B+W Community Insights
Our work rests on one idea: that communities get stronger when the organizations and people inside them can see their own conditions clearly. Most of what we do is for organizations. This page is the other half of the same idea, aimed at a different subject: not the organizations in a neighborhood, but the people who live there.
Why this subject, and why now
This is research on food retail in Philadelphia: how far households are from a supermarket, what kinds of stores their neighborhoods actually have, and what changes when a new federal rule takes effect on November 4.
Three things brought us to it.
A federal funding change is arriving, and it lands locally.
SNAP benefits have been fully federally funded since the program was made permanent in 1964. Under the law enacted in July 2025, two things change. Beginning October 1, 2026, every state pays 75 percent of the program's administrative costs, up from 50 percent. And beginning October 1, 2027, states pay a share of the benefits themselves for the first time — how much depends on each state's payment error rate, the share of benefits issued above or below what a household should have received.
States below a 6 percent error rate pay nothing; nine states are there now. Above it, the share is 5, 10, or 15 percent of benefit costs. Pennsylvania’s rate for 2025 came in at 9.21 percent, released by USDA in June 2026, which places it in the 10 percent tier — an estimated $410 million a year. That figure is not fixed: the state can use its 2026 error rate instead if it is lower, and legislation to delay the start by a year was drafted in the Senate this summer but not advanced. What is fixed is the direction. Whatever a state does in response, the consequences land on particular blocks, in particular stores, for particular households.
And Philadelphia has asked itself the right question without yet answering it.
In June 2026, City Council's Food and Nutrition Security Task Force recommended that every Council district develop a community-led food access strategy, and that the city publish metrics on where the geographic gaps are. Both require agreeing on what to measure. That is a research question, and it is the kind we can answer.
Thirty-Three Stores, One Supermarket
What counting stores misses in Philadelphia's highest-poverty neighborhoods — and what changes on November 4
We looked at every SNAP-authorized food store in Philadelphia and asked two questions about each of the city's neighborhoods: how far is the nearest supermarket, and what kind of stores are actually there. One neighborhood — Hartranft, in North Philadelphia — is used as the example throughout, because it is the sharpest case in the city: one of the highest poverty rates among neighborhoods with at least 5,000 residents, and the most lopsided store mix anywhere. It is an example, not a diagnosis of Philadelphia.
Philadelphia's highest-poverty neighborhoods are not far from a supermarket. In census tracts where 30 percent or more of residents live below the federal poverty line, the average distance is 0.44 miles. In the rest of the city, 0.43. By the distance test used to designate food deserts, there is no difference worth naming.
What they are short on is choice. In the North Philadelphia neighborhood of Hartranft, there are twenty-two convenience stores for every one supermarket. Hartranft is the extreme. Among the city's other highest-poverty neighborhoods with a supermarket, the ratio runs from two and a half to ten. In Chestnut Hill, Rittenhouse, and Society Hill, it is one or less.
And there are more stores, not fewer. Per resident, these tracts have about 50 percent more SNAP-authorized food retailers than the rest of the city — and about a third fewer supermarkets.
Those three facts only look contradictory until you ask what each kind of store is for. Then they describe one condition: many places to buy a few things, almost nowhere to buy a week of groceries. Neither distance nor a store count can see it.
Hartranft
Hartranft has thirty-three places to spend SNAP benefits. One of them is a supermarket.
Twenty-two are convenience stores. Eight are grocery stores. Two are other retailers that take SNAP. The supermarket is a Save A Lot at 701 West Lehigh Avenue, and it is the nearest supermarket for every census tract in the neighborhood — 16,320 people.
Hartranft's official boundary also contains Temple University's main campus, where four in ten residents are enrolled in college. Students living off campus report little income and few cars, which would register in this data as neighborhood poverty. We set that tract aside. The figures above describe the five tracts where people live year-round.
What the store types mean
These are USDA's own categories, and the distinctions carry the whole argument:
What that mix means
A reasonable person could look at thirty-three stores and conclude Hartranft is well served. Three pieces of federal data say otherwise.
