Cities
Why the rural-urban divide is reshaping politics in many democracies, what successful cities do, and what is changing about urban life globally.
(UN harmonized definition; 36% live in towns and 19% in rural areas - national definitions count 58% as urban)
(despite cities covering roughly 3% of the world's land surface)
(up from 8 in 1975 on the UN's new count; 37 projected by 2050)
A note on framing. Cities are one of the central organizing features of modern life and one of the topics where political polarization runs deepest. The rural-urban divide shapes voting behavior in nearly every developed democracy. The page below tries to walk through the underlying picture - what cities actually do, why they have become more politically salient, and what the next decades will reshape - without taking sides in the cultural argument. There are real trade-offs and real benefits in either direction; the data illuminates them better than the slogans typically do.
The world is now an urban world
On countries' own definitions, more than half of humanity became urban around 2007; the share is now about 58% and is projected to reach 67% by 2050. Those definitions vary, with population thresholds from 200 inhabitants in Denmark to 50,000 in Japan. The UN's 2025 satellite-based method sorts the same population differently: 45% in cities, 36% in towns and semi-dense areas, and 19% in rural areas. It is a different cut of the same reality rather than a contradiction, and it is worth knowing which one a given statistic is using. Different regions are at very different points on the curve. On national definitions, North America is about 80% urban, Europe about 75%, and Latin America about 82%. East Asia is about 65% and still rising. South Asia is about 36%, although the UN's new method puts 40% of Indians in cities and another 44% in towns. Sub-Saharan Africa is about 45%, with the fastest urban growth of any region.
The reasons cities have grown are reasonably consistent across regions. Cities offer higher wages on average than rural areas. They concentrate education, healthcare, and cultural opportunities. They reduce the cost of finding work, a partner, a community. They allow specialization that small populations cannot support. They cluster the kinds of economic activity (services, manufacturing for export, finance, technology) that rural economies cannot produce. The pull factors have been stronger than the push factors in most cases, with specific exceptions where rural-to-urban migration was forced by war, drought, or land dispossession.
What this looks like at the megacity level is harder to state than it sounds. In November 2025 the UN changed how it measures a city, and the ranking rearranged itself. Earlier revisions used each country's own definition, so a city was as large as its national statistical office said it was. The 2025 revision instead sorts one-square-kilometer grid cells by population density and contiguity, using satellite imagery to find where the built-up area sits rather than where the administrative line falls.
On that basis Jakarta is the world's largest city, at 41.9 million, followed by Dhaka at 36.6 million and Tokyo at 33.4 million. Under the old country-specific definitions Jakarta ranked thirtieth, at 12 million. The gap is not migration. Indonesian official statistics drew the boundary around the administrative core and left out the dense settlement pressed against it. Tokyo, measured the old way, was first at 37 million; it is now third. Seoul moved from thirty-seventh to tenth and Guangzhou from twentieth to sixth, without a single person changing address.
It is worth sitting with that for a moment, because it is a general lesson about population statistics rather than a quirk of one report. There are 33 cities above 10 million on the new count. Which of them is "the biggest" depends on a decision recorded in a methodology document. The underlying fact is not in dispute: the largest cities of the developing world have become larger than most countries in the world. Their internal economies, infrastructure, governance, and political dynamics increasingly operate at a scale that traditional city governance was not designed for.
The rural-urban political divide
One of the most consistent political patterns of the last 20 years has been the deepening divide between urban and rural voters in developed democracies. Cities have become more left-leaning, more cosmopolitan, more pro-immigration, more concerned with cultural-progressive issues. Rural areas have become more right-leaning, more nationalist, more skeptical of immigration, more concerned with cultural-traditional issues. In many countries the divide now rivals the older divides based on income or class, and it explains substantial fractions of recent election outcomes.
