GiveBetter x SF

Five Keys Schools and Programs

Accredited secondary-credential pathway at Five Keys Independence High School

Research time: ~15 min on GPT-5.6 Sol Medium
  • Research — organization and evidence review.
  • Modeling — cost-effectiveness analysis.
  • Historical estimate for research done before time tracking.

Published: 31 August 2026.

Donate

Summary

What do they do? Five Keys operates accredited public charter schools in custody and community settings. This report isolates the secondary-credential pathway at Five Keys Independence High School: re-engaging a learner without a diploma, assessing prior credits and needs, delivering flexible instruction and support, and helping the learner earn a diploma or equivalency. It does not evaluate Five Keys housing, shelters, workforce programs, or the broader organization as one intervention. More

Why this approach interests us

  • Accredited, flexible instruction is brought into jails and community settings, with transcript review, individualized graduation maps, GED/HiSET preparation, counseling, and reentry-linked support for learners whom traditional schools did not retain.

Our main reservations

  • ADA is not unique enrollment or dosage; reported GED completions lack an eligible denominator; graduation cohorts are nonstandard; and the public record provides no comparable nonparticipant group or clean San Francisco-only instructional cost.

What do you get for your dollar?

Five Keys Independence High School reported $37,213,618 of FY2024 expense and 2,232.74 annual ADA, or about $16,667 per ADA-equivalent year. If a year increases credential completion by our 10-point midpoint, the conditional estimate is about $166.67K per additional secondary credential. Five Keys has not measured that effect, so zero remains plausible.

gross historical benchmark
≈ $16,700FY2024 total FKIH expense per ADA-equivalent year; not a marginal price
our conditional best guess
≈ $167,000per additional credential if one ADA-equivalent year raises completion by 10 points
cost per better life
≈ $4.9Mper 10 QALYs; very-low-confidence education-to-health transfer with an 80% central discount
funding room
Not publishedthe $100,000 gift is a scenario, not a verified added-seat plan

1. What do they do?

Five Keys operates accredited public charter schools in custody and community settings. This report isolates the secondary-credential pathway at Five Keys Independence High School: re-engaging a learner without a diploma, assessing prior credits and needs, delivering flexible instruction and support, and helping the learner earn a diploma or equivalency. It does not evaluate Five Keys housing, shelters, workforce programs, or the broader organization as one intervention.

Re-engage an eligible learner

A learner age 15 or older without a diploma enrolls voluntarily in a jail, community setting, or supported online pathway.

Map the credential path

Staff review transcripts and skills, identify remaining credits or equivalency requirements, and create an individualized plan.

Deliver flexible instruction

Credentialed teachers provide classroom, independent-study, synchronous, GED/HiSET, literacy, and related support around custody and life constraints.

Verify completion and follow-up

A recommendation-grade cohort would reconcile every entrant, dosage, withdrawal, transfer, diploma or equivalency, and education, employment, and custody status through at least 36 months.

Scope of this review. The native model covers an ADA-equivalent year in FKIH and an additional accredited secondary credential. It excludes other Five Keys schools and services, credits without completion, selected-graduate recidivism comparisons, employment and earnings, and avoided custody. The separate comparison bridge models health utility only.

2. Monitoring and information sharing

Five Keys students across accredited charter schools, community sites, and 17 county jails in California

Organization and school administrative reporting. Five Keys reports more than 2,100 graduates to date and says its schools operate at more than 51 community settings and in 17 jails. Its 2024–25 FKIH LCAP reports 56 people passed all GED portions in 2022–23, while its FY2024 audit reports 2,232.74 annual ADA.

Our assessment. These figures establish delivery and credential outputs, not impact. ADA is an attendance-based funding denominator rather than unique learner flow, and the public record does not publish a San Francisco-only cohort, eligible or tested GED denominator, comparable nonparticipant group, or complete post-release follow-up.

