GiveBetter x Boston

Community Servings

Medically tailored meals, groceries, nutrition services, training and policy

Research time: ~3 min on GPT-6 Astra Low + ~6 min on GPT-6 Astra Light
  • Source audit — checked claims, assumptions and calculations.
  • Research — reviewed programs, finances and impact evidence.
  • Modeling — estimated costs, QALYs and uncertainty.

Updated: 2026-09-29

Donate

Summary

What do they do? Community Servings delivers medically tailored meals and groceries to people with serious illness across Massachusetts and Rhode Island. Registered dietitians tailor food to clients’ medical needs and provide nutrition counseling and education. The organization also supports clients’ families and runs workforce training and food-policy programs.

Why we’re interested in this organization:

  • The intervention has a randomized trial conducted with the actual provider, plus a larger recent Medicaid evaluation.

  • Original tax returns and a consolidating audit permit recipient-level cost and financing review.

  • County-level service counts establish substantial metro reach; a reported waitlist supports unmet demand.

Our main reservations:

  • Mental-health days are not QALYs: the utility-day bridge remains an explicit prior.

  • Current meals include family support and varying duration, not thousands of independent completed trial courses.

  • Waitlist size, geographic service counts and historical outcomes do not identify what an additional unrestricted gift buys. The approximately $314 million per 10 Boston-metro QALYs figure is a very weak-confidence partial-pathway scenario, not a whole-portfolio or incremental-gift price.

What do you get for your dollar? $31.4M per DALY in Boston. 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. Conditional partial diabetes mental-health pathway; whole-recipient cost, very weak-confidence utility prior.

FY 25: 1,221,750 meals; 7,973 individuals; 15% primary diabetes diagnosis. Trial: 3.94 fewer days when mental health interfered with quality of life per 30 days during 12-week on/off periods. These populations and denominators are not interchangeable.

1. What do they do?

Community Servings, Inc., EIN 22-3154028, is the recipient, not merely its meal program. It also provides groceries, nutrition counseling, workforce development and policy work. Its service area includes Massachusetts and Rhode Island, so all health cannot be assigned to Boston. Who we serve

2. Monitoring and information sharing

A 44-person randomized crossover diabetes trial compared 12 weeks of 10 meals/week against usual care plus a healthy-eating brochure. The secondary endpoint was days when mental health interfered with quality of life: 5.65 versus 9.59 days per 30 days during the 12-week on/off periods. This is not direct preference-based utility. Trial The 2026 Medicaid evaluation compared 1,866 recipients with 1,372 eligible nonrecipients; adjusted hospitalization and ED rates were lower during a mean 6.7 months of enrollment. It is observational and does not establish mortality or QALY gains; no savings offset is used. Study

3. Qualitative assessment

Tailored food can relieve food insecurity and treatment-related dietary burden. The current population differs from the historical trial's food-insecure adults with poorly controlled diabetes. The FAME-F trial found no meaningful mental-health or utility differences between delivery/dose variants, but every arm received meals; it cannot estimate meals versus no meals. FAME-F CDC utility crosswalks require more than a single group-average mental-health-days contrast. We therefore expose, rather than disguise, a .20 utility-day prior. CDC methods

4. What do you get for your dollar?

The central partial-pathway model retains $22,289,498 whole-recipient FY 2025 gross expense. It uses 610.875 realized 12-week diabetes course-equivalents, not meals as independent people. With the observed 3.94-day mental-health contrast, .20 utility-day prior, .5 transfer, .5 financing additionality and 6139/7973 geographic share, it yields 0.710333 metro QALYs, or $31.4M per DALY. Benefits stop after 12 weeks; no avoided hospitalization, mortality, hypoglycemia, caregiver or policy credit is added. All other portfolio net health is provisionally zero, not a lower-bound claim. This is a very weak-confidence analytical scenario accepted only as a conditional alpha scenario, not a precise causal estimate.

Model, assumptions and sensitivity

Entire legal recipient gross FY 2025 Form 990 expense; noncash/depreciation retained. Does not subtract healthcare savings; not full social cost including unrecognized volunteer opportunity cost.

