GiveBetter x Los Angeles

Urban Peace Institute

Violence-prevention infrastructure, training, regranting and justice reform

Research time: 25 min on GPT-6 Astra Light + GPT-6 Astra Medium
  • Research — reviewed programs, finances and impact evidence.

Updated: 2026-09-14

Donate

Summary

What do they do? UPI trains and supports community violence-intervention organizations and works on justice reform. Its LA Peacemakers initiative funds independent partner organizations; their entire impact is not automatically UPI's impact. A deliberately partial participant-health model suggests roughly $15 million per 10 LA QALYs, with wide uncertainty, zero-effect and harm cases.

Why we’re interested in this organization:

  • A concrete delivery network reaches people exposed to serious violence.

  • Original accounts verify a substantial local regrant network; a dated funding cliff creates a plausible, but unquantified, continuation opportunity.

  • Outside randomized evidence supports the possibility of violence reduction for well-targeted intensive services.

Our main reservations:

  • UPI's own declining injury counts lack a credible causal comparison.

  • Participant engagement, risk, service effect and private-funding response remain judgments; newly verified public support belongs in the counterfactual.

  • The model charges all recipient costs but counts only one LA pathway; it is neither a complete portfolio estimate nor a guaranteed lower bound.

What do you get for your dollar? $15.4M per better life: ten additional quality-adjusted life years in Los Angeles. Whole annual UPI recognized operating expense, including regrants and non-LA work, charged to a conditional partial-health finite LA Peacemakers participant-health branch. External partner/public resources and other portfolio benefits are unmeasured; external resources are stress-tested separately. This is not measured unrestricted portfolio return, full social-cost effectiveness, a lower bound or a donor recommendation..

Central expected net additional avoided victimizations 1.5404999999999998, including 0.3081 fatal events. These are fractional model expectations, not observed people saved.

1. What do they do?

UPI is the independent nonprofit Connie Rice Institute for Urban Peace, EIN 36-4816075. Its ordinary support funds training, technical assistance, regranting, community safety and justice reform, including work beyond Los Angeles. The official donation page identifies this recipient and EIN. No restricted Peacemakers-only gift is assumed. Recipient

LA Peacemakers braids $9.6 million of public and private support across 18 independent organizations during 2024–2026. Its first-year report covers February 2024–January 2025 and reports 1,027 new participants. These are observed operational counts, not 1,027 people whose lives were saved. Year-one report

FY2025 Schedule I lists 17 legal grant recipients totaling $3,625,415, including fiscal-sponsor routes. The 18-member initiative and a single fiscal-year grant table are different denominators. The largest listed recipient, Chapter Two, received $1,035,767. These transfers are already included in UPI expense and are never added a second time. Actual regrants

2. Monitoring and information sharing

The year-one report describes 13% fewer fatal injuries and 24% fewer nonfatal injuries between its two half-years. Its graph does not clearly tabulate the exact totals used in alpha, so those totals are removed. Injury definitions are not sufficiently specified to equate every reported event with a shooting. This before–after comparison has no controlled counterfactual, consistent person-time denominator or independent outcome linkage. It is not used to estimate e or r. Year-one report, p15

The year-one 1,027 new enrollments inform a prospective annual scale, not measured engaged person-years. The newer progress report separates 47 frontline positions from 53 operational positions; neither is a verified full-time equivalent. Cumulative enrollment subsequently approached 2,000. Reported prevented incidents are staff judgments, not observed counterfactual injuries. The model neither annualizes cumulative counts nor adds citywide homicide declines. Progress, p4 February 2026 statement

Remaining evidence priorities are unique engaged person-years by risk and service intensity, incident definitions, independent linkage to fatal/nonfatal outcomes, matched comparison trajectories and the exact staffing/service loss after already-committed grants and reserves. Prospective reporting must protect participants from identification, coercive information sharing and retaliatory exposure.

