Cities
in a Hurry

The Whole Thing in One Page

Cities are usually shown from above. The skyline stands for the place, the boundary contains it, and a population figure tells you how much city there is. None of these describes what city life feels like from the pavement. From there, a city is a set of things you can reach and things you cannot avoid.

An address gives access to jobs, schools, customers, friends, hospitals, parks, shops and strangers. It also brings rent, noise, traffic, heat, pollution, surveillance, delay and risk. The distance to any of them matters less than the time, money, reliability and safety required to cross it. Five kilometres on a frequent railway can be nearer than one kilometre across a motorway. A clinic beside your home is still distant if it is unaffordable or excludes you.

This helps explain why people gather in cities. Putting people and organisations close together lets them share expensive facilities, match specialised needs and learn from repeated contact. A deep labour market helps a rare skill find an employer. A large customer base supports a restaurant, theatre or repair shop that a small settlement could not. Ideas move through meetings, imitation and staff changing jobs. Proximity creates options.

It also creates competition for the same ground. Useful locations become valuable, and the value appears in land prices and rents. Households then trade space against access. Firms outbid homes in some districts. Wealth buys a shorter journey, a quieter street or an escape from danger. Rules about what may be built decide how many people can share a valuable location, while transport decides how far that location extends.

Density therefore needs machinery. Water must arrive clean, sewage and rubbish must leave, power must hold, streets must carry movement, and public authority must settle conflicts among neighbours. When these systems work, millions can occupy a small area and gain from one another. When they fail, the same closeness spreads disease, fire, flooding and breakdown. The advantage and the hazard come from the same compression.

Places also shape social life. Neighbourhoods influence whom people meet, what they see as normal, where they feel recognised, which institutions they can use and how much danger or support surrounds them. Yet people select neighbourhoods, and governments, landlords, lenders and past discrimination help select for them. Place changes probabilities rather than issuing commands. The strongest evidence shows effects that vary by age, duration, person and setting.

The system has no single physical form. It can be a compact rail city, a dispersed motorway region or a settlement built house by house before pipes and title arrive. Formal plans are one source of urban order; markets, custom and residents' improvisation are others.

Finally, cities remember. A railway station raises nearby value; that value attracts investment; investment changes residents and politics; politics protects or redirects the gain. A road, sewer, zoning rule or excluded district can govern behaviour long after its authors are gone. Residents keep remaking the system through moving, building, trading, voting, resisting and using spaces in ways planners did not expect.

People make places. Places narrow and enlarge what people can do. Then those people alter the place inherited by whoever arrives next.

That is the book.

Why You Should Care

The voucher paid towards a different home. The question was whether it could also buy a different future. Between 1994 and 1998, more than 4,600 low-income families joined the United States government's Moving to Opportunity experiment in five cities. Some were offered help to move to lower-poverty neighbourhoods. Others were offered a standard voucher without that neighbourhood restriction or received no additional assistance through the experiment. A lottery allocated the offers, not the addresses. Families could decline, and people in every group could move. But comparing the randomly assigned groups let researchers test what offering a route to a different neighbourhood changed.

The answer depended on whose future was being measured. The lower-poverty offer improved adult safety and several health outcomes without the hoped-for gains in employment and earnings. Later, linked tax records revealed benefits further down the family tree. Children younger than thirteen when their families were assigned that offer were more likely to attend college and earned more as young adults. Older children did not show the same gains; disruption was one possible explanation. Thirteen was an analytical dividing line, not a biological deadline. A neighbourhood was no medicine with one dose and one effect.

The experiment gets past two lazy answers. One says a postcode determines a life. The other says place is scenery and people carry their futures entirely inside themselves. Both miss the repeated influence of surroundings. A place changes the schools, peers, journeys and dangers encountered over years. It alters the cost of choices without making them for you.

You make urban decisions constantly, even when you do not call them that. The extra rent for a flat near a station is a price for time and options. A road that makes one journey faster may make walking less safe and another district noisier. A new park can improve daily life while raising nearby land values enough to exclude some of the people it was meant to serve. A bus route can be more consequential than a landmark because it changes which jobs can be accepted without buying a car. And not every reason to stay can be costed as a fare. Leaving can mean losing the neighbour with a spare key or the shop where speaking your language needs no explanation.

Employers, investors and governments make the same decisions at larger scale. They choose whether opportunity will be concentrated or connected, whether growth will be absorbed by more homes or higher prices, and whether the costs of traffic, heat, waste and flooding will fall on those with the least ability to leave. These choices become physical. Once poured in concrete or written into property rights, they can outlive the policy that produced them.

Even the size of a city depends on the question. A mayor's boundary may end halfway through a commuter's daily life. Population figures can count the municipality, the continuous built-up area or the wider labour market and give different answers without anyone making an arithmetic error. Before comparing cities, we need to decide which city we are talking about. The skyline will not settle it.

This book follows those connections from early settlements to the self-built neighbourhood and the metropolitan railway. It cannot turn findings from particular United States and European studies into laws of human settlement. The same street can offer freedom to one person and danger to another. A useful model must explain that difference rather than wish it away.

Still, the central question travels. What does this place put within reach, what does it force people to absorb, and who gets to change either? Once you can answer that, a city stops looking like a collection of buildings. It becomes a distribution of possible lives.

The Core Ideas

A City Is a Pattern of Access

Ask where a city ends and the answer changes with the reason for asking. A mayor may point to a legal boundary. A commuter follows the labour market beyond it. A water engineer sees a catchment and a pipe network. A renter sees the area reachable within a tolerable journey. Satellite imagery sees continuous roofs and roads. All are describing a city, but none is describing the same object.

This matters because municipal lines are often accidents of annexation, law and politics. One built-up area may contain dozens of governments. Another authority may include fields and distant villages inside a boundary carrying the name of a famous city. Comparing their populations, densities or budgets as though the units were equivalent creates precise nonsense.

Urban researchers therefore use several maps. A built-up area follows continuous development. A functional urban area joins a dense centre to the surrounding places tied to it by commuting and other flows. A metropolitan area is usually an official attempt to capture that wider system. The United Nations now publishes harmonised estimates based on population concentration alongside national definitions because the question cannot be solved by accepting labels alone.

Scale and density should also be separated. A million people spread across a wide motorway region and a million packed into a compact rail city create different possibilities. Density describes how much is concentrated in a given area. Scale describes how much population or activity the wider system contains. A small dense centre may offer intense street life but a shallow labour market; a large dispersed region may offer many jobs but make them costly to reach.

The useful common feature is access. A city puts many destinations within possible reach of one another: workers and employers, buyers and sellers, patients and hospitals, audiences and performers, pupils and schools. Density helps because shorter physical distances can support more contact. Scale helps because a larger population can sustain rare services. Networks matter because a railway, bridge or broadband connection can join separated concentrations into one practical system.

None of these is enough by itself. A dense district cut off by a river crossing, a fare or fear may provide poor access. A dispersed region linked by fast, reliable transport may function as one city despite low density in many parts. Personal ability matters too. A staircase changes the city for a wheelchair user. A late shift changes it for someone whose bus stops at midnight. Safety, language, legal status and care duties alter the map without moving a road.

Exposure is the reverse side. The same arrangement determines who breathes traffic pollution, hears aircraft, reaches shade, faces floodwater or lives beside a waste site. A household can enjoy access to the metropolitan economy while carrying a disproportionate share of its costs. Calling both conditions urban does not make them equal.

So the city is best treated as a changing field rather than a container. Every address occupies a position in several networks at once. Its value depends on what can be reached, by whom, at what cost and with what reliability. Its burden depends on what arrives uninvited. The skyline is visible. The city that governs daily life is made of travel times, prices, permissions and risks.

Proximity Creates Increasing Returns

Imagine a workshop that repairs an obscure kind of machine. Cheap premises are no bargain if the nearest customer is a day's drive away. Put the workshop near enough factories, spare-parts suppliers and trained mechanics, and its expensive address begins to pay for itself. This is an illustration of agglomeration economies: benefits that arise because other activity is close by. Financial districts, garment quarters and manufacturing belts embody different versions of the bargain.

The main mechanisms are sharing, matching and learning. Sharing lets many users support an input too expensive for one alone: a specialist laboratory, for example. Matching gives workers more possible employers and firms more possible recruits, making a rare skill less likely to go unused. Learning happens through observation, conversation and people moving between organisations. These are distinct advantages. A city may offer a deep recruitment pool without being especially good at helping knowledge cross between firms.

Clustering also shares risk. A worker in a one-employer town can lose both job and reason to remain when the plant closes. In a large labour market, another employer may need the same skill. A firm can experiment with a narrow product because suppliers, finance and customers are already nearby. Failure remains costly, but the surrounding system can reuse people, premises and knowledge rather than losing the whole investment.

Learning is easy to romanticise. Ideas do not drift through urban air. They travel with people and through relationships. Spanish employment records allowed researchers to follow workers across jobs and cities: experience accumulated in a larger city continued to pay after workers left it. The finding helps distinguish a skill carried by a person from a wage premium attached only to an address. It does not promise that every resident of a large city gets the same education.

Cities also allow specialisation in consumption and care. A modest town may support broad general care; a large urban region can support teams for rare conditions. A larger audience can sustain minority-language media, experimental music, specialist food and niche retail. The gain is not that every resident uses every option. It is that an option can exist because enough scattered demand has been brought together.

Increasing returns create cumulative growth. Once a cluster forms, suppliers arrive, trained workers move nearby and institutions adapt. The place becomes attractive partly because it is already attractive. Yet the process has limits. High rents, congestion, crowding, pollution and weak governance can erase the benefit. Some activities need cheap land more than dense contact. Digital communication substitutes for parts of proximity, especially routine exchange, while leaving other work dependent on trust, tacit knowledge, equipment or face-to-face coordination.

Agglomeration can widen differences inside the same city. The people who own scarce land or hold skills complementary to a growing cluster may capture large gains. Others face higher rents without matching wage growth. A booming economy can therefore create more opportunity and sharper exclusion at the same time.

Sorting also clouds the evidence. Productive people and firms may choose successful cities, making the place look more powerful than it is. High wages may compensate for high housing costs rather than represent a higher standard of living. Output per worker can rise while cleaners, carers and service staff are pushed into longer commutes. An average urban premium says little about its distribution.

Proximity changes what a cluster can do because one person's presence alters other people's options. The workshop gains customers, the mechanic gains possible employers, and a failed firm need not take every local use for its skills down with it. Cities persist despite their costs partly because they make such connections easier to complete.

