Mutual Aid Networks During Crises — Lessons From Every Major Disaster

Rebecca Solnit's A Paradise Built in Hell (2009) synthesized decades of disaster sociology into a single uncomfortable thesis: the dominant cultural narrative about disasters — that civilization is thin, that people panic, loot, turn feral — is empirically false. The actual pattern, documented across the 1906 San Francisco earthquake, the 1917 Halifax explosion, the London Blitz, 9/11, and Hurricane Katrina, is the opposite. Ordinary people self-organize. They share. They take extraordinary risks for strangers. They demonstrate what Solnit calls "improvised, provisional communities" that function, often, better than the formal systems that collapsed.

Disaster sociologists have had a term for this since the 1950s: "therapeutic community" (Fritz, 1961). The myth of the panicked mob was invented by elites, often deployed to justify militarized responses that killed more people than the disaster itself. (Katrina's "looter" narrative led to police and vigilantes shooting Black residents who were wading through water looking for food. See the ProPublica/Frontline investigation into the Danziger Bridge killings.)

What replaces the panic myth, once you look at actual data, is this: formal institutional response is slow, hierarchical, and optimized for legibility to donors and politicians. Informal horizontal response is fast, context-aware, and optimized for actually meeting needs.

Occupy Sandy (2012). The formal response to Hurricane Sandy in the Rockaways and Red Hook was a catastrophe. FEMA distribution was delayed, Red Cross visibility was low, and residents in public housing were stuck in buildings without power or water for weeks. Occupy Sandy — an emergent network of people who had been through the Occupy Wall Street encampments and had already learned how to run kitchens, medical tents, and working groups without hierarchy — activated within 48 hours. Key facts from the Homeland Security report (yes, DHS studied them):

- Set up at St. Jacobi Church in Sunset Park as primary hub - Coordinated roughly 60,000 volunteers over several months - Distributed aid via Amazon wedding registry — one of the first uses of e-commerce for disaster relief - Served over 120,000 meals - DHS and FEMA later studied them as a model

Why did they work? Three reasons. One, pre-existing trust networks from Occupy meant they could move without vetting every volunteer. Two, horizontal decision-making meant the hub manager could reroute supplies in real time instead of waiting for regional approval. Three, they operated outside the liability logic that paralyzes formal response — they just did the thing.

The Black Panther Survival Programs (1969–1982). Before we call it mutual aid in 2020 language, the Panthers called it Survival Programs Pending Revolution. The Free Breakfast for Children Program started in Oakland in January 1969. Within a year it was operating in 19 cities and feeding over 20,000 children a day. Before the federal School Breakfast Program was expanded in 1975 — which happened in no small part because the Panthers had demonstrated the need and the method — kids went to school hungry as a matter of course.

The Panthers also ran: - People's Free Medical Clinics in 13 cities - Free ambulance service in Winston-Salem - Sickle cell anemia testing (the Panthers were the first to bring this to national attention) - Free food giveaways (tens of thousands of bags) - Free clothing programs - Legal aid clinics

J. Edgar Hoover called the Free Breakfast Program "the greatest threat to internal security of the country." He wasn't wrong about the threat — a demonstration that communities could provide for themselves without the state was, and is, politically destabilizing. COINTELPRO was mobilized explicitly to disrupt the breakfast programs (see the 1976 Church Committee report).

Puerto Rico After Maria (2017). The federal death toll was initially reported as 64. The George Washington University study later estimated 2,975 excess deaths. The gap between the official response and the need was the largest in modern American history. Into that gap stepped the Centros de Apoyo Mutuo (CAMs) — Mutual Aid Centers. The most studied is Casa Pueblo in Adjuntas, which had been running on solar power since the 1980s as an environmental justice project. When the grid failed, Casa Pueblo became the only source of power for a wide radius — running a pharmacy refrigerator, a radio station, phone charging, and refrigeration for community members on dialysis.

