# Food Distribution¶

This manual describes the step-by-step procedures by which ImpactMatters analyzes food distribution nonprofits. We recommend reading the program analysis methodology first.

## Number of meals served¶

### Selection of outcome figure¶

No annual figure: In some cases, nonprofits will not report an annual figure, but report something like meals served per day. In these cases, if they also report something like days they are open (or if we can infer this information), we multiply the two figures to generate an annual figure. If the nonprofit has not reported the number of days it is open, we multiply the daily figure by 365 days in a year (or weekly figure by 52 weeks in a year, and so forth). Another example is if a nonprofit reports how many meals it serves to one beneficiary, and how many (non-unique) beneficiaries visited the nonprofit that year.

No year-specific figure: Sometimes a nonprofit will not report the meals it served in a given year, but instead state a general figure on the website. For example, “We serve 10 million meals a year.” Note that in this example, no year is specified. In these cases, we assume this number is representative of the meals they served in the year for which we have financial information. We apply the general outcomes figure to the year-specific cost figure, and make a note that we did so.

Meals vs. pounds for base figure: If a nonprofit reports both meals and pounds distributed, we choose the meals figure unless there is a reason that the pounds figure is a better option. In this case as well as generally, we take a nonprofit’s reported figures at face value. Some reasons we would choose the pound figure: if the meals figure is clearly a “ballpark” approximate and the pounds figure is more precise (e.g., “1,589,471 pounds” is preferred over “approximately more than a million meals”); if a nonprofit reports using pounds across most of its programs, which we aggregate; if we have reason to believe the nonprofit miscalculated meals. If we choose the pounds figure over the meal figure, we explain our reasoning. If using pounds reported by the nonprofit, we convert pounds to meals by dividing by 1.2 (the U.S.D.A. estimates the average weight of a meal at 1.2 pounds), unless the nonprofit reports a different conversion rate.

Converting from dollars to pounds: We occasionally need to calculate the number of meals by converting a nonprofit’s food expenses to pounds, and then to meals, using a dollar per pound conversion rate supplied by the nonprofit. This is rare, however, since we only do so in cases where the conversion rate clearly applies to all types of food included in the cost figure. For example, community donated food is often valued at a different rate than U.S.D.A. commodities, and purchased food may be valued using a different method altogether. We would only convert from dollars to pounds if the nonprofit explicitly stated “we value all food — including donated, U.S.D.A. commodities, and purchased — at \$1.60 per pound” or if it supplied different rates for different sources of food. A statement along the lines of “we value food at \$1.60 per pound” would not be sufficient. This information is usually found in the notes to the nonprofit’s audited financials.

No meals or pounds figure: Sometimes a nonprofit will not report meals or pounds of food distributed, but something like the number of backpacks of food distributed to children. In many of these cases, we are not able to come to a usable meals figure. In some cases, however, we can. Take the backpacks example. If a nonprofit specifies that each backpack contains five meals, then we can generate a final meals figure. If it does not specify number of meals in a backpack, we cannot. Another example is if a nonprofit only reports the number of Commodity Supplemental Food Program boxes it distributes. We know that Commodity Supplemental Food Program boxes are roughly 30 pounds.1 Therefore, we can convert the box figure to a pounds figure and thus ultimately to a meals figure. If any of these cases arise, we appropriately document.

### Irrelevant outcomes included¶

Background: In some cases, a nonprofit runs both a food distribution program as well as another program whose primary goal is not to distribute food but that does provide food to beneficiaries as a complementary activity, such as a job training program that serves lunch to trainees. Or, it may run another program that distributes food under very different conditions, such as disaster relief. Such programs are irrelevant to our cost-effectiveness calculation for the food distribution program(s) under analysis.

We continue to complete a cost-effectiveness estimate for the program of interest only if:

1. Both meals and costs of irrelevant programs can be excluded with very few or no assumptions, we complete a cost-effectiveness estimate for the program of interest.

2. Meals from irrelevant programs can’t be excluded from the total number of meals but likely aren’t substantial, and if costs from irrelevant programs can be excluded. We add those costs back in.

3. Meals from irrelevant programs can be excluded, and if costs from irrelevant programs can’t be excluded but likely aren’t substantial. We add those meals back in.

4. Meals from irrelevant programs can’t be excluded but likely aren’t substantial, and if costs from irrelevant programs can’t be excluded but likely aren’t substantial.

