New York City should not own grocery stores
is this your municipal socialism?
Andrew Cuomo will likely be the next mayor of New York City. This is unfortunate, as he appears to be uninterested in governing and is instead running to pursue a four-year revenge tour. The idea that many New York Democratic luminaries are certain that the man who was Governor for a decade is the optimal candidate for a city rapidly losing population and experiencing an acute cost of living crisis is quite strange to me!
Turning to the opposition, Zohran Mamdani and Brad Lander both have programs promising to make the city more affordable for its working class residents. Some of these proposals, such as Lander’s (insufficient) housing plan and Mamdani’s advocacy for the elimination of parking minimums, are attractive. There is one plank of the Mamdani campaign that I am enormously skeptical of: his proposal to open city-owned grocery stores. While I am sympathetic towards an agenda of a cornucopia of public options in housing, transportation, banking, and broadband, city-owned grocery stores are likely bad policy and would fail to efficiently address the concerns they are intended to ameliorate.
While Mamdani’s plan is not particularly detailed, he has made the rough outline clear. The city would open grocery stores on city-land, one in each borough. These stores would, according to the campaign, be able to sell groceries at reduced prices due to a lack of property taxes, rent, and profit, passing on savings to consumers and lowering the market price of food. Advocates of a municipal public option for groceries have provided two justifications for the intervention: food deserts and grocery costs. Below, I will discuss these two arguments in detail.
Justification #1: Municipal grocery stores would address urban food deserts.
Though not the core of Mamdami’s rhetoric, other efforts to establish public grocers have emphasized serving areas without convenient access to food retail. Food deserts, which can be defined as an urban tract with at least 500 people, or 33% of the population, living more than 1 mile from the nearest supermarket, supercenter, or large grocery store, are plausibly a problem for which municipal grocery stores would be a direct solution.
The literature on the health impacts of food deserts is spotty. While communities considered food deserts have higher rates of chronic diseases, I am very skeptical of the causal relationship here. Several studies have shown that interventions to expand access to healthy food, such as the opening of a new supermarket, had no impact on BMI or other health outcomes in food deserts. There is little evidence that food access drives negative health outcomes for low-income people and other risk factors associated with poverty drive unequal health outcomes.
Nutritional inequality is primarily driven by demand-side differences driven by costs of healthy food, time for preparation, and differential tastes. The most sophisticated study on this topic concludes that differential local supermarket density explains no more than 1.5% of the difference in eating between high and low-income households, but that a 15% increase in the SNAP budget would increase low-income household’s healthy eating to the level of high-income households. Why might food deserts not be particularly harmful? American consumers appear to have high willingness-to-travel, and most people living more than 1-mile from the nearest supermarket simply drive to said supermarket. Of course, for those with mobility impairments the ability to travel does not ameliorate a lack of access, but neither would a closer supermarket.
Regardless, should I be misreading the evidence on the harms of food deserts, food deserts are driven by factors that can be addressed with policy interventions other than public grocers, at lower cost.
The first narrative explaining the existence of food deserts comes from antitrust advocates, who argue that dollar-stores, scorched-earth covenants preventing other grocers from entering an area, and other noncompetitive arrangements drive abandonment of low-income areas by food retailers. They point to the lack of enforcement of the Robinson-Patman Act, which prohibited price discrimination by suppliers, as a reason why small, independent grocers have collapsed. I do not see how this explanation accounts for why small grocers protected from price-discrimination and illegal covenants would be able to sustainably earn profit in an area where a supermarket, benefiting from economies of scale, would be unable to. However, I would love to see some policy experimentation on this front, and would be convinced by enough evidence that antitrust enforcement may revitalize areas with low access to food retail.
In my reading of the evidence, urban food deserts are overwhelmingly a problem of housing, land use, and transport. Urban sprawl is strongly associated with risk of low access to food retail. This is initiative: density provides the demand for supermarkets; more consumers in a given tract means a higher likelihood that such an enterprise can remain profitable. The other cause of a plausible under-supply of urban supermarkets is high rents, which squeeze the already tiny profit margins of grocers and make operating a supermarket non-viable in low-income, but high-rent areas.
