GiveBetter x Houston

Daya

Culturally specific survivor safety, counseling and practical support

Research time: ~8 min on GPT-6.1 Sol
  • Research — reviewed programs, finances and impact evidence.
  • Modeling — estimated costs, QALYs and uncertainty.
  • Source audit — checked claims, assumptions and calculations.

Updated: 2026-10-06

Donate

Summary

What do they do? Daya supports South Asian survivors of domestic and sexual violence in the Houston region. It combines confidential crisis support, licensed counseling, housing assistance and economic empowerment. Culturally specific advocacy and legal referrals help survivors pursue safety and practical stability.

Why we’re interested in this organization:

  • Culturally specific advocacy addresses language and access barriers.

  • Freshly checked program counts support distinct counseling, housing and work-resource mechanisms.

  • Comparative primary evidence informs integrated safety scenarios without treating services as causal outcomes.

Our main reservations:

  • Service groups overlap and completed dose, households and local causal effects remain unobserved.

  • Current costs and newer outputs differ in vintage; monetary assistance is not net new income.

  • The limited modeled reference is nonpositive under retained burden priors; broader safety and whole portfolio remain uncertain.

What do you get for your dollar? The current model estimates net harm after health and household-resource effects; no positive cost per DALY applies.

Additional integrated safety/mental-health, stable housing and net disposable work/assistance resources; output counts not causal outcomes.

1. What do they do?

Daya supports South Asian survivors of domestic and sexual violence in the Houston region. It combines confidential crisis support, licensed counseling, housing assistance and economic empowerment. Culturally specific advocacy and legal referrals help survivors pursue safety and practical stability.

2. Monitoring and information sharing

Freshly fetched and visually checked 2025 official report: 683 clients, 564 crisis/safety,75 counseling,75 housing,109 economic users; these overlap and are not distinct households or completed effective doses. County shares 63% Harris+20% Fort Bend+4% Brazoria/Galveston/Montgomery give .87 local share once; other Texas/out-of-state excluded. Assistance spending includes $110,000 housing, $28,000 legal, $12,000 transport and $47,000 essentials, not measured causal net resources. Request an overlap/dose ledger, local unique households, housing retention, symptom/utility follow-up, after-tax earnings and benefits lost, alternative services, exposure cost and retaliation/privacy outcomes.

3. Qualitative assessment

The original 2023 JAMA DVHF study compared 219 housing-first and 125 usual-service participants in a weighted nonrandom analytic cohort of 344 over two years. Housing, safety and mental-health advantages support integrated mechanisms, but confounding, Washington setting and few Asian participants limit transfer. Symptom scales are not measured QALYs or Daya income. PATH's original 2018 randomized trial compared specialist psychological advocacy with usual advocacy; improved distress/depression at 12 months but no demonstrated further-abuse difference, with 64% retention. Dose 8 sessions plus 2 follow-ups is not Daya's unspecified counseling dose. These results motivate overlap-bounded counseling/housing and broad-safety scenarios, not assigning full DVHF to all 683 or summing depression, anxiety and housing utility. Current services include licensed therapy, practical assistance, job-search support and legal referral, not staff attorneys. Existing victim agencies, public aid, family and employer support remain alternatives. Culturally appropriate access may add value without proving universal response.