Land Turns Access into Price

The city creates access, but access does not remain free. It is attached to locations, and locations cannot be copied. When a station opens, a school improves or a district becomes safer, nearby land often becomes more valuable. The benefit is partly converted into rent, sale prices and development pressure.

Urban land models begin with a trade. A household can pay more for a central or well-connected location, or pay less and spend more time and money travelling. Firms make their own bids according to how much they value customers, workers, freight links or contact with other firms. The resulting pattern is never as neat as the textbook version, because cities have many centres, households differ, remote work exists and history leaves awkward parcels.

That conversion explains why an urban improvement can produce two honest reactions. A new railway, park or safer street makes a place better for existing residents. It can also raise the cost of remaining there. Owners may gain an asset windfall. Renters may face higher rents or a landlord seeking a different tenant. Future low-income households can be excluded even if few current residents are physically evicted. The question is not whether improvement or displacement is the real story. Both can follow from the same change.

Housing supply affects how the pressure is divided. Where more homes can be added in desirable areas, population can grow alongside demand. Where construction is constrained by physical limits, slow approval, restrictive zoning, finance, infrastructure or political resistance, more of the demand appears as higher prices and overcrowding. Regulation can protect safety, light, heritage and public goods. It can also reserve access for incumbents by preventing others from sharing a valuable location. Treating every rule as either benevolent planning or needless obstruction avoids the hard work of asking what it prevents, whom it protects and what alternative it permits.

Public finance enters at the same point. A railway or park paid for by everyone can create gains captured mainly by nearby owners. Taxes on land value or development obligations attempt to return part of that gain to the system that produced it. They can fund infrastructure or affordable homes, but their effects depend on timing, design and whether charges reduce construction.

Land markets do not start from equal claims. Property law, colonial allocation, racial exclusion, inheritance, public housing, mortgage systems and informal tenure determine who can capture rising value and who can be removed. A family occupying land without recognised title may have built a home, paid local charges and lived there for decades while remaining vulnerable to clearance. Formal ownership gives some residents the ability to wait for appreciation; insecure tenure can make the same improvement a threat.

Prices also transmit information imperfectly. A high rent signals strong demand relative to available space, but it does not reveal whether the demand comes from jobs, speculation, school access, tax treatment or a shortage created by rules. Nor does willingness to pay equal social value. A nursery, workshop or long-standing market may be outbid despite providing something residents consider irreplaceable.

The land beneath a city is therefore an allocation system. It turns collective investments and social change into private costs and gains. Any policy that changes access without considering this conversion risks improving a place while excluding the people whose lives were meant to improve.

Movement Makes the Map

A city is often measured in kilometres and experienced in minutes. Transport turns physical separation into practical distance, and every mode draws a different map.

A rail line creates corridors of access around stations. A frequent bus can join districts too dispersed for rail. Walking makes short, mixed trips possible without a timetable, provided crossings, gradients and safety permit it. Cars offer flexible point-to-point movement while demanding road space and parking at both ends. Freight has its own geography of ports, depots, delivery windows and loading bays. The city that works for a commuter travelling to a central office may fail a carer linking a school, a clinic, a supermarket and two homes.

Transport and land use then change each other. Faster travel makes more land reachable, encouraging homes and jobs to spread. New development creates trips that can consume the capacity which enabled it. A station attracts construction; construction fills trains; crowding produces demands for another line. A motorway permits longer journeys and decentralised warehouses; those locations become hard to serve without vehicles. Infrastructure does not merely respond to a fixed pattern of demand. It helps create the pattern.

The geometry of the network matters beside its headline speed. A fine-grained street grid offers many routes and makes short trips direct. Cul-de-sacs and barriers can place two nearby homes a long walk apart. Transfers impose waiting and uncertainty that a map's straight line hides. A service every five minutes is a different form of access from the same route every hour, even though both appear as one coloured line.

This is why road widening often disappoints. Gilles Duranton and Matthew Turner found that, across United States metropolitan areas in their study period, vehicle travel on interstate highways rose roughly in proportion to lane kilometres. Existing residents drove more, transport-intensive activity expanded and people relocated. The result is setting-specific, but the mechanism is general: when a scarce route becomes cheaper in time, some travellers change route, time, mode, destination or frequency until part of the gain is used.

The opposite mistake is to assume that public transport automatically reduces congestion. A successful railway may free road space that other drivers occupy. Its main achievement can be larger: allowing far more people to reach valuable destinations without requiring each one to store a vehicle near them. Capacity, access and environmental effect are different questions.

Prices and rules allocate movement too. A congested road offered free at the busiest time rations scarce space through delay. Charging can reduce demand and improve flow, but willingness to pay is not the same as urgency: a low-paid care worker may have less freedom to postpone a trip than a wealthy shopper. Alternatives and the use of charge revenue matter. Parking requirements can hide another cost by embedding car storage in housing and shops. A cheap fare is useless when service is unreliable; a fast journey remains inaccessible to someone who cannot reach the platform.

Good transport analysis therefore begins with destinations rather than speed. How many suitable jobs, schools, customers or services can different people reach within a given time and budget? The answer joins mobility to land use, frequency, transfers, operating hours, disability and safety. Movement makes the urban map, but the map differs for every body and every purpose.

Infrastructure Makes Density Liveable

Put a million people together without a water system and you have not built a city. You have scheduled an emergency.

Urban density concentrates demand for drinking water, power, streets, drainage, communications and waste removal. It also makes networked provision efficient. One treatment plant, sewer, substation or transit line can serve many users across a small area. The same concentration raises the cost of failure. Contaminated water reaches thousands. A blocked drain floods basements. A power cut stops lifts, pumps, refrigeration and payment systems at once.

The nineteenth-century industrial city exposed this bargain. Factories and railways pulled populations into districts whose wells, cesspits and streets had been designed for far fewer people. Disease was biological, but its transmission became an urban systems problem. An infected water supply could deliver danger with the convenience of a household service. Sewers, filtration, chlorination, refuse collection and public health administration helped make dense settlement survivable. Their precise shares in the historical fall in mortality remain disputed. Removing pathogens from shared drinking water saves lives; assigning every improvement in a city's death rate to its new waterworks is a different claim.

Infrastructure is political because networks distribute service. A pipe can pass a settlement without connecting it. A neighbourhood may have electricity on paper and lose it daily. Informal workers may collect rubbish where municipal service does not reach, while carrying health risks the formal accounts ignore. Affordability matters beside engineering: a metered connection that a household cannot pay for does not provide secure access.

Collective provision is necessary because one household's waste, fire risk or drainage does not stay on its own plot. Markets can supply pipes, power and transport, but standards, rights of way, coordination and obligations determine whether separate components form a working network. Urban government acquired many of its modern powers partly from the need to manage spillovers among people living close enough to harm one another without ever meeting.

Maintenance is the hidden half. Opening a bridge is an event; inspecting bearings for fifty years is administration. Pipes leak, drains silt, rails wear and software becomes obsolete. Capital budgets reward visible construction, while routine maintenance is easy to defer until breakdown makes it dramatic. Poor districts often receive weaker service because failure there carries less political cost, which then discourages investment and confirms the appearance that the district is unviable.

Infrastructure changes behaviour once it becomes dependable. Piped water permits bathrooms and taller buildings. Sewers allow waste to be separated from streets and wells. Street lighting extends usable time. A reliable railway lets households own less transport and firms draw from a wider labour pool. Each service alters the demand placed on the others, so capacity planning is a moving target rather than a final calculation.

Networks also create dependence. A high-rise apartment relies on lifts, pumps, fire systems and power in a way a one-storey house does not. A just-in-time food system depends on roads, refrigeration and data. Redundancy costs money, so efficient systems can become brittle when every spare route or reserve has been removed. Climate pressure intensifies familiar failures by loading drainage, cooling and power networks beyond past assumptions.

The lesson is not that centralised systems are always best. Large dense settlements cannot endure without coordinated systems, whether centralised or distributed. The visible city rests on continuous collective work: promises about flow, removal, repair and response. Whether those promises are kept shapes health and freedom more directly than most architecture does.

Neighbourhoods Sort and Socialise

Neighbourhood statistics are seductive. One district has higher income, longer life, better school results and less violence than another. The map seems to prove that the place produced the people. It may have done part of the work, but the people and institutions that selected who could live there also produced the map.

Households sort by income, family needs, ethnicity, tenure, taste, information and constraint. Landlords choose tenants. Lenders choose borrowers. Schools and public housing assign addresses. Discrimination, exclusionary rules and violence narrow choices further. A wealthy neighbourhood may have good outcomes because wealthy people arrived with resources, because the place adds advantages, or because both processes reinforce each other.

The housing-voucher experiment helps separate these processes, but it also shows why a better neighbourhood is an awkward package. A move can change safety, schools and social ties together. A child may spend years encountering a different set of possibilities; an adult brings skills and obligations formed elsewhere. That makes the duration and timing of exposure part of the mechanism. It also means the experiment cannot tell us which single ingredient to bottle and distribute.

Mechanisms operate through repeated contact. Schools, peers, pollution, violence, policing, job information, public services and expectations can accumulate over years. Social ties can provide childcare, credit, protection and knowledge that no formal programme records. Moving to a district with higher average income can improve some opportunities while severing the network that made daily life possible.

Neighbourhoods can also develop collective capacities. In a study of 343 Chicago neighbourhood clusters, Robert Sampson, Stephen Raudenbush and Felton Earls linked lower violence to collective efficacy, meaning social cohesion joined to a willingness to act for the common good. The evidence came from one city and observational data, but it identifies a plausible property of place that is more than the sum of individual incomes.

Segregation emerges through several routes. Thomas Schelling's famous model showed how modest individual thresholds could produce strong aggregate separation through repeated moves. It was a demonstration of feedback, not a history of real segregation. Actual cities add restrictive covenants, lending discrimination, public investment, school boundaries, unequal wealth and coercion. A neat model of preferences becomes misleading if it erases organised power.

City life also changes the social range of ordinary experience. Dense, mixed districts can expose residents to occupations, languages, beliefs and ways of living absent from a more homogeneous settlement. They can support anonymity, allowing people to escape constant observation by kin or neighbours. The same anonymity can feel like freedom, loneliness or danger depending on resources and trust. Urban contact broadens possible identities without guaranteeing understanding.