## Program costs¶

### Total program costs¶

Choice of summary line item: We are careful to ensure that we always select the correct total figure for a given line item. Three protocols we follow:

1. The program expenses are taken after depreciation is accounted for, if specified.

2. We take the total of both restricted and unrestricted revenue.

3. When audited financials include two years, we make sure we are using numbers for the year under analysis.

990 vs. financial statement: We aim to only use one financial document for the base of the analysis. Generally, we choose the financial statement over the 990. Occasionally we need to take pieces of information from both financial documents; if we do, we document accordingly. Note that discrepancies between the 990 and financial statement are reported and explained by the nonprofit on Schedule D of the 990, part XI (which explains revenue discrepancies) and part XII (which explains expense discrepancies.)

We prefer the document that, in rough order of prioritization:

1. Has dates that line up better with the dates of the outcomes.

2. Separately reports expenses by program.

3. Has more detailed information on line items, such as the value of donated food distributed or of administrative grants from the government.

4. Includes (excludes) more relevant (irrelevant) costs detailed in Schedule D of the 990.

5. Is itself the source of the outcomes figure.

Exclusion of unrelated programs: As described in the Cost-Effectiveness Methodology for Food Distribution Programs, many nonprofits run supplementary programs in addition to their core food distribution operation. Unless otherwise specified by the nonprofit, we assume that 5 percent of total program cost was spent on each unrelated program. The most common unrelated programs are:

1. Nutrition education (e.g., Cooking Matters)

2. SNAP outreach

3. Social services, case management

4. Job training in culinary services

5. Capacity building for partner agencies, if significant

6. Food voucher programs

7. Disaster relief

The following are generally not eligible for cost adjustment:

1. Food rescue

2. Food insecurity advocacy, if not a substantial portion of the nonprofit’s work

3. Community garden if we believe the garden is supplying food for the food bank or beneficiaries directly

### Donated food¶

Background: We assign a cost of \$0 to donated food distributed. The ideal method relies on the nonprofit reporting the value of donated food on its expense statement (rather than on its revenue statement). This implies that the donated food was distributed. When this information is available (and it often is for food banks), we deduct that line item from total program expenses.

Note that nonprofits filing Form 990-EZ are not required to report donations in-kind. Unless the nonprofit indicates otherwise, we assume no in-kind donations are included in expenses. We therefore do not adjust their expenses in this regard.

Donated food contributions versus distributed food expenses: When that information is not available, we have two options: Deduct the value of donated food reported as revenue (not expenses); or Deduct the value of distributed (not just donated) food reported as an expense. We compare the two numbers and deduct the smaller one.

If the smaller one is the distributed food expense, then we check to see whether the nonprofit reports what proportion of distributed food was donated versus purchased. If that proportion is available, we multiply the proportion by the value of distributed food. Note that this assumes that the dollar value per pound of donated food is equal to the dollar value per pound of purchased food.

Information for option 1 (donated food received) can be found on a few places in the 990. It can be found on Schedule M Noncash Contributions as the line item “Food inventory.” It can also be found on the Statement of Revenue, line 1g, which records the total value of noncash contributions, though this option is less preferred since it may include the value of nonfood items. On financial statements, the value of donated food is often reported on the statement of activities as revenue or support.

If a nonprofit reports noncash contributions on line 1g of the Form 990, it almost always breaks those contributions down on Schedule M. In these cases, we refer to schedule M. In the rare cases that the nonprofit records noncash contributions on line 1g but not Schedule M, we cannot know whether the noncash contributions contain more than the value of donated food. If this is likely, for instance if a nonprofit runs other sizable programs besides a soup kitchen, we estimate the portion of noncash contributions that is likely to be food by multiplying total noncash contributions by the ratio of food distribution-related expenses to total program service expenses. We deduct this amount from total program expenses.

Information on option 2 (food distributed) can be found on the Statement of Functional Expenses on both the 990 and financial statement.

Spoiled food: If specified as a line item and included in total program expenses, we deduct the value of donated food that spoiled. If there is only a line item that reports spoiled food and doesn’t specify whether it was donated, we deduct all of it on the assumption that the nonprofit purchased food to fill a real need in inventory, and so purchased food would unlikely be left to spoil. The exception to this is if the nonprofit operates its own farm or garden, where the chance of spending money on food that eventually spoils is higher. In that case, we do not deduct the spoiled food line item.