4. What do you get for your dollar?

Initial $26.5M is retained as the complete historical health-and-income diagnostic: health +0.076447 and income −0.003090. Burden-only correction gives combined −0.852807; additional native-program restriction gives current health −0.077592, income −1.134818, total −1.212410, so no finite positive current price. This is not a measured harm finding about Daya or its whole portfolio. Full FY 2024 annual $1,946,855 expense is matched conditionally to FY 2025 service outputs; unmodeled legal, safety, other housing and prevention benefits remain unknown, not zero. Clinical envelope 75 assumes maximum overlap of 75 counseling/75 housing; range 75–150 is tested, replacing rather than adding two utility streams. Retained .8 household mapping, .35 additionality, .35 clinical response, .05 utility for .5 year are judgments, not trial coefficients. Broader 564 safety/crisis utility replaces that stream; low-response and high-intensity hypotheses bound possibilities but none establish that every brief-contact user got a housing-first dose. Central gain groups are bounded to 109 economic and 75 housing users with 75 assumed overlap; .8*.35*.25 resource-response scale gives 5.25 joint resource-success and 2.38 work-only proxies, not successes on all 683. Net work $300 and assistance $500 after alternative aid, taxes and benefit withdrawals remain hypotheses; positive overlap .5 before joint log, then full $100 participation/safety/opportunity costs. Other exposed household proxies retain negative costs; clinical harm .0005 applies all 546.4 exposures even when b=0. These unmeasured burdens are decisive; weighted burden (109 intensive proxies at$100, remaining at$10) still has nonpositive combined reference. The $100 uniform prior is not an observed average for brief helpline users. Distinct transfer counterparties and provider losses are localized once, with full losses, no overlap haircut. No grants' face value, gross salary or ordinary survivor earnings is counted as new social income. Payer zero is unknown-neutral, not measured zero. No-exposure/no-gift, financial-only, clinical-null, adverse, alternative, finite duration, overlap and volunteer-cost regimes are explicit. Broad high-dose positive sensitivity remains a hypothesis rather than a favorable central chosen to rescue price. Work resources are after childcare, commute, taxes and benefit withdrawal; the separately modeled participation cost must exclude any work expense already deducted. The income model uses one welfare-equivalent adult per household at $20,000 baseline, not every family member.

Model, assumptions and sensitivity

Full annual recipient expense versus overlap-bounded counseling/housing and restricted economic channels; other portfolio and next gift unknown

Reported 2025 county shares imply .87 Houston MSA; applied once

Exposed=N*unique; union=exposed*b; clinical=(clinicalN*unique*b*clinicalResponse*q*finite discounted T − full exposed*harm)*g. Work/assistance success partitions=program counts*unique*b*success with intersection once. One joint log per disjoint household partition of overlap-adjusted positive resources minus full costs; all other exposure retains costs. Distinct transfer payer loss local once. Price=10*C/(health+signed income) only if positive.

C
1946855 USD annual recognized resources (observed). Annual recognized tax expense plus netted direct event costs once [services] [finance]
N
683 annual reference service opportunities (judgment). 2025 683 overlapping service clients; household mapping and causal response judged, counseling subgroup 75 [services] [finance]
unique
0.8 households per patient proxy (judgment). Distinct household mapping prior, not clinical efficacy. [services] [finance]
b
0.35 incremental annual response share (judgment). Additional annual implementation versus public/insurance/provider alternatives, not an additional marginal-gift haircut. [services] [finance] [dvhf] [path]
q
0.05 QALY per person-year (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance] [dvhf] [path]
T
0.5 finite clinical years (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance] [dvhf] [path]
g
0.87 Houston MSA share (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
delay
0.1 years (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
discount
0.03 annual discount rate (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
success
0.25 additional resource success share (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
medicalGain
0 USD annual otherwise-paid care avoided (unknown). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
workGain
300 USD annual net disposable work resources (judgment). Judged usable work resources after taxes, childcare, commute, benefits withdrawn and alternative work; do not subtract the same expense again in participation cost. [services] [finance]
transferGain
500 USD one-time net additional household housing resources (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
overlap
0.5 independent positive benefit share (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
cost
100 USD one-time incremental household burden (judgment). Retained uniform100 incremental time/safety burden on all exposed; not observed brief-contact average. Weighted-intensity sensitivity tests this prior. [services] [finance]
baseline
20000 USD annual household consumption (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
harm
0.00 QALY loss per exposed person (judgment). Full .0005 one-off exposure privacy/safety harm judgment, independent of benefit response. [services] [finance]
payerLoss
0 USD loss per resource success (unknown). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
payerPeople
100 distinct payer households (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
payerBaseline
50000 USD annual payer household consumption (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
externalCost
0 USD external loss per exposed household (unknown). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
resourceYears
1 fraction of one-year gains realized (judgment). Explicit finite local hypothesis, not an empirical recipient coefficient. [services] [finance]
clinicalN
75 counseling clients within union (judgment). Maximum-overlap counseling/housing envelope75; no observed overlap. Range75–150 sensitivity replaces utility, not additive. [impact] [dvhf] [path]
clinicalResponse
0.35 share (judgment). Reported subgroup mapped to incremental clinical responders conditionally; not a measured causal effect. [impact] [services] [dvhf] [path]
resourceN
109 clients (observed). Observed economic-program output, not causal gain. [impact] [services]
transferN
75 clients (observed). Observed housing-program output, not causal net assistance. [impact] [services]
intersectionN
75 clients (judgment). Maximal program overlap judgment; zero-overlap sensitivity. [impact] [services]
intensiveN
683 clients (judgment). Central conservative all-exposure100 burden; weighted-intensity109 sensitivity. [impact] [services]
otherCost
100 USD per household proxy (judgment). Full one-off lower-intensity exposure burden prior; weighted sensitivity10. [impact] [services]

