GiveBetter x New York City

Common Justice

Serious-violence diversion, survivor recovery and justice-system change

Research time: ~6 min on GPT-6 Astra Light + ~2 min on GPT-6.1 Sol
  • Research — reviewed programs, finances and impact evidence.
  • Source audit — checked claims, assumptions and calculations.
  • Modeling — estimated costs, QALYs and uncertainty.

Updated: 2026-10-07

Donate

Summary

What do they do? Common Justice offers restorative alternatives to incarceration for serious violent felonies and support for survivors in Brooklyn, Manhattan and the Bronx. It also funds advocacy, national partnerships, training and narrative work, so an ordinary gift is not restricted to local diversion. The conditional price covers violence, trauma and incarceration-related health effects of its core direct-service pathway using the entire recipient’s cost; it is not a complete valuation of its wider portfolio.

Why we’re interested in this organization:

  • Current original accounts permit a three-year gross-cost reconciliation.

  • Randomized evidence supports a restorative-conferencing mechanism, although not this exact replacement-for-prison program.

  • The model prices several material direct-service health pathways and explicitly tests worse violence outcomes.

Our main reservations:

  • Annual unique cases and episode costs are not disclosed in the current sources inspected; the capacity denominator is a judgment.

  • Very low reconviction among selected graduates is not a causal estimate and does not identify harm among all referrals.

  • National replication and advocacy may dominate the true return, but their incremental local effects remain unpriced.

What do you get for your dollar? $14.1M per DALY in New York City. DALY is our one-year comparison unit. QALY-based health gains and income-welfare equivalents retain their original assumptions; they are not measured clinical DALYs. Retained direct-service clinical reference; household work/transfer resources assessed unknown.

Complete restorative case-pair episodes, inferred from program expenditure and assumed episode cost. Net violent victimizations prevented, split into fatal and nonfatal outcomes. Incremental trauma recovery among existing survivors and health-only effects of avoiding incarceration.

1. What do they do?

Common Justice’s 2025 impact report describes work with people aged 16–28 accused of violent crimes and those they harmed. Participation depends on survivor agreement and cooperation from the legal system. Its principal direct-service pathways are accountability, prevention of repeat violence, survivor recovery and avoidance of damaging incarceration. The full recipient also works on victim compensation, civic engagement, replication and public narratives.

2. Monitoring and information sharing

The report describes agreements in all completed circles, 8% terminated for a new crime since inception, and one new violent-felony conviction in a study of graduates enrolled during 2012–2022. These are selected, differently timed observations, not annual treatment effects: graduates exclude failures, convictions miss unreported harm, and no matched counterfactual is supplied. The developing healing scale is not yet a QALY measure. Unique entrants, completion, victim outcomes, alternative sentences and all-referral follow-up would materially improve the estimate.

3. Qualitative assessment

A randomized-evidence Campbell review supports the possibility that consensual restorative conferences reduce subsequent offending and victim distress. Those trials often supplemented normal justice and are not a direct experiment in Common Justice’s diversion model. The central estimate therefore uses explicit priors, not the review’s standardized effect as an injury rate. Consent, survivor autonomy, new victimization and loss of incarceration’s incapacitation effect matter. The adverse scenario permits increased violence and net health deterioration.

Household net work and survivor-transfer resources assessed unknown, not zero or a90%-health finding.

4. What do you get for your dollar?

Signed household resources are assessed unknown and potentially material. The retained partial direct-service clinical reference remains about $14.1M per DALY, with 0.570643082634484 MSA QALYs at $8,027,684 of full annual recipient expense. This is not a complete combined valuation, zero-income finding or empirical90%-health share. All eight original worlds remain unchanged.

The current diversion model describes12–15-month intensive participation and agreements that may include restitution, community service, school and work. Avoided custody can enable work or benefits, but actual counterfactual sentences, net retained resources and paid outcomes are unidentified. Public incarceration savings are not participant cash; liberty is not salary. Full program, community-service, access and time burdens must be assessed independently of clinical response.