This point becomes obvious when we glance at the food deserts in New York City itself. According to the USDA Food Access Research Atlas, there is not a single census tract in New York City that is both low income and is more than 1 mile from the nearest supermarket. Note that this categorization of food desert excludes access to ethnic markets, bodegas and corner stores, dollar stores - which now often carry fresh produce, farmers markets, or any other food retail location.
When we narrow the search to census tracts that are both low-income and more than ½ mile from a supermarket, 31 census tracts appear. 12 are in Staten Island, and are thus overwhelmingly in R1, R2, R3, R4, and R5 zoning districts, that is, in single family and low-density neighborhoods. The same is true with a small census tract in Fort Hamilton, the three tracts in Canarsie, another in Rockaway Beach, and so on. The food desert in South Ozone Park is almost entirely Resorts World New York City and the Aqueduct Race Track, with the remainder R4 zoning. In the Bronx, Hunts Point has often been labeled a food desert, and ironically so given the presence of the Hunts Points Market, the largest food producer in the city. It thus seems worth noting that most of the neighborhood is zoned for heavy industry or light industry!
As we will return to later, it is far too difficult to build supermarkets in light industry districts in New York City. Up-zoning neighborhoods with low food retail access is the lowest-cost intervention to lower the cost to open a supermarket, and should be attempted before engaging in the costly step of opening a public grocer.
Municipally-owned grocery stores also seem like a poor solution to this problem on their own terms. While they promise to address food deserts directly by opening a store in the areas most in need, this faces the a priori observation that private supermarkets have already identified such locations as being unsuitable for food retail. Furthermore, it is unclear to me that there are viable city-owned lands for siting a supermarket also in areas in need of food retail, or that a lack of demand would not simply nullify any possible benefits a public grocer could have on nutritional outcomes.
Given the lack of evidence that food deserts have negative health impacts, and the much more efficient solution of upzoning the neighborhoods considered to have low access to food retail, public grocery stores will need to find another raison d'etre.
Justification #2: Municipal grocery stores would lower grocery prices.
The central argument for municipal grocery stores from Mamdani’s campaign is that a public option for groceries would lower grocery prices. From 2019 to 2023, the consumer price index for food at home increased 25%, higher than inflation on the core goods index. Grocery prices are particularly salient to consumers, as similarly to gas prices consumers encounter price increases more frequently than for other goods and thus experience inflation more vividly. In order for public grocers to reduce prices efficiently, the rise in the cost of groceries must be driven by concentrated market power, or public grocers must be able to provide groceries at lower cost than private supermarkets. The campaign has made both arguments.
First, let’s address the market power argument. The overwhelming evidence on the post-Covid price surge in groceries concludes that supply-side factors drove grocery inflation, including supply chain disruptions, labor shortages and wage costs, transportation and energy costs, and to a lesser extent, farm commodity costs. The demand-side also contributed to grocery inflation. During the pandemic, households shifted consumption from restaurants to grocery stores, constituting a demand shock to food-at-home.
Studies of market power in the food retail sector repeatedly demonstrate that while there has been consolidation, it is primarily driven by fixed costs and product differentiation, rather than market power. Some of the rise in gross retail margins is attributable to increased local concentration, with some evidence of market-power effects, but pass-through to consumers is limited. While during the pandemic supply-chain shocks may have enhanced retailer market power, food retail margins have returned to pre-pandemic levels, a narrow average of 1.6%. Local concentration, especially in rural areas, may still play a significant role in rising grocery prices, but I remain unconvinced that public grocers are the optimal solution, rather than enforcing or extending existing antitrust statutes.
Municipal grocery stores would represent limited competition to the quite competitive New York food retail market. Furthermore, rather than force private supermarkets to lower prices to meet the public option, municipal grocery stores may just force private retailers to shut down, crowding out the market and leading to the closure of independent grocers, leading to further market consolidation. There is also a possible effect of entry deterrence, where private firms refuse to enter the market because the subsidized store is present.
