- Children who grow up in areas with lots of mixing between income groups earn more as adults – around an additional £5,000 per year
- Friendships in the UK are less divided by class than in the US, but former industrial areas have fewer cross-class friendships
- Findings are part of early research from a study looking at 6 billion friendships among UK adults aged 25-64
London, 24 March – A major study into social connection in the UK has revealed that friendships across socioeconomic lines have a strong relationship with future earnings and social mobility.
The early research – based on aggregated, anonymised Facebook data for over 20 million people aged 25 to 64 in the UK – suggests that children who grow up in areas where people from different economic backgrounds become friends experience a significant earnings bonus in adulthood.
According to the study, funded by the Nuffield Foundation, children from low-income (below the median) families raised in areas with many cross-class friendships earn 38% more as adults than those from similar backgrounds who lived in areas with fewer such friendships. This translates to approximately £5,100 per year more in earnings on average.
The financial advantage remains strong, even after accounting for other area characteristics. Economic connectedness is the second most important independent factor (after median income) in predicting upward mobility within an area even when controlling for income, education and health outcomes.
The research was carried out by an international consortium including BIT (Behavioural Insights Team), together with Meta, Stripe Partners, the RSA, Neighbourly Lab, and researchers from Stanford and Opportunity Insights.
To understand the relationship with social mobility, the research team used UK administrative data that tracks the adult earnings of young people. It follows a 2022 US study led by Harvard economist Raj Chetty which found that children from poorer backgrounds typically earn more as adults if they grow up in areas with more friendships between high and low earners – a concept called ‘economic connectedness.’
The team analysed six billion Facebook friendships, covering roughly 58% of the UK population aged between 25 and 64, to see how connected people are across income levels.
More cross-class mixing in the South East
This new study, published today as a working paper, finds that the link between economic connections and social mobility also holds in the UK. However, the researchers discovered that social connections in the UK are less divided by income than in the US. In Britain, 47% of low earners’ friendships are with people who earn above the median income, compared with 39% in the US. Even among the poorest 10% of UK society, nearly 5% of their friendships are with people from the richest 10%, meaning that the poorest in our society are still friends with the richest.
The research team found significant differences in economic connectedness by region. Cross-class friendships are relatively common in South East England, particularly in London and its surroundings. By contrast, post-industrial regions such as northern England, South Wales, the Scottish Central Belt and Northern Ireland show much lower rates of economic connectedness.
Although areas with more high-income residents generally have greater economic connectedness, a high concentration of affluent individuals doesn’t always result in more cross-class friendships. For example, Kingston upon Thames and Canterbury have similar levels of affluence, yet the share of high-income friends among low-income residents is 10% higher in Kingston upon Thames. This difference is associated with varying outcomes too – disadvantaged children from Kingston upon Thames earn about £2,800 more annually as adults than those from Canterbury.
Hobby groups and local neighbourhoods
The researchers also looked at where cross-class connections are formed. They found that hobby groups, including sports teams such as five-a-sides, can be a rich place to forge friendships across groups. In hobby groups, people from low-income backgrounds form friendships with high-income peers at approximately 5% higher rates than if they were formed proportionately.
However, most of our friendships are actually formed in neighbourhoods and schools, which remain segregated by income. This presents challenges for forming connections across economic lines.
Universities
The study also found that universities serve as crucial environments where low-income individuals encounter peers from more advantaged backgrounds. In most UK universities, lower-income students still form fewer friendships with higher-income students than random chance would predict, but considerably more than in their local neighbourhoods. The share of high-income friends that low-income students have at university is over 40% greater than in their home neighbourhoods. The University of Cambridge, University of Salford, Kingston University, University of East London and the University of Reading are where students from low-income backgrounds are most likely to form cross-class friendships, relative to their numbers on campus.
Happiness and trust
The researchers also conducted a survey of nearly 4,000 UK adults on Facebook to examine if having more friends from different backgrounds was linked to a better quality of life. Generally, people with more tightly-knit social circles, particularly with more friends from high-income backgrounds, reported higher levels of life satisfaction, feelings of belonging and trust in others.
After accounting for the effects of income, the researchers compared individuals with the highest proportion of very high-income friends (about 1 in 5) to those with the lowest (around 1 in 40). The results showed that those with more very high-income friends reported 5% higher happiness levels and 23% greater trust. In short, people with richer friends tend to have higher well-being in ways that personal income alone does not fully explain.
These findings represent early-stage research from a wider programme of work. The next phase will look in more detail at the role of social capital on a wider set of life outcomes, including education, health and crime.
Ends
Quotes from researchers and partners:
Antonio Silva, Principal Investigator and Head of Social Cohesion at BIT, said:
“While we find that people in the UK make more friends from their own class, we still find surprising levels of cross-class friendships. Even among people who earn the least, nearly 5% of their friendships are with people from the highest earning group. As expected, cross-class friendships are more likely to happen in more well-off places but we still find a substantial amount of variation between places.
“While this is early research, it seems that cross-class friendships matter for future earnings. And indeed they may make us happier generally and more trusting of others. As most friendships are made in neighbourhoods and schools, these settings should be where we really focus on reducing segregation. We need to develop new ways of promoting connectedness across economic lines, whether by changing school catchment areas or building more mixed housing.”
Andy Haldane, CEO of the RSA, said:
“We live in a time characterised by great division and disunity. This landmark research offers still more proof that social connection and social mobility are inextricably linked, and core to the income and happiness prospects of individuals and nations alike. Fostering greater connection between people needs to become the golden thread running through all public policy, and we will be working with our partners across government, business and civil society to grow this evidence base and put it to work in communities up and down the country.”