First, USDA's own finding. Its Economic Research Service states that super stores and supermarkets generally have a wider variety of foods and lower prices than smaller stores. That is the agency whose store data this analysis uses, describing the difference in its own words.
Second, where the money actually goes. USDA's most recent study of SNAP redemption patterns, covering fiscal 2022 and published in December 2025, found that 78 percent of all SNAP benefits were redeemed at supermarkets and super stores. Convenience stores took 5.3 percent. Large and medium grocery stores took 3.6 percent; small grocery stores, 0.6 percent.
Third, the number that makes it concrete. In that same study, the average SNAP transaction at a supermarket or super store was $53. At a convenience store, $10.
This is not a complaint about small stores
One in five SNAP transactions in the country happens at a convenience store. They are open late, they are close by, and reaching them does not require a car — which matters in a neighborhood where nearly half of households do not have one. They do real work. They were simply never built to hold a family's weekly grocery shopping, and counting them as though they were is the measurement error this piece is about.
The supermarket is also a particular kind of store
Save A Lot is a limited-assortment discount grocer. A typical store runs about 15,000 to 16,000 square feet and carries roughly 2,000 to 3,000 different products, with prices the company reports at 20 to 40 percent below conventional supermarkets. A conventional supermarket runs 40,000 to 50,000 square feet and carries 15,000 to 60,000 products.
That is a trade, not a failing, and for a household stretching a fixed benefit it may be the right trade. But it is worth being precise about what it means: in these neighborhoods, the nearest supermarket is usually the cheaper store with fewer choices. In the rest of the city, both kinds are within reach.
The pattern across the city
Hartranft is the sharpest case. The pattern is citywide.
Six neighborhoods — three of the city's highest-poverty, three where poverty is under 10 percent:
| Neighborhood | Residents | Poverty | Supermarkets | Convenience stores | Per supermarket | All SNAP retailers |
|---|---|---|---|---|---|---|
| Hartranft | 16,320 | 44.9% | 1 | 22 | 22 | 33 |
| Upper Kensington | 26,493 | 42.5% | 3 | 25 | 8.3 | 46 |
| Strawberry Mansion | 17,929 | 37.3% | 2 | 10 | 5 | 18 |
| Rittenhouse | 17,491 | 9.7% | 2 | 1 | 0.5 | 5 |
| Society Hill | 6,485 | 9.0% | 1 | 1 | 1 | 3 |
| Chestnut Hill | 10,672 | 6.8% | 3 | 0 | 0 | 4 |
Poverty rankings are among the 92 Philadelphia neighborhoods with at least 5,000 residents. "All SNAP retailers" includes specialty stores, pharmacies, and dollar stores.
Chestnut Hill has three supermarkets that take SNAP and no convenience stores. Hartranft has one supermarket and twenty-two convenience stores. Per resident, Hartranft has more than five times as many SNAP-authorized food retailers as Chestnut Hill.
Citywide, across 366 census tracts, the same split — 30 percent poverty or above, against everyone else:
| Highest-poverty tracts | Rest of the city | |
|---|---|---|
| Tracts | 85 | 281 |
| Residents | 332,648 | 1,174,020 |
| Households without a vehicle | 37.9% | 22.8% |
| SNAP-authorized retailers per 10,000 residents | 12.7 | 8.3 |
| Supermarkets and super stores per 10,000 | 0.75 | 1.11 |
| Convenience stores per 10,000 | 5.7 | 3.1 |
More stores. Fewer supermarkets. About one supermarket for every 13,300 residents in high-poverty tracts, against one for every 9,000 everywhere else.
The measure that misses it
Philadelphia's Department of Public Health made a version of this argument in 2019. Its Neighborhood Food Retail report found that neighborhoods with higher poverty rates have more stores selling mostly unhealthy food, and asked the city to think past food deserts and supermarket access. That work measured what was on the shelf.
This adds the two things it did not: what kind of store is nearby, and what that kind of store is for. Both come out of public records that are refreshed every year.
And it explains the distance finding rather than contradicting it. These neighborhoods are not waiting for a closer supermarket. They already have one nearby. What they don't have is a second one, or a third, or the choice between them.