Why the divide deepened. Several factors have compounded since the 1990s. Manufacturing decline hollowed out industrial small cities and rural towns where good non-college jobs had been concentrated. The technology and finance sectors that grew in their place clustered in already-large cities. University graduates moved to cities for jobs and stayed for the lifestyle, producing self-reinforcing concentration of educated, younger, more progressive populations in urban areas. Immigration concentrated in cities, both because that is where the jobs were and because pre-existing migrant communities were there. Cultural and media institutions clustered in cities, shaping what cosmopolitan and rural communities saw as normal. Each effect was modest individually; together they produced a measurable cultural and political separation.
What the divide produces politically. In countries with first-past-the-post electoral systems (US, UK, Canada), urban populations have become "wasted" votes - concentrated in fewer geographic units than their numbers would suggest, producing election outcomes that tend to over-represent rural voters. In proportional systems (Germany, Netherlands, the Nordic countries), the divide produces more directly proportional representation, but cultural polarization around urban-rural lines persists. Either way, the divide has become a central organizing fact of contemporary politics, and some analysts find it more predictive of voting behavior than income, race, or religion.
What successful cities do
Cities vary enormously in how well they deliver the gains from concentration. Some cities (Tokyo, Singapore, Vienna, Zurich, Helsinki) consistently rank near the top of livability and functionality measures. Others (Lagos, Karachi, parts of Mumbai and Mexico City) struggle with chronic infrastructure failure, traffic, pollution, and security problems despite their economic dynamism. The differences are not random. Specific policy choices matter.
Public transit. Cities that invest in functional public transit (Tokyo's rail, Singapore's Mass Rapid Transit, Vienna's U-Bahn, Zurich's tram network) are dramatically more livable than cities that have not. The reduction in private-vehicle dependence improves air quality, reduces traffic congestion, expands the geographic range of jobs accessible to non-driving residents, and makes streets usable for purposes other than parking. The investment is enormous and slow; the cumulative payoff over decades is among the largest urban-policy returns available.
Housing supply. Cities that build housing in proportion to demand (Tokyo's permissive zoning is the canonical case) avoid the affordability crisis that has hit most major Western cities (covered on the Housing page). Cities that restrict construction (San Francisco, London, Sydney, Toronto, Vancouver) produce chronic unaffordability that compounds across decades. The choice between accommodating growth and preserving existing neighborhood character is a real trade-off; the cities that have chosen accommodation have produced substantially better outcomes for new residents and younger generations.
Public safety. Cities that manage crime and disorder competently produce cumulative quality-of-life advantages that are hard to overstate. The combination of effective but constrained policing, social services that reach vulnerable populations, careful urban design (lighting, foot traffic, mixed-use development), and political leadership that takes both safety and civil rights seriously produces better outcomes than cities that lean too far in either direction. The cities that get this right (most Western European, Japanese, Korean, Singaporean cities) are markedly safer than their economic peers in less-well-managed contexts.
Public space. Parks, libraries, community centers, plazas, walkable streets, public swimming pools, public sports fields - the "third places" that are not home and not work - shape how a city is actually lived in. Cities that invest in these (Vienna, Copenhagen, Singapore again, parts of Boston, Barcelona) produce communities that feel different from cities where the public realm has been left to decay or has been privatized. The investment is unglamorous and the returns compound over generations.
Governance. Cities with reasonable autonomy from national government, professional civil services, accountable elected leadership, and the financial resources to actually execute their plans tend to perform much better than cities that lack any of these. The institutional architecture matters; the same city under different governance arrangements has produced very different outcomes.
The Western downtown crisis since 2020
The COVID-19 pandemic and the subsequent shift to remote and hybrid work have disrupted the central business districts of nearly every major Western city. Office occupancy in San Francisco, downtown Los Angeles, Manhattan, Chicago's Loop, central London, and similar central business districts has fallen substantially. Retail that depended on lunchtime and after-work foot traffic has struggled. Public transit ridership has not fully recovered. Tax revenue from office buildings has dropped. The "doom loop" framing - falling foot traffic produces falling business activity, which produces falling tax revenue, which produces reduced services, which produces further falling foot traffic - has been a real concern in several specific cities.