Five Keys graduates in San Francisco and participants receiving education plus intensive case management and cognitive-behavioral programming in Los Angeles

Organization-described in-house study. A Five Keys media kit reports 26% recidivism for San Francisco graduates and 17% for the Los Angeles bundled program, compared with a 60%–70% statewide figure.

Our assessment. The media kit does not publish cohort dates, sample sizes, eligibility, a common recidivism definition, follow-up completeness, risk adjustment, or a contemporaneous comparison. Graduates are selected completers, and the Los Angeles result combines multiple services. The comparison cannot be interpreted as a Five Keys causal reduction.

Incarcerated adults in studies of adult basic, secondary/GED, postsecondary, and vocational education published from 1980 through 2011

Systematic review and meta-analysis of U.S. correctional-education evaluations. RAND found 30% lower odds of recidivism in the high-school/GED subset. Across the full correctional-education evidence base, its modeled three-year reincarceration risk difference was 12.9 points, but programs, outcomes, follow-up, and study quality varied; RAND identified curriculum, dosage, and program quality as unresolved evidence gaps.

Our assessment. This supports correctional education as a promising intervention class, not Five Keys' effect size. The studies predate current Five Keys delivery, mix prison contexts and program types, and do not establish San Francisco implementation quality, marginal cost, or philanthropic additionality.

3. Qualitative assessment

Accredited, flexible instruction is brought into jails and community settings, with transcript review, individualized graduation maps, GED/HiSET preparation, counseling, and reentry-linked support for learners whom traditional schools did not retain.

Key reservations

  • Five Keys combines accredited schools with reentry, workforce, housing, restorative-justice, and other services across multiple counties.
  • Cumulative graduates and annual GED completions are outputs without a reconciled eligible-student denominator or counterfactual.
  • The published 26% and 17% recidivism figures use an in-house comparison whose cohort, follow-up, and outcome definitions are not disclosed.
  • Comparing selected graduates with a statewide incarcerated population is vulnerable to selection, geography, calendar-time, and risk differences.
  • RAND's favorable intervention evidence is heterogeneous and does not identify Five Keys' curriculum, dosage, or local implementation effect.
  • The FY2024 audit isolates FKIH expense and ADA, but dividing those values yields only a historical cost per ADA-equivalent year—not cost per unique learner, marginal seat, or additional credential.
  • Public school funding and large service contracts may fund overlapping inputs; the extra activity and displacement created by a private gift are unknown.

Benefits not included in our estimate

  • The reported 26% or 17% recidivism rates as a causal Five Keys effect
  • The 56 GED completions as if all were caused by Five Keys or generated in San Francisco
  • Credits earned, attendance, instructional contact, and graduation ceremonies as final beneficiary outcomes
  • RAND's 12.9-point pooled reincarceration difference as if it were measured for Five Keys
  • Employment, earnings, avoided incarceration costs, public safety, family stability, confidence, and community spillovers
  • QALYs, WELLBYs, and mortality benefits inside the native credential model; the separate 10-QALY bridge applies an explicit external health transfer

4. What do you get for your dollar?

Our current model: about $167,000 per additional credential—conditional on a 10-point effect we cannot yet verify.

The cost anchor divides audited FY2024 FKIH total expense by audited annual ADA. It is a schoolwide historical intensity, not a marginal cohort budget. The 5%, 10%, and 20% credential-effect cases represent one additional credential per twenty, ten, or five ADA-equivalent years. They are decision scenarios, not confidence bounds. We do not use the reported 56 GED completions or Five Keys recidivism claims as causal effects.

A better life is our comparison unit of 10 additional quality-adjusted life years (QALYs), potentially spread across people. These are uncertain estimates, not measured returns or verified donation offers.