What changes on November 4
On May 8, 2026, USDA published a final rule updating what SNAP-authorized retailers must keep in stock. It took effect on July 7. Every SNAP retailer other than specialty food stores — butchers, fish markets, produce stands — must comply by November 4, 2026.
The requirement: seven distinct varieties of staple foods in each of four categories — dairy, fruits or vegetables, grains, and protein — stocked continuously.
There is a second requirement that gets less attention and is harder for a small store. At least three of those four categories must include a perishable variety — something that would spoil within two to three weeks at room temperature. Refrigeration, turnover, and absorbed spoilage are the constraints a corner store faces that a supermarket does not.
USDA's own regulatory impact analysis projects that roughly 5,000 retailers a year could be denied SNAP authorization under the new standard, against about 2,000 a year now — a rise from roughly 3.5 percent of applicants to 5 percent. USDA expects most of them to restock and reapply. Small stores are about 71 percent of SNAP-authorized retailers nationally; convenience stores alone are about 45 percent, the largest single category. The Food Research & Action Center has argued that without more flexibility and real technical assistance, the rule could push small retailers out of SNAP in precisely the places where full-service grocery is already scarce. The rule also removes butter, jerky, cheese dips, snack bars, and fruit spreads from the list of foods that count toward a store's qualification — items a corner store is likely to stock. They remain purchasable with SNAP; they simply no longer count toward staying authorized.
This cuts both ways, and it would be dishonest to pretend otherwise. If Hartranft's thirty small stores meet the new standard, they will carry more staple varieties than they do today, and the gap this piece describes narrows. If some cannot, they drop out of SNAP, and it widens. Our data cannot tell you which will happen.
What it can do is establish the before. The numbers in this piece describe Philadelphia's SNAP retail mix as of September 3, 2026 — before the deadline that will change it. Anyone can pull the same file in December and see what moved.
What to ask
This analysis measures access and store composition. It does not measure hunger, and a better measure is a starting point rather than a solution. But it does turn a few things into answerable questions, each one attached to something the city has already committed to.
How is my Council district measuring food access — by counting stores, or by looking at the mix?
The Task Force recommended a community-led food access strategy in every Council district. What is ours, and who is writing it?
The Task Force also asked the city to publish metrics on geographic gaps. Which metrics, and how often?
After November 4, how many SNAP-authorized retailers does my neighborhood still have — and is anyone checking?
What this data cannot tell you
It counts SNAP-authorized stores only. A store that does not accept SNAP is not in here. For the question of where SNAP households can shop, that is the right universe. For a full picture of a neighborhood's food retail, it is not.
Store type is USDA's label for format, not an inspection. The categories describe size and setup, not what is actually on the shelf or what it costs. We checked the 25 stores USDA classifies as supermarkets in high-poverty tracts; at least 24 are what the label says. The Health Department's 2019 work on what stores stock is the complement to this.
This measures access, not hunger. There is no direct measurement of food insecurity at the neighborhood level in Philadelphia. Everything here is about what stores exist and what kind they are.
Carless is not one condition. Rittenhouse has a higher share of households without a vehicle than Hartranft — 54.9 percent against 48.5 — at a poverty rate of 9.7 percent against 44.9. Not owning a car in Center City is not the same circumstance as not owning one where a single supermarket serves the neighborhood. Vehicle access should never be read alone.
College students are screened out, and where to draw that line is a judgment. Tracts where more than 30 percent of residents are enrolled in college are excluded before any figure is calculated — nine in all, including tracts at Temple and near the University of Pennsylvania. Tracts below the line can still have student residents; one of Hartranft's five is about one-fifth college students, near where a stricter screen would fall. Excluding it, Hartranft still has one supermarket, nearest for every remaining tract.
Neighborhood boundaries are the city's, and they are contested. These follow OpenDataPhilly's definitions. They are the common reference, not a settled answer. A neighborhood with no supermarket inside its boundary may have one directly across the street.
We cannot predict which stores will meet the new stocking standard.