The 2025 impact report documents victim-compensation reforms, including a higher funeral-expense maximum. Statutory limits are not actual awards or new gift-caused payments. Existing enacted reforms, ordinary victim services, court alternatives and public implementation remain baseline. Restitution can credit survivor households while debiting responsible households: household sharing, transfers and losses must be jointly modeled rather than counted as pure gains.

A numerical bridge requires current funding-responsive unique resource-sharing participants and survivors, actual alternative custody/community conditions, retained earnings and benefits after taxes, withdrawals, childcare/commuting and other work costs, paid restitution/compensation, finite delay/duration and MSA residence. Positive overlap applies only to gains; negative resource burdens cannot be discounted by health improvement. Public payer transfers and disjoint real resources must not duplicate recipient or external-cost diagnostics. Broader policy, housing, survivor and national resources remain unquantified, not zero; no gross salary, retail restitution value or ordinary survivor earnings is credited.

FY2025 gross recipient expense is $8,027,684: Form 990 Part IX $7,972,026 plus Schedule D donated services $19,475 plus separately netted event expense $36,183. FY2024 is $8,472,083 ($8,441,733 plus $30,350 donated services); FY2023 is $6,770,078 ($6,758,628 plus $11,450). All are July–June accrual periods. The model retains the whole recipient cost. Direct-service spending of $2,973,529 divided by an assumed $60,000 per complete case episode gives a steady-state annual equivalent of 49.56 case pairs, not an observed enrollment count. National programs are not removed from the numerator.

The conditional central price is approximately $140.68 million per ten NYC MSA QALYs. The $60,000 complete-case cost is a capacity judgment, not an observed fee or funding offer. Effects are approximated at cohort onset; delayed prevention would lower positive present value. No-prison-health and adverse cases expose dependence on the largest health assumption.

Model, assumptions and sensitivity

Entire recipient gross annual cost; partial direct-service health denominator. Wider policy and national replication effects are unpriced, not claimed absent.

Only health accruing within the official 22-county NYC MSA is priced. Three-borough delivery motivates a 98% central share; no national chapter or expenditure-share proxy.

N=D/c. F(r,T)=(1-exp(-(r+ln(1+d))*T))/(r+ln(1+d)). Q_all=N*a*b*[v*((1-z)*ui*F(ri,Ti)+z*ud*F(md,Td))+ps*us*F(rs,Ts)+up*F(0,Tp)-h]. Q_NYC=g*Q_all. Price=10*C/Q_NYC when positive, otherwise null. Violence effects over two years are valued at onset for simplicity; a delay sensitivity can only reduce their positive value. Fatal and nonfatal victimization are mutually exclusive; survival is counted once per victim.

C
8027684 USD annual gross recipient expense (observed). Part IX plus donated services plus separately netted event costs. [return2025] [d2025]
D
2973529 USD direct-service annual expenditure (observed). Part III direct services; used only to infer capacity, not to shrink whole cost. [return2025]
c
60000 USD direct-service cost per complete case-pair episode (judgment). $60,000 central, $30,000–$120,000 sensitivity; assumes steady-state spending finances complete multiyear episodes. Not a quoted service price. [return2025] [impact]
v
0.05 additional violent victimizations prevented per case over two years (judgment). Net of incapacitation, repeat offending and other prevention; not reconvictions. Signed in adverse case. [rjc] [impact]
z
0.005 fatal share of prevented victimizations (judgment). 0.5% event mix assumption, not observed homicide rate. [impact]
ui
0.2 utility loss per nonfatal injury (judgment). Physical and psychological burden of a newly prevented victimization. [rjc]
ri
0.5 annual nonfatal injury recovery rate (judgment). Exponential recovery, no permanent injury claim. [rjc]
Ti
3 years nonfatal injury follow-up (judgment). Finite three-year horizon.
ud
0.8 utility among survivors of a prevented fatal victimization (judgment). Health-adjusted survival, not a full-quality life.
md
0.02 annual competing mortality hazard (judgment). 2% limits survival in a high-risk population.
Td
30 years survival follow-up (judgment). Finite 30-year survival cap.
ps
0.1 additional existing survivors improving per case pair (judgment). 10% gain beyond alternative support. Existing-survivor recovery excludes new injuries in v. [rjc] [impact]
us
0.15 utility improvement per responding survivor (judgment). Not a validated conversion of the healing scale. [rjc]
rs
0.5 annual survivor recovery-effect decay (judgment). Rapid decay to avoid indefinite trauma-recovery credit.
Ts
2 years survivor follow-up (judgment). Finite two-year horizon.
up
0.02 net annual health-utility gain from avoided incarceration (judgment). Small 0.02 health-only prior; excludes liberty, income and fiscal savings. Can be zero or negative. [impact]
Tp
3 years incarceration-related health difference (judgment). Three-year cap, not an observed sentence. [impact]
a
0.5 fraction of case equivalents additional to other services/diversion (judgment). 50% after counterfactual alternative-service substitution; no second causal discount.
b
0.25 recipient funding-response fraction (judgment). 25% of proportional historical delivery changes with ordinary funding; recognizes reserves, grants and referral constraints. [return2025] [impact]
g
0.98 NYC MSA share of priced health (judgment). 98% because the priced direct service is in three NYC boroughs, allowing migration; does not assert 98% of national portfolio. [current]
h
0.00 additional QALY harm per marginal case pair (judgment). Zero central, 0.01 adverse; separate from increased victimization.
d
0.03 annual discount rate (judgment). 3% continuous-equivalent discount log(1+d).