### Donated labor¶

Background: We assign a cost of \$0 to unskilled volunteer labor contributed to the nonprofit. There is no need to subtract the cost of volunteer services from the program expenses since Generally Accepted Accounting Principles dictate that the value of volunteer services need only be recorded as income under two conditions: (1) if “the volunteer services create or enhance a nonfinancial (physical) asset — for example, your volunteers build a shelter for the homeless” or (2) “the volunteer services involve specialized skills that would otherwise have to be paid for by your nonprofit if volunteers were not available — for example, volunteer services performed by lawyers, accountants, doctors, electricians, and other professionals.”2 We do aim to include the cost of specialized donated labor. By Generally Accepted Accounting Principles, it should already be included as an expense in total program expenses.

### Earned revenue¶

Background: We aim to deduct from total program expenses any revenue that is generated by any program included in our analysis, whether that included program is a food distribution program or an unrelated program for which we could not exclude costs. To determine whether a form of revenue is generated by an included program, we ask ourselves: if this program didn’t exist, could the nonprofit still generate this revenue? If not, then we deduct that revenue. We never deduct revenue along the lines of “unrelated business income.”

Membership dues: Given sufficient information, when calculating the cost-effectiveness of a food bank, we deduct so-called “membership dues” from total program expenses. Membership dues are fees paid by soup kitchens and food pantries to receive food from food banks, and are often labeled “shared maintenance fees” or “handling fees” in financial statements (usually in Statement of Activities) and Form 990s (usually in Statement of Revenue).

When calculating the cost-effectiveness of a soup kitchen or food pantry, we deduct from total program expenses any membership fees paid by the pantry or kitchen. This is to avoid double counting food bank partner costs when we add partner costs back in to our estimate using a different method described below. This information is not typically available.

Background: The U.S. Department of Agriculture (U.S.D.A.) runs three food distribution programs: Commodity Supplemental Food Program (C.S.F.P.); Food Distribution Program on Indian Reservations (F.D.P.I.R.); and The Emergency Food Assistance Program (TEFAP). These Federal programs provide State and local agencies — including food banks, pantries and kitchens — with packages of groceries for special populations (called commodities) as well as grants to cover associated administrative costs. We call these administrative grants.

Note that while we assign a cost of \$0 to general donated food from individuals and corporations, we count the full cost of U.S.D.A. commodities and associated administrative grants. See Cost-Effectiveness Methodology for Food Distribution Programs for more. We aim to count the cost of U.S.D.A. commodities and administrative grants exactly once, as a cost to the government (not the nonprofit). To do so, we calculate the cost to the government to operate these programs and apply it to partner costs, described in more detail under Partner Costs. To avoid double counting these costs, we deduct the value of U.S.D.A. commodities and the expended administrative grants from the nonprofit’s costs. We first discuss our method for deducting administrative grants, then for deducting U.S.D.A. commodities.

We are only concerned with government administrative grants and commodities when we analyze food banks. We do not account for government costs when we analyze soup kitchens and food pantries. This is because kitchens and pantries usually do not report enough information for us to account for government costs. Additionally, most kitchens and pantries do not receive U.S.D.A. commodities directly from the government. Therefore, we make the simplifying assumption that — unless otherwise specified — that kitchens and pantries receive no U.S.D.A. commodities.

Subtracting U.S.D.A. administrative grants: If a nonprofit has spent more than \$750,000 in federal awards in a fiscal year, it must report this spending in its audited financial statement (Schedule of Expenditures of Federal Awards). This is the best source of information for the amount of TEFAP, C.S.F.P. and F.D.P.I.R. administrative grants expended by the nonprofit. If available, we subtract these administrative grants from total program expenses. Note that the Schedule of Expenditures of Federal Awards typically breaks out the value of the commodities themselves and the associated administrative grants on separate lines. We only subtract administrative grants.

If a nonprofit does not have a Schedule of Expenditures of Federal Awards and we have no other source of information on U.S.D.A. administrative grants, we don’t subtract anything. Other common sources of information on administrative grants are revenue line items reported on a nonprofit’s Statement of Activities. We generally avoid subtracting line 1e from the 990’s Statement of Revenues if there is any reason to believe line 1e also includes non-food-related government grants or if we think that by subtracting both line 1e and the value of donated food, we would be subtracting government commodities twice. However, if we believe line 1e only includes administrative grants, then we do subtract it from the total.

Other Federal administrative grants: We only subtract Federal administrative grants that are associated with U.S.D.A. commodity distribution. That means that we do not deduct administrative grants for programs that are (1) unrelated to food distribution (such as grants related to the Supplemental Nutrition Assistance Program) or (2) related to food distribution but do not have a component in which the government also donates food (such as grants for the Summer Food Service Program).