Belonging is another accumulated effect. A market where a trader recognises a customer, a place of worship, a familiar route and a language heard in public can make a district usable in ways a list of services misses. Their value is relational: moving the same facilities elsewhere does not reproduce the network. This is why relocation can improve housing quality and still damage a life. Attachment is part of how place becomes legible, supportive and worth defending.

A movable chair makes the relationship almost comically visible. William H. Whyte's observations of Manhattan public spaces helped make the case for seating people could rearrange. A fixed bench tells you where to sit and whom to face. A chair lets you adjust the encounter. It is a small piece of authority handed back to the user, not a guarantee of sociability. People still read a space through safety, policing, custom and whether they expect to be welcomed. Good design leaves something for its inhabitants to decide.

A neighbourhood is made through repeated encounters, and changes as people invest in it or withdraw. A postcode records none of that work.

Places Remember Decisions

Cities look permanent because buildings and roads change slowly. Their uses and values are less stable, but they still carry the pressure of choices made decades earlier.

A transport line lays down an initial route. Homes and firms gather around it. Land values rise, political constituencies form and later investment follows the established corridor because replacing it would be expensive. A zoning rule limits what can be built; the resulting scarcity raises prices; owners who gain from scarcity acquire a reason to defend the rule. A school boundary influences demand for homes, and the housing market then hardens the boundary into a social division. Each step makes the next one more likely.

Economists call this path dependence when earlier events alter the range or cost of later options. It does not mean history dictates one future. It means reversing a pattern requires more than removing the original cause. A motorway driven through a neighbourhood may decentralise residents and jobs, divide local streets and change investment for generations. Taking down the road does not restore the displaced households, lost businesses or old property structure by itself.

American highway construction after 1950 illustrates the scale of the feedback. Research by Nathaniel Baum-Snow found that new radial highways contributed materially to the movement of population away from central cities. Rising incomes, cars, housing policy and racial exclusion mattered too. The road was neither sole cause nor neutral response. It changed which locations became practical, and households and firms then built a metropolitan pattern around that new accessibility.

Environmental exposure stores decisions too. A 2020 study of 108 United States urban areas compared historical lending grades with satellite measurements of summer land-surface temperature. Areas once redlined, marked as hazardous for lending, were commonly hotter than better-graded areas. The measurement concerns surfaces, not the air a resident breathes or their individual heat exposure. Nor does the association prove that a federal map caused every later difference. It shows a persistent correspondence between past exclusion and present physical conditions, with vegetation and paved surfaces among the links.

Government boundaries can deepen the memory. The housing market, road network and labour market may operate across an entire region while tax bases, schools and planning powers stop at municipal lines. Wealthy jurisdictions can protect services and restrict growth while poorer ones carry infrastructure or social costs. The metropolitan problem then persists because every authority can make a locally rational decision that leaves the shared system worse.

Memory can be useful. Street grids allow countless buildings to change while routes remain legible. Old industrial structures can support new firms and homes. Established communities hold knowledge, care networks and identity that a clean plan would destroy. The same persistence can make cities adaptable as well as stubborn: durable frameworks reduce the need to rebuild everything at once.

A city begins as a distribution of access and exposure. People react to that distribution by moving, bidding, building, voting and organising. Their reactions alter prices, institutions and networks, which reshape the distribution faced by the next residents. The place is never an independent force pressing on passive people. It is accumulated human action made costly to reverse.

To change a city, therefore, identify the feedback that keeps reproducing the pattern. A new facility without affordable access may raise nearby prices. A transport improvement without housing may widen the commuter field rather than let more people live close by. A rule changed without finance or infrastructure may remain theoretical. Durable repair often means changing several linked incentives and capacities together.

How It Actually Works

Before the skyline

Grain had to reach people who were no longer growing all their own food. A potter needed customers who could feed him; builders needed supplies while they worked. Before a city could impress anyone with its skyline, it had to organise these dependencies. Storage, exchange and administration made larger concentrations possible. Defence and ritual also drew people together, giving them reasons to remain near others they did not know.

Early Mesopotamian cities made this visible in walls, temples, workshops and written accounts. Other urban traditions developed along the Nile, in the Indus region, northern China, Mesoamerica and the Andes, among other places. Some were royal capitals, some trading centres, some ceremonial concentrations, and many combined these roles. There was no universal sequence from village to market to state.

The threshold was not a particular number of roofs. It was the reach of the relationships between them. Officials could collect grain from households they never met. Workshops could depend on distant supplies. Markets brought languages and customs into contact, while rulers and religious institutions gathered labour for walls, drains and monuments. Cities increased what people could accomplish together, and made it easier for some people to command the work of others.

Urban concentration also changed politics. Shared walls, wells, markets and roads required rules about contribution and use. Records made taxes and obligations legible to authorities. Public works could protect a population while demonstrating who commanded its labour. The city was an economic concentration and a machine for making collective power visible.

The bargain was immediate. Specialisation raised output and variety, while dependence increased. A failed harvest, blocked route, siege or political collapse could interrupt supplies for people no longer producing their own food. Waste and infection accumulated where many bodies and animals shared water and space. Inequality could be built into the plan, from elevated compounds to peripheral labour quarters. Urban life began as both expanded possibility and organised vulnerability.

The walking city

For most of urban history, ordinary daily movement was constrained by the distance a person, cart or animal could cover. This made cities compact without making them orderly. Houses, workshops, markets, shrines and storehouses mixed closely because separation cost time. Streets followed terrain, property lines and accumulated use as often as a ruler's design.

Some cities used grids, broad avenues or segregated precincts. Others grew through narrow lanes and irregular plots. Planning and apparent disorder coexisted. A formal ceremonial axis could meet an alley negotiated house by house. The street was route, drain, workplace, sales floor and social space. Front doors, windows and stalls put domestic and commercial life in contact.

The compact city made encounter frequent without making society open. Guild, caste, citizenship, gender and legal status could determine who entered a street, market or occupation. Proximity placed different groups within sight while institutions kept their rights far apart. Crowding and social mixing should never be treated as the same condition.

Walking compressed class as well as distance, but did not erase it. The wealthy could occupy safer ground, larger compounds or streets near power. The poor often lived beside workshops, walls, flood-prone land or sources of smoke and smell. Servants, enslaved people and porters supplied much of the movement that elite accounts treat as effortless.

The city was never enclosed by its wall in economic terms. Rome depended on grain fleets. Chang'an sat inside routes of tribute and exchange. Baghdad joined court, scholarship and commerce to waterways and caravan networks. Tenochtitlan drew food and goods across causeways and canals into vast markets. Political citizenship might stop at a wall while food, disease and labour crossed it. The dense centre existed because a much larger territory supplied it. Urban and rural were partners joined by unequal flows, not separate worlds.

Scale remained limited by supply and removal. Water had to be carried or channelled. Fuel and food consumed transport capacity. Fires moved rapidly through crowded timber buildings. Authorities regulated markets, street widths, building materials and waste with mixed success. The walking city could achieve immense density and sophistication, but every additional resident intensified demands on systems that relied heavily on human and animal effort.

Industry outruns the old city

Coal and steam loosened production's dependence on water power, while factories and railways broke parts of the old transport limit. Goods could move in volumes and at speeds beyond carts. Firms gathered near stations, docks and labour. Migrants arrived from rural districts and abroad because a city offered wages, anonymity and a chance to enter markets closed at home.

The nineteenth-century industrial city expanded faster than its housing and public systems. Rooms were subdivided, courtyards filled and speculative terraces erected close to mills and workshops. Smoke settled over districts that also contained homes. Rivers and drains received industrial waste, sewage and rubbish. The urban poor were blamed for conditions produced by wages, property markets and infrastructure that left them few alternatives.

Manchester became a symbol because industrial growth was unusually exposed there: mills, warehouses, canals, railways and dense worker housing assembled within a few decades. London showed the same mechanisms at a greater scale and with more layers of wealth. Chicago connected grain, timber, meat and finance through rail and lake transport, then rebuilt repeatedly after fire. The cases differed, but all demonstrated that transport and production could enlarge a city faster than inherited institutions could make it safe.

Cholera revealed the cost. The disease spread through contaminated water, although nineteenth-century officials debated its cause through the language of miasma, filth and contagion. John Snow's investigation around the Broad Street pump in 1854 helped show that cases clustered around a water source, but one map and one removed handle did not create modern sanitation. The outbreak was already declining, and the decisive urban change required filtration, sewers, clean supply, laboratory knowledge, taxation and public administration built over decades.

Housing became part of labour discipline. A worker living near a mill could be summoned early and replaced from a large local pool. Employers sometimes built housing, charities built model dwellings and municipalities cleared districts, each attaching different rules to shelter. Rent absorbed part of the urban wage, while eviction made the promise of city opportunity insecure.

Industrialisation also changed time. Factory bells and railway timetables coordinated thousands of people who no longer worked by local daylight or task. Commuter services separated home from work for some households. Cheap urban newspapers, music halls, department stores and political meetings gathered large audiences. A person could join an occupation, movement or subculture unavailable in a village, while living among strangers whose names they never learned.

The city became a mass social environment before governments learned to measure it. Censuses, sanitary maps, mortality registers and housing surveys turned streets into administrative evidence. Reformers exposed overcrowding and disease; landlords and employers resisted costs; officials demolished some districts and improved others. The same information that supported services also enabled surveillance, classification and clearance.

Making density governable

One response, especially in industrialising Europe and North America, was a stronger urban state with pipes, inspectors, engineers, public transport and powers over land. Its growth was uneven and contested. Cities borrowed heavily to build waterworks and sewers. They paved roads, collected refuse, created fire services and imposed building rules. These systems raised land values and public health while shifting costs through taxes, charges and compulsory purchase.

Large projects carried political visions. Baron Haussmann's rebuilding of Paris under Napoleon III produced boulevards, sewers, parks and a more connected capital. It also demolished districts, displaced residents and increased the state's ability to move traffic, officials and troops through the capital. Ildefons Cerdà's 1859 extension plan for Barcelona used a grid of broad streets and chamfered blocks intended to improve light, ventilation and circulation. Later development filled many courtyards and increased densities beyond the original scheme. Plans entered the city, then owners, residents and markets edited them.

Public transport loosened the walking limit. Horse trams, suburban railways, electric streetcars and underground lines let people live farther from work. The change created new land values around stops and allowed cities to expand along corridors. It also separated populations. Middle-income households could seek larger homes beyond industrial centres, while workers with lower wages remained near jobs or paid a greater share of income for travel.

