Some US cities now have worse inequality than Mexico, a study has shown

Skid Row, Los Angeles. Image: Getty.

The cities of the Americas are unequal places. US census data and recent American Community Surveys show that in most modern American metropolises, resources are unevenly distributed across the city – think New York City’s lower Manhattan versus the South Bronx – with residents enjoying unequal access to jobs, transportation and public space.

In 2014, New York City’s GINI inequality index was 0.48, meaning that income distribution was less even in New York City than in the US as a whole (0.39). It was also higher than the most unequal OECD countries, Chile (0.46) and Mexico (0.45).

Latin America, which is the world’s most unequal place, is also by far the most urbanised region of the globe. More than 80 per cent of its population lives in large cities. Between 1950 and 2005, the region’s big cities grew precipitously. Both Mexico City and São Paulo jumped from just under three million people to, in both cases, nearly 19m.

Data on urban inequality is largely unavailable, but it is clear that this rapid urbanisation has been far from equitable. According to a 2012 UN Habitat report, the large majority of Latin America’s non-poor population lives in major metro areas, while the poorest live in rural areas.

What does inequality look like?

No matter where you live, measuring inequality is tricky, because its incidence and extent changes in different parts of the city.

Sure, there are rich neighbourhoods and poor ones: high-income and low-income households sort themselves across cities according to preference (for local public goods and neighbourhood composition) and needs (according to budget, job location and housing prices).

But not every neighbourhood is comprised fully of households with the same income. Income sorting across space is often “imperfect”, meaning that rich and poor households might live in the same neighbourhood and share common social ties and local amenities.

As a result, a very specific and local kind of inequality emerges within neighbourhoods. This phenomenon is sizeable in US metro areas, Census Bureau data shows. Not only do unequal households live very close together, but neighbourhoods also represent small communities where local inequality, on average, seems to track overall urban inequality.

For example, New York City, Chicago and Los Angeles all have neighbourhood income inequality at least 20 per cent larger than Washington’s, which matches the difference in the cities’ GINI indices. We found that inequality within individual neighbourhoods has also been rising precipitously over the past 35 years (even in very small neighbourhoods), indicating an increase of income heterogeneity at the community level.

This unexpected finding is likely related to the comeback of North American cities over the past decade – the so-called great inversion. Across the Americas, jobs and firms are moving back into major metro areas, attracting more skilled people, who are generally young, receive higher wages and prefer to settle down where their jobs are.

As high-income young couples buy up homes in historically distressed neighbourhoods long dominated by the working and renting class – and gentrify them – they push up income heterogeneity in those places. This is happening in cities across the Americas.

Gentrification has occurred in many North American cities, increasing local income inequality and, in some cases, tensions. Image: Michael Premo/Flickr/creative commons.

Keeping up with the Joneses

We wanted to better understand this phenomenon. Why is local income inequality rising? How can we quantify it? What are the trends in uber-localised inequality? And what does it all mean for city dwellers?

Those were the questions driving our study – So close yet so unequal: Reconsidering spatial inequality in US cities – which focused on US cities. Our preliminary findings were recently published in a Catholic University of Milan Working Paper.

Unlike traditional assessments of inequality, which accept administrative partitions of the city as the unit of analysis and measure income inequality in those neighbourhoods, we look at inequality among neighbours, putting people at the centre of our analysis.

The underlying thought experiment consists of asking individuals to compare their income with that of neighbours living within a given distance range (from few blocks to entire census areas), thus quantifying income inequality in that particular person’s neighbourhood.

In doing so for every person in a city – any city – one should be able to measure two aspects of spatial inequality: the average income inequality within individual neighbourhoods (is my neighbour richer than me?), and inequality among the average incomes of each neighbourhood (is that neighbourhood richer than mine?).

We found that these two indices define a typology of cities that mirrors what urban planners have found at the city level. Some places are “even cities”. Like Washington DC, they display relatively low income inequality everywhere.

Other metro areas, among them Miami and San Francisco, show high urban inequality, but high and low-income households are rather evenly distributed throughout the city. These are so-called “mixed cities”.

