When Law Makes Crime Worse (Part II): From Diagnosis to Design

TL;DR

Part I showed how criminal law can be criminogenic—through perverse incentives, legitimacy erosion, and displacement. Part II is a design manual: a checklist, stress-tests, and concrete alternatives that reduce harm without amplifying crime.

The Thesis

Badly designed criminal law manufactures the very markets, identities, and conflicts it seeks to suppress. Good design doesn’t start with “What can we criminalise?” but with “What problem are we solving, and is criminalisation the least criminogenic tool available?”

Five Pathways Law Makes Crime Worse

1. Elasticity Blindness: Prohibitions that ignore supply–demand elasticity (e.g., drugs, sex work) tend to raise prices, increase violence, and enrich organised actors while not shrinking demand.
2. Enforcement Capacity Mismatch: Rules that cannot be consistently enforced breed selective enforcement, corruption risks, and delegitimise the law.
3. Criminogenic Burdens: Collateral consequences (records, exclusion from housing or work) can lock people into illegal markets.
4. Legitimacy Erosion: Overbreadth, vagueness, or visibly unfair penalties tank procedural justice and voluntary compliance.
5. Displacement & Adaptation: “Crackdowns” shift crime into darker corners (or online), increase harm severity, and reduce reporting.

The P.R.O.P.E.R. Criminalisation Test

Before drafting or enforcing, run this six-part stress test. If any step fails, re-design or don’t criminalise.

1. Problem clarity: Define the specific harm (victim, context, measurable outcome). Ban proxy-moralising.
2. Rights & proportionality: Is coercion necessary and least restrictive? Are penalties proportionate to actual harm (not symbolic outrage)?
3. Operational capacity: Can police, courts, prisons actually deliver this at scale—without crowding out serious crime?
4. Predictable incentives: Model displacement, black-market price effects, and identity labelling (does “offender” status entrench crime?).
5. Evidence baseline: What’s the counterfactual? Pre-register metrics and track them (see Tripwires below).
6. Review by default: Add a sunset clause and mandatory impact review with public data.

Tripwires: Metrics That Force Reconsideration

• Violence multiplier: If the prohibition correlates with a > X% rise in violence severity in targeted markets, review triggers.
• Enforcement disparity index: If stop/search, charging, or sentencing disparities exceed a preset threshold controlling for exposure, independent audit triggers.
• Dark-market premium: If unit price or purity volatility spikes post-law, assess black-market strengthening.
• Reporting collapse: If victim reporting drops materially in affected communities, pause and re-design.
• System strain: If dockets or prisons exceed capacity targets, triage to non-penal tools.

Design note: Tripwires need public dashboards, independent oversight, and consequences (automatic review, narrowed scope, or sunset).

Prefer These Tools Before You Criminalise

1. Administrative & Civil Regimes: Licensing, inspections, targeted fines with income-graduation, and swift adjudication.
2. Restorative & Problem-Solving Pathways: Conferencing, community courts, and treatment-linked dispositions for low-harm, high-volume offenses.
3. Narrow, Elements-Rich Offences: If criminal law is necessary, narrow the conduct, add mens rea and harm thresholds, and cap penalties.
4. Procedural Justice First: Visibility, reason-giving, respectful treatment—these increase voluntary compliance and reduce reoffending.
5. Target the Profit Node: Use confiscation, follow-the-money tools, and platform liability calibrated to avoid over-removal and rights breaches.

Mini Case Snapshots (What Usually Goes Wrong)

• Youth Curfews: Often displace gatherings, increase antagonistic police–youth contact, and reduce trust. Design for place-based guardianship and positive night-time economy instead.
• Street-Based Sex Work Bans: Push work into riskier, isolated spaces; reduce screening; increase violence. Consider decriminalised + regulated models with labor and safety enforcement.
• Drug Possession Crackdowns: Raise black-market margins and violence; clog courts; deter help-seeking. Pivot to health-led responses and focused deterrence on violent actors.
• Anti-Social Orders as Status Symbols: Overbroad orders invite non-compliance and contempt churn. Use narrowly tailored conditions with procedural fairness and service links.
• Three-Strikes Escalators: Distort plea incentives and inflate sentence length without clear marginal deterrence. Replace with graduated, evidence-based sentencing.

EU/ECHR Guardrails You Can Use (Not Fear)

Proportionality (EU Charter art. 52; ECHR case law) is your friend: it demands narrow tailoring and evidence. Legality (nullum crimen, art. 49 Charter) militates against vague “catch-all” offences. Non-discrimination and due process principles reinforce procedural justice. Build them in upfront.

Prosecutor’s Playbook: Fast Re-Design on the Ground

1. Divert by Default: First-time low-harm offenses → restorative/administrative tracks.
2. Focus on Violence: Prioritise cases that actually reduce serious harm; de-prioritise symbolic possession crimes.
3. Charge Narrowly: Avoid stacking marginal counts that only increase churn and resentment.
4. Review Disparities Monthly: If disparities breach your red-line, change screening criteria.

Legislator’s One-Page Pre-Flight Checklist

• What non-penal tools have we tried, and with what data?
• How will offenders exit the system (record relief, employment, housing)?
• What’s our sunset + review clause text? (Include metrics and an external auditor.)
• How will we detect displacement, corruption risk, and black-market premiums?
• What is the narrowest offence definition that still addresses the harm?

Template clause: “This Act expires on [date] unless renewed after independent review confirming net harm reduction against pre-registered indicators.”

Design Culture, Not Just Laws

Criminal law should be a last-mile tool, not our first reflex. Build a policy culture that prizes evidence, reversibility, narrowness, and dignity. If a rule can’t survive the P.R.O.P.E.R. test and tripwire metrics, it shouldn’t survive at all.