Zoning and planning emerged partly to control conflict among adjacent uses. A tannery, apartment and school impose different costs on neighbours. Rules could keep dangerous industry away from homes, protect light and organise infrastructure. They could also exclude cheaper housing, racial groups or businesses disliked by existing owners. The power to separate nuisance from residence became a power to decide which people and activities belonged.

Sanitary reform could protect and coerce in the same operation. Inspectors entered homes, condemned buildings and classified districts through standards written by middle-class authorities. Some rules removed lethal conditions; some criminalised survival practices without providing an affordable alternative. Capacity to intervene was necessary, but the distribution of voice determined how intervention was experienced.

Parks and civic buildings offered another answer to crowding. They provided air, recreation, spectacle and places where classes encountered one another under rules set by authorities. Their creation sometimes required clearing occupied land. Public space was never merely empty ground. Seating, policing, opening hours, vendors and acceptable behaviour determined who could use it and for what.

By the early twentieth century, many wealthy cities had turned density into a managed system. Clean water, sewers, electricity, lifts, mass transit and steel construction allowed populations to concentrate without the same mortality cost. The achievement encouraged confidence that experts could design social order from streets and buildings. The next phase would test that confidence at metropolitan scale.

The metropolitan machine

The motor car changed access by offering flexible movement across a widening region. Governments laid roads, required parking and redesigned streets for speed. Shops, factories and offices followed cheaper land. Housing spread beyond old transit corridors. The city became harder to recognise from its centre because daily life was now organised across municipalities, ring roads and business parks.

Suburbanisation was not one voluntary flight from density. Rising incomes and demand for space mattered. So did subsidised roads, mortgage systems, school boundaries, public housing policy, racial violence and exclusion. In the United States, highways enabled decentralisation while often cutting through Black and low-income neighbourhoods whose political resistance carried less weight. Urban renewal cleared districts labelled blighted and replaced some with offices, public projects or roads.

Research comparing planned and built highway routes indicates that postwar interstate construction contributed to central-city population loss. It did not create every suburb or explain every household move. It changed the feasible geography: a distant home became reachable by car, a central block became noisy or divided, and firms gained new sites near interchanges. Once homes, warehouses and retail were built around that system, driving shifted from an option towards a requirement.

Parking made the transformation less visible. A vehicle spends most of its time stored, so homes, offices and shops devoted valuable land to spaces that appeared free to the driver. Minimum parking rules bundled the cost into rents, prices and larger distances between destinations. Roads moved cars; the wider land system made car use necessary.

Other cities expanded through rail rather than motorway dominance. Greater Tokyo's dense network of public and private railways joined multiple centres, with development and retail clustered around stations. This did not eliminate long commutes, crowding or housing pressure. It produced a different metropolitan machine in which large numbers could reach dispersed centres without each trip requiring road and parking space.

Daily time became one of the system's main inequalities. A household priced out of a job-rich district might remain attached to its economy through a two-hour journey, paying in sleep, care time and reliability rather than rent. Metropolitan growth could enlarge the number of jobs in theory while placing the best matches beyond a person's feasible schedule.

The post-industrial transition rearranged the map again. Manufacturing employment declined in many older centres, leaving vacant land, polluted sites and workers whose skills and homes were tied to vanished industries. Finance, education, medicine, technology, tourism and culture concentrated in selected districts. Some central neighbourhoods became desirable after decades of disinvestment. Rising demand restored buildings and tax bases while producing new exclusion and conflict over who had earned a right to remain.

Metropolitan government rarely matched the system it was meant to govern. Jobs, housing and pollution crossed boundaries, while tax revenue and planning powers remained local. One municipality could restrict homes while benefiting from a regional labour market. Another could host lower-cost housing without the same commercial tax base. Local decisions accumulated into regional congestion, segregation and shortage without any one authority choosing the final pattern.

The self-built city

During the late twentieth and early twenty-first centuries, much urban growth occurred where formal housing, land records and infrastructure could not keep pace with population and demand. Residents obtained plots through purchase, occupation, customary claims or political brokerage, then built incrementally as money and materials became available. Streets, shops, workshops, religious buildings and transport services often appeared before official plans or pipes.

The label informal can hide more than it reveals. Some settlements are tolerated but technically unauthorised. Others have recognised land but unapproved buildings, rental markets outside regulation or services supplied through unofficial connections. They range from precarious shacks to multi-storey neighbourhoods with mature businesses. What joins them is incomplete access to formal rights and systems, not one physical form.

By 2022, United Nations estimates already put more than a billion people in slums or informal settlements. The scale marks a vast housing and service failure, but residents are not evidence of an absence. They plan routes, pool money and negotiate connections because waiting for complete formal provision is not an available strategy. A settlement can be short of pipes without being short of organisation.

Transport and commerce are often self-organised too. Minibuses, motorcycle taxis, street vending and home-based workshops connect demand that formal networks overlook. These systems can be flexible and indispensable while exposing workers to unstable income, unsafe conditions and arbitrary enforcement. Their existence is evidence of unmet demand, not proof that public provision is unnecessary.

Improvement can take several paths. Clearance may remove a hazard while destroying livelihoods and social networks if residents are relocated far from work. Secure tenure can encourage investment, but a title alone does not provide drainage, transport or affordable credit and may expose land to market pressure. In-situ upgrading can add water, sanitation, streets and legal protection while preserving location, yet it requires space, finance and institutions willing to work with residents rather than treating them as temporary.

In Karachi, the Orangi Pilot Project's sanitation programme turned that organisation into a division of work. Its approach asked residents to finance and manage household and lane-level sanitation, with technical support, while government remained responsible for larger sewers and treatment. The distinction matters: a lane can organise a pipe, but that pipe needs somewhere safe to empty. The programme's design was cooperation across scales, not a demonstration that residents could replace the state. A household's effort and the city's investment had to meet underground. That is the self-built city's predicament in miniature: people can make much of their neighbourhood, but they cannot make every wider system it needs.

The city after distance changed

Telephones, the internet and remote work removed the need for some exchanges to occur in the same room. Predictions that distance had died followed each change. Instead, cities rearranged. Routine information travelled easily, while specialised labour markets, universities, laboratories, care, entertainment, logistics and face-to-face trust continued to cluster. Digital services relied on physical warehouses, cables, power and delivery workers occupying particular land.

The Covid-19 pandemic showed that many office tasks could move home and also exposed who could not leave place behind. Nurses, drivers, cleaners, builders and shop workers still travelled. Homes became workplaces with unequal space, quiet and connectivity. Central business districts lost some daily demand; neighbourhood centres gained some. The long-term balance differs by occupation and city, and no single post-pandemic model has replaced the old one.

Climate pressure now tests systems laid out for previous conditions. Heat follows building materials, shade, ventilation and access to cooling. Flood damage follows topography, drainage, land cover and who can afford protection or recovery. Retrofitting a mature city is harder than drawing a new one because streets, rights, debts and populations already occupy the ground.

Ageing changes the same map. A district usable by a healthy adult may become isolating when pavements, crossings, benches and nearby services fail an older resident. Climate adaptation, disability access and care are linked by one practical question: can bodies with different limits still reach what daily life requires?

Future urban growth will therefore involve extension and repair at once. New districts need land reserved for streets, services and public use before every plot is fixed. Existing districts need more homes, cleaner movement, shade, drainage and maintained networks without treating current residents as removable obstacles. Technology can measure flows and coordinate service, but it cannot decide whose time, privacy, home or risk should count most.

Digital connection changes the value of proximity. Bodies, land and infrastructure remain stubbornly located.

How we know

Cities resist clean experiments. People choose where to live, firms choose where to invest and governments place infrastructure where growth is already expected. A correlation between a neighbourhood and an outcome can therefore mix selection, treatment and measurement.

Researchers use several imperfect tools. Random assignment, as in Moving to Opportunity, can identify effects for participating families and offered moves, but cannot represent every population or destination. Natural experiments compare places changed by routes, boundaries or policies for reasons partly outside local trends, though the identifying assumptions remain contestable. Longitudinal records follow people across moves. Surveys reveal experience that administrative data miss. Street observation, sensors, mobile traces and satellite images show behaviour and built form while raising questions about privacy and who is invisible to the system.

Definitions remain a basic hazard. Municipalities, built-up areas and labour markets produce different populations and densities. Results also travel badly when law, transport, tenure and family structure differ. The strongest urban claim therefore names its place, period, population and mechanism. A city can teach another city. It cannot provide a universal template.

What People Get Wrong

“A city is whatever lies inside the boundary”

The boundary is persuasive because governments publish data, elect mayors and collect taxes within it. The line looks official, so the object looks real.

Urban life is rarely so obedient. A continuous built-up area can cross many municipalities, while an official city can include large rural tracts. Commuters, freight, housing searches, pollution and water move across legal lines. Counting only the centre can make a growing metropolitan region look as though it is shrinking. Comparing two municipal densities can compare an old core with an entire county-sized authority.

Boundary choice also changes accountability. A prosperous suburb can depend on jobs and infrastructure in the core while keeping its tax base and restricting new housing. A central authority can bear costs generated by a wider region it does not control.

The correction is to match the boundary to the question. Use a municipality for legal power and finance, a built-up area for physical development, and a functional urban area for connected labour and housing markets. Before comparing cities, ask which relationships the chosen line includes and which it cuts.

“Density causes urban misery”

Crowded tenements, traffic, disease and high-rise isolation made density an easy villain. The images are real, but they combine several different conditions: many people per unit of land, too little floor space per person, weak infrastructure, poverty and exclusion.

Density can support frequent transport, nearby services and efficient networks. It can also magnify infection, noise and heat when water, ventilation, drainage and public space fail. Low-density regions can contain long commutes, severe pollution, unaffordable housing and social isolation. The relevant question is what kind of density, connected by which systems and occupied under what standards.

A well-serviced apartment district with small households can be dense without crowding. A low-rise peripheral settlement can be overcrowded inside each home while appearing spacious from a satellite. The denominator changes the diagnosis.

Confusing density with overcrowding produces bad policy. Clearing dense districts or limiting homes may reduce the number of residents without reducing demand, pushing people farther out or into tighter accommodation elsewhere. Density is a capacity that can be managed well or badly. Misery comes from the relation among people, space, income, rights and infrastructure, not from a population count divided by hectares.

“More roads cure congestion”

The belief follows ordinary experience. Add a lane to a queue and there is more room. Immediately after opening, traffic may move faster, which appears to settle the case.