On the ability of municipal grocery stores to offer groceries at lower cost, this is the primary reason for municipal ownership advanced by the Mamdani campaign. It is predicated on the claims that as public grocers would not have to pay property taxes, would not have to make a profit, would centralized warehousing and distribution and buy wholesale, and would partner with local neighborhoods for sourcing, they would be able to offer groceries at lower prices than private supermarkets. Let's address each of these points in turn.
i) Property taxes
While a public grocery store on city owned land would not have to pay property taxes, it would be subject to a wide variety of additional siting costs, including the Uniform Land Use Review Procedure and CEQR environmental review. The selection of a site would likely take months, if not years, of community board and city review. Furthermore, if property taxes inhibit private supermarkets from offering lower prices, perhaps we should reform the NYC property tax system such that commercial properties do not pay an effective tax rate several times that of single-family homeowners. This is probably the best argument for the ability of public grocers to offer lower prices and pass on savings to consumers, but all other considerations greatly outweigh it.
ii) Profit
Though public grocery stores would not have to make a profit, it is unlikely they would in the first place, given the lack of access to economies of scale and the intent to open them in communities where a private supermarket is not economically viable. Thus, it is technically true that New York City could open public grocery stores, run them at a severe loss, and sell at below marginal cost to drive private supermarkets to lower prices or close, it is unclear to me what the point of doing so would be. Without the benefits of economies of scale or the variety of other advantages of state enterprises, we would just be inefficiently subsidizing artificially low grocery prices. Is our goal simply to reduce the prices of groceries at the expense of other goods by funding a ‘national champion’ grocery store with tax revenue? TANSTAAFL.
iii) Centralized warehousing and distribution, wholesale purchasing
Private supermarkets already do this, at a far larger scale than five city-owned grocery stores would be able to access. Public grocery stores, without significant administrative reform, would face enormous constraints to procurement, including competitive sealed bidding or RFPs regardless of urgency or vendor relationships, Minority and Women-Owned Business Enterprise procurement goals (e.g. 30% of total contract value), conflict of interest rules, and the ban on discretionary buying. Fixed price contracts with 1-year duration, the standard for municipal procurement, would be a significant cost for a public grocery store, as supermarkets frequently have much shorter contracts with suppliers.
Supermarkets often engage in spot buys, exploiting market dips - public grocery stores would be prohibited from doing so. Now, one of the motivating principles of this newsletter is that such regulatory constraints on public enterprise must be lifted as a precondition to the success of said public enterprise, but until Mamdani proposes such dramatic changes to New York City’s procurement regime these costs must be taken into account.
Regardless, it still remains impossible for municipal grocery stores to access economies of scale. With such narrow profit margins, supermarket chains rely scale from on hundreds of stores, volume discounts with national order volumes, slotting fees and trade allowances, private-labels / store-brand goods, large fleets of trucks for backhaul optimization, global sourcing leverage, real-time inventory management systems, dynamic pricing software, standardized training, shared back-office functions, and management training programs. Without the scale of international or national chains, these cost-savings will not be accessed, and attempts at software procurement will face additional regulatory barriers.
Given the tiny margins of food retail and the fact that Mamdani’s proposal necessarily only envisions a small number of stores, it is ludicrous to suggest that public grocery stores would be able to access even a fraction of the economies of scale necessary to sell groceries at lower prices than the private sector. That is of course, even if you do not place additional requirements on procurement as part of the proposal itself.
Mamdani’s proposal also calls for partnering with local neighborhoods for products and sourcing. Whether this is an attempt at increasing public buy-in, supporting local neighborhoods, or simply a reflex on the left at this point, it would surely drastically increase the costs of public grocery stores if they were burdened with the obligation to utilize “local” sources, negotiate with local organizations, and so on.
It would constrict the total contract volume negotiated with national suppliers and create an inefficient subsidy for local food production in one of the largest cities in the world for no apparent reason. While I am happy to be proven wrong, I would expect to see many more requirements added to such a proposal, including requirements for vegan and vegetarian options, as well as ethnic and cultural foods, restrictions on the sale of unhealthy food options, etc.
v) Labor costs
The black hole in the middle of the budgeting for municipal grocery stores is labor costs. Always a core tension in social-democratic public enterprise proposals, they are particularly salient here - in an industry almost always paying minimum wage or near-minimum wage, with low barriers to firing and hiring. Meanwhile, municipal grocery stores would have to pay $25-$35/hour. Workers would likely be represented by municipal unions, driving up costs further with public-sector benefits packages and pay.
Hiring constraints, including the rigid job posting process, background checks, city HR onboarding, civil service eligibility lists, equal employment opportunity policy, as well as firing constraints, including the “just-cause” standard, due process hearings, union protection, and back pay for wrongful determination all would increase costs by raising the cost of hiring and firing store employees.