central: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

historical alpha central: Cost: $1.9M; Houston health QALYs: 0.08; all-population health QALYs: 0.09. Income-equivalent healthy years: 0.00. Complete initial diagnostic.

burden accounting only: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -0.78. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

maximum distinct clinical envelope: Cost: $1.9M; Houston health QALYs: 0.08; all-population health QALYs: 0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

broad safety integrated hypothesis: Cost: $1.9M; Houston health QALYs: 0.17; all-population health QALYs: 0.20. Income-equivalent healthy years: -1.13. Unsupported broad safety envelope replaces counseling/housing utility, not additive; no claim all564 received housing-first.

high intensity broad safety hypothesis: Cost: $1.9M; Houston health QALYs: 2.16; all-population health QALYs: 2.48. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

service intensity weighted burden: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -0.24. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

resource no overlap: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

clinical null: Cost: $1.9M; Houston health QALYs: -0.24; all-population health QALYs: -0.27. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

financial only: Cost: $1.9M; Houston health QALYs: 0.00; all-population health QALYs: 0.00. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

full unknown: Cost: unknown; Houston health QALYs: unknown; all-population health QALYs: unknown. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

current cost unknown: Cost: unknown; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

zero expansion: Cost: $1.9M; Houston health QALYs: -0.24; all-population health QALYs: -0.27. Income-equivalent healthy years: -1.19. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

no exposure: Cost: $1.9M; Houston health QALYs: 0.00; all-population health QALYs: 0.00. Income-equivalent healthy years: 0.00. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

no gift no exposure: Cost: $0; Houston health QALYs: 0.00; all-population health QALYs: 0.00. Income-equivalent healthy years: 0.00. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

true zero: Cost: $1.9M; Houston health QALYs: 0.00; all-population health QALYs: 0.00. Income-equivalent healthy years: 0.00. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

adverse: Cost: $1.9M; Houston health QALYs: -2.47; all-population health QALYs: -2.83. Income-equivalent healthy years: -1.26. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

higher net work: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.04. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

no resource success: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.19. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

no gains: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.19. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

negative overlap zero: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.24. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

negative overlap one: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.24. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

mixed sign: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.18. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

payer loss 500: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.16. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

external provider cost: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.61. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

full transfer incidence: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.16. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

annual full response: Cost: $1.9M; Houston health QALYs: 0.22; all-population health QALYs: 0.25. Income-equivalent healthy years: -1.04. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

half geography: Cost: $1.9M; Houston health QALYs: -0.04; all-population health QALYs: -0.09. Income-equivalent healthy years: -0.65. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

all unique: Cost: $1.9M; Houston health QALYs: -0.10; all-population health QALYs: -0.11. Income-equivalent healthy years: -1.42. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

zero resource duration: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.19. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

one year health: Cost: $1.9M; Houston health QALYs: 0.08; all-population health QALYs: 0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

two year health: Cost: $1.9M; Houston health QALYs: 0.39; all-population health QALYs: 0.45. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

volunteer inclusive cost hypothesis: Cost: $2.5M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

weak clinical: Cost: $1.9M; Houston health QALYs: -0.23; all-population health QALYs: -0.26. Income-equivalent healthy years: -1.13. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

strong usual services: Cost: $1.9M; Houston health QALYs: -0.19; all-population health QALYs: -0.22. Income-equivalent healthy years: -1.17. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

no incremental participation cost: Cost: $1.9M; Houston health QALYs: -0.08; all-population health QALYs: -0.09. Income-equivalent healthy years: 0.05. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

Counterfactual: Other victim services, public benefits, legal aid, family support and employment alternatives retained

Attribution: Annual incremental recipient response is distinct from cohort deduplication and geographic share; no claim of whole annual portfolio or marginal donation causal effect.