Boston-Cambridge-Newton MA-NH CBSA 14460 under OMB 23-01. Numerator county clients: Middlesex 1294+ Essex 1102+ Suffolk 2468+ Norfolk 828+ Plymouth 447=6139. Denominator 7973 total individuals; mapped counties sum 7893, leaving 80 unresolved. No assumed New Hampshire benefit. Applying client share to diabetes meal health is a transfer judgment. Rockingham and Strafford are the two New Hampshire counties in the MSA; no verified clients there are credited.

N=min(M*primaryMealShare*diabetesShare/(mealsPerWeek*weeks),people*diabetesShare)*courseRealization; Q_all=financialAdditionality*(N*mentalDaysDifference/30*utilityPerAffectedDay*(benefitWeeks*7/365.25)*transfer+otherQalys); Q_metro=g*Q_all; price10=10*C/Q_metro only for positive net health. benefitWeeks=12 centrally.

C
22289498 USD (observed). Whole legal-recipient FY 2025 Part IX 21,894,207 plus 395,291 direct event expense. Includes capital planning and affiliate transfer; no meal-program cost carveout. [irs25]
M
1221750 meals/year (observed). FY 2025 meals prepared/delivered, not people or a disease-specific dose count. [annual25]
people
7973 individuals/year (observed). Reported total individuals; not assumed all primary diabetes patients. [annual25]
primaryMealShare
0.5 fraction (judgment). Half of meals assigned to primary patients. Family delivery share does not identify this fraction; leaves caregiver/dependent and ambiguous meal allocation uncredited. [annual25] [irs25]
diabetesShare
0.15 fraction (judgment). 15% primary-diagnosis diabetes observed; applying the person mix to meal exposure is a separate unverified allocation assumption. [annual25]
mealsPerWeek
10 meals/week (observed). Trial intervention dose, not evidence of dose completion by current clients. [rct]
weeks
12 weeks (observed). Trial intervention duration; model credits at most 12 weeks and no residual benefit. [rct]
courseRealization
0.8 fraction (judgment). 80% of dose-equivalent capacity assumed to represent adequate continuous trial-like courses; no current patient-duration distribution.
mentalDaysDifference
3.94 days per 30 days (observed). 9.59 minus 5.65 days when mental health interfered with quality of life per 30 days in the 12-week randomized crossover on/off periods; secondary self-report endpoint, not generic mentally unhealthy days. [rct]
utilityPerAffectedDay
0.2 utility decrement per affected day (judgment). Explicit .20 prior: a day no longer experiencing mental-health interference with quality of life gains .20 of a full-health day. Not measured EQ-5 D or a validated application of a CDC crosswalk; test 0,.05,.4.
transfer
0.5 fraction (judgment). 50% replication/current-severity transfer, including secondary-outcome uncertainty, small historical trial and different current diabetes severity/food insecurity.
financialAdditionality
0.5 fraction (judgment). 50% effective activity survives replacement by insurance, contracts, other donors and reserves. Waitlist indicates demand but does not identify this elasticity. [funding] [audit25]
g
0.7699736611062336 fraction (judgment). 6139/7973 client-count anchor: five Boston MSA MA counties; unallocated 80 and NH counties get no credit. Assumes diabetes exposure and health yield have same residence mix; actual health share unknown. [annual25]
otherQalys
0.00 net QALYs before financing (judgment). All other portfolio net health set to 0 for a partial pathway, not a claim of no benefit or a proven lower bound.

Weak-confidence diabetes mental-health pathway: Cost: $22.3M; Boston health QALYs: 0.71; all-population health QALYs: 0.92. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=313789229.8495234; no post-service persistence, hospital/mortality credit or cost savings offset.

Utility prior .05: Cost: $22.3M; Boston health QALYs: 0.18; all-population health QALYs: 0.23. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.05,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=1255156919.3980937; no post-service persistence, hospital/mortality credit or cost savings offset.

