Methodology

This information is based on an evidence-based methodology

Each figure on this site is one row in a structured ledger: a venture, a metric, a year, a value, a source, and a confidence grade. This page is the whole rulebook.

The ledger

The ledger contains 383 metric rows across 26 public projects. Each row records one year of one metric in its native unit: pounds of food, program visits, volunteer hours, kilograms of CO2e, dollars raised, etc. Projects have their own source and one of four confidence grades: Verified (published or audited), Probable (derived from a verified figure by a standard conversion), Estimated, or Modeled. The site refreshes from the ledger frequently. If a number changes here, it changed in the ledger first, in version history.

How a dollar figure is made

Modeled social value for a project is calculated:

value × multiplier × CPI factor = modeled social value (2026 USD)

Social return on investment (SROI) states an outcome in dollars. It takes a count of something that happened, such as pounds of food delivered or hours volunteered, and multiplies it by an evidence-based price for that outcome. Each price comes from a published source and applies the same way to every row that uses it.

Each metric carries one evidence-based outcome price, the multiplier in the formula above, with a vintage year (the year its source estimated it) and a citation. Nothing else touches the number: no growth assumptions, no projections, no rounding up.

The Consumer Price Index for All Urban Consumers (CPI-U) is the Bureau of Labor Statistics measure of what urban households pay for goods and services over time. The CPI factor uses it to restate each price from its vintage year in 2026 dollars, referenced to May 2026 (335.123). That puts a 2019 price and a 2023 price in the same total.

The multipliers

What is priced and what is counted

Reach metrics counted without a dollar value: Active Chapters, Certified Brands, Certified Products, Charitable Contribution, Companies Transitioned to EO, Countries with Certified Brands, Events Held, Festival Selections, Grocery Value of Food Distributed, Growers Participating, Healthy Choice Food Boxes, Local Distribution, Member Companies, Miles Biked by Volunteers, NCGPs (No-Cost Grocery Programs), Number of Volunteers, People Served, Program Food Waste Diverted (Upcycled Certified), Research Fellows Funded, Self-Sufficiency and Nutrition Program Participants, Soft Commitments to the Fund (before the SPV pivot), Studies and Reports Published.

Meals Provided derives from pounds distributed at 1.2 pounds per meal, the Feeding America standard.

New Hires (Low-Income Communities) vintage 2012 · CPI x1.4596 · low confidence · citation
$90,000 per permanent jobs
EO Beneficiaries vintage 2013 · CPI x1.4386 · medium confidence · citation
$14,000 per worker
Health Care Costs Avoided vintage 2015 · CPI x1.4139 · low confidence · citation
$1,863 per person-years of improved health
Companies Educated about EO vintage 2023 · CPI x1.0998 · medium confidence · citation
$125 per company
Audience (Cultural Reach) vintage 2022 · CPI x1.1451 · medium confidence · citation
$38.50 per attendees / viewers
Cultural Attendees vintage 2022 · CPI x1.1451 · high confidence · citation
$38.50 per attendee
Volunteer Hours vintage 2024 · CPI x1.0683 · high confidence · citation
$33.49 per hour
Books Sold vintage 2019 · CPI x1.3108 · low confidence · citation
$12.00 per copy
Audience (Passive Media Reach) vintage 2026 · CPI x1 · medium confidence
$3.99 per viewer
Meals Provided vintage 2023 · CPI x1.0998 · high confidence · citation
$3.58 per meal
Attributable Food Waste Prevented (Upcycled Certified) vintage 2024 · CPI x1.0683 · low confidence · citation
$2.70 per pound
Food Waste Prevented vintage 2024 · CPI x1.0683 · medium confidence · citation
$2.70 per pound
Upcycled Food Production vintage 2021 · CPI x1.2368 · medium confidence · citation
$1.50 per pounds of upcycled product
Dollars Raised vintage 2025 · CPI x1.0409 · n/a confidence
$1.00 per USD
Modeled Value (Direct) vintage 2026 · CPI x1 · n/a confidence
$1.00 per USD (modeled, current dollars)
Wealth Created vintage 2024 · CPI x1.0683 · high confidence
$1.00 per USD
CO2e Avoided vintage 2020 · CPI x1.2949 · high confidence · citation
$0.21 per kg CO2 equivalent
CO2e Emitted vintage 2020 · CPI x1.2949 · high confidence · citation
$0.21 per kg CO2 equivalent
Public Art Viewers (Incidental) vintage 2025 · CPI x1.0409 · low confidence · citation
$0.00 per viewer passes

The CPI factors

How the factor is computed

Factor = 335.123 (May 2026) divided by the vintage year's annual average. Source: BLS CPI-U all items, series CUUR0000SA0. The 2025 annual average is the mean of the 11 published months; October 2025 was not published because of the federal funding lapse. Factors are refreshed when the reference month is re-based.