Travel demand is not fixed. A faster route attracts drivers from other roads and times, encourages trips previously avoided, changes destinations and makes road-dependent development more attractive. In a major study of United States metropolitan highways, vehicle travel rose roughly with interstate lane capacity over the period examined. That result does not mean every safety improvement or missing link is pointless. It warns against treating today's traffic as all the traffic an improved road will ever carry.

The adjustment can take years, which is why a project may appear successful at opening and disappointing later. It also varies with growth, network structure and available alternatives, so the study of one road is not a universal forecast.

Roads provide access, freight movement and redundancy. Congestion is the signal that scarce space at a busy time is being rationed by delay. Durable relief requires some combination of pricing, alternatives, changed land use and realistic acceptance that popular routes will fill. The goal is useful access, not a permanently empty road.

“Neighbourhoods determine destiny”

Maps of life expectancy, income or school attainment can make postcode look like fate. Advocates sometimes sharpen the claim because visible inequality demands action. Critics then answer that people select where they live, so the map proves nothing about place.

Moving to Opportunity showed that offering access to lower-poverty areas could improve younger children's later outcomes, while older children and adults showed different patterns. It refutes destiny without reducing place to scenery. The timing and circumstances of a move matter.

Even a measured effect rarely identifies one ingredient. Lower poverty may arrive with different schools, safety, peers, housing quality and public services. Researchers can estimate a package more readily than they can separate every mechanism inside it.

A neighbourhood is neither a spell nor scenery. It changes repeated exposure to schools, peers, violence, pollution, institutions and information. Families bring resources and networks with them, and moving can destroy support as well as open opportunity. Policy should improve places and expand genuine mobility without pretending that relocation is a universal treatment or that residents must leave for their lives to improve.

“Gentrification is either rescue or removal”

One story describes investment returning to neglected districts: safer streets, repaired buildings, new shops and a stronger tax base. The other describes wealthier arrivals displacing poorer residents and erasing a community. Each side can produce photographs and testimony because both processes occur.

Research has struggled to compress them into one displacement rate. Results differ by city, period, tenure and definition. Some studies find modest increases in out-migration; others find that vulnerable residents are no more likely to move than comparable residents elsewhere. Change can happen through who moves in, who cannot enter later, rising rent burdens, cultural displacement and the loss of businesses or institutions, even when most original residents stay. Owners and protected tenants may gain while insecure renters face pressure.

Administrative data can miss forced moves between temporary addresses, doubling up with relatives, business closure and the feeling of becoming unwelcome in one's own public spaces. Testimony can reveal these losses while remaining unable to measure their prevalence. Both forms of evidence matter.

The useful question is disaggregated. Who remains, who leaves, where do movers go, who is excluded, and who captures the rising land value? Improvement without protection can remove its intended beneficiaries. Protection without enough housing can freeze scarcity. The urban task is to permit better places while widening the right to remain and to enter.

“Informal settlements are cities that failed”

From an official map, an unplanned district can look like absence: no approved plots, no complete street plan, no formal utility network. The language of slums then turns millions of households into a single defect.

Informality usually contains extensive order. Residents buy, rent, subdivide, build, lend, trade and organise services through rules that may be customary, political or partly recognised. The settlement often exists because formal land and housing are inaccessible, slow or too expensive near work. Its narrow lanes and incomplete drainage can create severe risk, while its location and social networks support livelihoods that distant replacement housing would destroy.

Recognition can itself change value. Once tenure is secured and services arrive, outside buyers may see an opportunity that residents could not realise under insecurity. Upgrading therefore needs safeguards against the improvement becoming a route to later removal.

Calling the place a failure hides who failed to provide affordable formal access. Romanticising self-help hides fire, eviction, disease and exploitation. The practical response cannot be inferred from the label. Where location and social networks are valuable and hazards can be reduced, negotiated upgrading may outperform distant relocation; where danger cannot be made tolerable, moving people safely may be necessary. In either case, secure enough tenure, service connections, safer routes and room for residents to shape the work matter. Informality is constrained city production, not a population waiting for urban life to begin.

“A clever plan can fix the city”

Plans are seductive because a city seen from above resembles an object that can be rearranged. Straight roads, separated uses, towers in open space or a digital control room promise to remove conflict through design.

A plan changes incentives and possibilities; it does not suspend them. Landowners respond, residents adapt, budgets shrink, maintenance slips and uses appear that the drawing did not anticipate. Cerdà's open Barcelona blocks were built more densely. Motorways intended to improve movement encouraged dispersal and traffic. Public spaces designed for everyone can exclude through policing, price or discomfort.

Local knowledge is no guarantee of justice either. Incumbents can oppose newcomers, informal leaders can extract payments and consultation can be captured by the people with time to attend. Expertise and participation each correct failures in the other.

The alternative is not to stop planning. Streets, sewers, rights of way and flood protection require decisions before private building closes the options. The correction is to treat plans as rules and capacities inside a changing system. Test who gains, who pays, how behaviour may respond and who can revise the scheme. A city cannot be finished. It can be given better feedback and fairer room to change.

Use It

Measure access in minutes

When judging a home, office, school or public project, stop at neither distance nor speed. Count suitable destinations reachable within a realistic time, at the hours they are needed, by the people who must make the trip.

A station two minutes away has little value if trains are infrequent, unaffordable or inaccessible. A job ten kilometres away may be reachable by one worker and impossible for another collecting a child before nursery closes. Add waiting, transfers, reliability, walking conditions and safety. Then ask what happens when one link fails.

This changes how transport success is seen. A faster road can improve one journey while reducing access for people crossing it on foot. A slower street can increase access if shops, homes and services are brought close enough together. Compare peak and off-peak conditions, and repeat the exercise for a child, an older person and someone without a car. One network can contain several different cities.

The useful urban unit is not kilometres travelled. It is worthwhile opportunities made feasible without consuming the day.

Follow the price of access

Whenever a place improves, look for where the gain goes. A new railway, park, school or safer street may appear in higher land values before it appears in household income.

Separate owners, renters, businesses and future entrants. An owner can gain from appreciation while a renter faces a higher renewal. A protected tenant may stay while a new low-income household is excluded. A shop can receive more customers and still lose its lease. The same project can help current users, enrich landowners and narrow later access.

Then inspect the mechanism for sharing the gain. Can more homes or premises be added? Is part of the land value returned through tax, public ownership or development obligations? Are existing residents protected without closing the district to newcomers? Who controls any extra tax revenue, and where does the next investment go?

Urban improvement becomes distributive policy the moment it changes what people will pay to be nearby.

Separate selection from place effect

Imagine that a district's average income rises after regeneration. In one version, existing residents get better jobs. In another, they leave and richer residents arrive. The map brightens by the same amount; the human stories are different.

Follow people as well as places. Who could enter, who stayed, who left and where did they go? Compare residents' circumstances before and after change, not just the average address. A new school or railway may help, but it may also attract households already better equipped to succeed. The useful test is what happened to similar people without the same change.

This is not an excuse to dismiss neighbourhood effects. It is a way to ask what an improvement accomplished. A child gaining a safer route to school is a benefit even before it appears in an earnings table. A displaced household does not vanish from the assessment because it vanished from the district. Good evidence keeps the people in view when the boundary stops containing them.

An improving statistic is a question to investigate, not yet a verdict on an improving life.

Trace the network behind the surface

A functioning street conceals flows. Water arrived, waste left, power remained stable, deliveries found space, data moved and somebody maintained the pavement. The visible place is the final interface of several systems.

When a district appears dirty, congested or unreliable, follow each network before blaming residents. Who owns it? Where does its authority stop? How is maintenance funded? Which users are connected formally, and which buy substitutes at a higher price? What single failure would stop several services at once?

This lens also exposes false economy. Deferred pipe repair, an unreliable bus or a missing pavement may save a budget line while charging households in lost time, damaged goods, illness or private alternatives. Infrastructure decisions distribute daily effort, including the quiet failures that never make the news.

The person carrying water, waiting for a bus or climbing broken stairs is completing work the urban system declined to do.

Ask who absorbs the exposure

Urban benefits and burdens rarely fall on the same map. A motorway serves regional travellers and exposes nearby homes to noise and pollution. A waste site serves the whole city and occupies one district. A flood barrier protects one bank and can alter water elsewhere.

List the beneficiaries, then list those who cannot opt out of the cost. Pay attention to renters, children, older people, informal workers and residents whose political authority stops at a boundary crossed freely by traffic or pollution. Average exposure can improve while the worst burden remains concentrated.

The next question is whether compensation changes the harm or merely prices it. Money cannot replace health, identity or a functioning local network in every case. Yet refusing to quantify costs can make them disappear from appraisal. Exposure also accumulates: repeated noise or heat belongs in the account alongside rare disasters, with attention to its intensity, duration and effects rather than its ability to make headlines.

Use numbers where they clarify scale, and keep distribution visible where an average would hide it.

Look for stored decisions

When an urban pattern appears natural, search for the decision it stores. A neighbourhood boundary may follow an old railway, lending rule or school catchment. A treeless district may reflect decades of disinvestment. A retail strip may exist because a tram once stopped there, even after the tracks disappeared.

Then identify the feedback preserving it. Do high values give owners an incentive to block construction? Does unreliable service push people into cars, weakening political support for the service? Does a poor reputation reduce investment, producing the neglect used to justify the reputation?

Changing the original rule may be necessary and insufficient. The accumulated prices, debts, networks and expectations can continue the pattern. Effective repair acts on several links: legal permission, finance, infrastructure, operating capacity and protection for people exposed during transition. Timetables and school choices can preserve a geography after its physical cause has weakened.

History matters most where it has become an incentive in the present.

The limits

Cities shape people through opportunities, costs and repeated exposure. They do not explain a person completely. Family, class, health, law, culture, discrimination, technology and national institutions enter every urban outcome. A street layout cannot compensate for poverty wages, and a housing voucher cannot supply every relationship lost in a move.

Evidence is uneven. The cleanest causal studies often examine narrow programmes in wealthy countries because records and policy variation are available there. Informal tenure, extended households, gendered mobility and weak municipal capacity can change how an intervention works. A scheme that succeeds where rights are enforceable may fail where staying in the home itself is uncertain.

Planning also has hard trade-offs. More housing can reduce scarcity while changing a district. Faster movement can enlarge access while increasing noise or emissions. Strong local participation can protect knowledge and incumbents while excluding future residents who are absent from the meeting. There is no urban design that removes politics because land and access cannot be allocated without winners, costs and limits.