Administrative grants associated with State-originating commodity programs: Though U.S.D.A commodities are passed through States, they are Federal commodities and not State commodities. State commodity programs, by contrast, are programs where the commodities originate at the State level and thus the State acts as more than just a pass-through agent. The Federal government is not involved. An example is the Massachusetts Emergency Food Assistance Program. We generally do not deduct from total program expenses the administrative grants associated with State-originating commodity programs. The only reason we would do so is to avoid double-counting when the nonprofit provides the amount of both Federal and State commodities it distributed as a single figure — see Government costs below.

We can tell if a commodity program originates at the State or Federal level:

1. Commodity programs labeled TEFAP, C.S.F.P. or F.D.P.I.R. are Federal.

2. Any program listed in the audited financials under the Schedule of Expenditures of Federal Awards is Federal. Separately, the audited financials may include a Statement of Expenditures of State Awards, which can be assumed to include only State-originating programs.

If we cannot find information to confirm the origin of the commodities, we assume all of them are Federal.

### Government commodities¶

Background: Since we aim to count the cost of U.S.D.A. commodities exactly once, we check to make sure that the value of U.S.D.A. commodities has been excluded from the nonprofit’s costs. Most frequently, it will already have been excluded when we subtracted donated food distributed using either of these sources:

1. A line item on the nonprofit’s financial statements for “donated food received,” “donated food distributed” or simply “food distributed”

2. Schedule M of the 990 under “Food inventory.”

[On the rare occasion that we suspect U.S.D.A. commodities were not included the above sources, we search the nonprofit’s audited financial statements for a line item like “commodities distributed” or “commodities received” on the Statement of Functional Expenses, Statement of Revenues or Schedule of Expenditures of Federal Awards.]

Complete exclusion of commodities: Note that we only deduct administrative grants that are linked to programs used in our estimates. For example, a food bank operates both the C.S.F.P. and TEFAP programs. We know how much the C.S.F.P. program cost the food bank, but we don’t know how many C.S.F.P. meals it served, so we exclude the C.S.F.P. program from our analysis entirely. Because we’ve already excluded the cost of the C.S.F.P. program, we need not deduct its associated administrative grants. Therefore, we would only deduct TEFAP administrative grants.

[If the nonprofit did not report C.S.F.P. outcomes and also did not separate out C.S.F.P. expenses, then we would not be able to deduct the C.S.F.P. cost. In that case, we would still deduct the C.S.F.P. grants because the associated cost of the program is included in our analysis.]

## Partner costs¶

### Government costs¶

Estimated government cost per pound/meal: To calculate total government costs, we multiply the number of pounds of U.S.D.A. commodities distributed by the nonprofit by our estimate of the government’s cost per pound of commodities. We calculate the government’s cost per pound using the U.S.D.A.’s annual summary of Food and Nutrition Services programs. We do so by dividing the total cost of TEFAP by the total pounds distributed as part of TEFAP. [TEFAP tends to be the largest of U.S.D.A.’s food distribution programs. We assume that C.S.F.P. and F.D.P.I.R. also cost the government a similar amount per pound.] By this calculation, we find that the government spends \$0.93 in total per pound of commodities distributed in 2018, for instance. The \$0.93 represents both “commodities distributed (entitlement and bonus) and the Federal share of State admin. expenses.”

We always use this method for calculating government costs. We do not simply add the value of U.S.D.A. commodities distributed by the nonprofit (perhaps reported in its annual report or on the Schedule of Expenditures of Federal Awards). We prefer the above method because it allows us to capture the government’s full costs at the Federal level, not just those that have been passed through the State and to the nonprofit.

There may be cases when the nonprofit provides the amount of both Federal and State commodities it distributed as a single figure. When this is the case, we take the single percentage or poundage presented and multiply it by the government’s cost per pound (e.g., \$0.93 as above). We must then ensure that any administrative grants associated with State-originating commodities programs are subtracted from the nonprofit’s total program expenses. The value of those administrative grants can usually be found in the nonprofit’s audited financials.

Selecting U.S.D.A. commodities figure: We multiply U.S.D.A. commodities by estimated government cost per meal/pound. The following describes the process by which we determine the amount of U.S.D.A. commodities, ordered by preference.

1. If a nonprofit reports the pounds of U.S.D.A. commodities that it distributed, then we multiply this pounds figure by estimated government cost per pound. The result is total government cost.

