I just started my summer vacation, which tends to be a great time to reflect. Especially changing one's surroundings sparks a lot of thoughts, some of which are worth expanding - which is what led me to writing this blog post.
Cities used to have a third category, one that was neither home nor work: a place that let you exist near other people without a transaction attached. Ray Oldenburg named it in 1989, in The Great Good Place: the café, the church hall, the pub, the place that asks nothing of you, except that you show up. By the time he wrote the book, third spaces were already disappearing, and they had been disappearing for longer than my generation (millennials) has been alive. Most of Europe and the West spent the second half of the twentieth century quietly demolishing it, and did such a thorough job that the generation now old enough to notice is also the generation that's supposed to explain why everyone feels so alone.
Try to list Helsinki's third places today and you run out fast. A library, maybe. A shopping mall, maybe? But if you're honest about what counts, the mall is more geared towards conspicuous consumption and trading convenience for purpose. Not much else, and Helsinki isn't the exception, it's closer to the median outcome of a five-decade trend that started before millennials were born and accelerated entirely on Gen X's watch.
When it started, and who was driving
Oldenburg published his book in 1989 as an elegy, not a warning. The places he was describing were already going. Robert Putnam's Bowling Alone picked up the data trail a decade later and found attendance to recreational clubs down 58 percent over 25 years, family dinners down 33 percent, and friends visiting each other's homes down 45 percent, a decline that traces back to the 1960s and 70s, squarely on the Boomer and Gen X watch, not ours.
The mechanism was never mysterious. Postwar suburbanization moved American, and later European life away from dense downtowns and into car-dependent sprawl. The white picket fence, chalet adosado or omakotitalo became the dream that was sold to and bought by families in most of the West. But, it gets really difficult to have thriving third places if there is nowhere to walk to. Zoning did the rest, sorting commercial, residential, and civic space into separate zones that only a car can connect. Then came the specific accelerant in the 2008 Great Financial Crisis: austerity. A case study out of Coventry University tracked how post-crash austerity urbanism in England closed libraries and community centers even as new commercial substitutes, like Fab Labs, tried and mostly failed to replace what civic infrastructure used to do for free. A 2025 analysis of US census-tract data from Axios found a rapid decline in third-place availability just between 2019 and 2021, concentrated in exactly the communities that could least afford to lose them.
But the starting position wasn't the same everywhere, and it's worth being precise about why, because it explains a comparison I'll come back to later. Spanish towns spent centuries walled and defensible, a legacy of Roman fortification and then the Reconquista, which for hundreds of years meant a hostile frontier ran through the peninsula and a town that wasn't dense enough to defend didn't survive to have a history. People and dwellings packed inside the wall, and the fields that fed them stayed outside it, worked from town rather than lived on. That geometry outlasted the walls themselves by centuries: density first, agriculture at a commute. Finland never had the same pressure. Settlement there followed the arable land itself, dispersed across glacial ridges and forest clearings won one farm at a time through burn-beating cultivation, with each household sitting on or next to the fields it depended on rather than walking out to them. Spain built density for defense and inherited a walkable urban core by accident. Finland built dispersal for agriculture and inherited the opposite. 20th century car-dependent suburbanization then landed on two very different starting points, thinning out what density Spain had and further scattering what Finland never concentrated in the first place.
By the time millennials were adults with disposable time, the physical stock of places to spend it had already been destroyed by two generations of planning and budget decisions. What was left had mostly gone commercial: a café that expects a €6 latte purchase every forty minutes, a gym that charges for the privilege of standing near other people. Social media gets blamed for finishing the job, but I believe that framing gets the sequence backwards. Social media didn't create the vacuum. It moved into a hole that policy and urban planning had already dug, a full generation before the first smartphone. The platforms found an audience that had nowhere left to gather in person, and heavily monetized the gap. Blaming the thing that filled the hole, rather than whoever dug it, is a pattern worth calling out now, because it's going to come up again in this piece.