Utility prior .4: Cost: $22.3M; Boston health QALYs: 1.42; all-population health QALYs: 1.85. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.4,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=156894614.9247617; no post-service persistence, hospital/mortality credit or cost savings offset.

No causal mental-health change: Cost: $22.3M; Boston health QALYs: 0.00; all-population health QALYs: 0.00. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":0,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=null; no post-service persistence, hospital/mortality credit or cost savings offset.

Gift replaces existing funding: Cost: $22.3M; Boston health QALYs: 0.00; all-population health QALYs: 0.00. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=null; no post-service persistence, hospital/mortality credit or cost savings offset.

Two additional days/month of mental-health interference with quality of life: Cost: $22.3M; Boston health QALYs: -0.36; all-population health QALYs: -0.47. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":-2,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=null; no post-service persistence, hospital/mortality credit or cost savings offset.

Five other-portfolio QALYs harm: Cost: $22.3M; Boston health QALYs: -1.21; all-population health QALYs: -1.58. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":-5}; realized 12-week course equivalents=610.875; price10=null; no post-service persistence, hospital/mortality credit or cost savings offset.

Only one quarter of equivalent courses realize trial exposure: Cost: $22.3M; Boston health QALYs: 0.22; all-population health QALYs: 0.29. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.25,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=190.8984375; price10=1004125535.5184749; no post-service persistence, hospital/mortality credit or cost savings offset.

Only half health accrues in metro: Cost: $22.3M; Boston health QALYs: 0.46; all-population health QALYs: 0.92. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.5,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=483218884.245886; no post-service persistence, hospital/mortality credit or cost savings offset.

90% metro health: Cost: $22.3M; Boston health QALYs: 0.83; all-population health QALYs: 0.92. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.9,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=268454935.6921589; no post-service persistence, hospital/mortality credit or cost savings offset.

Only 10% financing additionality: Cost: $22.3M; Boston health QALYs: 0.14; all-population health QALYs: 0.18. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.1,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=1568946149.2476172; no post-service persistence, hospital/mortality credit or cost savings offset.

Higher allocation, realization and transfer: Cost: $22.3M; Boston health QALYs: 9.52; all-population health QALYs: 12.36. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.67,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":1,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.4,"transfer":1,"financialAdditionality":1,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=1023.215625; price10=23417106.70518832; no post-service persistence, hospital/mortality credit or cost savings offset.

Effect only last four weeks: Cost: $22.3M; Boston health QALYs: 0.24; all-population health QALYs: 0.31. {"C":22289498,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0,"benefitWeeks":4}; realized 12-week course equivalents=610.875; price10=941367689.5485704; no post-service persistence, hospital/mortality credit or cost savings offset.

25% additional uncounted resource cost: Cost: $27.9M; Boston health QALYs: 0.71; all-population health QALYs: 0.92. {"C":27861872.5,"M":1221750,"people":7973,"primaryMealShare":0.5,"diabetesShare":0.15,"mealsPerWeek":10,"weeks":12,"courseRealization":0.8,"mentalDaysDifference":3.94,"utilityPerAffectedDay":0.2,"transfer":0.5,"financialAdditionality":0.5,"g":0.7699736611062336,"otherQalys":0}; realized 12-week course equivalents=610.875; price10=392236537.3119043; no post-service persistence, hospital/mortality credit or cost savings offset.

Counterfactual: RCT contrast already includes usual care and brochure; do not apply a second ordinary-care subtraction. Transfer covers current population/trial uncertainty. Financing additionality separately discounts replacement by insurance/contracts/donors. Courses are annual-flow equivalents; no lifetime or repeated-year credit.

Attribution: Primary-patient allocation, disease-dose allocation and course realization are unverified; count-based geography is not directly health-weighted. Utility-day prior is not an observed utility gain. Central other portfolio 0 is neither total-health estimate nor proven lower bound.

Very weak-confidence deterministic scenarios, not probability-weighted expectation or confidence limits. Source evidence supports native outcomes more strongly than the QALY price. Independent review accepts conditional alpha publication with these scope limits; donor status remains HOLD.