3. Qualitative assessment

READI randomly offered intensive employment, cognitive-behavioral programming and support. Only 55% began programming. Its 12% lower victimization estimate is treatment-on-the-treated relative to control compliers, not a significant intent-to-treat reduction or a UPI coefficient. The primary violence index was not significantly improved; outreach-referred participants had a stronger 43% victimization estimate. The 11 events per 100 controls over 20 months is a rate benchmark, not the same estimation denominator. We retain r=.04 and e=.10 as uncertain transport judgments for heterogeneous UPI-supported services, not study-derived effects. Original trial

A closer local study models nonrandomized GRYD incident response in 2014–17, using spatial-temporally sampled controls. It estimates roughly 14–18% lower retaliatory triggering, with meaningful model and inferential uncertainty. This is neither all violence nor the incremental effect of UPI, training, or Peacemakers. Existing GRYD is part of the baseline; the study supports mechanism plausibility without supporting direct transplantation of a headline reduction. Original local study

Mortality value is now an explicit finite survival calculation: .8 health utility, a 47-year cap, 1% annual all-cause mortality and 3% discounting produce 16.7227 QALYs per initially avoided death. The annual death probability and cap are judgments, not a recipient life table; future victimization risk has not disappeared. Nonfatal value is .1 acute QALY plus .05 annual utility loss for at most ten years, with the same survival/discounting, totaling .5047 QALY. The small Boston cohort supports ongoing morbidity but does not estimate these parameters. Fatal and nonfatal branches are mutually exclusive. Survivor study

Only one service year is central. Extra downstream health is not another program year. Continued exposure to violence, displacement to other people or neighborhoods, worker injury/burnout and privacy harms can erode or reverse benefit. No arrest, crime-cost dollar or reported mediation is directly converted to QALYs. The Richmond evidence previously reviewed remains contextual, not an additional credited channel.

4. What do you get for your dollar?

Three original accrual returns cover consecutive July–June fiscal years ending 2023, 2024 and 2025. Full operating expenses are $3,185,128, $5,293,277 and $8,897,684 (mean $5,792,030). The FY2024 return includes an additional $100,000 nonoperating leasehold-disposal loss; Schedule D reconciles its $5,393,277 total to audited operating expenses. Both series are disclosed rather than silently mixed. No reported event, gaming, inventory or donated-service adjustment requires an extra latest-year addback. FY2025 Schedule D independently repeats $8,897,684 as reported audited expense, but does not substitute for the missing correctly matched audit package. FY2023 FY2024 reconciliation FY2025 reconciliation

The latest full cost includes $7,075,093 program, $1,264,399 administration and $558,192 fundraising. Regrants are included once. The revised model generates 5.7743 modeled additional LA QALYs from the selected participant branch, giving approximately $15.4 million per 10. This change from alpha is mainly the explicit residual mortality assumption, not new measured clinical effectiveness. Positive joint stress cases and a zero/harm case are not confidence limits. Original FY2025 return

Full recipient annual cost is a ratio normalization, not an observed marginal gift or guaranteed purchase of 1,027 engaged person-years. National work and unpriced justice reform stay in the denominator. Partner/public resources outside UPI are not included in this donor-cost scope; a separate sensitivity adds $3 million of hypothetical external delivery resources. Attribution a does not make external resources costless, and no social-cost-effectiveness claim is inferred.

Model, assumptions and sensitivity

Full annual recipient expense including regrants, support costs and non-LA work, normalized to a prospective service cohort. This is not verified marginal cost. External partner costs unmeasured; explicit resource sensitivity only.

g=1 for the modeled LA Peacemakers participant branch within LA County. This is not a claim that all UPI activity or benefit is local; Orange residents qualify but no separate Orange benefit is invented. National training and statewide justice work remain unquantified.

A(T,m,d)=sum(t=1..T)[((1-m)/(1+d))^t]; qd=u*A(T,m,d); qn=qa+du*A(Tn,m,d). E=sum(t=0..ceil(D)-1)[min(1,D-t)*((1-m)*rho/(1+d))^t]. V=N*r*e*a*b*E. Qall=V*(f*qd+(1-f)*qn)-H; QLA=V*(f*qd+(1-f)*qn)*g-H. price=10*C/QLA only if positive. Central D=1; downstream health starts at the avoided-event time, modeled as time zero of its exposure period.