What the data actually says it costs
Public health academics have actually done very interesting research in this respect. The US Surgeon General's 2023 advisory put the mortality risk of chronic loneliness on par with smoking fifteen cigarettes a day. A 2024 systematic review in Health & Place tied the built environment directly to loneliness outcomes, meaning this isn't just a mood problem, it's an infrastructure problem with a measurable physiological signature: Cacioppo's work on lonely individuals found altered gene expression tied to inflammation, and a 2024 cross-sectional study in Scientific Reports modeled urbanization and social infrastructure access as direct predictors of loneliness and downstream mental health, not correlates of it.
Coupling has followed the same curve. A large European Social Survey analysis tracking cohorts born between 1930 and 1989 across 30 countries found singlehood after leaving home rose sharply, especially among women, up eleven percentage points to 52 percent. Population Europe calls it a relationship recession: one in seven young EU adults now live alone, and even couples who do move in together are less likely to marry or have a child than earlier cohorts. None of this is about people falling out of love faster. It's what happens to coupling rates when there are fewer unstructured places to meet a partner in the first place and fewer social contexts left that put strangers in the same room without a transaction attached.
Pontevedra, and the weird version of third places I actually grew up in
Writing about this feels at the same time relatable and odd to me, because I grew up inside some sort of experiment about third spaces.
My hometown of Pontevedra banned cars from its historic center in 2000. It wasn't just a gesture: traffic through the center went from roughly 80,000 vehicles a day to less than 7,000, while the population and vehicle ownership both grew. Motorized journeys within the city dropped by two thirds, speed limits came down to 30 km/h, then 20, and 10 on the narrowest streets. Sidewalks got much wider, benches went in, and the mayor, who everyone knows as Lores, made the same decision every year for a quarter of a century since: prioritize the pedestrian over the cyclist, the cyclist over the bus, the bus over the car, and hold the line against exceptions. Today close to 70 percent of all journeys in Pontevedra happen on foot, against a European average nowhere close to that for a city its size. No pedestrian has died in traffic there since 2011 (!), and traffic accidents fell by 2/3.
The demographic and health numbers move in the same direction, against the regional trend rather than with it. Galicia as a whole has been shedding population for over a century, but Pontevedra has grown by roughly 1,000 people a year since 1999, making it the fastest-growing city in the region, and it holds the lowest urban unemployment rate in Galicia, the youngest urban population in Galicia, the least outbound emigration in the province, and the highest urban birth rate of any city its size, against a regional backdrop where Galicia's overall fertility rate sits well below the Spanish average). None of that proves pedestrianizing a city center raises the birth rate, but it's a nice directional data point. Pontevedra also benefits from a functioning local economy that other pedestrianized towns don't have. But a place that made the built environment easier to be young, mobile, and social in has ended up, over the same 25 years, both healthier by the ordinary measures (air quality good 70 to 80 percent of the time, zero pedestrian deaths) and more demographically alive than its neighbors, bucking the trend of a region where the surrounding towns are aging and emptying out. That correlation is consistent with the mechanism, even if it can't carry the whole causal weight on its own.
Most mid-sized cities never made that choice. They chose the status quo, kept the arterial roads, the lower building density, the assumption that getting from A to B means a car, and by extension they kept the isolation that comes bundled with it. I notice the difference every time I go back. I meet more people by accident in one afternoon of walking around Pontevedra, stopping to tomar algo with whoever I run into, than I do in a month of errands in Vantaa. That's not nostalgia talking, but the result of a very intentional choice of urban policy.
Research backs the anecdote up more than I expected it to. Leyden's 2003 study in the American Journal of Public Health compared residents of walkable, mixed-use neighborhoods to residents of car-dependent suburbs and found the walkable group scored higher on every social capital measure going: they knew more neighbors, trusted more people, and were more civically engaged. A more recent review on neighborhood walking specifically, separate from the built form itself, found that the physical act of walking through a neighborhood is what drives the social connection, not just the design that makes walking possible. And on the outcome I actually care about right now, as a new parent: a built-environment study on maternal mental health found that women in multi-purpose, walkable communities had a measurably lower rate of postpartum depression than women in low-walkability, single-use neighborhoods, even after adjusting for income and education.