2. If a nonprofit reports the percentage of U.S.D.A. commodities out of total food sourced/distributed, then we multiply this percentage by total meals/pounds distributed. The decision on whether to use meals or pounds depends on whether the conversion rate from meals to pounds is exactly 1.2 or not. If the conversion rate is 1.2, then we use the meals figure and multiply it by government cost per meal. If the conversion rate is not 1.2, then we find the original pounds figure, and multiply it by government cost per pound.

3. If a nonprofit reports the value of U.S.D.A. commodities that it received or distributed as well as its valuation rate for U.S.D.A. commodities, then we divide value of U.S.D.A. commodities by this reported valuation rate and multiply the result by the government’s cost per meal. However, we only convert from dollars to pounds if we have clear and specific information on U.S.D.A. commodity valuation. For example, the statement “For 2018, we valued all donated food, including government commodities, at \$1.57” is clear. The statement, “For 2018, we valued donated food at \$1.57” is not clear. If the information is unclear, then we do not use this method and instead skip to the next option, described below.

4. If a nonprofit does not clearly report any of the above information, then we apply a standard proportion of U.S.D.A.-sourced food to the total outcomes (15 percent), and multiply this by estimated government cost per pound or meal. Whether we use pounds or meals follows the logic described in point 2 above.

### Distribution and sourcing partners¶

Background: Getting food to a beneficiary involves both sourcing and distributing it. The sourcing is typically undertaken by food banks, while the distribution is typically undertaken by food pantries and soup kitchens. Both sourcing and distribution is essential to the process, and so both costs need to be taken into account when estimating the cost of providing a meal.

Therefore, when creating an estimate for a food bank, we include the costs to its distribution partners, which are largely food pantries and soup kitchens. When creating an estimate for a pantry or kitchen, we include the costs to its sourcing partners, which are typically food banks.

The food sourcing and distribution process can be complex and involve many actors. One food bank may source food from another food bank and later distribute food to a third food bank. We simplify the process and assume — in all cases — that a food bank does not source from or distribute to another food bank. We assume this even for the cases in which a food bank does state that it sources from and/or distributes to another food bank. We acknowledge that this simplification could underestimate the true cost of a meal, but we opt for the simplification because we do not have enough information to consistently apply a more complex standard to all nonprofits.

We make a similar simplifying assumption for pantries and kitchens. We assume — in all cases — that pantries and kitchens only distribute to beneficiaries, i.e., they do not distribute to other pantries, kitchens, and food banks.

### Distribution partners¶

Estimated pantry and kitchen cost per meal: In order to calculate the costs of food banks’ distribution partners, we generated an estimate of the average pantry or kitchen’s cost per meal. To do so, we reviewed a sample of 10 food pantries and 10 soup kitchens filing Form 990. For each program in the sample, we first calculated program service expenses, net of the value of in-kind food donations: Expenses (reported on one of lines 4a–e on page 2, Part III, Statement of Program Service Accomplishments) minus Food inventory (line 19, Schedule M, Part I, Types of Property, Noncash Contributions). Next, we divided this cost by the food pantry or soup kitchen’s reported number of meals or meal-equivalents (usually found in Additional Data on Form 990, as an addendum to lines 4a–e of Part III, Statement of Program Service Accomplishments). The resulting figure is the food pantry’s or soup kitchen’s cost per meal. We calculated the weighted average cost-per-meal across all 20 sampled pantries and kitchens, where the weights are the share of food programs run by the 32,677 agencies surveyed by Feeding America that classify as food pantries (67 percent) and the share of food programs that classify as soup kitchens (33 percent).3 The result is \$1.81. We then apply this weighted average to all partner-distributed meals by the food bank under analysis.

Weighing based on share of pantry versus kitchen partners: If there is any indication that a food bank’s distribution partners are only pantries (or only kitchens), for partner costs we only use the pantry per-meal cost of 51 cents (or kitchen per-meal cost of \$4.45), not the weighted average of the two.

Share of partner-distributed food: Some food is distributed through partners, while other food is distributed via the food bank itself, in collaboration with feeding sites like schools and senior centers. We need to know what share of food is distributed through partners to accurately apply partner costs to that share.

When share is known: If meals are distributed through partner organizations like pantries and kitchens and that proportion is clearly specified, then we apply the \$1.81 cost per meal to just those meals.

If a food bank runs its own pantry and/or kitchen through which it distributes food and those meals are clearly specified as such, then we assume that partner costs for those meals are zero.

Sometimes meals are distributed through non-kitchen or pantry organizations, such as a school, hospital or senior center. If the suspected marginal cost of distributing food is close to zero for these organizations, then we assume those partner costs are zero. If the suspected marginal cost of these organizations is not close to zero, e.g., if the food bank delivers groceries and these partners then prepare meals, then assume full partner costs for these organizations.

