Source and Claims Ledger

Reentry Financial Literacy Programs (shared across all three programs)

Last updated: June 24, 2026

This is the master list of every number and claim used in the program materials. Every claim is labeled as one of:

• SOURCED: backed by a named, credible source.
• NEEDS SOURCE: believable but not yet backed. Do not publish as fact until sourced.
• INTERNAL HYPOTHESIS: our own working assumption or program design choice. Label it as such.
• MARKETING TO SOFTEN: language that overpromises. Replace with the softer version shown.

Rule we follow everywhere: we do not use the claim “financial literacy reduces recidivism by 40%.” No direct, credible source was found for that exact claim. We replace it with softer language (see below).

1. Claims we WILL use (sourced)

Recidivism, general. About 68% of people released from state prison in 2005 (30 states) were arrested again within 3 years; 79% within 6 years. A later cohort (34 states, released 2012) showed 62% within 3 years.

• Label: SOURCED
• Source: U.S. Bureau of Justice Statistics (BJS), “2018 Update on Prisoner Recidivism: A 9-Year Follow-Up Period (2005–2014)”; and “Recidivism of Prisoners Released in 34 States in 2012.”
• Careful wording: say “arrested again,” not “returned to prison.” Re-arrest is not the same as re-conviction or re-incarceration. Note the range (62%–68%) rather than one fixed number.

Employment and reincarceration (correlation). People who could not hold a job after release had higher re-arrest rates than peers who stayed employed for a year. One analysis cites about 52% versus 16% over three years.

• Label: NEEDS SOURCE (citation not firm)
• Status: The exact 52% versus 16% figure traces to secondary summaries, not a confirmed primary source. It is removed from all public and funder-facing copy until a primary source is confirmed. The grant one-pager and sales page do not use it.
• Careful wording: this is a correlation. Employment and lower recidivism move together. Do not claim our program causes this gap.

Unemployment after release. The unemployment rate among formerly incarcerated people has been estimated around 27%, far above the general population, with wide gaps by race and gender.

• Label: SOURCED
• Source: Prison Policy Initiative, “Out of Prison & Out of Work.”

Banking access. Formerly incarcerated people are more likely to be unbanked than people who were never incarcerated. One survey found about 29% of justice-involved respondents were unbanked before incarceration, and unbanked status is common after release even among those banked before.

• Label: SOURCED
• Source: Financial Health Network survey of justice-involved consumers; Rice University / Social Science Research study, “Locked out of banking” (2024).
• Careful wording: use “more likely to be unbanked” and the ~29% figure. Do NOT use “78% without stable banking access” (see softened claims).

General unbanked rate, U.S. In 2023, about 4.2% of U.S. households (around 5.6 million) were unbanked.

• Label: SOURCED
• Source: FDIC, 2023 National Survey of Unbanked and Underbanked Households.

2. Claims we will SOFTEN (marketing to soften)

“Financial literacy reduces recidivism by 40%.”

• Label: MARKETING TO SOFTEN: do not use.
• No direct credible source found for this exact figure. The current FSA correctional page uses “40% lower recidivism with financial literacy programs” in its stats row. Remove it.
• Replace with: “Financial education can support reentry by improving money skills, reducing reliance on predatory services, and helping participants navigate employment, debt, banking, and benefits. Research links stable employment and financial stability to lower re-arrest rates, though no single program guarantees a specific drop in recidivism.”

“78% of returning citizens without stable banking access.”

• Label: MARKETING TO SOFTEN: no direct source found.
• Replace with the sourced banking claim above: formerly incarcerated people are more likely to be unbanked, with around 29% unbanked even before incarceration, and unbanked status common after release.

“Over two-thirds return to prison within 3 years.”

• Label: NEEDS SOURCE as written / SOFTEN wording.
• The two-thirds figure is for re-arrest, not return to prison. Replace “return to prison” with “are arrested again.”

“Break the financial cycle” / “Reduce recidivism through financial education” (headline promises).

• Label: MARKETING TO SOFTEN.
• Keep as motivation, not as a measured promise. Pair any recidivism mention with the softened language above. Never state a guaranteed percentage outcome.

3. Claims that NEED a source before we publish them as fact

• Specific local recidivism rates for any single facility or county. NEEDS SOURCE (use the facility’s own data or state DOC reports).
• Any dollar figure for “money lost to check cashers / payday loans per year per person.” NEEDS SOURCE (cite CFPB or state data if used).
• Claims about how many returning citizens have court-ordered debt or child support arrears. NEEDS SOURCE (cite a named study before stating a percentage).
• Any claim that our specific program improved outcomes. NEEDS SOURCE: we have no outcome data yet. Until a pilot is measured, describe the program as designed and evidence-informed, not proven.

4. Internal hypotheses (our design assumptions, labeled honestly)

• A 6-week banking and credit program is the most urgent first step because banking and ID problems block almost everything else after release. INTERNAL HYPOTHESIS.
• Peer facilitation (formerly incarcerated co-facilitators) improves trust and attendance. INTERNAL HYPOTHESIS, supported in general reentry literature but not yet measured in our program.
• Pre/post knowledge gains plus a 90-day behavior survey are enough to show early value to funders. INTERNAL HYPOTHESIS / design choice.

5. Standing language rules for all materials

1. No em dashes anywhere. Use periods, commas, or “and.”
2. Plain language, roughly 8th-grade reading level.
3. Respectful. No savior tone. No shame. No hype.
4. Every factual number carries a source or a label.
5. We say “educational content only, not financial, legal, or tax advice” on student-facing and funder-facing materials.
6. We never promise a specific recidivism reduction.