Sensitivity

  • Utility prior .05 quadruples price versus .20; .4 halves it. Zero utility or no gift response eliminates a finite positive price.
  • A reversal to two additional affected days/month yields negative health; negative results are retained.
  • Halving metro health share does not halve expense. Unknown 80 residents and unequal diabetes/meal distribution create additional geographic uncertainty.
  • One 12-week course maximum is credited per modeled person; dose-equivalent caps prevent treating meals or recurring courses as independent beneficiaries.
  • A 5-QALY other-portfolio harm before financing reverses the central sign. Uncredited clinical benefits could substantially increase value; the central pathway cannot rank the whole portfolio.
  • Four-week duration credits only a third of central health; 25% additional resource costs raises price 25%. No long-term persistence is assumed.

Unresolved inputs

  • Current primary diabetes patient count, meals delivered/consumed, continuous 12-week course completion and county-linked dose.
  • Preference-based utility measurements or valid patient-level crosswalk inputs, replication and adverse outcomes.
  • Current dated waiting list, costed unfunded capacity plan, financing replacement, and uptake per incremental gift.
  • Whole-portfolio health and harms, volunteer opportunity costs and metro-specific disease/utility shares.

5. Funding and previous grants

FY 2025 Part IX expense 21,894,207 plus netted direct event 395,291 gives 22,289,498; FY 2024 similarly 19,432,858+379,657=19,812,515. These retain the legal recipient's whole cost, including capital planning and transfers; consolidated affiliate spending is not added again. FY 25 return FY 24 return The audit reports $8.195 million consolidated financial assets available within one year and an increased credit line; net assets are not all liquid. Audit A 2025 management update reported 215+ waitlisted people, critical/end-stage-only enrollment and an expected $500,000 local-food grant canceled. It does not establish today's waitlist or a priced unfunded tranche. Funding update Donor recommendation HOLD until capacity, replacement and local beneficiary response are verified. Official donation page exists; checkout was not transacted.

Annual expenses

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

  • FY 2024: $19.8M; Community Servings, Inc., 12-month period, Original Form990 Part IX plus Part VIII direct event expenses; gross legal-recipient expense, including noncash expense, capital planning and affiliate grants. Source
  • FY 2025: $22.3M; Community Servings, Inc., 12-month period, Original Form990 Part IX plus Part VIII direct event expenses; gross legal-recipient expense, including noncash expense, capital planning and affiliate grants. Source

6. Sources

  1. FY2025 original Form 990 (tax year 2024). Community Servings / IRS. Published: not stated; retrieved: 2026-09-29.
  2. FY2024 original Form 990 (tax year 2023). Community Servings / IRS. Published: not stated; retrieved: 2026-09-29.
  3. FY2025 audited consolidating financial statements, FY2024 comparative. Community Servings and affiliate / independent auditor. Published: not stated; retrieved: 2026-09-29.
  4. FY2025 annual report. Community Servings. Published: not stated; retrieved: 2026-09-29.
  5. Medically Tailored Meal Delivery for Diabetes Patients with Food Insecurity: Randomized Cross-over Trial. Berkowitz et al. / Journal of General Internal Medicine. Published: not stated; retrieved: 2026-09-29.
  6. Medically tailored meals receipt and healthcare utilization and costs in Massachusetts Medicaid demonstration. Hager et al. / Nature Medicine. Published: not stated; retrieved: 2026-09-29.
  7. Designing Medically Tailored Meal Interventions: FAME-F Factorial Randomized Clinical Trial. Berkowitz et al. / Circulation Cardiovascular Quality and Outcomes. Published: not stated; retrieved: 2026-09-29.
  8. Diabetes State Burden Toolkit Technical Report. US Centers for Disease Control and Prevention. Published: not stated; retrieved: 2026-09-29.
  9. Update: Growing Demand Amid Funding Challenges. Community Servings. Published: not stated; retrieved: 2026-09-29.
  10. Who We Serve. Community Servings. Published: not stated; retrieved: 2026-09-29.
  11. Donate. Community Servings. Published: not stated; retrieved: 2026-09-29.