The largest US metro areas also have the most unequal neighbourhoods. In New York and Los Angeles, the way high and low-income households are distributed across the urban footprint reflects what planners call the “unstable city” model.

The Great Gatsby in the ‘hood

Such substantial and increasing inequality appears to imply several contradictory things for cities and their residents.

As shown in Figure 1, lower neighbourhood inequality is associated, on average, with large upward mobility gains for young people who grew up in poor families, a phenomenon reported in recent work by Stanford University’s Raj Chetty.

Figure 1: upward mobility in America’s urban neighbourhoods. Upward mobility gains/losses for children living in poor families in 2000, by Commuting Zone. Image: author provided.

Children of better-off families benefit, too, from living in a homogenous local community, thanks to “positive contagion” facilitated by social interaction among wealthy young peers.

Both findings are evidence of a “Great Gatsby Curve” in America’s neighbourhoods. That is, greater income inequality in one generation amplifies the consequences of having rich or poor parents for the economic status of the next generation.
Yet greater income inequality within individual neighbourhoods may actually be a good thing for poorer locals. Figure 2 shows that they experience life expectancy gains, perhaps due to positive health modelling and increased aspirations among poor adult residents.

Figure 2: Life expectacy in America’s urban neighbourhoods. Image: author provided.

Addressing inequality

For policy makers, then, our findings create an intergenerational trade-off. A “mixed city” model would seem to promote life expectancy gains for poor adults who live there, while the “even city” ideal furthers economic mobility of young people who grow up poor.

Lessons learned from such a policy debate in the US could have important international consequences.

No one has yet applied our neighbourhood-based inequality analysis to Latin America’s unequal cities. But we can see that in metropolises such as Mexico City, and São Paulo in Brazil, as well as in smaller cities, uncontrolled sprawl and lack of urban planning has increased the distances between high, middle and low-income households.

The view from the Rocinha favela, in Rio de Janeiro, where ‘urban renewal’ is now encroaching on some of the poorest parts of the city. Image: AHLN/Flickr/creative commons.

This is the “polarised city” model, and our paper found little evidence of it in US cities (with the exception of Detroit and Washington). Such places have substantial heterogeneity in income across neighbourhoods and relatively little heterogeneity within neighbourhoods.

In Latin America’s polarised cities, the poor are separated from the rest of the population. As a result, they have lower access and opportunities for education, employment and services. This inequality has been exacerbated by gentrification and by the region’s growing global economic engagement. This has strengthened urban elites’ connections to the world while relegating Latin America’s poor further into the periphery.

In such cases, increasing the urban income mix seen in New York City might actually have beneficial effects for the city’s neediest residents. This is a relevant area for future study. It would be interesting, for example, to plot cities across the Americas on the same graph, examining regional trends in longevity and mobility based on neighbourhood-level inequality.

The ConversationSuch hyper-local analysis would offer both policymakers and international agencies the kind of information they need to improve the lives of today’s city dwellers, both now and in the future.

Eugenio Peluso is associate professor of economics at the University of Verona. Francesco Andreoli is a post-doctoral researche at the Luxemburg Institute of Socio-Economic Research (LISER).

This article was originally published on The Conversation. Read the original article.


Everything you ever wanted to know about the Seoul Metro System but were too afraid to ask

Gwanghwamoon subway station on line 5 in Seoul, 2010. Image: Getty.

Seoul’s metro system carries 7m passengers a day across 1,000 miles of track. The system is as much a regional commuter railway as an urban subway system. Without technically leaving the network, one can travel from Asan over 50 miles to the south of central Seoul, all the way up to the North Korean border 20 miles north of the city.

Fares are incredibly low for a developed country. A basic fare of 1,250 won (about £1) will allow you to travel 10km; it’s only an extra 100 won (about 7p) to travel every additional 5km on most lines.

The trains are reasonably quick: maximum speeds of 62mph and average operating speeds of around 20mph make them comparable to London Underground. But the trains are much more spacious, air conditioned and have wi-fi access. Every station also has protective fences, between platform and track, to prevent suicides and accidents.