C
8897684 USD per annual full recipient budget (observed). Original FY2025 expense, audited-total reconciliation and grants retained once. [upi-99025] [upi-sd25] [upi-si25]
N
1027 prospective engaged participant-equivalents for one full service year (judgment). 1,027 first-year enrollments anchors scale only; engagement and person-time not observed. [upi-y1] [upi-progress]
r
0.04 serious shooting/homicide victimizations per participant-year without the modeled increment (judgment). Below annualized READI control benchmark; local mix and background declines uncertain. [upi-readi-full] [upi-current]
e
0.1 relative victimization reduction (judgment). 10% working transport judgment, not the READI TOT or GRYD retaliation coefficient. [upi-readi-full] [upi-gryd]
f
0.2 fatal share of avoided modeled victimizations (judgment). 20%; not estimated from the broadly defined UPI injury graph. [upi-y1]
u
0.8 health utility while alive after initially avoided death (judgment). .8 constant simplified utility, not measured recipient health.
T
47 maximum downstream years after initially avoided death (judgment). 47-year finite cap, not a verified life expectancy.
m
0.01 annual all-cause death probability conditional on being alive (judgment). 1% annual death probability in discrete survival (1-m)^t; includes continuing violence and other mortality. This is not a continuous hazard or an age-specific recipient life table.
d
0.03 annual discount rate (judgment). 3% normative health and future-cost discount.
qa
0.1 acute QALYs lost per nonfatal injury (judgment). .1 immediate-area approximation; not derived from PTSD prevalence. [upi-survivors]
du
0.05 annual chronic utility decrement per nonfatal injury (judgment). .05 across all survivors as an expected average, not everyone permanently disabled. [upi-survivors]
Tn
10 maximum chronic morbidity years (judgment). 10 years; beyond the observed follow-up and varied as a judgment. [upi-survivors]
a
0.75 UPI-enabled fraction after partner continuation (judgment). .75, separate from private gift response; no measured attributable fraction. [upi-y1] [upi-si25]
b
0.5 effective private-funding capacity response (judgment). .5 after secured funding, reserves, substitution and allocation; no verified gift-sized gap. [upi-99025] [upi-bscc-contract] [upi-city-cao] [upi-city-final] [upi-current]
g
1 LA share of modeled branch benefit (judgment). 1 for LA resident participant branch; local residence still to verify, no headquarters allocation. [upi-y1]
H
0 additional LA QALYs lost outside modeled victimizations (judgment). 0 central is not safety evidence; additive worker/privacy/displacement harm retained in stress.
D
1 finite years of risk reduction for the same cohort (judgment). 1 central; fractional bridge and two-year cases cannot be read as repeated new cohorts.
rho
1 annual persistence of relative effect after first year (judgment). 1 irrelevant for central D=1; .5 in durability-only second-year stress.

Conditional partial participant-health estimate: Cost: $8.9M; Los Angeles QALYs: 5.7742936466569965; all-population QALYs: 5.7742936466569965. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Lower untreated participant risk: Cost: $8.9M; Los Angeles QALYs: 1.4435734116642491; all-population QALYs: 1.4435734116642491. {"C":8897684,"N":1027,"r":0.01,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

READI-like risk, not a local observation: Cost: $8.9M; Los Angeles QALYs: 9.527584516984048; all-population QALYs: 9.527584516984048. {"C":8897684,"N":1027,"r":0.066,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Only 500 annual engaged participant-equivalents: Cost: $8.9M; Los Angeles QALYs: 2.8112432554318394; all-population QALYs: 2.8112432554318394. {"C":8897684,"N":500,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Twenty-percent effect judgment: Cost: $8.9M; Los Angeles QALYs: 11.548587293313993; all-population QALYs: 11.548587293313993. {"C":8897684,"N":1027,"r":0.04,"e":0.2,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

New funding mostly replaces other resources: Cost: $8.9M; Los Angeles QALYs: 1.1548587293313994; all-population QALYs: 1.1548587293313994. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.1,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Fully responsive new private funding: Cost: $8.9M; Los Angeles QALYs: 11.548587293313993; all-population QALYs: 11.548587293313993. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":1,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

All modeled capacity financed anyway: Cost: $8.9M; Los Angeles QALYs: 0; all-population QALYs: 0. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Most benefit survives without UPI: Cost: $8.9M; Los Angeles QALYs: 1.9247645488856657; all-population QALYs: 1.9247645488856657. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.25,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Only a three-month service gap closed: Cost: $8.9M; Los Angeles QALYs: 2.8871468233284983; all-population QALYs: 2.8871468233284983. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":1,"g":1,"H":0,"D":0.25,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Same cohort: one extra year at half effect without extra spending: Cost: $8.9M; Los Angeles QALYs: 8.549318263254293; all-population QALYs: 8.549318263254293. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":2,"rho":0.5}. Finite model equations above; stresses are not confidence bounds.

