance security, and challenger procurement. Proposal 4 contains no vendor product selection, software benchmark, lease, migration obligation, or commercial lock-in; its prize is continuation support for registered prevention activity. Proposal 3 governed competing investigative hypotheses within a case through equal evidence access, preregistered predictions, and bounded validation slots. Proposal 4 does not rank explanations, allocate case tests, or influence any individual investigation; it compares organizational prevention plans using aggregate regional outcomes while placing all case authority outside the arena. The league is therefore independently adoptable by a municipal oversight and budget body without implementing the laboratory contest, tool procurement process, or hypothesis gate.","revision_record":{"parent_version":null,"progress_targets_addressed":["Materially different problem from proposals 1, 2, and 3","Distinct actors, prize, intervention, externality structure, and causal path","Operational completeness matching proposal 1","Explicit authority limits, rollback, falsifiers, and bounded evidence","Diversity explanation covering every earlier sealed proposal"],"conceptual_changes":["Initial proposal centered on geographic performance rivalry and exported public-safety harms rather than forensic quality, vendor procurement, or within-case hypothesis selection."],"operational_changes":["Specified a protected noncontestable service floor, registered prevention plans, boundary-adjusted scoring, equal contest-resource caps, audit-triggered provisional awards, an externality reserve, and portfolio continuation."],"evidence_changes":["Defined a retrospective nonpublic shadow evaluation with classification audits, sensitivity analysis, and no effect on operations, budgets, personnel, or cases."],"claim_changes":["Limited the claim to recoupling an existing marginal-resource contest with net regional harm and reporting integrity; made no claim of novelty, prevalence, demand, effect size, or established efficacy."]}}