The model in this book is therefore a diagnostic, not a recipe. It tells you where to look: access, exposure, price, networks, selection and feedback. It cannot decide what distribution is fair or which loss a community should accept.

The one thing to keep

Keep the address, but stop treating it as a pin.

An address looks like a point on a map. In lived terms it includes the job reachable before a shift starts, the person available in an emergency, the rent, the noise and the right to remain. Two homes with identical rooms can offer radically different lives.

That package was made, and people keep remaking it. A railway changes the value of a home. A rent rise changes who can stay. A neighbour who knows when to knock makes living alone a different proposition. None of these appears in the floor plan, but each belongs to the place. The city is working on a life long before anyone calls the result an urban outcome.

So a person late for work may need more than a lecture about punctuality; a city celebrated for its prosperity may still make an ordinary wage impossible to live on. Neither observation abolishes personal responsibility. It puts responsibility back among the conditions in which people have to exercise it. We can ask better questions once the buildings stop taking all the credit, and the residents stop taking all the blame.

Places do not write destinies. They edit the menu of feasible choices, every day, often before choice becomes visible. The next time you look down a street, look past what has been built to the lives it makes possible, and the lives it keeps out.

Terms

City. A settlement concentrating people, activities, institutions and connections beyond household or village scale. No single population threshold works everywhere because administrative boundaries, density and economic reach differ.

Urbanisation. An increase in the share of a population living in urban settlements, accompanied by changes in settlement and social life. Urban population can grow in number without that share rising.

Administrative city. The territory governed by a municipal or equivalent authority. It matters for law, tax and service responsibility, but may include rural land or exclude suburbs that form part of daily urban life.

Built-up area. Land covered by buildings, streets and other constructed surfaces forming a physically continuous settlement. It ignores municipal borders but may join places whose governments or labour markets remain separate.

Functional urban area. A dense urban centre plus its surrounding commuting zone. It often describes the lived labour and housing market better than a legal boundary, though the selected flow threshold matters.

Metropolitan area. A large city and surrounding settlements integrated through commuting or other economic and social ties. Definitions vary, so population figures require checking the geography used.

Density. The amount of population, housing, employment or floor space within a stated area. Averages can hide sharp local variation, while a changed boundary can alter the result without moving anyone.

Crowding. Insufficient private space, privacy or facilities relative to the number of occupants, usually measured within a dwelling. It is a housing condition rather than a synonym for neighbourhood density.

Agglomeration economy. A benefit created when firms and workers locate near one another, allowing better matching, shared suppliers and faster learning. Congestion, high land costs and pollution are corresponding agglomeration costs.

Externality. A cost or benefit imposed on people outside the decision producing it. Traffic delay, noise, cleaner air and street activity can cross property lines without complete payment.

Accessibility. The useful destinations a person can reach within constraints of time, money, reliability, safety and eligibility. It measures opportunities rather than movement for its own sake.

Land rent. The price paid for control of a location, conceptually separated from its building. Access, permission to build, amenities, scarcity and expectations all enter urban land value.

Bid-rent. The pattern created when households and firms compete for locations according to what proximity is worth to them. High bids near valued access can displace lower-paying uses.

Zoning. Rules permitting, restricting or separating land uses and building forms within mapped districts. Zoning can protect residents, provide predictability or preserve scarcity, depending on design and enforcement.

Floor-area ratio. Total permitted or built floor area divided by plot area. A ratio of four could mean four full storeys or a taller building covering less ground.

Mixed use. Several activities, such as homes, shops, schools and workplaces, within one building or walkable area. Mixture can shorten trips and spread activity across the day.

Sprawl. Low-density outward growth in which destinations are widely separated and travel often depends on private vehicles. The label should not substitute for measuring access, land use or emissions.

Induced demand. Additional travel appearing after road capacity makes driving easier or cheaper. People change routes, times, destinations, modes and locations, so early congestion relief may shrink.

Infrastructure. Durable networks and facilities providing streets, water, sewers, power, transport, communications and waste services. Their value depends on maintenance, connection and operating institutions as well as construction.

Network effect. A change in value caused by the number or arrangement of connected users and nodes. More connections can enlarge usefulness, while disruption can spread further through the same system.

Public space. Streets, pavements, squares, parks and other collectively used places, whether publicly or privately owned. Design, management, policing, custom and price affect who can remain there.

Social infrastructure. Places and organisations supporting repeated contact, care and collective life, including libraries, playgrounds, cafés, schools and religious buildings. Their social effect depends on use and access.

Neighbourhood effect. A causal influence of local conditions beyond characteristics residents brought with them. Establishing one requires separating exposure from residential selection, family circumstances and wider city effects.

Residential sorting. The process through which households choose or are channelled into locations according to income, preferences, discrimination, tenure, schools, transport and available housing. Sorting can imitate a place effect.

Segregation. Spatial separation of social groups across homes, schools, work or public life. It may arise through law, markets, discrimination, preferences and feedback, and persist after explicit rules disappear.

Collective efficacy. Shared expectations that neighbours can act together and intervene for common purposes, especially informal social control. It is a group property, not a synonym for friendliness.

Gentrification. Socioeconomic and physical change in a lower-income neighbourhood as investment and higher-income demand rise. It can bring services while increasing rent pressure, exclusion and displacement.

Displacement. Involuntary movement or exclusion caused by demolition, eviction, rising costs, harassment or loss of suitable housing. It includes households prevented from entering an area they once could afford.

Informal settlement. Housing or neighbourhood development lacking some recognised tenure, planning permission, infrastructure or building compliance. Informality is an institutional relationship, not proof that residents created no order.

Path dependence. Earlier choices shaping later options because networks, property, habits, expectations and sunk costs accumulate around them. A pattern can persist without being efficient, inevitable or impossible to change.

Go Deeper

The accessible economic case

Edward Glaeser, Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier (Penguin Press, 2011). Glaeser makes the clearest popular case for dense urban life as a machine for exchange, learning and lower resource use. His examples move quickly from New York and Boston to Bangalore, Lagos and ancient cities, and his explanation of why poor people move to cities is especially useful. Read it for agglomeration, housing supply and the difference between helping people and preserving buildings. The optimism is deliberate, and the title overstates what any city guarantees. Keep testing who captures the gains and which institutions make density tolerable.

The classic street-level challenge

Jane Jacobs, The Death and Life of Great American Cities (Random House, 1961). Jacobs attacks the habit of treating the city as a diagram to be simplified. Her streets contain shops, children, strangers, old buildings, informal supervision and uses changing by the hour. The book explains why a district can look untidy from above yet work coherently for people inside it. It is also a document of mid-century North American cities, strongest on dense mixed neighbourhoods and less reliable as a universal design code. Read it for observation and argument, then resist turning her preferred blocks into another master plan.

The evidence on neighbourhoods

Robert J. Sampson, Great American City: Chicago and the Enduring Neighborhood Effect (University of Chicago Press, 2012). Sampson follows Chicago across surveys, administrative records, observations and history to ask why neighbourhood differences persist even as residents move. Collective efficacy, concentrated disadvantage, organisations and reputations become measurable parts of place rather than atmosphere. The book is long and methodologically demanding, but it is the best continuation of this book's warning about selection and causation. Chicago is the case, not a stand-in for every city. Its reward is a model of how to take neighbourhoods seriously without treating them as fate.

The corrective beyond formal planning

AbdouMaliq Simone, For the City Yet to Come: Changing African Life in Four Cities (Duke University Press, 2004). Through Winterveld near Pretoria, Douala, Pikine near Dakar and Jeddah, Simone examines urban life sustained by provisional alliances, informal work, movement and social improvisation where official systems are partial or unreliable. The prose is theoretical and sometimes difficult, but the cases make a necessary correction to accounts built from formal plans, property records and municipal services. Read it to see people themselves operating as connective infrastructure. It does not romanticise insecurity; it asks how urban residents create room to act when stable institutions cannot be assumed.

Notes and Sources

The book uses city as an analytical object rather than a universal legal category. Urban definitions, causal evidence and institutional arrangements vary sharply, so claims below are bounded by place and period where that changes their meaning. Current official material was checked on 5 September 2026.

The opening argument

Access and exposure. The framing draws on the long urban and transport tradition of treating accessibility as the opportunities reachable from a location rather than movement alone. Hansen's 1959 paper supplied an early operational statement. Later accessibility research adds travel time, land use and the characteristics of users. This book extends the frame to unavoidable exposure because access to urban benefits and exposure to urban costs are produced through many of the same networks and land decisions.

Moving to Opportunity. HUD's programme recruited more than 4,600 volunteer families in five United States cities between 1994 and 1998. Randomisation allocated offers: a voucher restricted initially to a low-poverty area with counselling, a standard voucher, or no additional MTO assistance. Existing assistance continued for controls, who could also move. Comparing assigned groups avoids treating households that accepted an offer as a random sample of movers. The 2011 evaluation found improvements in safety and several adult health outcomes without significant adult earnings gains. Chetty, Hendren and Katz's 2016 article found gains in later college attendance and earnings for children under thirteen at random assignment to the low-poverty offer. Older children's estimates differed; disruption is a proposed explanation, not an isolated experimental finding. Age thirteen is an analytical cut-off, not a demonstrated developmental threshold. These results concern a bundled intervention in the participating families, not every move or neighbourhood.

What is urban. World Urbanization Prospects 2025 reports national urban definitions and a harmonised spatial classification based on the Degree of Urbanization. The harmonised method distinguishes cities, towns and semi-dense areas, and rural areas using population concentration and grid cells. It improves comparison but does not abolish judgement about commuting regions, seasonal populations or administrative power. The book therefore avoids a single global urban share and uses the definitional change itself as the important fact.

The Core Ideas

Boundaries, scale and density. The distinctions among administrative city, built-up area, functional urban area and metropolitan region follow United Nations and European Commission methodological work. Population and density comparisons are sensitive to boundary choice. A municipality is still the correct unit for some questions, including taxation, elections and legal authority; the warning is against treating it as the natural unit for every urban process.

Agglomeration. Duranton and Puga's review organises agglomeration mechanisms as sharing, matching and learning. These are families of models and evidence rather than three effects that appear in every city. The Spanish worker evidence comes from De la Roca and Puga, who use administrative employment histories to distinguish sorting, a current-location premium and learning accumulated in larger cities. Their result is retained with its national setting visible. The book does not use a single percentage urban wage premium because estimates change with skill, industry, housing cost, selection and spatial unit.