All of this would likely result in at least a doubling of labor costs versus private supermarkets. Though one may justify this as a program of public works creating high-paying, dignified positions in a municipal workforce, I could think of many other possible projects for such a program - such as public housing, passive building retrofits, subways, light rail, street repair, solar installation, modernization of utilities, undergrounding power lines, street tree planting, or broadband. Even more adventurous public enterprises, such as a city-owned modular housing factory, trade schools, or hydroponic farms would be more likely to be investments that would not simply serve as jobs creation programs. It would also help if Mamdani was committed to wholesale civil service reform, though this does not appear to be the case.
Thus, we can conclude that not only would public grocery stores not efficiently address concentrated market power, they would not be able to offer groceries at lower prices than the private sector without essentially acting as a colossal inefficient subsidy for groceries. If this is our goal, just pay people to buy lettuce instead of spending three years arguing with a Community Board to establish a public option in a sector without significant market power or shortages.
The poor record of municipal grocery stores
The above concerns are demonstrated by the experience of municipal grocery stores in the United States, which to say the least is not a catalogue of success.
The St. Paul supermarket is the sole notable success in the history of city-owned grocery stores in the United States. After the town had been without a grocery store for 20 years, St. Paul issued a zero-interest bond to the local rural electricity cooperative to build a grocery store, with distribution handled by a private firm. After the family running distributor retired in 2013, the city purchased the grocer and now runs it as a municipally owned business. The supermarket is now slightly more profitable than the average rural grocer, and is strongly supported by St. Paul residents.
The success of this enterprise is likely related to the impetus for its creation - the 17 miles between this town of 614 and the nearest grocery store. In such an extreme food desert, with such a small population to be served, local initiatives may be necessary. Given that New York City has no communities that isolated, I don’t believe that the lessons from this program can be generalized, especially given the many other small town public grocery store failures.
In Erie, Kansas, the municipal grocery store did not attract enough customers due to competition with a Walmart 15 miles away and a Dollar General across the street. In 2022, it only had a single profitable month, and was soon handed over to a private operator. In 2019, Baldwin, Florida opened a grocery store to address the lack of food retailers. While this did have the immediate positive effect of incentivizing the local dollar stores to begin selling fresh produce, the market was unable to offer competitive pricing and ran at a loss during the entirety of its operation. This culminated in its closure last year.
Chicago is the largest American city to seriously consider a public grocery store. Brandon Johnson’s proposal followed a string of failures to induce private supermarkets to open in neighborhoods with low food retail access, including Rahm Emmanuel’s $10 million tax incentive to a Whole Foods Englewood that closed six years later due to a lack of demand. After the closing of the Whole Foods, some residents, organized as the Resident Association of Greater Englewood, protested the opening of a new ‘sub-standard’ grocery-store, delaying its opening for several weeks. The store was eventually opened, benefitting from a $20 million city contract with the operator.
In 2023, Brandon Johnson’s office began considering a publicly-owned grocery store as an alternative to public-private partnerships. Its advocates hoped that “a Chicago-owned grocery store will become a case study that all cities can look into.” The city hired consultants to produce a feasibility report that recommended the policy. Unfortunately, the report has never been released. Given that the mayor appears to have dropped the proposal entirely, we can assume that the city decided that the endeavor was not a cost-effective way of increasing local availability of food retailers. Instead, Chicago is now considering a municipally-owned public market with individual vendors, similar to Reading Terminal in Philadelphia. This process, and the fact that Brandon Johnson is a leftist with one of the lowest approval ratings in the United States, ought to give Democratic Socialists pause.
Political capital and alternatives
Though Mamdani continues to argue that his proposal is supported by ⅔ of New Yorkers, and grocery prices are salient, I am not convinced that public grocery stores are worth arguing for given their failure as a policy intervention. Clearly - this isn’t working. Mamdani continues to trail Cuomo by huge margins among non-college primary voters, and continues to have low name recognition. I don’t have polling data that would confirm this, but anecdotal evidence while discussing the policy with others suggests that the grocery proposal reads as technocratic.
Perhaps Mamdani should focus on the portions of his agenda that most working class primary voters can immediately grok - such as raising the minimum wage, building public housing, or making buses free. A robust plan on public safety would also be worthwhile to cut into Cuomo’s non-college base.
