Notes — Chris Arnade on Walking Cities

Notes on Chris Arnade in conversation with Tyler Cowen — Conversations with Tyler (https://conversationswithtyler.com/episodes/chris-arnade/), 18 June 2025 (recorded 27 February 2025).

Four questions [Adler frame]

Q1 — What is it about as a whole? A working method — walking a city fifteen miles to read it off its streets — examined by its own practitioner. Arnade, a former physicist and bond trader who now walks the world’s cities and files a Substack about it, uses the conversation to lay out how he learns a place on foot, where the method works (Seoul, Istanbul, Amman, the Faroes, El Paso) and where it fails (Beijing, Shanghai, Phoenix, Dakar). The talk radiates from walkability into the politics of urban design, McDonald’s as public infrastructure, the case for cities over wilderness, and finally the priors the walking overturned — on religion, cultural relativism, and the power of elites.

Q2 — How is it argued? Almost entirely from accumulated first-hand observation, not theory. Arnade reasons from concrete cases — the six security checks into Tiananmen, the painted-mural shopfronts in Shanghai, an East LA McDonald’s, a Kampala bar — to general claims. Where he reaches for a framework it is borrowed and lightly worn: James C. Scott on legibility and top-down control, Walter Benjamin’s ‘aura’, the ‘1,000 true fans’ idea. Cowen supplies a running counterpoint (the partial-freedom theory of Chinese control, the romantic defence of museums), so several claims are pressure-tested in the exchange rather than asserted flat.

Q3 — Is it true, in whole or part? The observational claims are well-grounded in Arnade’s unusual sample — a decade of walking hundreds of cities — and he is candid about the method’s limits, conceding that some cities defeat walking and that he has ‘gone in too blind’, under-reading political background. His larger generalisations are impressionistic by nature. [?] The ‘98 per cent of people are not political’ figure is a rhetorical estimate, not a measurement. [?] His claim that some cultures are simply better than others is asserted as conviction, not argued. His attribution of the ‘1,000/2,000 true fans’ idea wavers on air between Kevin Kelly and Stewart Brand (the essay is Kevin Kelly’s).

Q4 — What of it? A portable epistemology: pick a single, embodied method (here, walking), push it until you find its domain and its edges, and treat the failures as data about the place rather than about the method. It also supplies the wiki’s first sustained treatment of urban walkability and street-level sociology, and — through the ‘elites build culture’ turn and the Scott-inflected reading of control — a link to the wiki’s cluster on power, legibility, and how societies are governed.

Glossary

Totalitarian anarchy — Arnade’s coinage for a state that intends tight control but is too incompetent to achieve it: the apparatus of surveillance and checkpoints is real, yet porous enough to walk through. [§ On a Scottian view of Chinese cities]

Organic street life — the low-regulation, spontaneous, shop-and-crowd life of a lived-in street (Taipei, Istanbul’s Asian side); its deliberate removal — replaced in one Shanghai case by painted murals of shops — is Arnade’s marker of top-down control. [§ On a Scottian view of Chinese cities]

Walkability metric — Arnade’s roughly eight-category score for a city, weighing climate, crime, pollution, density, topography, and local distribution; used to rank places and to defend counter-intuitive calls like a high score for Los Angeles. [§ On dimensions of urban walkability]

Local distribution — the single biggest walkability factor in Arnade’s account: the guarantee that a shop is always somewhere within reach, so a walk is never a dead stretch. [§ On Istanbul and India]

Stochastic experimental learning — Arnade’s name for learning by wandering and chance encounter, as opposed to book-learning; he holds that both are needed and admits he has leaned too far into the former. [§ On what long walks teach you]

Front row / back row America — the class distinction from Arnade’s book Dignity: the ‘front row’ being those with post-graduate degrees from elite institutions, the ‘back row’ everyone else; El Paso is his case of back-row optimism. [§ On where not to go]

Key claims by section

On a Scottian view of Chinese cities [§ On a Scottian view of Chinese cities]

Beijing and Shanghai felt strikingly similar and ‘inscrutable’ to Arnade at the granular, on-foot level; he suspects the uniformity is intentional — a design meant to remove difference — and likens both to a cartoon whose background keeps repeating.
Reading James C. Scott, he interprets Chinese urban form as top-down regulation for control: small winding neighbourhoods replaced by fifty-storey towers and malls, which are easier to govern and to surveil.
He coins totalitarian anarchy — a regime that intends control but is too incompetent to deliver it (he cleared several Tiananmen security checks without the required QR code; the near-universal VPN makes the firewall a filter on the incompetent rather than a wall, and a monitoring channel besides).
The signature image: a Shanghai district bought for redevelopment where real shops had been replaced with painted murals of shopfronts — organic street life removed and substituted with a cartoon of itself. Cowen adds that the looseness may be partly deliberate, since partial freedom makes people easier to control. [?]