Conditional core-pathway central: Cost: $8.0M; New York City health QALYs: 0.57; all-population health QALYs: 0.58. {"C":8027684,"D":2973529,"c":60000,"v":0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.1,"us":0.15,"rs":0.5,"Ts":2,"up":0.02,"Tp":3,"a":0.5,"b":0.25,"g":0.98,"h":0,"d":0.03}

Low throughput and small health gains: Cost: $8.0M; New York City health QALYs: 0.00; all-population health QALYs: 0.00. {"C":8027684,"D":2973529,"c":120000,"v":0.005,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.02,"us":0.15,"rs":0.5,"Ts":2,"up":0,"Tp":3,"a":0.25,"b":0.1,"g":0.98,"h":0,"d":0.03}

Higher throughput and larger gains: Cost: $8.0M; New York City health QALYs: 8.89; all-population health QALYs: 9.07. {"C":8027684,"D":2973529,"c":30000,"v":0.15,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.25,"us":0.15,"rs":0.5,"Ts":2,"up":0.05,"Tp":3,"a":0.75,"b":0.5,"g":0.98,"h":0,"d":0.03}

No marginal funding response: Cost: $8.0M; New York City health QALYs: 0.00; all-population health QALYs: 0.00. {"C":8027684,"D":2973529,"c":60000,"v":0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.1,"us":0.15,"rs":0.5,"Ts":2,"up":0.02,"Tp":3,"a":0.5,"b":0,"g":0.98,"h":0,"d":0.03}

Diversion increases violence and harms health: Cost: $8.0M; New York City health QALYs: -0.34; all-population health QALYs: -0.35. {"C":8027684,"D":2973529,"c":60000,"v":-0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0,"us":0.15,"rs":0.5,"Ts":2,"up":-0.01,"Tp":3,"a":0.5,"b":0.25,"g":0.98,"h":0.01,"d":0.03}

No net health benefit from avoided incarceration: Cost: $8.0M; New York City health QALYs: 0.22; all-population health QALYs: 0.23. {"C":8027684,"D":2973529,"c":60000,"v":0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.1,"us":0.15,"rs":0.5,"Ts":2,"up":0,"Tp":3,"a":0.5,"b":0.25,"g":0.98,"h":0,"d":0.03}

Additional external service resources: Cost: $8.1M; New York City health QALYs: 0.57; all-population health QALYs: 0.58. {"C":8089632.520833333,"D":2973529,"c":60000,"v":0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.1,"us":0.15,"rs":0.5,"Ts":2,"up":0.02,"Tp":3,"a":0.5,"b":0.25,"g":0.98,"h":0,"d":0.03}

25% additional unmeasured resources: Cost: $10.0M; New York City health QALYs: 0.57; all-population health QALYs: 0.58. {"C":10034605,"D":2973529,"c":60000,"v":0.05,"z":0.005,"ui":0.2,"ri":0.5,"Ti":3,"ud":0.8,"md":0.02,"Td":30,"ps":0.1,"us":0.15,"rs":0.5,"Ts":2,"up":0.02,"Tp":3,"a":0.5,"b":0.25,"g":0.98,"h":0,"d":0.03}

Combined resources unknown; clinical reference retained: Cost: $8.0M; New York City health QALYs: 0.57; all-population health QALYs: 0.58. Original clinical model and full annual recipient scope retained; net household resources unknown, not zero.

