The Adaptation Infrastructure No One Counts

When I was thirteen, a grassfire came through the Oakland hills and pushed smoke down into Fruitvale. Not the 1991 firestorm; my family wasn't here yet for that one. But it closed streets, and it filled the air with a smell that still tightens my chest when I catch it.

I don't really remember the fire. I remember Mrs. Delgado's van.

She kept a mental list of who on our block didn't have a car. My mother's phone rang before any official alert reached us, a woman from her organizing group who lived closer to the hills calling in Spanish to say the smoke was thickening and which streets were going down. My grandmother stood on the back steps, looked at the sky, touched a leaf on her herb planter, and said we had time but not much. She'd grown up farming in the Guatemalan highlands. Reading weather and vegetation was how you ate, and how you stayed alive.

I didn't have language for it then. I only knew that the hill neighborhoods got emergency resources within the hour, and our block got Mrs. Delgado driving people out four at a time, and that the phone tree reached families who would never have called 911 because they were undocumented and afraid.

Years later, sitting through federal grant workshops where officials described adaptation infrastructure as sirens, sensors, and hardened power lines, I kept thinking about that van.

Who lived and who didn't in Chicago

The clearest documented case of social ties producing a measurable difference in survival is the 1995 Chicago heat wave. Sociologist Eric Klinenberg compared two adjacent West Side neighborhoods: North Lawndale, 96 percent Black with a 44 percent poverty rate, and South Lawndale, also called Little Village, 85 percent Latino with a 22 percent poverty rate. North Lawndale's heat-wave death rate was 40 per 100,000. South Lawndale's was 4. Ten times the difference between two places that share a border.

Klinenberg's explanation had less to do with the heat than with what had happened to daily life in each neighborhood over the previous twenty years. North Lawndale had lost more than half its population between 1970 and 1990, and its commercial streets had emptied out. Older residents stayed inside because the block outside felt dangerous. South Lawndale had gained population over the same period, and it still had foot traffic, storefronts, churches running programs for seniors. An older person who walks two blocks to a store with air conditioning survives. An older person who sits alone in a hot apartment does not. A separate case-control study found that living alone more than doubled the odds of heat-related death, while a working air conditioner or access to transportation each cut those odds by roughly 70 percent.

The complications are real and I don't want to skip them. Mitchell Duneier later argued that Klinenberg risked inferring individual isolation from neighborhood-level numbers. Duneier found that only two elderly North Lawndale victims were actually living alone, which raises a fair question about whether the neighborhood comparison explains the circumstances of particular deaths. The tenfold mortality ratio is measured. How much of it comes from social cohesion rather than poverty, housing stock, or decades of disinvestment is still unsettled among people who study this.

The pattern turns up elsewhere, though. After Hurricane Katrina, Daniel Aldrich documented how the Vietnamese American community in New Orleans East used organizational ties it already had, mostly through Mary Queen of Vietnam Church, to coordinate house-gutting crews and pass recovery information around. When the power company said it needed 500 signatures before restoring electricity to the neighborhood, the church collected more than a thousand in a day. About 90 percent of residents were back within two years while much of the city lagged. Longer-term research confirms the strong recovery but credits a mix: shared identity, information exchange, targeted programs, and material advantages some households had going in. The ties were part of the explanation. Not the whole of it.

What the forms measure

If community ties function as adaptation capacity, you would expect the systems that evaluate risk and allocate recovery money to account for them somewhere.

They mostly don't. FEMA's benefit-cost analysis counts avoided deaths, injuries, displacement, lost productivity, and utility outages. Its "social benefits" category turns out to mean the mental stress of displacement and the economic cost of lost work time. HUD's formula for disaster recovery funds calculates unmet need from damaged housing units, repair costs, income categories, and expected assistance from other federal programs. Standard residential appraisals do analyze neighborhood characteristics, so a stable school or a busy commercial street can show up indirectly, if buyers demonstrably pay for it.

None of these systems is blind to everything beyond the structure. FEMA counts several harms that aren't property damage. HUD grantees can fund community organizations and case management. Appraisals pick up some neighborhood attributes through comparable sales. But there is no input anywhere for a household's network, for what it cost to build, or for what severing it takes away. When a buyout offer arrives, the thing that got Mrs. Delgado's neighbors out has no field on the form.