Primary comparative evidence supports mechanisms, not local utility or net income coefficients. Household/program intersection and full exposure burdens unmeasured; no finite positive current conditional reference. Unmodeled channels and marginal-gift EV remain unknown.

Sensitivity

  • historical alpha central: $26.5M per DALY. Complete initial diagnostic.
  • burden accounting only: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • maximum distinct clinical envelope: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • broad safety integrated hypothesis: no finite positive estimate. Unsupported broad safety envelope replaces counseling/housing utility, not additive; no claim all564 received housing-first.
  • high intensity broad safety hypothesis: $1.9M per DALY. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • service intensity weighted burden: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • resource no overlap: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • clinical null: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • financial only: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • full unknown: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • current cost unknown: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • zero expansion: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • no exposure: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • no gift no exposure: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • true zero: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • adverse: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • higher net work: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • no resource success: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • no gains: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • negative overlap zero: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • negative overlap one: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • mixed sign: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • payer loss 500: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • external provider cost: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • full transfer incidence: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • annual full response: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • half geography: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • all unique: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • zero resource duration: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • one year health: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • two year health: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • volunteer inclusive cost hypothesis: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • weak clinical: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • strong usual services: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.
  • no incremental participation cost: no finite positive estimate. Conditional finite annual partial channel, not measured QALYs, complete portfolio or next gift.

Unresolved inputs

  • Current unique completed household cohort and dose
  • Causal utility/duration and signed net resources versus alternatives
  • Full public/payer/partner opportunity incidence
  • Marginal donor response and complete portfolio outcomes

Health and income breakdown

For the modeled $1.9M: -0.08 health QALYs + -1.13 household-resource-equivalent years = -1.21 combined years. A nonpositive total has no positive-benefit price. Income equivalents are welfare comparisons, not observed clinical QALYs or DALYs.

Method and caveats.

5. Funding and previous grants

FY 2024 $1,928,451 tax expenses plus $18,404 netted direct event cost gives $1,946,855. FY 2023 $1,694,184 plus $18,861 gives $1,713,045; FY 2022 $1,417,570 plus $16,537 gives $1,434,107. Schedule D reconciles expense totals without a populated additional donated-resource adjustment. This is not evidence that volunteer resources cost zero. Newer FY 2025 outputs versus FY 2024 cost are a conditional scope mismatch. Reserves do not demonstrate an expansion bottleneck. Fresh root verification of the FY 2024 original confirms the accepted full-cost numerator. Fresh 2025 output PDF has no matching audited FY 2025 expense or current marginal-gift bottleneck. Recipient assistance expense is already in C; separate payer-incidence scenarios represent displaced public/private alternatives, not adding the same recipient grant expense twice.

Annual expenses

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

  • FY 2022: $1.4M; org:daya, 12-month period, Annual recognized tax expense plus netted direct event costs once. Source
  • FY 2023: $1.7M; org:daya, 12-month period, Annual recognized tax expense plus netted direct event costs once. Source
  • FY 2024: $1.9M; org:daya, 12-month period, Annual recognized tax expense plus netted direct event costs once. Source

6. Sources

  1. Current culturally specific survivor services. Daya. Published: not stated; retrieved: 2026-10-06.
  2. Original 2025 annual report. Daya. Published: not stated; retrieved: 2026-10-06.
  3. Original FY2024 IRS Form 990. IRS. Published: not stated; retrieved: 2026-10-06.
  4. Original FY2023 IRS Form 990. IRS. Published: not stated; retrieved: 2026-10-06.
  5. Original FY2022 IRS Form 990. IRS. Published: not stated; retrieved: 2026-10-06.
  6. Current recipient donation route. Daya. Published: not stated; retrieved: 2026-10-06.
  7. Coefficient Giving income welfare conversion. Coefficient Giving. Published: not stated; retrieved: 2026-10-06.
  8. Original comparative DV Housing First study: two-year safety and housing mechanisms. JAMA Network Open / Sullivan et al.. Published: 2023-06-26; retrieved: 2026-10-06.
  9. Original PATH psychological advocacy randomized trial. PLOS One / Ferrari et al.. Published: 2018-11-27; retrieved: 2026-10-06.