On favorite McDonald’s locations [§ On favorite McDonald’s locations]

In the US, McDonald’s is used because it is functional relative to a poor neighbourhood; in Beijing, Arnade found people used it because it is dysfunctional relative to the regulatory tightness around it — somewhere you can simply sit and relax.
Abroad he usually avoids McDonald’s because other cultures supply their own organic food spaces (Japan’s 7-Elevens and izakaya); China was the exception.
Much of Dignity was written in an East LA McDonald’s; his addiction-and-poverty stories came from sitting there typing notes, and the recognition that he kept meeting everyone at McDonald’s — a genuine community centre and site of social mixing — was itself the story.

On Seoul [§ On Seoul]

Seoul is a favourite for its ‘functional dysfunction’ — Tokyo’s safety and efficiency, but quirkier, less uptight, less known; he does a beloved ten-mile daily walk along a reclaimed former drainage ditch.
On food he rates Japan’s median higher than Korea’s (and Japan’s worst worse — chicken sashimi), while agreeing Korea’s scene is less understood and less precious.

On dimensions of urban walkability [§ On dimensions of urban walkability]

Arnade’s roughly eight-category metric weights climate, crime, and pollution heavily, plus density, topography, and local distribution.
He defends ranking Los Angeles high — great weather and skies, a ubiquitous bus system, the best US food scene — over online objections; Phoenix is nearly unwalkable and defeated his deliberate attempt to prove otherwise.
Dakar and Kampala rank low on heat, absent walking infrastructure, and the risk of a language-barrier miscommunication (not robbery) spiralling; topography alone can decide it (Rio’s mountains, tunnels, and no-go zones).

On physical bothers [§ On physical bothers]

Heat and dehydration are his chief physical problems, because the places he finds most interesting cluster near the equator, and he tolerates heat less well with age.
He packs very light on the principle that a third of what one brings goes unused and anything can be bought on the road; his only wants are more clean shirts and a second pair of shoes for Sunday mass.

On what long walks teach you [§ On what long walks teach you]

Arnade cut the ‘boring parts’ — wheat fields, the Gobi, pine forest — from an original plan to walk around the world, doubting he would learn much on the fourth day in a forest; the object is people and culture, not landscape.
He distinguishes book-learning from stochastic experimental learning (his own method) and says both are needed — he has leaned too far into walking and plans a six-month reading break.
What he learns least well is political background; he reads too little of it, though he notes most people everywhere (his estimate: ~98 per cent) are not political and do not raise politics unprompted. [?]

On the Faroe Islands [§ On the Faroe Islands]

The Faroes earn Arnade’s one exemption from being ‘two with nature’: sublime, never-tiring beauty plus a functionality he prizes; he extended a five-day visit to two weeks.
His method there is to take a bus out, walk four or five hours, and catch a timed bus back, island-hopping on foot.

On Istanbul and India [§ On Istanbul and India]

Istanbul is among the most walkable cities (Tokyo takes the top award); its strength is local distribution — always a shop nearby — plus weather, beauty, diversity, and lived-in history (a thirteenth-century wall used as a car park), especially on the Asian side.
India, Jakarta, and Hanoi rate only ‘okay’, dogged by pavement potholes that force constant downward attention; much of the developing world was not built to be walked, and locals rightly prefer scooters.

On where not to go [§ On where not to go]

He endorses deliberately visiting a place you are sure you do not want to go (subject to safety) — the strategy that ‘failed’ with Phoenix, which has no intention of being walkable and, he grants, should not be.
He is burned out on Latin America from his banking years and unmoved by the Caribbean and Scandinavia; he is curious about Suriname/Guyana (drawn partly by Guyana’s ~40 per cent GDP growth and oil), Russia, and Belarus, tempered by risk.
El Paso is his study in optimism: a low-crime, high-optimism working-class Mexican-American city where the American dream is alive; he credits the slope of the curve over its height — residents can see across to Juárez how far their lives have improved.

On the good and bad of working on Wall Street [§ On the good and bad of working on Wall Street]

Best: smart people, and the enforced serial expertise (three months on oil, then palladium) of a macro emerging-markets trader.
Worst: the narrow ‘view from the Ramada’ — interchangeable wealthy neighbourhoods, an entire national bond-investor base fitting into one restaurant each night.

On the role of pigeons and snapping turtles [§ On the role of pigeons and snapping turtles]

The addiction project began through pigeons: financial-crisis-era walks around New York led Arnade to the Black and Puerto Rican pigeon keepers of the Bronx and Brooklyn, and following them to Hunts Point started the addiction work.
His Substack, Chris Arnade Walks the World, pays him to walk the world; he invokes the ‘1,000/2,000 true fans’ idea as a liberating model (you need only a niche and a few thousand people) and has no plan to turn the newsletter into a book. [?source — he attributes the idea uncertainly to Kevin Kelly, then Stewart Brand; the essay is Kevin Kelly’s.]

On AI-assisted travel [§ On AI-assisted travel]

Cowen and Arnade disagree about museums: Cowen (the self-described romantic) values the aura of the terracotta warriors or the Hagia Sophia — the splendour no reproduction captures; Arnade poses ‘the Walter Benjamin question’ and shrugs at museums, visiting only military-history ones for their propaganda.
Cowen makes the practical case for AI in travel — photograph any painting, bird, or building and interrogate it beyond any guidebook; Arnade, so far using AI only as a copy-editor, is starting to consider it.