The network

The  service has a complex system of ownership and operation. The Seoul Metro Company (owned by Seoul City council) operates lines 5-8 on its own, but lines 1-4 are operated jointly with Korail, the state-owned national rail company. Meanwhile, Line 9 is operated jointly between Trans-Dev (a French company which operates many buses in northern England) and RATP (The Parisian version of TfL).

Then there’s Neotrans, owned by the Korean conglomerate Doosan, which owns and operates the driverless Sinbundang line. The Incheon city government, which borders Seoul to the west, owns and operates Incheon Line 1 and Line 2.

The Airport Express was originally built and owned by a corporation jointly owned by 11 large Korean firms, but is now mostly owned by Korail. The Uijeongbu light railway is currently being taken over by the Uijeongbu city council (that one’s north of Seoul) after the operating company went bankrupt. And the Everline people mover is operated by a joint venture owned by Bombardier and a variety of Korean companies.

Seoul’s subway map. Click to expand. Image: Wikimedia Commons.

The rest of the lines are operated by the national rail operator Korail. The fare structure is either identical or very similar for all of these lines. All buses and trains in the region are accessible with a T-money card, similar to London’s Oyster card. Fares are collected centrally and then distributed back to operators based on levels of usage.


The Korean government spends around £27bn on transport every year: that works out at 10 per cent more per person than the British government spends.  The Seoul subway’s annual loss of around £200m is covered by this budget.

The main reason the loss is much lower than TfL’s £458m is that, despite Seoul’s lower fares, it also has much lower maintenance costs. The oldest line, Line 1 is only 44 years old.

Higher levels of automation and lower crime rates also mean there are fewer staff. Workers pay is also lower: a newly qualified driver will be paid around £27,000 a year compared to £49,000 in London.

New infrastructure is paid for by central government. However, investment in the capital does not cause the same regional rivalries as it does in the UK for a variety of reasons. Firstly, investment is not so heavily concentrated in the capital. Five other cities have subways; the second city of Busan has an extensive five-line network.

What’s more, while investment is still skewed towards Seoul, it’s a much bigger city than London, and South Korea is physically a much smaller country than the UK (about the size of Scotland and Wales combined). Some 40 per cent of the national population lives on the Seoul network – and everyone else who lives on the mainland can be in Seoul within 3 hours.

Finally, politically the biggest divide in South Korea is between the south-west and the south-east (the recently ousted President Park Geun-Hye won just 11 per cent of the vote in the south west, while winning 69 per cent in the south-east). Seoul is seen as neutral territory.  


A driverless train on the Shinbundang Line. Image: Wikicommons.

The system is far from perfect. Seoul’s network is highly radial. It’s incredibly cheap and easy to travel from outer lying areas to the centre, and around the centre itself. But travelling from one of Seoul’s satellite cities to another by public transport is often difficult. A journey from central Goyang (population: 1m) to central Incheon (population: 3m) is around 30 minutes by car. By public transport, it takes around 2 hours. There is no real equivalent of the London Overground.

There is also a lack of fast commuter services. The four-track Seoul Line 1 offers express services to Incheon and Cheonan, and some commuter towns south of the city are covered by intercity services. But most large cities of hundreds of thousands of people within commuting distance (places comparable to Reading or Milton Keynes) are reliant on the subway network, and do not have a fast rail link that takes commuters directly to the city centre.

This is changing however with the construction of a system modelled on the Paris RER and London’s Crossrail. The GTX will operate at maximum speed of 110Mph. The first line (of three planned) is scheduled to open in 2023, and will extend from the new town of Ilsan on the North Korean border to the new town of Dongtan about 25km south of the city centre.

The system will stop much less regularly than Crossrail or the RER resulting in drastic cuts in journey times. For example, the time from llsan to Gangnam (of Gangnam Style fame) will be cut from around 1hr30 to just 17 minutes. When the three-line network is complete most of the major cities in the region will have a direct fast link to Seoul Station, the focal point of the GTX as well as the national rail network. A very good public transport network is going to get even better.