Land and housing. Alonso's bid-rent model supplies the basic location trade between accessibility and land cost. Duranton and Puga review how multiple centres, durable structures, transport and development alter the simple monocentric picture. Gyourko and Molloy review evidence that land-use regulation affects the amount, location and form of housing supply, while also stressing measurement and political endogeneity. The text therefore says restrictive rules can push demand into price rather than assigning every high price to regulation.

Publicly created land value. The discussion of stations, parks, development obligations and land taxation is an institutional mechanism rather than a claim that one instrument works identically everywhere. Who captures an increase depends on ownership, tenancy, tax design, development response and expectations formed before construction. No universal incidence estimate is implied.

Movement and accessibility. Hansen provides the early accessibility concept. The practical definition used here also reflects later transport appraisal that separates mobility from access and requires attention to land use, time, cost and user characteristics. The carer making linked trips is illustrative rather than a reported individual case.

Induced traffic. Duranton and Turner estimate that vehicle kilometres travelled on interstate highways in United States metropolitan areas rose approximately one for one with interstate lane kilometres over their study period. They identify more driving by existing residents, commercial traffic and population change as contributing channels. The book does not claim that every road project produces the same elasticity, or that safety works, network completion and freight access have no value. It retains the narrower inference that road capacity changes behaviour and land use, so demand cannot be held fixed in appraisal.

Infrastructure and mortality. Cutler and Miller's 2005 study of United States cities attributed a large share of mortality decline to clean-water technologies. Anderson, Charles and Rees's 2022 analysis challenged the size of that contribution. The accompanying Comment and Reply locate disagreements in data transcription, intervention-year coding and population denominators, not merely in abstract model choice. Cutler and Miller corrected errors but continued to find a substantial contribution to total mortality decline; their infant-mortality estimate fell. The body does not adopt a percentage or treat either analysis as the last word. Pathogen removal and the share of a historical mortality decline attributable to particular interventions are different claims. Melosi supplies the longer institutional history of water, sewerage and waste services.

Neighbourhood exposure. The MTO evidence establishes causal effects for the families, treatments, destinations and periods studied. It does not isolate one ingredient called neighbourhood quality. Schools, safety, peers, environmental conditions and social networks moved together. Observational neighbourhood research faces selection because residents arrive with different resources and constraints.

Belonging and attachment. The discussion of recognition, familiar routes, language, worship and local networks is an interpretive synthesis rather than a claim that one measurable form of belonging applies across cultures. It corrects the access model where a list of destinations would miss relational value. The text does not claim that attachment makes harmful conditions acceptable or that every resident experiences the same district as home.

Collective efficacy. Sampson, Raudenbush and Earls surveyed 8,782 residents across 343 Chicago neighbourhood clusters in 1995. Their multilevel analysis linked collective efficacy, defined as social cohesion combined with willingness to intervene for the common good, to lower violence after accounting for several individual and neighbourhood factors. The study is observational and Chicago-specific. It supports a plausible group-level mechanism, not a universal intervention effect.

Segregation. Schelling's 1971 models show how repeated local choices can generate aggregate separation more extreme than individual preferences suggest. Schelling explicitly discussed organisational practices and correlated constraints beside individual choices. The manuscript uses the model to explain feedback while drawing the history of actual United States segregation from work on law, covenants, lending, public investment, violence and wealth. The model is not offered as a sufficient causal history.

Public space. Whyte's Street Life Project used observation and time-lapse film, largely in Manhattan, to investigate public-space use. His argument for movable seating is also described by Andrew M. Manshel, a former Bryant Park Restoration Corporation officer, in a 2009 essay republished by the park in 2017. The chair illustrates user control; it is not presented as a guaranteed causal intervention in sociability or safety. Jacobs supplies the wider argument for street-level observation. Neither author's favoured spaces become universal design laws.

Path dependence and highways. Baum-Snow compares planned highway routes with later construction to estimate the contribution of United States interstate highways to central-city population decline from 1950 to 1990. The design aims to address the fact that roads are often built where decentralisation is already expected. The result remains one component of a broader history involving income, automobiles, mortgages, public housing and racial exclusion. The manuscript therefore states that highways contributed materially rather than caused suburbanisation alone.

Redlining and urban heat. Hoffman, Shandas and Pendleton's 2020 study compares summer land-surface temperatures with historical Home Owners' Loan Corporation grades in 108 United States urban areas. It concerns satellite-observed surfaces, not a measured air-temperature difference or an individual's heat dose. The association is consistent with persistent inequalities in vegetation, paved surfaces and investment, but does not identify the maps as the sole cause. The book keeps the study date and measured variable visible rather than describing its observations as a new measurement in 2026.

Operating history

The first cities. Monica L. Smith provides the broad archaeological synthesis behind the account of multiple urban traditions, specialisation, public space, infrastructure and durable dependence among strangers. Early cities differed in state form, ritual, trade and ecology, so the manuscript rejects a single ladder from village to market to city. Mesopotamia appears only as one well-documented example because it has its own series title.

Informal-settlement data vintage. The UN's 2024 Goal 11 account identifies 1.12 billion people in slums or informal settlements in 2022. The 2025 account repeats 1.12 billion; the Goal 11 progress page available in September 2026 reports about 1.16 billion without specifying a common observation year in that summary. The body uses the explicitly dated 2022 threshold, not a spurious exact 2026 census. Slum, informal settlement, poverty, illegality and housing inadequacy are not interchangeable. Different national observations and definitions also prevent treating successive headline figures as a clean growth series. Sources: https://unstats.un.org/sdgs/report/2024/goal-11/ ; https://unstats.un.org/sdgs/report/2025/goal-11/ ; https://sdgs.un.org/goals/goal11 .

The walking and industrial city. Clark's long European history and Hall's planning history support the transition from compact walking cities to industrial and networked urban regions. The statements about mixed streets, peripheral support and status restrictions are comparative syntheses, not claims that all pre-industrial cities had one morphology. Manchester, London and Chicago are selected examples of rapid transport and production growth outrunning inherited housing and services.

Cholera and John Snow. Snow's 1855 second edition sets out the waterborne argument and Broad Street investigation. Later histories show that the pump-handle story is often compressed into a lone-hero account. The outbreak was already declining when the handle was removed, and metropolitan sanitation emerged through multiple waterworks, sewers, filters, laws and administrative capacities. Snow remains important because he changed the explanation and assembled local evidence, not because one map mechanically ended cholera.

Haussmann and Cerdà. Hall supplies the wider history of nineteenth-century planning. Haussmann's Paris joined circulation, sanitation, parks, property development, state display and displacement. Aibar and Bijker examine how Cerdà's Barcelona extension changed through political conflict, property interests and implementation. The reference to filled courtyards describes divergence from the original open-block ideal rather than failure of the entire grid.

Zoning and public space. The book treats zoning as an instrument with several histories: nuisance control, development coordination, property protection and exclusion. Its effects depend on what is permitted, the alternative sites available and the institutions enforcing it. Jacobs, Hall, Gyourko and Molloy, and United States segregation histories inform this account. No claim is made that all zoning systems share American racial origins or that removing rules alone supplies safe, serviced housing.

The metropolitan machine. Baum-Snow, Jackson and Shoup support the account of highway-led decentralisation, mortgage and land policy, and parking as an embedded land cost in the United States. The manuscript keeps the United States setting explicit. Sorensen supplies the contrasting Japanese planning and rail history. Greater Tokyo is used to demonstrate a polycentric rail-oriented metropolitan form, not as proof that its institutions can be copied intact.

Informal settlement. Roy treats informality as a mode of urbanisation and governance rather than a residual category. Simone examines provisional connections through work, movement and association in specific African and Middle Eastern settings. Neither work warrants one global account of tenure or upgrading. The Orangi Pilot Project Research and Training Institute's account of its low-cost sanitation programme describes household and lane-level provision organised by residents, technical support from the programme, and external network responsibilities assigned to government. This is a participant organisation's account of the approach, not an independent evaluation proving universal delivery, treatment performance or transferability: https://www.opp.org.pk/low-cost-housing-program/ .

Upgrading and tenure. Secure tenure can reduce eviction risk and support investment, but legal title alone does not provide drainage, transport, credit or affordability. Clearance, relocation, in-situ upgrading and regularisation have different effects according to hazard, location, land value and resident power. The discussion is deliberately conditional and does not claim that negotiated upgrading is feasible in every dangerous site.

Digital work and located systems. Barrero, Bloom and Davis document the persistence and uneven occupational distribution of home working after the pandemic in the United States. The text makes no global share claim. It uses remote work to show that some information exchange detached from offices while care, logistics, construction, service work and digital infrastructure remained located.

Heat, flooding and ageing. The IPCC assessment of cities, settlements and infrastructure supports the account of climate risk acting through existing land, housing and network inequalities. Hoffman and colleagues provide the bounded heat example. Accessibility for older and disabled residents is framed as a design and service question rather than a biological claim. The illustrative list of pavements, crossings, benches and nearby services is consistent with age-friendly planning guidance but is not a tested package with one effect size.

How we know. Randomised offers, natural experiments, longitudinal records, surveys, observation, administrative data, sensors and remote sensing answer different questions. None creates a complete urban counterfactual. Mobile-phone and platform data can omit people without devices, misclassify purpose and create privacy risks. The book's rule, to name setting, period, population and mechanism, is a claim-control standard rather than a description of universal practice.

What People Get Wrong and Use It

Density and crowding. Density can refer to residents, dwellings, jobs or floor area over a chosen land area. Crowding usually concerns insufficient living space within homes or rooms. Infrastructure, income and ventilation mediate health and comfort. The correction does not claim high density is harmless; it separates a spatial capacity from the failures often correlated with it.

Gentrification. Freeman and Braconi's study of 1990s New York found slower turnover among disadvantaged households in gentrifying neighbourhoods than elsewhere. Ding, Hwang and Divringi found that vulnerable Philadelphia residents were generally no more likely to leave gentrifying neighbourhoods, but vulnerable movers were more likely to reach lower-income destinations. Brummet and Reed's national longitudinal work found modestly increased out-migration, benefits for many stayers and neighbourhood change driven substantially by altered in-migration. The studies use different periods, units and definitions. They do not make displacement negligible. Direct eviction, exclusion of future residents, rent burden, business loss, cultural change and informal moves require separate evidence.