What is actually happening six years on. Keycard data from Kastle Systems put office attendance in ten large US cities at about 56% of the pre-pandemic level in mid-September 2026, close to its 2025-26 high, with San Francisco at about 47% and New York at about 68%. National office vacancy, 18-20% depending on the broker, has begun to edge down. Manhattan is recovering strongly: prime Midtown vacancy is down to 2.2% (CBRE). San Francisco's vacancy (29.2% in mid-2026, CBRE) is still among the highest, though Cushman & Wakefield ranks its fall over the past year among the largest in the country. Office-to-residential conversion is happening but is technically and financially harder than the abstract case suggests. Retail mixes have shifted toward different uses (more food, less office-supply). The overall shape of central business districts (CBDs) is being remade, with substantial uncertainty about the end state.
What is sometimes missed. The crisis is concentrated in specific kinds of cities (large US coastal metros, parts of London) and is not uniform across cities. Mid-sized US cities that did not have monoculture-CBD economies have largely recovered. European cities have generally fared better than US cities because their downtowns were always more mixed-use. Asian cities returned to office work much faster and have not seen comparable disruption. The "future of cities" framing that some commentary uses generalizes from a specific subset of cities to a global story that does not match the underlying data.
How major cities compare
The cities below are a small sample chosen to illustrate variation. None of them is "average." Populations are the UN's 2025 city figures, which leave out lower-density suburbs (national figures for Singapore and Vienna).
The takeaway: the variation among cities is among the largest of any social comparison. Some cities deliver enormous quality-of-life advantages; others struggle to deliver basic functionality at scale. The differences are not primarily about wealth - Vienna is poorer per capita than San Francisco but delivers measurably better outcomes on most metrics - but about specific structural choices made consistently over decades.
The paths from here
Dated September 2026, with US office vacancy edging down and growth in the biggest US cities slowing as immigration fell. Urban life globally is being reshaped by several forces simultaneously. Each path below is one realistic shape the next two decades could take.
Continued urbanization in the developing world
African and South Asian cities continue to absorb hundreds of millions of new residents over the next 25 years. The infrastructure investment required is enormous; the share that actually gets built will determine whether these cities deliver the productivity gains historical urbanization has produced or get stuck in chronic infrastructure failure.
Will it happen? The demographic pressure is locked in: the UN expects cities to gain 986 million people by 2050, more than half of them in seven countries, including India, Nigeria, and Pakistan. Whether the infrastructure keeps up depends on choices being made now about transit investment, housing policy, and governance design in countries with limited fiscal space. Most are under-investing relative to the scale of growth.
Western downtown reinvention
Office-heavy central business districts adapt by converting to residential, mixed-use, and entertainment functions. The recovery is uneven - some cities reinvent successfully, some experience long-term decline. The shape of "downtown" by 2040 looks different from the 2010s pattern, with implications for transit, retail, and tax bases.
Will it happen? Pieces are happening. NYC and several other cities have begun office-to-residential conversions at scale, and converting or repurposing obsolete buildings took about 33 million square feet out of the US office stock in the five quarters to mid-2026 (Cushman & Wakefield). Mid-sized cities have largely already adapted. The most-stressed cities (San Francisco specifically) face harder paths. The base case is selective reinvention with substantial city-by-city variation.
Climate adaptation reshapes which cities thrive
Heat, water stress, sea-level rise, and storm intensity reshape which cities are pleasant and functional places to live. Some highly successful current cities (Phoenix, parts of Florida and the US Gulf Coast, parts of South Asia and the Middle East) face increasing climate stress. Some currently-secondary cities (Buffalo, Minneapolis, parts of Northern Europe) become more attractive. Migration patterns within and between countries adjust over decades.
Will it happen? Partly under way. Insurance markets have begun pricing climate exposure into specific markets, but people have not followed yet: the South dominated the Census Bureau's lists of the fastest-growing and highest-gaining US cities in 2024-25. The pattern may accelerate as climate effects intensify; the speed depends partly on how quickly information about climate exposure becomes salient in housing-purchase decisions.