This is a September 2026 snapshot. Stores open, close, and change hands.
Why this matters
The city has said it wants food access strategies in every district, metrics on geographic gaps, and residents with lived experience of food insecurity in the room where decisions get made.
All three depend on the same thing: agreeing on what to count. Count storefronts and Hartranft looks like one of the better-served neighborhoods in Philadelphia. Count what kind of store, and it is one of the worst. The difference is not a technicality. It decides which neighborhoods get attention and which get passed over.
Retail data. USDA Supplemental Nutrition Assistance Program retailer file, snapshot September 3, 2026. Store types are USDA's classification. Deauthorized retailers drop out of the file, so presence reflects active authorization on the snapshot date. Where the file lists the same store at the same address more than once, it is counted once; separate vendors sharing an address, such as those in Reading Terminal Market, are counted separately.
Store checks. The format of the 25 supermarkets in high-poverty tracts, and the operating status of every store named here, were verified against operator websites, trade coverage, and current business listings in September 2026.
Distance. Straight-line from each tract's 2020 population-weighted center to the nearest SNAP-authorized supermarket or super store. Straight-line distance understates walking distance. It is also how USDA's Food Access Research Atlas has measured distance for food-desert designations, which is why it is used here; the newest version of that atlas adds a driving-distance measure alongside it. Medians: 0.42 miles in high-poverty tracts, 0.37 elsewhere.
Demographics. U.S. Census Bureau, American Community Survey, tract level. All rates are population-weighted. Single-tract estimates can carry margins of error above ten percentage points; neighborhood totals are more reliable than any one tract.
Geography. Philadelphia census tracts with more than 500 residents. Excluded: tracts where group quarters — correctional facilities, dormitories, institutional housing — are the dominant population, and tracts where more than 30 percent of residents are enrolled in undergraduate or graduate school (ACS table B14007). 366 tracts remain: 85 at or above 30 percent poverty, 281 below.
Neighborhoods. OpenDataPhilly neighborhood boundaries joined to 2020 census tracts. Rankings and the neighborhood map cover the 92 neighborhoods with at least 5,000 residents.
Where SNAP benefits are redeemed. USDA Food and Nutrition Service, Benefit Redemption Patterns in the Supplemental Nutrition Assistance Program, Fiscal Year 2022, published December 2025.
Store variety and prices. USDA Economic Research Service, Food Access Research Atlas documentation. Save A Lot store format figures: company licensing materials and trade coverage, 2024–2026.
Stocking standards. USDA Food and Nutrition Service, Updated Staple Food Stocking Standards for Retailers in the Supplemental Nutrition Assistance Program, final rule, 91 FR (May 8, 2026); effective July 7, 2026; compliance date November 4, 2026; and FNS retailer guidance, updated September 4, 2026. Retailer projections: USDA's Regulatory Impact Analysis. Analysis of the risk to small retailers: Food Research & Action Center, March 2026.
Policy context. Philadelphia City Council Food and Nutrition Security Task Force report, June 2026. Philadelphia Department of Public Health, Neighborhood Food Retail in Philadelphia, September 2019.
SNAP cost-sharing and state error rates. H.R. 1 / P.L. 119-21; Congressional Research Service, IN12723; Food and Nutrition Service proposed rule, June 24, 2026; USDA fiscal 2025 payment error rates, released June 24, 2026; Food Research & Action Center, Why the SNAP Administrative Cost Shift Threatens State Budgets, August 2026.
Interpretation is Brown + Wheaten's own and labeled as such. Where a cause cannot be determined from the available data, this piece says so rather than choosing one.
Take it with you
One Supermarket
The full piece, formatted to print or forward.
Also available: the full store list (Excel) — every SNAP-authorized food retailer in the 85 census tracts at or above 30 percent poverty. 423 stores by name and address, with the store type USDA assigns each one, broken out by neighborhood and by Council district.
Brown + Wheaten Strategy Lab is a public benefit business in Philadelphia. We do market and community research for small businesses, founder-led organizations, and community-rooted nonprofits. Food access is where we concentrate the community side of that work.