2012 dollars CPI-U annual average 229.594
x1.4596
2013 dollars CPI-U annual average 232.957
x1.4386
2019 dollars CPI-U annual average 255.657
x1.3108
2021 dollars CPI-U annual average 270.97
x1.2368
2022 dollars CPI-U annual average 292.655
x1.1451
2023 dollars CPI-U annual average 304.702
x1.0998
2024 dollars CPI-U annual average 313.689
x1.0683
2025 dollars CPI-U annual average 321.943
x1.0409

Why one pound of food carries more than one price

The same 7,007,257 pounds of rescued food are measured four ways on this ledger, and three of those are priced and summed. The rest is reported and deliberately left at zero. Which venture a pound belongs to decides where its grocery value lands: Denver Food Rescue's is reported at zero, and Fresh Food Connect's is still priced. That is the most important thing to know about the portfolio total, so it sits here rather than in a footnote.

The priced lenses come to $6.31 a pound and $44,220,786, which is about 47.3 percent of the portfolio. They are separate routes to one underlying number, the retail price of the food: the cost per meal is household grocery spending, ReFED's figure is foregone retail sales built from point-of-sale data, and the grocery value is that same shelf price again. Under the strict test, which asks whether the same value is counted twice for the same stakeholder, pricing them together does not pass.

I report them anyway, and I would rather say so here than bury it. Each answers a different question a reader might genuinely have, and collapsing to one throws away the others. What I will not do is imply they are independent. If you want the conservative reading of this portfolio, use the larger single lens rather than the sum: $22,990,498 against $93,473,860.

Two things sit outside this overlap on purpose. CO2e Avoided prices the greenhouse cost, which ReFED reports separately from the financial value, so it is a genuinely different quantity. Health Care Costs Avoided prices a health improvement as a cost avoided by insurers rather than as value delivered to the household, so it lands on a different payer and is not another slice of the same pound.

Put all of it together and one pound of rescued food carries $12.32 of modeled value across 5 lines:

Of that $12.32, the $6.31 described above is the overlap: one value counted twice. The rest lands on people other than the household, so it adds rather than repeats. I publish the whole stack because it is the clearest single measure of how much this ledger rests on one physical pound, and because a reader is entitled to check my arithmetic against the same 7,007,257 pounds each lens above is measured over. The home page's pounds figure is larger, because it adds the attributed slice of the Upcycled Certified program described below; these lenses deliberately exclude it, since none of them was measured on that tonnage. Health Care Costs Avoided is recorded on Denver Food Rescue rows only, so on that venture alone the stack runs slightly higher than this portfolio-wide figure.

That last line is the largest priced outcome here. It also holds the lowest confidence grade on the page. Its multiplier is an association in the research between food insecurity and health care spending. No study here measures what rescued food does to a person's health. Booking it as an avoided cost assumes the association is causal. It assumes as well that food moves a person from food insecure to food secure. Neither is established. A reader who wants the most cautious reading of this site should set the line aside.

Denver Food Rescue records 1,340,344 visits to its no-cost grocery programs. Visits are not people. The model divides them by 46.36, the organization's own 2023 ratio of visits to unique participants. That ratio comes from 241,186 visits and 5,203 people. What remains is 28,915 person-years. The model keeps 85 percent of those, the share of participants reporting better health from eating more fruit and vegetables. That share comes from Denver Food Rescue's participant survey. It then multiplies by 0.80 for attribution, the provider-level fraction NEF Consulting used for FareShare. It multiplies by 0.75 for deadweight, because some of these households would have found produce elsewhere. The remainder prices at $1,863.17 per person-year, from Berkowitz et al. (2018). That figure is in 2015 dollars and adjusts here to $2,634.37.

Two of those four inputs are measurements. Two are judgments. The visit count and the 46.36 divisor come from Denver Food Rescue's records. The 85 percent comes from its survey. The attribution fraction comes from a UK study of a different organization. The deadweight figure is mine, with no published number behind it. Each one prints on its own line here, so a reader who doubts one of them can redo the multiplication.

The Upcycled Certified line

The Upcycled Food Association wrote the Upcycled Certified standard and launched the program in 2021. Member companies do the diverting. Certified companies diverted 772,467 tonnes in 2021 and 2022. Those are the two years when the program existed and I ran the association. It sold the program to Where Food Comes From in 2024. Everything diverted after that stays unpriced here.