Other CAMs emerged in Caguas, Mariana, Las Marías, Utuado. They ran community kitchens, cleared roads the government hadn't reached, organized medical visits. Researchers like Marisol LeBrón and Yarimar Bonilla have documented how these networks were the actual infrastructure that kept people alive.

COVID-19 Neighborhood Pods (2020–2022). Within weeks of the March 2020 lockdown, Town Hall Project, Mutual Aid NYC, and hundreds of independent pods had mapped entire cities. The Hunter College Urban Policy Lab counted over 800 distinct mutual aid groups in New York City alone by June 2020. The pattern was remarkably consistent: a shared spreadsheet or Signal group, a handful of organizers, a "needs" column and a "resources" column, and someone doing the match. Bed-Stuy Strong delivered groceries to over 20,000 households. Invisible Hands in Manhattan organized thousands of errands for elderly and immunocompromised neighbors.

Flint, Michigan (2014–present). The state of Michigan switched Flint's water source to the Flint River in 2014 to save money, didn't treat the water for corrosion, and poisoned the pipes. For over 18 months, officials denied the lead contamination. The filling of the gap was entirely community: faith institutions (Greater Holy Temple Church became a major hub), Black fraternities and sororities, Concerned Pastors for Social Action, and a web of volunteers running a distribution system that lasted years. Dr. Mona Hanna-Attisha, the pediatrician who broke the story, wrote in What the Eyes Don't See that the community networks did the work the state refused to do.

Dean Spade's Mutual Aid: Building Solidarity During This Crisis (and the Next) (2020) is the most rigorous contemporary framework. The three-part definition:

1. Mutual aid projects work to meet survival needs and build shared understanding about why people do not have what they need. It's not just soup. It's soup plus a political education about why there's no soup.

2. Mutual aid projects mobilize people, expand solidarity, and build movements. The relationships built in crisis become the organizing base for the next fight.

3. Mutual aid projects are participatory, solving problems through collective action rather than waiting for saviors. No clients and providers. Everyone does and everyone gets.

Spade draws the hard distinction between mutual aid and charity. Charity: - Flows one direction (giver to receiver) - Requires a deservingness test (drug test for benefits, sober house, documentation) - Preserves the underlying conditions that created the need - Extracts gratitude as a form of payment - Is tax-deductible, which means it is state-sanctioned

Mutual aid: - Flows multiple directions - Assumes everyone has something to give - Names and organizes against the conditions that created the need - Builds political power - Is largely invisible to the state

The distinction matters because the charity model has colonized a lot of what people call mutual aid. A food bank that requires ID and a means test is charity. A fridge on the sidewalk with "take what you need, leave what you can" is mutual aid.

Why does horizontal coordination outperform hierarchy in disaster? Three reasons, with research backing.

1. Information latency. Hierarchical systems require information to travel up and decisions to travel down. In a fast-moving crisis, by the time a need is reported, approved, and a response is authorized, the situation has changed. Horizontal networks make decisions at the point of contact. (See Malone & Bernstein, 2015, on the communication costs of hierarchy; Clay Shirky's Here Comes Everybody on the collapse of coordination costs in networked systems.)

2. Trust and vetting costs. Hierarchical aid organizations spend enormous resources on vetting — preventing fraud, managing liability, maintaining brand integrity. This is rational for them but disastrous in crisis, where speed matters more than fraud prevention. Horizontal networks run on pre-existing relational trust, which was already paid for in the quiet times. (See Robert Putnam's Bowling Alone on social capital; Elinor Ostrom's Governing the Commons on trust-based resource management, for which she won the Nobel in Economics in 2009.)

3. Context specificity. A national org sending identical boxes of supplies to every site is optimizing for logistics, not need. A neighbor knows that Mrs. Rodriguez on the third floor is diabetic and needs a specific kind of shelf-stable food. You cannot get that resolution from the top down. You have to already know.

Here's where the essay gets practical. If the research says pre-existing relational networks are the thing that saves lives, the question becomes: how do you build one?

