Destruction As A Service

Robert Putnam's Bowling Alone came out in 2000. The argument is familiar even to people who haven't read it: civic participation collapsed across the second half of the twentieth century, social capital eroded, Americans stopped joining things. Putnam blamed television, generational succession, suburbanization, and a vague drift in cultural attention. The book is a serious work of measurement. The diagnosis has aged poorly.

Putnam saw a real phenomenon and reached for the explanations a political scientist had available in 1995. People were not bowling in leagues anymore. The leagues had been there in 1960 and they were gone by 1995, so something must have changed about people. The simpler explanation, available to anyone who tried to start a bowling league in 2025, runs the other direction: the bowling alley got sold and turned into a self-storage facility, the parking lot got rezoned, the insurance got repriced, and the league was never coming back because the substrate it ran on no longer exists. Bowling alleys in the U.S. dropped from about 12,000 in the 1990s to under 4,000 today, a 67% decline that long predates any complaint about kids these days.

This works as a supply-side story. Putnam was looking at the demand curve and trying to explain why it had moved. The change he documented happened upstream. Civic infrastructure got dismantled and replaced with private substitutes that capture the same coordination function while extracting rent on it. Leagues didn't die because people stopped caring. They died because the institutions that produced league-making as a byproduct stopped producing it.

The pattern deserves a name, because the pattern is what matters. Call it Destruction as a Service.

The pattern works through a recognizable sequence. Identify a piece of shared infrastructure that produces a public good as a byproduct of its operation — proximity, coordination, trust, attention, memory, third-place social contact, generational knowledge transfer. Unbundle that good from its institutional context. Sell access to the unbundled component back as a service, priced individually, available on demand. Externalize the cost of the now-missing public good onto the public institutions obligated to absorb the failure modes — schools, libraries, mental health services, courts, parks departments. Book the revenue against the unbundled service. Let the institutions absorb the cost.

The pattern doesn't require a conspiracy and doesn't require any individual actor to understand what they're doing. The pattern requires only that the incentives point in the same direction across enough sectors that the aggregate effect produces disassembly of the civic substrate. Most of the people who built Soho House aren't thinking about Putnam. Most of the people who built DoorDash aren't thinking about the third place. The aggregate effect lands the same way as if they were.

Soho House and its peers — NeueHouse, the Wing before it died, the Assembly, the Ned, the proliferating "members club" category — represent the cleanest version of the pattern. They sell what civic infrastructure used to provide for free as a byproduct of being a citizen: regular gathering space, assumed commonality among the people present, low-friction socializing without performance pressure. The pricing tells the diagnostic story. With proximity abundant, no one would pay $4,500 a year for it. Proximity becomes sellable as a service because the alternative — public libraries, union halls, neighborhood churches, municipal rec centers, mainline civic clubs — has been hollowed out for forty years. Soho House represents the predictable next product in a market where civic infrastructure has been disassembled.

The pattern repeats across categories. Coaching used to happen because the high school had a budget and the dad next door volunteered. Private youth sports has grown into a $40 billion industry while the public version got means-tested out of existence in much of the country. Tutoring used to happen because teachers had time and parents had networks, and now happens through a billable service. Childcare used to happen because three generations lived within fifteen miles of each other, and now costs more than rent in 49 states when delivered through the kind of service that keeps a working family solvent. Eldercare, neighborhood watching, ride-sharing among friends, knowing-where-the-kids-are, getting-a-meal-when-someone-is-sick, picking-up-someone-from-the-airport — all of it has been unbundled from the civic substrate and is now available, at a price, from someone who used to do the same thing for nothing because they were embedded in a community that produced reciprocity as a default.

The temporal pattern stays consistent across cases. Private value extraction happens at velocity. Cost displacement happens at velocity. Residual public obligation arrives slow. By the time the school district is dealing with the kid whose parents both work three jobs because childcare cost too much, the company that sold the unbundled service has booked the revenue and the executives have moved on. Civil economics describes this directly: private benefit captured at velocity, public cost socialized over decades, no balance sheet anywhere reconciling the two.

The class dimension matters here, because the pattern operates as a sorting machine. Affluent families pay for the unbundled services and get something close to what civic infrastructure used to provide, except curated and selective and predictably high-quality. They aren't just buying community. They're buying social risk insurance: insulation from institutional failure, controlled peer environments, future opportunity pipelines. They're also buying the option to never encounter the people whose lives are getting worse because the substrate they're paying to replace has gone missing for everyone else. The purchase actively widens the gap because the people who can afford the unbundled version stop showing up at the public version, which then loses political support, which then degrades faster, which then makes the unbundled version more necessary.

That cycle runs as a self-reinforcing loop. Unbundling makes the public version look bad. The public version losing support accelerates the unbundling. Accelerated unbundling makes the unbundled version more profitable. The loop runs on its own once it starts, which explains why the dynamic has continued regardless of which party held power.

What Putnam couldn't see in 2000, because the platform-and-private-equity layer of this hadn't matured yet, runs at the level of business model rather than culture. Destruction of civic infrastructure functions as the operating principle of a particular kind of firm. The firm exists to identify a thing currently produced as a byproduct of public or civic infrastructure, separate it from that infrastructure, package it as a service, and sell it back to the people who used to receive it for nothing. The civic infrastructure doesn't need deliberate destruction. The infrastructure needs only to keep degrading while the substitute service captures the demand its degradation generates.

What the postwar consensus took to build — the public libraries, the school systems, the municipal rec leagues, the union halls, the mainline churches with their basement fellowship halls, the neighborhood third places — has been getting unwound by actors who face nothing remotely like the forcing function that produced the original construction. The original got built because a world war made the alternative unsurvivable. The unwinding has been happening because financial structures made the unwinding profitable and no political coalition has formed to stop it.

A note on what the argument is and isn't. Declension narratives flatten everything into a vague mood and end up nostalgic for a past most of the people writing them never lived in. Cultural symptoms run real and run downstream. The substrate sale matters. The substrate was civic infrastructure. The buyers were firms that figured out how to monetize its components. The sellers were governments and institutions that allowed the unbundling because they had no theory of why the bundling mattered.

Design as Repair names the response. The discipline treats the public version as worth keeping running, even when the unbundled private version comes cheaper and faster and shinier. The repair frame takes the leftover infrastructure — the schools, the libraries, the parks, the courts, the agencies that absorb the cost displacement — and treats them as the actual site of the work. Diagnosis without a posture toward action stays literary, and the moment doesn't need more literary work.

Putnam was looking at the right collapse. The mechanism wasn't culture. The mechanism was a business model. Twenty-five years have gone by while social-capital advocates argued about how to revive civic life and the firms that monetized civic disassembly got better at it. The tools to refuse the trade work at the infrastructural, institutional, regulatory, and procedural level. The tools require the unsexy work of refusing to let the unbundling continue, in specific transactions, against specific actors, in defense of specific public assets.

The civic infrastructure was built once, the decision to stop maintaining it was a decision.