Same cohort: second funded service year, both costs included: Cost: $17.5M; Los Angeles QALYs: 11.324342879851585; all-population QALYs: 11.324342879851585. {"C":17536212.155339807,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":2,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Shorter survival after the initially avoided death: Cost: $8.9M; Los Angeles QALYs: 4.320253375519606; all-population QALYs: 4.320253375519606. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.03,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Mortality-only health branch: Cost: $8.9M; Los Angeles QALYs: 5.152264004553359; all-population QALYs: 5.152264004553359. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0,"du":0,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Only ninety percent of branch benefit local: Cost: $8.9M; Los Angeles QALYs: 5.196864281991297; all-population QALYs: 5.7742936466569965. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":0.9,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Joint low-positive stress, not a confidence bound: Cost: $8.9M; Los Angeles QALYs: 0.006166745103691574; all-population QALYs: 0.006166745103691574. {"C":8897684,"N":500,"r":0.01,"e":0.02,"f":0.1,"u":0.7,"T":30,"m":0.03,"d":0.03,"qa":0.02,"du":0.01,"Tn":3,"a":0.25,"b":0.25,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Joint favorable stress, not a confidence bound: Cost: $8.9M; Los Angeles QALYs: 262.4910903452534; all-population QALYs: 262.4910903452534. {"C":8897684,"N":2000,"r":0.066,"e":0.4,"f":0.25,"u":0.9,"T":55,"m":0.005,"d":0.03,"qa":0.2,"du":0.1,"Tn":10,"a":1,"b":0.8,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

No additional victimization reduction: Cost: $8.9M; Los Angeles QALYs: 0; all-population QALYs: 0. {"C":8897684,"N":1027,"r":0.04,"e":0,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Increased victimization plus other incremental harm: Cost: $8.9M; Los Angeles QALYs: -3.8871468233284983; all-population QALYs: -3.8871468233284983. {"C":8897684,"N":1027,"r":0.04,"e":-0.05,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":1,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Add hypothetical external delivery resources: Cost: $11.9M; Los Angeles QALYs: 5.7742936466569965; all-population QALYs: 5.7742936466569965. {"C":11897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Three-year mean operating expense, other inputs held fixed: Cost: $5.8M; Los Angeles QALYs: 5.7742936466569965; all-population QALYs: 5.7742936466569965. {"C":5792030,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":0,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Central benefit with one additional QALY of other harm: Cost: $8.9M; Los Angeles QALYs: 4.7742936466569965; all-population QALYs: 4.7742936466569965. {"C":8897684,"N":1027,"r":0.04,"e":0.1,"f":0.2,"u":0.8,"T":47,"m":0.01,"d":0.03,"qa":0.1,"du":0.05,"Tn":10,"a":0.75,"b":0.5,"g":1,"H":1,"D":1,"rho":1}. Finite model equations above; stresses are not confidence bounds.

Counterfactual: Existing public GRYD/Safe Passage, separate BSCC training, committed philanthropy and spendable reserves continue. a discounts service benefit that persists without UPI; b discounts capacity not changed by a new ordinary gift. Central preserves one service year, not permanent program operation. The bridge scenario models a temporary service gap, not an extra annual cohort or permanent funding loss.

Attribution: Only serious shooting/homicide participant victimization is modeled. UPI monitoring injury counts have broader/unclear definitions and do not supply the causal rate. Benefits to other victims, policy and training recipients are omitted. V already includes partner and donor additionality; no regrant, staff, retaliation or citywide-decline credit is added again.

Partial-health judgment central retained despite unobserved marginal parameters. Finite competing mortality and health loss are explicit judgments. Scenarios vary scale, risk, effect, funding, attribution, exposure, survival, local share and resource scope; no probabilistic quantiles or recipient-wide effectiveness claim. For multi-year stress cases, the expected-event calculation is a low-risk linear approximation, not a joint competing-risk simulation of repeated injury in the same person. It cannot support large extrapolations or infinite cohorts.