However, public grocery stores do not appear to increase P(defeat Cuomo), and I can’t imagine any world where spending political capital on it after you enter Gracie Mansion makes any sense. I understand the ambition of municipal Democratic Socialism, but attempting to foray into new sectors rather than refining, reforming, or even radically expanding the city’s public investment in areas in which it already has capacity: housing, etc, is a luxury that Democratic mayors, increasingly facing sub-10% approval ratings - cannot afford.
If we actually want to reduce grocery prices, as well as the prices of food in general, there are a host of more efficient and less politically costly options. FRESH Zoning promised to incentivize the opening of new supermarkets by allowing supermarkets to be built in light-manufacturing districts, providing tax abatements and exemptions, and granting a density bonus for including a supermarket on the ground floor of a residential building.
In order to qualify for these benefits, a supermarket must meet several overly-stringent requirements, including mandating that half of the store’s area must be for home preparation and consumption, 30% of the store’s area (not total volume?) must be dedicated to selling perishable food, 500 square feet must be for fresh produce, etc. Due to these requirements, by 2016 - 7 years after the program was established, only ten new FRESH zoned supermarkets had been opened, in a city that in 2009 was estimated to be able to hold 100 additional supermarkets.
FRESH zoning does provide a decent basis to begin zoning reforms that facilitate food abundance. The requirements for FRESH Zoning should be greatly reduced, and all food retailers should qualify for the right to build in light-manufacturing districts, as well as the density bonus for residential buildings. Amazingly, FRESH Zoning doesn’t even lift parking minimums, it simply reduces them to 1 spot per 1000 square feet. All parking minimums on food retail should be lifted, perhaps as a first step to lifting them entirely city-wide.
The city could also identify and rezone areas considered food deserts, streamline the food retail licensing process, impose a statute of limitations on litigation and community board review for new supermarkets, allow small format grocers, bodegas, farmers markets, and other non-traditional food retailers in residential districts, shift to default approval for food retail, and lobby the state to allow grocers to sell liquor and wine.
We could also reduce costs for food away from home by restoring outdoor dining and lifting the cap on vending permits. Finally, we could implement demand flexibility for grocery store electricity use and encourage grocery stores to end the vestigial practice of leaving some fridges without doors, both reducing electricity usage and reducing costs. The above would address concerns around high food costs, and it is notable that Mamdani supports many of those policies. A particularly skilled candidate could even message seemingly piecemeal reforms as populist common-sense, but I will leave that question to pollsters and strategists.
But wait.
The United States federal government already owns a massive chain of publicly owned supermarkets, with enough locations to benefit from economies of scale. With this system, the government is able to sell groceries at low cost to underserved communities, passing on savings from dodging regulations and taxes faced by the private sector. I am of course referring to the 240 stores of the Defense Commissary Agency.
The commissaries could be a model for a workable public option for food, as the federal government is able to operate at extraordinary scale and access considerable bargaining power with suppliers, far beyond what a municipal operation would be capable of. The commissary system is more cost-efficient than an equivalent cash transfer and is cheaper than private sector competitors. The most important predictors of commissary success appear to be the density of military customers, which is trivial as they are the only ones with access, but also the price differential between commercial grocery stores and commissary stores. This strongly suggests that the commissaries are a successful program that is operating as an efficient, scaled public enterprise. Annual revenue is roughly $5 billion, two-thirds that of Sprout’s. The Commissary system has even had fascinating knock-on effects, such as the innovations of the Subsistence Research Laboratory, which was largely responsible for the emergence of dehydrated foods and much of the methodologies for the consumer taste testing.
Municipal social-democrats face a crucial constraint: they are running a city. The ability of municipalities to engage in public enterprises is limited by the inability to access economies of scale - as well as path-dependence from existing undesirable regulatory regimes. While cities ought to stick to what they do best and focus on public investments that are net welfare gains on a local level, Project 2029 or what have you may be able to think bigger. Perhaps municipal grocery stores are not viable, but I can’t help but hope for Matt Bruenig to expand on his proposal for the federal government to purchase Kroger. My prior is that a federal supermarket chain would also be inefficient, but after perusing the evidence it has a much better chance of being viable than one in the five boroughs.