Counterfactual: Existing courts, alternative diversion, survivor services and government-funded work continue. v is the signed net violence difference versus that baseline, including incapacitation; a addresses replacement delivery, b addresses marginal funding response, each once.

Attribution: No credit for every graduate avoiding reconviction, previously enacted laws or nationwide media reach. Direct service is a central operational mechanism but only about half of program spending; wider institutional change remains a material omitted pathway.

Scenario ranges are subjective alternatives, not confidence intervals. Capacity and all clinical coefficients are unobserved priors. The central is donor-useful only as an explicitly partial-health estimate, not a complete portfolio ranking.

Sensitivity

  • Episode cost drives capacity inversely; the report supplies no verified annual cohort.
  • The no-prison-health scenario removes the most contestable direct participant benefit.
  • Violence can increase if diversion loses protective incapacitation or survivor safeguards fail.
  • External service costs and unrecognized resources worsen the price; no incarceration savings offset them.
  • A three-year delay would discount all positive health by (1.03)^-3; longer-term national spillovers remain unpriced.

Unresolved inputs

  • Unique annual referrals, entrants and completions with full episode resource costs.
  • Comparison-group offending, injury and survivor health outcomes including non-completers.
  • Counterfactual sentences and health in custody versus community, including mortality and access to treatment.
  • Marginal donor-response evidence, restricted reserves and funded referral capacity.
  • Local incremental effects and costs of victim-compensation implementation, replication and narratives.
  • Current gift-responsive unique case/household cohorts, actual counterfactual sentences/diversion, retained net work/benefits versus alternatives, paid restitution/compensation and payer mix, taxes/withdrawals/full program work/access/time costs, MSA residence and finite delay/duration.

5. Funding and previous grants

The ordinary donation page confirms Common Justice, Inc., EIN 82-0993366. FY2025 reported revenue is $7,405,997, government grants $1,985,121 and net assets $15,357,301. Its impact report appeals for help amid dwindling resources, but neither a deficit nor that appeal proves that the next gift opens a case. The model discounts alternative services and prosecutions once and funding response once; existing government commitments and enacted compensation reform belong in the baseline. It deducts no prison-budget savings and includes an induced-resource sensitivity.

Annual expenses

Organization-level spending, including programs, administration and fundraising. The research list averages three comparable, consecutive full fiscal years when available.

  • FY 2023: $6.8M; Common Justice, Inc., 12-month period, Accrual Form 990 gross expenses including donated services and separately netted event costs. Source
  • FY 2024: $8.5M; Common Justice, Inc., 12-month period, Accrual Form 990 gross expenses including donated services and separately netted event costs. Source
  • FY 2025: $8.0M; Common Justice, Inc., 12-month period, Accrual Form 990 gross expenses including donated services and separately netted event costs. Source

6. Sources

  1. FY2025 original Form 990. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  2. FY2025 original Schedule D reconciliation. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  3. FY2024 original Form 990. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  4. FY2024 original Schedule D reconciliation. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  5. FY2023 original Form 990. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  6. FY2023 original Schedule D reconciliation. Common Justice / IRS. Published: not stated; retrieved: 2026-09-14.
  7. 2025 impact report. Common Justice. Published: 2025-12-18; retrieved: 2026-10-07.
  8. Diverting serious violence. Common Justice. Published: not stated; retrieved: 2026-10-07.
  9. Ordinary recipient donation and legal entity. Common Justice. Published: not stated; retrieved: 2026-09-14.
  10. Randomized restorative-conferencing systematic review. Strang et al. / Campbell Systematic Reviews. Published: 2013-11-04; retrieved: 2026-09-14.