Sensitivity

  • Central qd=16.722700436719766, qn=0.5047303165397902, net QLA=5.7742936466569965, price/10=15409129.747239782.
  • At $1 million per 10 QALYs, the branch requires 23.74 net additional avoided victimizations versus 1.5404999999999998; holding other central assumptions fixed would require e=1.541, greater than 1.
  • Each of N, r, e, a and b proportionally changes gross benefit; a better clinical conversion alone cannot resolve weak donor response. No threshold is a giving recommendation.
  • A two-year durability claim discounts second-year effect and survival. If a second funded service year is required, the second annual cost is included too; price does not simply halve.
  • Attribution/private response a=b=1 gives price/10=5778423.655214918; this is a diagnostic, not default.
  • Using comparable mean operating expense gives price/10=10030716.05711163; it is not a claim that current capacity can be bought at historical cost.
  • Zero/full replacement and net harm have no finite beneficial price. Joint favorable/low-positive cases carry no assigned probabilities.
  • The quarter-year bridge also raises b from 0.5 to 1: its benefit is one-half of central, not one-quarter. It represents a fully responsive three-month gap; retaining central b=0.5 would yield one-quarter of central benefit. No permanently unfunded year is inferred.

Unresolved inputs

  • Correctly matched FY2025 audit and evidence recurring revenue-recognition controls were implemented; current cash restrictions, payables and board commitments.
  • Peacemakers post-2026 renewal and termination status by award, remaining cash obligation, marginal gift allocation, trained-worker recruitment/retention and funded versus unfunded person-time.
  • Coalition-to-Schedule-I reconciliation (18 initiative members versus 17 annual legal grant routes), fiscal sponsors and partner cofunding/overlap with approved Safe Passage contracts.
  • Unique annual engaged participants, intensity, incident definitions, untreated risk and residence; independently linked person-time outcomes with credible comparisons.
  • UPI-specific causal contribution rather than all GRYD/CVI effectiveness, displacement and other-victim effects, worker health/privacy burden and diminishing returns.
  • Age-specific survival and finite nonfatal utility loss; no empirical probability distribution or complete portfolio price.

5. Funding and previous grants

FY2025 revenue was $12,539,904, government grants $5,292,475, cash $6,898,638, unrestricted net assets $6,213,733 and total liabilities $2,716,158. Cash equals about 9.3 months of FY2025 expense before commitments; that is not a current runway. The balance includes restrictions and obligations, and the correct FY2025 audit is still needed. FY2024 audited notes document concentrated funding and accrued external grants, illustrating why gross cash is not a spendable gap. Latest accounts

UPI announced an $8 million Ballmer commitment over two years in July 2024 and says Peacemakers funding ends in 2026. This is historical committed support, not a fresh 2026 renewal or an extra amount to add to annual revenue. No verified remaining Peacemakers gap, renewal probability or marginal hiring budget was found. Ballmer announcement Dated cliff claim

The cliff is not all-recipient shutdown: BSCC identifies a separate UPI frontline-training contract from August 2025 through June 2029. Los Angeles also approved a $5.087 million Safe Passage award, with $3,438,380 in provider contracts including Community Partners and Alliance for Community Empowerment, recipient routes also appearing in UPI grants. Council authorization and appropriation do not by themselves prove every contract has been executed or replaces identical services. They do establish a material funded baseline to investigate, not zero public support. BSCC Named allocation Adopted recommendations Final appropriation

Central b=.5 remains an explicitly subjective effective capacity response to ordinary private support after reserves, replacement donors, secured grants and current commitments; a=.75 separately represents the UPI-enabled share of service benefit. These discounts must not be estimated from the same loss twice. A short-bridge case credits only a quarter-year gap and then assumes other funding resumes. Complete replacement produces zero. UPI itself describes trust and information-sharing concerns with federal funding; the report treats these as documented organizational constraints, not an independent legal finding. UPI statement

The filing labeled FY2025 contains FY2024 statements plus a February 2026 corrective-action attachment about recurring contribution-recognition controls. The period mismatch remains unresolved; obtain the correctly matched audit, implementation evidence and a current unrestricted cash/commitments schedule before treating this as donor-ready.