None of that is a claim that Pontevedra has solved fertility, or that pedestrianizing a city center is sufficient on its own. Some of that social capital effect is self-selection: people who already want a socially dense life may simply choose to live somewhere that offers it. But the mechanism runs in a meaningful direction. Chance contact requires proximity, and proximity requires a design that makes walking the default instead of the exception, and chance encounters minimize isolation and loneliness while increasing life satisfaction. You can't app your way to what a plaza does for free.
The convenient villains
Which brings me to the part that actually annoys me.
Whenever these numbers come up culturally, the explanation on offer is never "we spent fifty years dismantling the physical infrastructure of ordinary social contact and then charged rent on whatever replaced it." The explanation on offer is dating apps, millennials being too picky, or, increasingly, remote work: the one material win our generation actually extracted from a labor market that otherwise gave us worse wages, worse housing, and less job security than Gen X or Boomers got at the same age.
There's a real thread of evidence behind the remote work critique, and I'll give it its due without letting it carry the argument. A 2026 study in Science did find a genuine net increase in loneliness tied to the post-pandemic rise in full-time remote work, and the effect scales with frequency rather than showing up at one or two days a week. Fair enough. But look at the timeline next to it. Putnam's decline data starts in the 1960s. Oldenburg was writing an elegy for the third place in 1989. Remote work at scale is a five-year-old phenomenon. A five-year-old practice did not cause a sixty-year-old trend, and the same research that flags its cost also credits it with keeping parents, especially mothers, in the labor force who would otherwise have had to leave it entirely. Remote work is an amplifier layered on top of an already-collapsed social infrastructure. It didn't build the vacuum. It's the second thing, after social media, to get blamed for moving into one that was dug by other people's zoning maps and austerity budgets. Curious about who is driving that narrative? You don't have to guess very hard: it's mostly employers who are hell-bent on physical presence at the expense of productivity and workforce wellbeing (with little data to back up their thesis); and commercial real estate interests who stand to lose if the value of office space goes down.
There's a name for the underlying mechanism, and it's worth stating precisely instead of just gesturing at it. In 1968, the ecologist Garrett Hardin published an essay in Science arguing that any resource held in common, a shared pasture, a shared fishery, a shared civic square, gets degraded when individuals are free to extract value from it without bearing the cost of that extraction themselves. His conclusion was blunt: "Freedom in a commons brings ruin to all." A third place is a commons in exactly this sense. A city council that closes a library saves money it can point to immediately. A developer who replaces a plaza with retail captures rent it can bank immediately. The cost of both decisions, the lost chance encounters, the coupling that didn't happen, the postpartum depression that showed up in a neighborhood with no third place to walk to, is diffused across everyone else and doesn't arrive on a balance sheet for another twenty years. Nobody who made the cut has to pay for what it produced. That's not a millennial character flaw. That's an externality, on a decades-long fuse.
Hardin's own prescription for this, coercive population control, was as bad as the diagnosis was sharp, and he was rebutted in his own lifetime by the economist Elinor Ostrom, who showed that real-world commons are usually managed well by the communities that depend on them, not doomed to collapse by some law of human nature. Ruin isn't inevitable. It's what happens by default when nobody is actively governing the commons, which is a very different claim, and a more useful one, because it points to something that can actually be built rather than just mourned.
The angrier cost
There's a version of this argument that extends past loneliness and low birth rates into something with sharper teeth: politics and polarization.
The easy explanation for why cities vote one way and the countryside votes another is sorting, the idea popularized by Bill Bishop's The Big Sort that people with certain values simply move to where people like them already live. The data doesn't fully support that as the whole story: large studies using individual-level survey data have found that population density itself predicts party support even after controlling for many individual factors. This is driving us further and further apart from each other, not just physically, but also ideologically: a European study published in 2025 found substantial affective polarization along the urban-rural line, tied to real resentment on both sides, correlating with support for green and liberal parties among urbanites and for traditional, authoritarian, and nationalist parties among rural populations.