Step 1: Know your block by name. This sounds trivial. It is not. Most Americans, per a 2018 Pew study, cannot name any of their neighbors. The baseline act of mutual aid infrastructure is memorizing the names and door numbers of the 20–40 people physically closest to you. This is the unit of rescue.

Step 2: Map capacity, quietly. Who has a car? Who has medical training? Who has a generator, a pressure cooker, a ham radio, keys to the church basement? Who speaks Spanish, Mandarin, Haitian Creole, Arabic? Who is undocumented and must not appear in any formal list? Who is a registered nurse? Who has a printer, a pickup truck, a well? The map lives in a notebook, not a Google Doc. There are reasons.

Step 3: Practice low-stakes coordination. Mutual aid atrophies without practice. Block parties, potlucks, tool libraries, baby clothes swaps, neighborhood WhatsApp groups. These are the reps. The muscle is relational trust, and it only grows through repeated small exchanges.

Step 4: Identify the physical infrastructure. Every neighborhood mutual aid network needs a hub — a church, a mosque, a community center, a bar, a laundromat, somebody's garage. Somewhere with a roof, a bathroom, an outlet, and a sympathetic custodian. Know where yours is before you need it.

Step 5: Make the redundant comms plan. Cell networks fail. Internet fails. Power fails. Every neighborhood mutual aid plan needs a layer that works when the grid doesn't: a phone tree that's on paper, a meeting place at a specific time each day, someone with a battery-powered radio, ham operators on known frequencies. This is the thing most pods skip and then regret.

Step 6: Prepare politically. Dean Spade's point: mutual aid without political analysis just becomes seasonal volunteering. The network should know why the state failed, who it failed on purpose, and what the network's posture is when officials show up trying to co-opt or shut down the operation. (This happened to Occupy Sandy, to Food Not Bombs in multiple cities, to the Panthers. It will happen to you.)

Exercise 1: The 20-Door Map. Draw a rough map of the 20 dwellings physically closest to yours. Write in the names of the people who live there. If you cannot fill in more than five, your mutual aid infrastructure is functionally zero. Start there. This week, meet three new people on your map.

Exercise 2: The Capacity Inventory. For the people you do know: what do they have that would matter in a 72-hour blackout, a flood, a wildfire evacuation, a health emergency? Write it down. What do they lack? Write that down too.

Exercise 3: The Low-Stakes Rep. Organize something small this month. A potluck. A tool share. A block cleanup. The content doesn't matter. The reps do.

Exercise 4: Find the hub. Walk to the nearest space that could function as a community hub if your power were out for a week. Introduce yourself to whoever runs it. Find out how decisions get made there.

Exercise 5: Read the case files. Pick one: Occupy Sandy, the Panther Breakfast Program, Casa Pueblo, Mutual Aid NYC. Spend an hour reading primary sources — not summaries. Notice what they actually did, in what order.

- Spade, Dean. Mutual Aid: Building Solidarity During This Crisis (and the Next). Verso, 2020. - Solnit, Rebecca. A Paradise Built in Hell: The Extraordinary Communities That Arise in Disaster. Viking, 2009. - Hanna-Attisha, Mona. What the Eyes Don't See. One World, 2018. - Bonilla, Yarimar and LeBrón, Marisol (eds). Aftershocks of Disaster: Puerto Rico Before and After the Storm. Haymarket, 2019. - Nelson, Alondra. Body and Soul: The Black Panther Party and the Fight Against Medical Discrimination. University of Minnesota Press, 2011. - Homeland Security Studies and Analysis Institute. "The Resilient Social Network: @OccupySandy #SuperstormSandy." 2013. - Fritz, Charles. "Disasters and Mental Health: Therapeutic Principles Drawn from Disaster Studies." 1961. - Ostrom, Elinor. Governing the Commons. Cambridge University Press, 1990.

Every disaster that ever happened was answered by neighbors first. The neighbors who showed up were the ones who already knew each other. Mutual aid is not a hashtag or a volunteer shift. It's the standing, relational, political infrastructure of a people who have decided they will not wait to be saved.

Build it now. The next one is already on the way.