Mid-sized cities capture growth
Remote and hybrid work, plus continued affordability stress in the largest cities, drive sustained migration to mid-sized cities (Austin, Nashville, Raleigh, Boise, Lyon, Manchester, Bristol, Bilbao, etc.). These cities grow into more mature commercial and cultural hubs. The geographic distribution of economic activity becomes meaningfully more dispersed.
Will it happen? Partly already happening. Several mid-sized US cities grew substantially during the pandemic. In 2024-25 Austin passed one million residents and Raleigh 500,000, while New York City lost more people than any other US city (Census Bureau). Whether the pattern sustains depends on whether remote work patterns continue to allow it and on whether the mid-sized cities maintain the affordability advantages that made them attractive in the first place.
The political divide between cities and rural areas continues to deepen
Urban and rural populations continue to diverge culturally and politically. Election outcomes increasingly reflect this geographic divide more than older income or class divides. Federal and state-level politics in countries with strong urban-rural divides produce repeated cycles of conflict, with regional autonomy and devolution becoming larger themes in politics.
Will it happen? Already happening. The divide has been deepening for two decades. In the 2024 US election rural voters backed Trump 69%-29%, a wider margin than in 2016 or 2020, while urban voters backed Harris 65%-33% (Pew Research Center). Whether it produces constructive devolution (more regional autonomy) or destructive conflict (deeper polarization, weaker national institutions) depends on choices that have not been made yet.
In the base case, continued developing-world urbanization is the dominant force. Mid-sized city growth (path 4) and Western downtown reinvention (path 2) reshape the rich-world landscape. Climate adaptation (path 3) becomes a larger variable through the 2030s. The political divide (path 5) deepens absent specific reform energy. The cumulative effect: cities matter more than ever, in different ways than they did 20 years ago.
Where serious analysts disagree
Cities are one of the topics where careful researchers using similar data reach different conclusions. Each reading below is held by named scholars worth engaging directly.
Cities are the engines of human progress and need to grow more
Cities concentrate the human capital, networks, and institutional density that produce most economic growth, innovation, and cultural production. The constraints on urban growth - restrictive zoning, "not in my backyard" (NIMBY) politics, infrastructure under-investment - are some of the largest self-imposed barriers to prosperity in the rich world. The case for letting cities grow - through housing reform, transit investment, and appropriate density - is one of the strongest evidence-based policy arguments available.
Held by: Edward Glaeser (Harvard, "Triumph of the City"), Alain Bertaud (NYU), and most working urban economists. The data on agglomeration effects is robust; the political case for acting on it has been weaker than the evidence base would predict.
The rural-urban divide is the central political fact of our time
The cultural and political separation between cities and rural areas has become the dominant axis of contemporary politics, more important than income, race, or religion in shaping voting behavior. Understanding this divide - what drives it, what could narrow it, what makes it worse - is essential for anyone trying to make sense of contemporary politics. Many of the most heated political fights stem from this geographic-cultural separation.
Held by: Will Wilkinson, Ezra Klein, and a substantial body of contemporary political analysis. The voting data robustly show the divide; how much of it reflects place itself rather than who lives where (education, age, race, religion) is still debated. The harder question is what to do about it - the drivers (different economies, different demographics, different cultural reference points) are not easily addressed.
Western city decline is a real phenomenon, not a temporary disruption
The post-2020 decline of several major Western cities (San Francisco specifically, parts of Chicago, parts of New York, parts of London) is not just a pandemic-era disruption that is reverting. It reflects deeper trends: housing unaffordability driving out middle-class families, public-safety challenges driving out small businesses, declining trust in city governance, and the rise of remote work making city presence optional for high-earning knowledge workers. Some of these cities will recover; some will not. The political conversation has been slow to acknowledge the differences.
Held by: Joel Kotkin and a tradition of urban-policy analysis skeptical of the "cities always win" framing. Recent data cut both ways: average growth in big US cities slowed to 0.3% in 2024-25 and New York City has 2.5% fewer residents than in 2020 (Census Bureau), but office vacancy has begun to fall and the Economist Intelligence Unit credits New York with years of falling crime.