Separately, GreenLight Fund announced a $1 million multi-year Oakland Peace Academy investment in October 2025, with cohorts expected in 2027–2029. This supports the existence of funded non-LA activity, not an extra LA benefit or proof that Peacemakers is fully renewed. It should not be added again to historical revenue without a recognition and payment schedule. Funder announcement

Annual expenses

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

  • FY 2023: $3.2M; Connie Rice Institute for Urban Peace dba Urban Peace Institute, EIN 36-4816075, 12-month period, Accrual recipient operating expenses; excludes nonoperating asset disposal losses.. Source
  • FY 2024: $5.3M; Connie Rice Institute for Urban Peace dba Urban Peace Institute, EIN 36-4816075, 12-month period, Accrual recipient operating expenses; excludes nonoperating asset disposal losses.. Source
  • FY 2025: $8.9M; Connie Rice Institute for Urban Peace dba Urban Peace Institute, EIN 36-4816075, 12-month period, Accrual recipient operating expenses; excludes nonoperating asset disposal losses.. Source

6. Sources

  1. FY2025 original accrual Form 990. Urban Peace Institute / IRS. Published: 2026-04-06; retrieved: 2026-09-14.
  2. FY2024 audited statements with FY2023 comparative totals. Urban Peace Institute / Armanino LLP. Published: 2025-01-23; retrieved: 2026-09-14.
  3. Audit filing labeled 2025: FY2024 statements and 2026 corrective-action attachment. Urban Peace Institute / Federal Audit Clearinghouse. Published: not stated; retrieved: 2026-09-14.
  4. LA Peacemakers Initiative Year One Report. Urban Peace Institute. Published: not stated; retrieved: 2026-09-14.
  5. LA Peacemakers Initiative Progress Report. Urban Peace Institute. Published: not stated; retrieved: 2026-09-14.
  6. 2025 homicide trends and 2026 funding expiration. Urban Peace Institute. Published: 2026-02-19; retrieved: 2026-09-14.
  7. Current recipient donation page and EIN. Urban Peace Institute. Published: not stated; retrieved: 2026-09-14.
  8. READI Chicago randomized trial policy brief. University of Chicago Crime Lab. Published: not stated; retrieved: 2026-09-14.
  9. Peacemaker Fellowship and urban gun violence in Richmond. Bruno et al. / JAMA Network Open. Published: 2026-08-31; retrieved: 2026-09-14.
  10. Patient-reported outcomes 6–12 months after firearm injury. Herrera-Escobar et al. / Annals of Surgery. Published: 2021-12-01; retrieved: 2026-09-14.
  11. Coalition statement on new CalVIP funding. GIFFORDS. Published: not stated; retrieved: 2026-09-14.
  12. FY2023 original Form 990. Urban Peace Institute / IRS. Published: 2024-05-09; retrieved: 2026-09-14.
  13. FY2024 original Form 990. Urban Peace Institute / IRS. Published: 2025-04-16; retrieved: 2026-09-14.
  14. FY2024 Schedule D expense reconciliation. Urban Peace Institute / IRS. Published: 2025-04-16; retrieved: 2026-09-14.
  15. FY2025 Schedule D expense reconciliation. Urban Peace Institute / IRS. Published: 2026-04-06; retrieved: 2026-09-14.
  16. FY2025 Schedule I actual regrant recipients. Urban Peace Institute / IRS. Published: 2026-04-06; retrieved: 2026-09-14.
  17. READI Chicago original randomized study. Bhatt et al. / Quarterly Journal of Economics. Published: not stated; retrieved: 2026-09-14.
  18. Local nonrandomized GRYD incident-response study. Park et al. / Journal of the American Statistical Association. Published: 2021-04-22; retrieved: 2026-09-14.
  19. CalVIP capacity-building contracts and awards. California BSCC. Published: not stated; retrieved: 2026-09-14.
  20. Two-year Ballmer commitment announced in 2024. Urban Peace Institute. Published: 2024-07-17; retrieved: 2026-09-14.
  21. Safe Passage award and recipient allocation. Los Angeles City Administrative Officer. Published: 2026-05-27; retrieved: 2026-09-14.
  22. Adopted Safe Passage recommendations. Los Angeles City Council. Published: 2026-06-16; retrieved: 2026-09-14.
  23. Mayor concurrence: Safe Passage funds appropriated. Los Angeles City Clerk. Published: 2026-08-13; retrieved: 2026-09-14.
  24. UPI statement on constraints of federal replacement funding. Urban Peace Institute. Published: 2025-10-24; retrieved: 2026-09-14.
  25. GreenLight Fund announces Oakland Peace Academy investment. GreenLight Fund. Published: 2025-10-15; retrieved: 2026-09-14.