When share is unknown: When a food bank doesn’t specify what proportion of total meals were distributed via partners versus itself, we apply a standard proportion of 68.2 percent. We estimated this standard proportion by reviewing a sample of 10 food banks and taking their average share of food distributed through partners.

Note that we always apply this calculated estimated partner share if we do not know the true share. Even if we have reason to suspect the true share is lower or higher than the estimated share, we use the estimated share. The exception is if there is concrete information that suggests the partner-distributed share is close enough to 100 percent or 0 percent to use those values instead of the standard 68.2 percent.

When we apply the standard proportion of partner-distributed food, we do so using a reference to the “Literature Data” sheet, instead of manually typing it.

When we do not know the share of partner-distributed food, it is usually because the food bank only reported total meals distributed across all programs. However, in some cases, a food bank will report separate outcomes by program, but we still do not know what share of the total outcomes were partner distributed. For example, the food bank may report on outcomes for most but not all programs, resulting in uncertainty about what proportion of the remaining, uncategorized meals were partner versus bank-distributed. In all these cases, we apply the estimated partner share to the total meals distributed, regardless of the information we have on meals distributed program by program.

### Sourcing partners¶

Estimated food bank cost per meal: In order to find partner costs for food pantries and soup kitchens, we need to estimate a food bank’s cost per meal. To do so, we reviewed a sample of 10 food banks filing Form 990. For each food bank in the sample, we first calculated the cost the food bank incurred, net of the value of in-kind food donations and net of any dues it collected from food pantries and soup kitchens: Total program service expenses (line 25 on page 10, Part IX, Statement of Functional Expenses) minus Value of food distributed (found on lines 24a–d) minus Membership dues (line 1b of page 9, Part VIII, Statement of Revenue). Next, we divided this cost by the food bank’s reported number of meal-equivalents (usually found in Additional Data on Form 990, as an addendum to lines 4a–e of Part III, Statement of Program Service Accomplishments). [Many food banks report the estimated number of meals that can be created from the food items it collects and distributes. The calculation is often based on the U.S.D.A.’s estimates of the average weight of a meal, 1.2 pounds,4 which Feeding America has adopted as part of its own reporting of outcomes.] The resulting figure is the food bank’s cost per meal-equivalent. The average cost per meal-equivalent across 10 randomly selected food banks is 37 cents.

Share of partner-sourced food: Some food is sourced through partners — typically food banks — while other food is sourced through purchases and community donation. We need to know what share of food is sourced through food banks to accurately apply partner costs to that share.

When share is known: If a pantry or kitchen clearly reports what proportion of its food was sourced from food banks, we apply that proportion in our analysis.

When share is unknown: If a pantry or kitchen does not report the share of partner-sourced food, we apply an estimated share of partner-sourced food. For both meal and grocery programs, Feeding America’s 2014 Hunger in America report lists — on a national level — the average percentage of total food distributed by source. The average share of food bank-sourced food for meal programs is 47.3 percent, the average share for grocery programs is 74.4 percent and the weighted percentage across both programs is 65.7 percent.

For soup kitchens, unless otherwise specified, we assume 47.3 percent of its food was sourced from food banks. For food pantries, we assume 74.4. In cases where a nonprofit is primarily a soup kitchen but runs some food pantry programs, we apply the 47.3 percent share, and vice versa for food pantry organizations that run some soup kitchen programs. If it is difficult to tell whether an organization is primarily a soup kitchen or food pantry, we assume 65.7 percent of its food was sourced from food banks.

## Style¶

Converting from pounds to meals: When converting from pounds to meal using the 1.2 conversion rate, we do the calculation using a reference to the “Literature Data” sheet, instead of manually typing 1.2.

Referencing the Hunger in America report: When we use a number derived from the Hunger in America report, we do derive and calculate the number in the cells of the “Data - Food Programs” sheet, but instead reference the number suing the “Literature Data” sheet.

Referencing estimated partner costs: When we include a nonprofit’s estimated partner cost, we do so by referencing the “Literature Data” sheet, instead of manually typing it.

Footnotes

1

See U.S. Hunger Solutions: Best Practices for Connecting Seniors Participating in the Commodity Supplemental Food Program to SNAP by the Food Research & Action Center. For a description of typical items included in a box along with weights, see U.S.D.A Foods Åvailable List for the Commodity Supplemental Food Program.

2
3
4