How we calculate the estimate

CONDITIONAL COST PER ADDITIONAL SECONDARY CREDENTIAL: $16,667 ÷ 10%
= $166,672

Model inputs and assumptions
FY2024 Five Keys Independence High total expense
$37,213,618 (range: Fixed published value). Audited combining schedule: $30,984,984 education plus $6,228,634 management and general. This is schoolwide historical expense, not incremental philanthropic cost. high for audited accounting; low for marginal use.
FY2024 annual average daily attendance
2,232.74 ADA (range: Fixed published value). Audited Schedule of Average Daily Attendance. ADA is an attendance-based funding denominator, not unique enrollment, dosage, or completed student-years. high.
Historical gross expense per ADA-equivalent year
$16,667 (range: Fixed published value). $37,213,618 divided by 2,232.74 annual ADA. This covers the whole school across counties, custody and community settings, and multiple educational outcomes. moderate for historical intensity; low for marginal cost.
Reported GED completers
56 people (range: Fixed published value). The 2024-25 FKIH LCAP reports 56 people passed all GED portions in 2022-23. It does not publish the eligible or tested denominator and is not used as a causal effect. moderate; descriptive only.
MFI-modeled additional credential probability per ADA-equivalent year
10% (range: 5%–20%). A transparent decision range, not a Five Keys estimate or confidence interval. The midpoint assumes one additional credential for ten fully funded ADA-equivalent years; the endpoints assume one for twenty or five. No credible Five Keys comparison currently identifies this effect. very low.

What would $100,000 imply at the historical cost ratio?

  • Favorable positive scenario: 1.2 additional credentials. 6 historical ADA-equivalent years · $83.34K per credential
  • Best guess: 0.6 additional credentials. 6 historical ADA-equivalent years · $166.67K per credential
  • Cautious positive scenario: 0.3 additional credentials. 6 historical ADA-equivalent years · $333.34K per credential

Uncertainty. The conditional dollar range excludes zero and harm. Five Keys reports no comparable group for credential completion, and ADA is not a participant count. If learners would have obtained credentials through another school, if short jail stays prevent sufficient dosage, if a gift displaces public charter revenue, or if an added ADA-equivalent year does not increase completion, additional credentials approach zero and no finite positive impact price exists.

Our current best estimate: about $4.9 million per better life (10 QALYs).

Publish an explicit, very-low-confidence decision estimate of approximately $4.9 million per 10 QALYs. The model divides Five Keys' $16,667 gross historical expense per ADA-equivalent year by a modeled 10% additional-credential probability and 0.34 QALY per additional credential, then multiplies by 10 QALYs. The 0.34-QALY midpoint applies an 80% transfer and causal discount to a peer-reviewed U.S. model's 1.7 discounted lifetime QALYs per additional high-school graduate. The source model is old and largely association-based; Five Keys serves older, justice-affected learners and includes equivalency credentials, so this is a judgmental donor estimate rather than a measured Five Keys health effect. The joint positive-effect sensitivity is approximately $490,000 to $66.7 million, and a true null remains plausible.

EXPLORATORY COST PER 10 QALYS · ONE BETTER LIFE: $16,667 ÷ (10% × 0.34 QALY) × 10
= $4.9M

QALY conversion assumptions
Gross historical cost per ADA-equivalent year
$16,667 (range: Fixed FY2024 ratio). Audited school expense divided by annual ADA; not a unique learner, marginal seat, or donor price. moderate historical; low marginal.
Additional credential probability per ADA-equivalent year
10% (range: 5%–20%). Existing MFI decision scenarios; Five Keys has not measured its causal credential effect. very low.
Lifetime QALYs per additional credential
0.34 (range: 0.05–1.7). The midpoint retains 20% of the published 1.7-QALY estimate: an explicit 80% discount for residual confounding, a 2007 model and mortality assumptions, adult and justice-affected learner transfer, diploma-versus-equivalency differences, and uncertain persistence. The 0.05-1.7 range is a decision range, not a confidence interval. very low transfer.
  • Optimistic joint bound: $490.21K. 20% credential effect · 1.7 QALY per additional credential
  • Published health model; central credential effect: $980.43K. 10% credential effect · 1.7 QALY per additional credential
  • Central transfer-discounted estimate: $4.9M. 10% credential effect · 0.34 QALY per additional credential
  • Cautious positive joint bound: $66.67M. 5% credential effect · 0.05 QALY per additional credential

The numeric range is conditional on two positive effects: Five Keys must cause additional credentials, and those credentials must cause additional lifetime health utility. Neither is measured for Five Keys. If either effect is zero, or a private gift displaces public funding rather than adding instruction, the impact price has no finite upper bound. The paper modeled lifetime health gains associated with producing an additional high-school graduate through smaller elementary-school classes. It did not study Five Keys, adult education, custody settings, GED/HiSET pathways, or philanthropy. Its education-health coefficients do not establish that Five Keys causes 1.7 QALYs per credential.