Behind that is one of the more replicated findings in social psychology. Over 500 studies covering more than 250,000 people have found that contact with an outgroup reduces prejudice toward it, through reduced anxiety and increased empathy. Density forces incidental contact with people unlike you as a side effect of simply inhabiting a shared space.
Isolation does damage here on its own, independent of geography entirely. People who spend more time alone with smaller social networks score higher on measures of political extremism, an effect that research finds is mediated specifically by loneliness. A 2021 study of former extremists found loneliness was among the most commonly cited reasons people gave for adopting extremist views in the first place. Weak social belonging on its own predicts a higher probability of political radicalization. Note that none of that is about a specific ideology, but rather about what someone with no third place and no incidental contact left in their world has available to them, when they go looking for belonging, and who's waiting there to offer it to them. This can be seen across religious cults, the incel movement and other extremist organizations.
If the mechanism runs through contact, then rebuilding the physical conditions for contact is not just a nostalgia project, it's a clear lever against exactly the kind of polarization we all say we want less of. A plaza doesn't ask anyone's politics before it puts two people at adjacent tables, and gets them discussing. The infrastructure that fixes loneliness is, by the same pathway, the infrastructure that makes people harder to radicalize. It's one of the few interventions that isn't zero-sum between generations, and one of the few reasons to think the fix is more tractable than the problem looks.
What would it take to get third spaces back?
Ostrom's point, applied here, is the whole argument: the third place doesn't come back on its own, and it doesn't come back by scolding a generation for adapting to the environment it was handed. It comes back through deliberate governance of the commons, by people with the standing to act on it. A short, concrete list of who that is and what they could actually do:
Local government
Pontevedra is not a museum piece, it's a replicable policy: cap speed limits, remove through-traffic, widen sidewalks, legalize corner cafés and small mixed-use commerce inside residential zoning instead of restricting it to arterial roads. None of this requires new technology, just sustained political will over multiple terms, and a little faith from the population that a better urban model is possible and walking doesn't make you less of a man/woman.
Regional and national government
Treat libraries, community centers, and public squares as core infrastructure spending, the way Eric Klinenberg argues for in Palaces for the People, not as the first line item cut when a budget tightens. Invest intentionally and consistently in building and improving spaces that actually bring people together and add value to their lives. Build higher density closer to public transit, complete with a varied palette of public and private services (cafés, libraries, gyms, small independent retail) that have a low threshold to success. Tie transit and housing funding to walkability metrics the way health impact assessments already measure them, so the incentive runs toward density and mixed use instead of against it.
Private sector
Employers who've kept hybrid or remote policies could redirect some of what they used to spend on office perks toward the neighborhoods their people actually live in: sponsoring a community space, a co-working room with a childcare corner, a recurring open meetup, instead of a ping-pong table nobody commutes in to use. Commercial landlords could trade a fraction of retail square footage for a tax or zoning concession in exchange for genuinely non-transactional public space, a bench, a covered courtyard, a corner that doesn't require a purchase to sit in.
Civil society, meaning us
Infrastructure is necessary and not sufficient. A plaza with nobody in it is still just pavement, so if we don't use it, it loses its purpose. The unstructured hangout, the standing weekly walk, the tomando algo with no agenda, has to be practiced deliberately by people who were raised, correctly, to optimize every hour. That's the one piece of this that doesn't require a vote or a budget line, just a decision to show up somewhere with no purpose attached and stay long enough to be interrupted by someone else.
None of these fixes the underlying fifty-year deficit by themselves, and I'd be skeptical of anyone who tells you it does. But the commons doesn't require nostalgia to rebuild. It requires someone deciding, the way Pontevedra decided in 2000, that it's worth the political cost. I'd like my daughter to grow up with a version of the plaza, somewhere she can exist in public alongside her peers, her elders and her juniors. Right now I couldn't tell you where in Vantaa that would be.
Image credit
Photo of Praza da Ferrería, Pontevedra by Bene Riobó, licensed under CC BY-SA 4.0.