I price one thousandth of that tonnage. It comes to 1,702,998 pounds. Those price at $2.88 a pound after inflation adjustment, the same ReFED figure as every rescued pound on this site. They book avoided emissions at the same rate as the rest. Together the line is $5,710,334. The 0.001 is my judgment. No published figure supports it.

Published work puts the fraction higher. Garrett and colleagues (2016) measured potential additionality of two farm certification standards at 0.14. Probst and colleagues (2024) found under 16 percent of issued carbon credits represent real reductions. Keep a tenth of 0.14 for the association and the answer is 0.014.

Blackman and Naranjo (2012) explain why I go lower. Certification changes behavior where applicants do not already comply. It verifies behavior where they do. A company applies to certify a product it already makes.

The tonnage also runs larger than the food kept out of waste. Section 4.3.6 of the standard sets the counting rule. Input weight includes moisture removed later in drying or baking. The standard's own example diverts 35 tons of blueberry pomace, dries it to 10 tons, and reports 35. A certified ingredient's inputs count again inside a certified finished product. One mass of food can appear under two certificates.

This line uses one retention factor. The SROI Guide warns that attribution often sits inside a deadweight estimate. It says to take care not to take off more than you should. Splitting the 0.001 into two factors would deduct the same fact twice.

Two things cut against this line. Published social return studies of food rescue put deadweight at 60 percent or below, an order of magnitude from where I sit. Those studies price the work of collecting and moving food, and this prices a label on food somebody else diverted. Separately, of eighteen certification and convening bodies checked, none prices a share of its members' outcomes at all.

The Guide asks for a range. At 0.005 this line would be $28,551,671. At 0.014 it would be $79,944,678. The whole rest of this ledger models at $87,763,527.

What is never summed

The ratio

Modeled social value is stated in 2026 USD; dollars raised is stated at face value, per that metric's own methodology note. Dividing value by raised is the input-versus-outcome comparison the two figures exist for.

That works out to $93.5M of modeled value against $38M taken in: about $2.46 of modeled value for every dollar in.

Not every dollar in that denominator was given, and until September 2026 this page said it was. What the money actually is, across the portfolio: 91 percent a nonprofit total-revenue line from a Form 990, which splits donations from program income nowhere, so I cannot claim it as contributions even though most of it very likely is; 7 percent Fresh Food Connect's earned software revenue; 1 percent capital committed to Small Capital's fund; and well under one percent that I can name as a grant or a gift outright. Each venture page now states which of those its own figure is, rather than all of them saying "given" over whatever happens to be underneath.

Projections

Every public idea, built or not, can carry a researched projection: a conservative annual steady state, stated per year and per metric. Projections are a different class of data from records, and the site treats them that way: they are labeled projections everywhere they render, and they are never summed into the ledger's recorded totals, the leaderboard, or the ratio above.

A projection earns its number one of three ways. An idea that does not exist yet anchors to named comparables (capital to launch, projected annual value, projected people served), never a total addressable market. A built venture anchors to its own recorded track record: the average of its last few complete ledger years, carried forward flat unless a documented growth factor says otherwise. A venture with a stated goal reports the goal as a goal, with its target year, and the per-year rate derived from the two; reach goals stay unpriced where pricing them would overstate modeled value.

All three follow the ledger's evidence rule: every figure traces to a published comparable, to the venture's own ledger rows, or to a named anchor carrying a source, a vintage year, and a confidence grade. The first registered anchor is per-viewer cultural value for film and media work, $5 per viewer. That is a market-price floor set at standard digital rental pricing ($3.99 to $5.99), under half the 2024 average US cinema ticket of $11.51 (Cinema United). It runs about an eighth of the recorded cultural-reach multiplier above ($38.50 per attendee, vintage 2022), because an idea that does not exist yet has not earned the recorded rate. An idea I cannot anchor one of these ways ships without numbers until the research exists.

Sources

IRS Form 990 filings

For the nonprofits in the portfolio, the primary financial record is public. Each link goes to that organization's IRS Form 990 filings on ProPublica, built from the EIN on its ledger row.

Data refreshed Sep 24, 2026, 1:23 PM Mountain Time.

A note on what these numbers are

Every dollar figure on this site is a modeled estimate: a count of something that happened, priced with a published outcome price and restated in 2026 dollars. The sources and the arithmetic are on this page so anyone can reproduce a figure or argue with it. Nothing here is investment or financial advice, and nothing here is an audited statement of any organization's finances. The IRS Form 990 links above are where those live.

Keep reading

Always question the numbers.

If a figure looks wrong, it might be. Tell me and I will trace it.