5. Funding and previous grants

There is no gift-restricted marginal expansion plan, causal credential effect, unique learner flow, public-funding displacement rule, or verified long-run education, employment, and custody follow-up.

The $100,000 scenario is illustrative. Five Keys does not publish a current, gift-restricted credential expansion plan, the seats or teacher time a private gift would add, the public charter reimbursement those seats would generate, or how a gift would avoid replacing city, county, state, federal, or restricted funding.

This review does not establish a verified marginal funding offer or a complete history of grants.

Donate

Opens the organization’s giving page. A general donation may not fund the specific activity modeled here; confirm allocation with the recipient.

6. Sources

  1. Financial Statements and Supplementary Information, June 30, 2024. Five Keys Schools and Programs / Gilbert CPAs. audited financial statements and school combining schedules. Published: 2024-12-16; retrieved: 2026-08-31.
  2. 2024–2025 Local Control and Accountability Plan. Five Keys Independence High School. school administrative reporting. Published: 2024-09-20; retrieved: 2026-08-31.
  3. Education program and service locations. Five Keys Schools and Programs. organization program description. Published: current program page; publication date not stated; retrieved: 2026-08-30.
  4. 2023–2024 Local Control and Accountability Plan. Five Keys Charter School. school administrative reporting. Published: 2023–24 planning and reporting cycle; retrieved: 2026-08-30.
  5. What Are the Five Keys?. Five Keys Schools and Programs. organization-reported cumulative output. Published: publication date not stated; retrieved: 2026-08-30.
  6. Media Kit: Results and Outcomes. Five Keys Schools and Programs. organization-described in-house comparison. Published: publication date not stated; retrieved: 2026-08-30.
  7. Evaluating the Effectiveness of Correctional Education. RAND Corporation / U.S. Bureau of Justice Assistance. systematic review and meta-analysis. Published: 2013; retrieved: 2026-08-30.
  8. FY2024 Form 990. Five Keys Schools and Programs / Internal Revenue Service. organization tax filing. Published: filed 2025-05-14; retrieved: 2026-08-30.
  9. Nonprofit prime-contract ledger: Five Keys Schools and Programs. City and County of San Francisco. public-contract accounting. Published: data as of 2026-08-20; retrieved: 2026-08-30.
  10. Health and Economic Benefits of Reducing the Number of Students per Classroom in US Primary Schools. American Journal of Public Health. peer-reviewed cost-utility model. Published: 2007-11-01; retrieved: 2026-08-31.
  11. Programs to Increase High School Completion: A Community Guide Systematic Health Equity Review. American Journal of Preventive Medicine. systematic review of high-school-completion interventions. Published: 2015-12-01; retrieved: 2026-08-31.
  12. New evidence of the effects of education on health in the US: Compulsory schooling laws revisited. Social Science & Medicine. quasi-experimental education-health study. Published: 2015-02-01; retrieved: 2026-08-31.

Annual expenses: years and sources

Average annual expenses (three consecutive fiscal years): $119,811,440. Organization size is separate from the modeled cost-effectiveness of a donation.

Five Keys Schools And Programs

Whole legal entity Form 990 Part IX total functional expenses; latest original reconstructed filings checked, fiscal-end year from printed reporting dates. Older fallback rows from IRS extracted API.

Latest original filings checked; ProPublica extracted API lags these original returns. FY2023 taken from prior-year total expenses column in FY2024 original return.