The real urban story is in the developing world
Western urban debates (downtown decline, NIMBY politics, gentrification) are local stories about a tiny share of global urbanization. The actual urban story of the 21st century is in Lagos, Mumbai, Dhaka, Manila, Cairo, Karachi, Kinshasa, and the long list of similarly-situated cities absorbing hundreds of millions of new residents. Whatever happens in San Francisco or Manchester, the shape of global urbanization is determined by what happens in those cities.
Held by: Kim Dovey, parts of UN-Habitat, and a tradition of global urban scholarship. Their argument is that the political and academic attention to Western city issues is disproportionate to their share of the actual story. The data on global urban populations supports the framing.
Density and walkability are health and climate imperatives
The dispersed-suburb development pattern that dominated US growth from 1945 to roughly 2010 is incompatible with climate targets and public health, given the large chronic disease burden of car-dependent lifestyles. It also serves aging societies poorly, because suburbs do not work well for non-driving older adults. The future-of-cities conversation should be substantially about how to retrofit auto-dependent regions into more walkable forms, which has barely begun in most US metros.
Held by: Strong Towns (Charles Marohn), the Congress for the New Urbanism, and a body of public-health and transit research. The argument has been gaining traction; the political fights at municipal level have been intense and the change has been slow. Critics, Kotkin among them, reply that many households prefer suburban living, and Census data show smaller cities on the edges of US metro areas often outgrowing their cores.
What the evidence supports: cities are real engines of progress, and specific governance failures limit how much of it they deliver. The rural-urban divide is real and politically central. Western city decline is uneven, but real in some cases. The global urbanization story is dominated by developing-world cities. Density and walkability bring real health and climate gains, even as many households keep choosing suburban space. Each of these is worth holding at once rather than picking one as the right framing.
What this means for you
Cities touch everyday life through where you live, where you work, what your daily life looks like, and the politics of the place around you. A few practical observations:
If you are choosing where to live
Within one country, the differences between cities on the quality-of-life measures that matter to daily living can rival the differences between countries. A walkable, transit-served, affordable mid-sized city often delivers a better life than a more famous large city in the same country at substantially lower cost. The trade-offs involved (career opportunities versus daily quality of life) are worth thinking through deliberately rather than defaulting to the most prestigious option. Climate exposure of specific cities is now a real input for 30-year housing decisions.
If you live in a stressed city
Most "city is dying" framings are exaggerated; most "city has problems" framings are accurate. The constructive position is to be involved in the specific governance choices that determine whether the city improves or continues to decline. Local elections, planning commissions, transit board appointments, and ballot measures are the actual mechanisms; getting involved at this level produces more change than complaining about it on social media.
If you live in a small town or rural area
The data on rural strain in the US (closing hospitals, more deaths than births, fewer jobs for non-college-educated workers) is real and remains unresolved under both parties, even though migration has made the rural population grow again in recent years. Rural broadband, healthcare access, infrastructure investment, and economic-development support are all areas where specific federal and state policy matters. Voting on these specifics rather than on cultural-issue proxies is more useful than the politicized conversation suggests.
If you are voting on local policy
Most consequential urban policy is local: zoning, permitting, transit, public safety, schools, parks, public health. The local elections that determine these are usually decided by small numbers of engaged voters - and the engaged voters are typically older, longer-resident, and more property-owning than the general population. Younger and renter voters who actually engage in local politics have outsize effect. The single most under-invested civic action is consistent participation in local elections and hearings.
If you are thinking about long-term investment
Real estate values over 30 years correlate strongly with city quality - urban governance, infrastructure investment, climate exposure, employment trajectory, and demographic dynamics. The cities and neighborhoods that have invested well do measurably better over decades than the cities that have not. None of this is investment advice; it is observing where the underlying variables point. Mid-sized cities that have invested in transit, housing, and public realm are showing measurable advantages over large cities that have not.
The mechanics behind this
The cities story sits on top of three deeper mechanisms covered elsewhere on this site. If the analysis above depends on ideas you want to understand first, these fundamentals make the conversation more legible:
