How big data and local politics can make our cities more inclusive

The UK port city of Bristol, where one of us is mayor, is characterized by incredible concentrations of both wealth and poverty.

Some neighbourhoods offer the highest quality of life in the UK, but others are among the poorest in the country. In one area, Clifton, nearly every young person goes on to higher education; in Hartcliffe in the south, fewer than one in 12 do.

People from around the country flock to Bristol to work in a thriving economy, but the city has been identified as one of the worst in the UK for racial inequality in education and employment.

When local government seeks to make Bristol better, it is not working at the level of a single city, but a complex patchwork of poverty and prosperity. Just keeping track of this ever-changing landscape is a huge challenge – one that a current revolution in big data promises to address.

Since their emergence about 10,000 years ago, cities have been subject to two fundamental forces: agglomeration and segregation. They draw their power from pulling people together on a massive scale, but such collisions inevitably consolidate inequalities and divisions.

Entire groups end up on the margins, unable to fully participate in a city’s economic, social and civic life. For advocates of a more inclusive city, one of the great challenges is fighting segregation in all its forms. But to effectively bring a city back together, we must have an accurate map of the fault lines that push its residents apart.

Some segregation is obvious to the naked eye — some streets have sports cars and others have cracked pavements — but the visible disparities between neighbourhoods are only one link in a long chain of social difference.

We are segregated in the places we visit, the work we do and the people we see. These subtler fault lines are hard to track, but they have huge implications for policy-makers. How can we distribute city services if key beneficiaries never travel to certain spots? How can we form united communities when entire swaths of the population never encounter each other?

This is where big data can be revolutionary. Research from the Massachusetts Institute of Technology (MIT) has been able to track the subtlest forms of segregation with large, anonymous datasets of mobile phone activity.

In a series of analyses in different cities, it measured how often people from different backgrounds bump into each other on the street or talk to each other on the phone. These findings allow us to define the amorphous borders of what, together with sociologist Richard Sennett, we have started calling “liminal ghettos”.

The term liminal ghetto is used in contrast to the more rigid, official ghettos of history – such as the Jewish quarters of medieval Europe and the redlined inner cities of the post-war United States – and it takes new, technologically augmented eyes to see them.

What does it mean to live in a liminal ghetto? In Stockholm, Sweden, we found that people do not only live in segregated areas, but they are also more likely to visit neighbourhoods that have the same socioeconomic profile as their own.

In Singapore, we found that a stratified “rich club” are the most segregated group in terms of phone calls: they are more likely than anyone else to only call one another. Liminal ghettos do not only divide social groups; they also trap individuals into corners of loneliness.

In the Portuguese city of Porto, we found that the COVID-19 pandemic caused a 35% drop in the overall diversity of social interactions, with the most pronounced and lasting damage to the social ties of women and the elderly.

Liminal ghettos are not a new discovery: they have always been obvious to those living inside them. As one of us grew up in 1980s Bristol as a mixed-race boy from a low-income household, he knew there were places he should not go. Some places were safe and welcoming to people of his skin colour; others were subtly or overtly hostile.

These fractures were worsened by poor and expensive public transport, inequalities in house ownership, and separated schooling. While none of the insights one can gain through big data are surprising per se, the ability to quantify them precisely could be revolutionary.

Equipped with a detailed knowledge of how our cities are divided, governments can improve the targeting of social services and new developments. We can pierce through the walls of liminal ghettos, reaching groups who might otherwise be cut off.

Our research into liminal ghettos can also highlight their weak points, places that tend to bring people together despite the forces that divide them.

A study in New York City found that when parks are shut down for construction, encounters between Black and white residents fall. The study shows that certain kinds of spaces are integrators, capable of bringing people together. One simple way to make a better city could be to build more of them, as it is currently the focus in Bristol.

What does bridging liminal ghettos look like in practice? It means creating integrated spaces that are both accessible and desirable: venues where people from all walks of life want to go and feel welcome. Parks, plazas and schools fit this category. But it can also happen in unexpected venues.

Last winter in Bristol, rising fuel prices and falling temperatures were threatening the physical and financial well-being of the most vulnerable residents. To combat the crisis, we opened more than 100 free, warm spaces across the city.

We knew that these rooms needed to be both warm and welcoming; otherwise, they would be segregated and stigmatizing. To this end, we called the initiative Welcoming Spaces and focused on making each one a hub of social services and community.

When winter came these rooms — in churches, mosques, and community centers — did not only keep people safe from the cold, but they also thawed the frosty edges of our city’s liminal ghettos.

In a survey, 93% of Welcoming Spaces users said the biggest impact of the programme was not keeping them warm, but providing a place for people to socialize. This finding underscores just how divided our cities have become — communities and individuals have been cut off from each other — but also how hungry we are to reconnect.

Together, our studies at MIT and our experience in Bristol demonstrate the power of a data-driven, locally led approach to promote inclusion.

Big data can give us unprecedented quantifications of the problems we face, but the decision power needs to be in the right hands. National governments typically have the resources to lead large urban programmes, but only local governments possess an intimate knowledge of their surroundings.

By combining big data with “rich data” – the nuanced individual stories that make up the fabric of a city – we can merge local wisdom technical tools.

Cities have always been caught between segregation and agglomeration, but in this very moment we could shift the balance between them. If we can truly understand how urban citizens get pushed apart, we have a chance to finally pull them back together.

Carlo Ratti is co-chair and Marvin Rees is a member of Global Future Cities Council at the World Economic Forum.