Informality and planning. The correction follows Roy and Simone in rejecting the idea that lack of full formal recognition equals absence of organisation. It also rejects romanticisation. Fire, disease, unsafe structures, insecure tenure and arbitrary enforcement are material harms, while resident-built transport and commerce can be indispensable. The correct policy cannot be inferred from the label alone.

Planning and adaptation. Jacobs and Hall support the criticism of plans that abstract away existing use and behavioural response. Cerdà, highway investment and observed public-space use provide the concrete examples. The manuscript still defends advance reservation of streets, networks and public land because incremental private decisions can close options needed for collective systems.

Practical lenses. The six lenses are syntheses, not validated scoring systems. The workshop, regenerated-district comparison and unnamed households, travellers and journeys are illustrative or explicitly hypothetical, not reported cases. The sanitation programme and public-space research are documented examples with their provenance given above. No invented resident, dialogue or authorial experience is presented as fact.

Terms and Go Deeper

Definitions follow the sources named above and standard urban economics, sociology, planning and transport usage. Several terms remain contested. City and metropolitan area depend on jurisdiction. Gentrification and displacement vary by measure. Informal settlement describes legal and service relationships rather than one building type. Neighbourhood effect is reserved for a causal influence rather than a mapped association.

The four recommended works were checked against publisher and institutional records on 5 September 2026. They serve different purposes: Glaeser for the accessible economic case, Jacobs for close observation and the critique of orthodox planning, Sampson for sustained neighbourhood evidence, and Simone for urban systems made through informal and provisional social connections.

Bibliography

Official sources and original evidence

Aibar, Eduardo, and Wiebe E. Bijker. “Constructing a City: The Cerdà Plan for the Extension of Barcelona.” Science, Technology, & Human Values 22, no. 1 (1997): 3-30. doi:10.1177/016224399702200101.

Anderson, D. Mark, Kerwin Kofi Charles, and Daniel I. Rees. “Reexamining the Contribution of Public Health Efforts to the Decline in Urban Mortality.” American Economic Journal: Applied Economics 14, no. 2 (2022): 126-157. doi:10.1257/app.20190034.

Anderson, D. Mark, Kerwin Kofi Charles, and Daniel I. Rees. “Reexamining the Contribution of Public Health Efforts to the Decline in Urban Mortality: Reply.” American Economic Journal: Applied Economics 14, no. 2 (2022): 166-169. doi:10.1257/app.20210230.

Barrero, José María, Nicholas Bloom, and Steven J. Davis. “The Evolution of Work from Home.” Journal of Economic Perspectives 37, no. 4 (2023): 23-50. doi:10.1257/jep.37.4.23.

Baum-Snow, Nathaniel. “Did Highways Cause Suburbanization?” Quarterly Journal of Economics 122, no. 2 (2007): 775-805. doi:10.1162/qjec.122.2.775.

Brummet, Quentin, and Davin Reed. “The Effects of Gentrification on the Well-Being and Opportunity of Original Resident Adults and Children.” Working Paper 19-30. Philadelphia: Federal Reserve Bank of Philadelphia, 2019. doi:10.21799/frbp.wp.2019.30.

Chetty, Raj, Nathaniel Hendren, and Lawrence F. Katz. “The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment.” American Economic Review 106, no. 4 (2016): 855-902. doi:10.1257/aer.20150572.

Cutler, David M., and Grant Miller. “Reexamining the Contribution of Public Health Efforts to the Decline in Urban Mortality: Comment.” American Economic Journal: Applied Economics 14, no. 2 (2022): 158-165. doi:10.1257/app.20190711.

Cutler, David, and Grant Miller. “The Role of Public Health Improvements in Health Advances: The Twentieth-Century United States.” Demography 42, no. 1 (2005): 1-22. doi:10.1353/dem.2005.0002.

De la Roca, Jorge, and Diego Puga. “Learning by Working in Big Cities.” Review of Economic Studies 84, no. 1 (2017): 106-142. doi:10.1093/restud/rdw031.

Ding, Lei, Jackelyn Hwang, and Eileen Divringi. “Gentrification and Residential Mobility in Philadelphia.” Regional Science and Urban Economics 61 (2016): 38-51. doi:10.1016/j.regsciurbeco.2016.09.004.

Duranton, Gilles, and Diego Puga. “Micro-Foundations of Urban Agglomeration Economies.” In Handbook of Regional and Urban Economics, vol. 4, edited by J. Vernon Henderson and Jacques-François Thisse, 2063-2117. Amsterdam: Elsevier, 2004.

Duranton, Gilles, and Diego Puga. “Urban Land Use.” In Handbook of Regional and Urban Economics, vol. 5A, edited by Gilles Duranton, J. Vernon Henderson, and William C. Strange, 467-560. Amsterdam: Elsevier, 2015.

Duranton, Gilles, and Matthew A. Turner. “The Fundamental Law of Road Congestion: Evidence from US Cities.” American Economic Review 101, no. 6 (2011): 2616-2652. doi:10.1257/aer.101.6.2616.

European Commission, Eurostat, Food and Agriculture Organization of the United Nations, UN-Habitat, International Labour Organization, and Organisation for Economic Co-operation and Development. Applying the Degree of Urbanisation: A Methodological Manual to Define Cities, Towns and Rural Areas for International Comparisons. 2021 ed. Luxembourg: Publications Office of the European Union, 2021.

Freeman, Lance, and Frank Braconi. “Gentrification and Displacement: New York City in the 1990s.” Journal of the American Planning Association 70, no. 1 (2004): 39-52. doi:10.1080/01944360408976337.

Gyourko, Joseph, and Raven Molloy. “Regulation and Housing Supply.” In Handbook of Regional and Urban Economics, vol. 5B, edited by Gilles Duranton, J. Vernon Henderson, and William C. Strange, 1289-1337. Amsterdam: Elsevier, 2015. doi:10.1016/B978-0-444-59531-7.00019-3.

Hansen, Walter G. “How Accessibility Shapes Land Use.” Journal of the American Institute of Planners 25, no. 2 (1959): 73-76. doi:10.1080/01944365908978307.

Hoffman, Jeremy S., Vivek Shandas, and Nicholas Pendleton. “The Effects of Historical Housing Policies on Resident Exposure to Intra-Urban Heat: A Study of 108 US Urban Areas.” Climate 8, no. 1 (2020): 12. doi:10.3390/cli8010012.

Intergovernmental Panel on Climate Change. “Cities, Settlements and Key Infrastructure.” In Climate Change 2022: Impacts, Adaptation and Vulnerability, edited by Hans-O. Pörtner et al. Cambridge: Cambridge University Press, 2022.

Manshel, Andrew M. “A Place Is Better Than a Plan: Revitalizing Urban Areas Is Best Done through Small Improvements, Not Grand Designs.” City Journal, 19 October 2009. Republished by Bryant Park, 3 March 2017. https://bryantpark.org/blog/a-place-is-better-than-a-plan-revitalizing-urban-areas-is-best-done-through .

Orangi Pilot Project Research and Training Institute. “Low Cost Sanitation Program.” 2 April 2017. https://www.opp.org.pk/low-cost-housing-program/ .

Roy, Ananya. “Urban Informality: Toward an Epistemology of Planning.” Journal of the American Planning Association 71, no. 2 (2005): 147-158. doi:10.1080/01944360508976689.

Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277, no. 5328 (1997): 918-924. doi:10.1126/science.277.5328.918.

Sanbonmatsu, Lisa, Lawrence F. Katz, Jens Ludwig, Lisa A. Gennetian, Greg J. Duncan, Ronald C. Kessler, Emma Adam, Thomas W. McDade, and Stacy Tessler Lindau. Moving to Opportunity for Fair Housing Demonstration Program: Final Impacts Evaluation. Washington, DC: United States Department of Housing and Urban Development, 2011.

Schelling, Thomas C. “Dynamic Models of Segregation.” Journal of Mathematical Sociology 1, no. 2 (1971): 143-186. doi:10.1080/0022250X.1971.9989794.

Snow, John. On the Mode of Communication of Cholera. 2nd ed. London: John Churchill, 1855.

United Nations Department of Economic and Social Affairs, Population Division. World Urbanization Prospects 2025: Summary of Results. New York: United Nations, 2025.

United Nations Department of Economic and Social Affairs. The Sustainable Development Goals Report 2025. New York: United Nations, 2025.

United Nations Department of Economic and Social Affairs. “Goal 11.” Sustainable Development Goals progress reporting, consulted 5 September 2026. https://sdgs.un.org/goals/goal11 .

United Nations Statistics Division. “Goal 11: Sustainable Cities and Communities.” SDG Extended Report 2025. New York: United Nations, 2025.

United Nations Statistics Division. “Goal 11: Sustainable Cities and Communities.” The Sustainable Development Goals Report 2024. New York: United Nations, 2024. https://unstats.un.org/sdgs/report/2024/goal-11/ .

Modern works

Alonso, William. Location and Land Use: Toward a General Theory of Land Rent. Cambridge, MA: Harvard University Press, 1964.

Clark, Peter. European Cities and Towns: 400-2000. Oxford: Oxford University Press, 2009.

Glaeser, Edward. Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier. New York: Penguin Press, 2011.

Hall, Peter. Cities of Tomorrow: An Intellectual History of Urban Planning and Design Since 1880. 4th ed. Chichester: Wiley-Blackwell, 2014.

Jackson, Kenneth T. Crabgrass Frontier: The Suburbanization of the United States. New York: Oxford University Press, 1985.

Jacobs, Jane. The Death and Life of Great American Cities. New York: Random House, 1961.

Massey, Douglas S., and Nancy A. Denton. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press, 1993.

Melosi, Martin V. The Sanitary City: Environmental Services in Urban America from Colonial Times to the Present. Abridged ed. Pittsburgh: University of Pittsburgh Press, 2008.

Sampson, Robert J. Great American City: Chicago and the Enduring Neighborhood Effect. Chicago: University of Chicago Press, 2012.

Shoup, Donald C. The High Cost of Free Parking. Chicago: Planners Press, 2005.

Simone, AbdouMaliq. For the City Yet to Come: Changing African Life in Four Cities. Durham, NC: Duke University Press, 2004.

Smith, Monica L. Cities: The First 6,000 Years. New York: Viking, 2019.

Sorensen, André. The Making of Urban Japan: Cities and Planning from Edo to the Twenty-First Century. London: Routledge, 2002.

Whyte, William H. The Social Life of Small Urban Spaces. Washington, DC: Conservation Foundation, 1980.