There's a second consequence to that absence. Because no grant built these networks, nothing on any budget line maintains them. Physical adaptation at least makes it into a capital plan, even when the money to operate and repair it never materializes. Social infrastructure doesn't get that far.

What repeated activation costs

When Hurricane Beryl knocked out power across Houston in July 2024, community organizations moved on a timeline formal response couldn't match. CrowdSource Rescue, built during Harvey, got generators to a senior independent-living center and hundreds of meals out across the city. Roughly six dozen Interfaith Ministries drivers delivered 2,000 hot Meals on Wheels a day and checked on homebound clients while they did it. The Houston Food Bank pushed more than 400,000 pounds of supplies through over 1,600 community partners.

Interfaith Ministries had pre-positioned five shelf-stable meals per client back in June, before the season started. You can only plan like that if you already know who lives where and what each person needs, which is thousands of individual relationships between drivers and clients, accumulated over years.

The same reporting shows the cost of being called on this often. CrowdSource Rescue had 30 generators available, down from the 300 it bought after the 2021 freeze. Volunteers described 18-hour days. The Associated Press found organizations strained by serving as Houston's backup system for the fourth time in seven years.

Kathleen Tierney has argued that celebrating community resilience can fit comfortably with a politics that lets government off the hook, that praise for how well neighbors take care of each other becomes a reason not to bury the power lines. Ben Hirsch, co-director of West Street Recovery in Houston, said it more plainly after Beryl:

The equity problem cuts deeper than strain. Aldrich and Kevin Crook found that after Katrina, politically organized New Orleans neighborhoods successfully kept FEMA trailers out, which pushed temporary housing into areas with less capacity to object. Social capital protected the people who had it, and part of that protection came at the expense of people who didn't.

And these ties can be damaged by the same event they're supposed to carry people through. A study of two flooded neighborhoods found that in the Lower Ninth Ward, the typical respondent knew 100 neighbors by name. That density stopped being usable once the whole network scattered. Wealthier Lakeview residents had more contacts outside the city and more money to act on whatever those contacts offered. Poorer residents weren't short on relationships. Their relationships were all in the same floodplain.

My grandmother's weather reading, Mrs. Delgado's van, my mother's phone tree: all of it sat on one block in Fruitvale. If that block had flooded, the network would have gone under with the houses.

After the buyout

If community ties do this kind of work, then cutting them should show up afterward in the lives of people who relocated. It does, though not in a straight line.

Research following three communities after Hurricane Sandy, one that rebuilt, one where residents took buyouts, and one next door to the buyout zone, found that buyout participants reported lower place attachment and social capital three years on. The neighborhood adjacent to the new vacant lots declined too, which says something about what buyouts do to the households that stay. A longer follow-up tracked 111 people and found losses in place identity and close ties that hadn't recovered.

Set against that, another Sandy study found people who relocated were 75 percent less likely to report worsened stress than people who stayed, after adjusting for housing damage, disaster debt, and social support. Living in a damaged house with ongoing flood risk and no clear rebuilding timeline has its own weight. In the same study, respondents with friends or relatives nearby to lean on were about 85 percent less likely to report worsened stress. The ties mattered on both sides of the decision.

No federal agency systematically tracks where households go after a buyout. Whether people land somewhere safer, whether they rebuild a network, whether they end up isolated in a town where nobody knows their mother's dialysis schedule: unknown, mostly, because nobody asks after the closing.

The list people can already make

I'm not arguing anyone should stay in a dangerous place because the neighbors are good people. None of this stops floodwater or holds up a roof, and I've sat with families in FEMA trailers who would trade every relationship on their street for a house that doesn't flood.

But families facing these decisions can usually name what they'd be giving up with a precision that surprises them when they hear themselves do it. The doctor who knows the medication history without pulling the chart. A child's teacher who understands why she goes quiet during storms. The neighbor who will notice if they don't come outside for two days in August. The woman at church who drives their mother to dialysis on Tuesdays and Thursdays.

On my own block now, a retired nurse two doors down caught a kid's allergic reaction at a block party before his parents saw it. The family next door has a generator and runs an extension cord over when the power goes. A woman down the street texts me in Spanish when the air quality index crosses 150, because she knows my mother still hangs laundry outside. Each of those took years to arrive at. I can describe all of them and price none of them, which is roughly the position the people making relocation decisions find themselves in, holding a number for the house and nothing at all for the rest.