Marnie Freeman, Director of Neighbourly Lab, said:
“Our forthcoming qualitative research demonstrated that when people from different income levels have opportunities to mix on a regular basis, their lives are impacted positively in a range of ways. This included social mobility as well as community connections and overall well being. The settings we saw as most fertile for creating the conditions for sustained mixing included workplaces and sports and hobby groups. These create some exciting opportunities for developing both top down policy interventions and bottom up grass roots initiatives.”
Nick Clegg, President of Global Affairs at Meta, said:
“This work is a major contribution to our understanding of the relationship between social connections and economic opportunity. Studying Facebook friendships in this way can reveal a huge amount about the society we live in. And it shows how Meta’s data can be used for societally significant research when shared responsibly and in a way that protects people’s privacy. We’re all still grappling with the ways social media interacts with the world – both positive and negative – which is why we want to support high quality independent research.”
Shankar Iyer, Research Scientist at Meta, said:
“Large-scale social network data is opening up new research opportunities on the importance of social capital and the value of people’s connections. Our results show that the relationship of cross-class connections with people’s ability to rise in socioeconomic status holds in both the US and the UK. They also help identify specific social settings like hobby and recreation groups that can help support cross-class interaction and the nation’s social fabric.”
Bob Putnam, Political Scientist at Harvard and author of Bowling Alone:
“This research confirms American studies that in Britain too, social integration breeds life success, both economic and socioemotional. Its unique combination of qualitative and quantitative evidence illuminates just what sort of social integration is most important, especially hobby groups and sports teams. The insights from this important new study should help inform the social policies of the new Starmer government.”
David Halpern, President Emeritus at BIT, said:
“This is a breakthrough study. It sheds new light on the character and strength of our social networks, and how they differ across society. It also shows how consequential this ‘social capital’ is for our life chances, and for those of our children. Policymakers know that environmental and human capital matters. But they are much less aware of social capital – its impacts or even how to measure it. Official statistics capture places and individuals, but they don’t capture or see the importance of networks.”
Notes to editors
- For more information on the analysis or to speak to one of the experts involved, please contact Mark Byrne, BIT Press Office, on +4477452349096 or [email protected]. Spokespeople are available for broadcast interviews.
- Both the research summary and the associated charts / graphics are available here (link to folder)
Funding acknowledgement
The work of BIT, the RSA, Stripe Partners, and Neighbourly Lab on this project is funded by the Nuffield Foundation, based in the UK. However, the views expressed are those of the authors and not necessarily those of the Foundation. Contributions from Meta are self-funded.
Methodology
This research used several large-scale datasets to investigate social capital and its relationship with economic mobility and well-being in the United Kingdom. The primary dataset comprised friendship connections from approximately 20 million anonymised Facebook users in the UK, representing around 58% of the population aged 25–64. This rich network data enabled the construction of measures of economic connectedness at the local authority and postcode district level, quantifying the prevalence of friendships between individuals of different socioeconomic statuses (SES). SES was estimated based on anonymised and aggregated Facebook data. The key variables constructed using Facebook data referenced in this release are:
- Economic connectedness: The average percentage of friends from higher-income backgrounds (above median income) in the social networks of people from lower-income backgrounds (below median income) in a specific area. This can also be understood as a form of bridging social capital.
- Exposure: The average percentage of higher-income people (above median income) that lower-income people (below median income) encounter in a particular setting (e.g., neighbourhood, school, hobby group).
- Friending bias: A measure of homophily that describes the tendency of people to form friendships with others of similar income levels, measured as the difference between actual friendship rates across income groups and what would be expected based on exposure alone in a particular setting.
To assess economic mobility, the study used longitudinal earnings records from the UK Longitudinal Education Outcomes (LEO) dataset for England. This allowed for the estimation of upward mobility by tracking the adult income rank of children from low-income families in different areas. To extend mobility analysis beyond England, a methodology was developed using anonymised Facebook data, allowing for the estimation of intergenerational mobility across the entire UK, validated against the LEO data in England.
Finally, the research incorporated survey data on subjective well-being collected from almost 4,000 adult Facebook users in the UK. A survey gathered information on happiness, life satisfaction, trust, and social support, enabling the examination of the association between these factors and individuals’ social connections, particularly the proportion of high-income friends.
The analysis involved merging these distinct datasets at both aggregate (local authority) and individual levels, employing statistical methods such as regression analysis to identify correlations and patterns between social capital metrics, economic mobility outcomes, and subjective well-being indicators.
About BIT
BIT (the Behavioural Insights Team) is a global research and innovation consultancy which combines a deep understanding of human behaviour with evidence-led problem solving to improve people’s lives. BIT works with all levels of government, nonprofits and the private sector, applying behavioural science expertise with robust evaluation and data to help clients achieve their goals.
BIT has more than 200 staff, and operates from seven offices around the world. Together our staff provide unrivalled behavioural science expertise, amassed through the delivery of more than 1,700 projects across hundreds of countries.
BIT was established by the UK government in 2010. In 2014 BIT became an independent social purpose company, part owned by the Cabinet Office and the innovation charity Nesta. Since 2021 BIT has been entirely owned by Nesta and is part of the Nesta group.
For more information visit bi.team and follow BIT on LinkedIn.
About the RSA
We are the RSA, the royal society for arts, manufactures and commerce. We are a social impact charity that has been at the forefront of social change for 270 years. Through our Design for Life mission, we turn world-leading ideas into world-changing actions. We’re committed to a world where everyone can fulfil their potential and contribute to more resilient, rebalanced, and regenerative futures. Central to all our work is our global network of over 31,000 Fellows, who work collectively to enable people, places and the planet to flourish in harmony.
We invite you to be part of this mission by joining our fellowship community. Together, we’ll unite people and ideas in collective action to unlock opportunities and regenerate our world.
Find out more at thersa.org
