Unpacking the Pathways to Wellbeing: The Promise of Mediation Analysis in Wellbeing Economics
by Jasmine Kazantzis, Senior Economist and Data Analyst at State of Life
When using wellbeing economics to value social impact, we often focus on the what: what interventions improve life satisfaction, what activities make people happier, what policies deliver the greatest wellbeing benefits. But another interesting and important question is how.
How exactly does an intervention improve wellbeing? What are the pathways through which it works? Does it foster a greater sense of belonging? Build trust? Reduce loneliness? Improve health? Or does it strengthen the relationships that give life meaning?
At State of Life, we’ve been exploring these questions through a recent piece of work for the Belong Network, the UK’s leading membership organisation on social cohesion. We applied mediation analysis to unpack how participation in the community and voluntary activities offered by some of Belong’s member networks (e.g. grassroots sports teams, community theatre initiatives, youth groups), leads to improvements in wellbeing.
The findings hint at something both intuitively familiar and empirically powerful: that the wellbeing benefits don’t just come from doing the activity but also from improvements in feelings of belonging.
What is mediation analysis and why does it matter?
Mediation analysis is a statistical technique that helps identify the mechanisms through which one variable influences another. Instead of simply showing that community participation is linked to higher wellbeing, mediation analysis asks how. It breaks down the overall effect into direct and indirect pathways, allowing us to see how much of the impact flows through key channels such as trust or belonging.
This approach has been widely used in the health and social sciences to uncover hidden mechanisms (for example, whether a treatment improves outcomes through behavioural changes). But it has not been applied in wellbeing economics - until now.
By applying mediation analysis to data from Understanding Society, the UK’s largest household panel study, we were able to quantify not just whether community activities are associated with higher life satisfaction, but also whether those benefits work through feelings of trust and belonging in this case.
What we found: belonging as a powerful pathway
Across activities such as community groups, social/working men’s clubs, and sports clubs, participation was linked to higher life satisfaction. Crucially, much of that benefit appears to flow through belonging.
For example, belonging explained 40% of the wellbeing impact of participating in sports clubs and around 30% of volunteering. Trust also played a role, but a smaller and less consistent one.
In short: participation matters, but feeling that you belong seems to be what makes it matter most.
Why this matters for wellbeing economics
Wellbeing economics is increasingly central to how we think about social value and whether policies work. But so far, much of the focus has been on measuring high-level outcomes, such as life satisfaction, happiness, or anxiety, rather than the processes that lead to them.
Mediation analysis opens up new possibilities. It allows us to unpack the black box between intervention and impact, revealing why certain policies or programmes work better than others.
Imagine being able to show that an employment programme improves wellbeing not just because people find jobs, but because those jobs reduce loneliness foster and increase trust. Or that a youth sports initiative enhances life satisfaction largely because it improves mental health, reduces loneliness and builds identity.
This kind of evidence doesn’t just make for better evaluation - it makes for better design. By identifying the active ingredients of wellbeing, we can target interventions more effectively, replicate what works, and tell richer, more human stories about social value. And because these pathways may differ for different people and communities, mediation analysis could also help tailor approaches to what matters most in different contexts.
What are some of the limits?
While mediation analysis offers exciting potential, it comes with important caveats.
First, it can reveal pathways but not prove causation. Our analysis used cross-sectional data, meaning we can’t yet say whether belonging causes greater wellbeing, or whether happier, more connected people are simply more likely to take part. Longitudinal data would help untangle this.
Second, mediation analysis assumes that all relevant variables have been measured. In reality, factors like personality, local environment, or genetics might also shape both belonging and wellbeing, but aren’t asked about in the data and cannot be adjusted for, creating hidden biases.
Where next?
Our work for the Belong Network is an early step, a proof of concept rather than definitive evidence. The next frontier is to apply mediation analysis with longitudinal data to better identify causal pathways and to widen the lens beyond belonging and trust to include loneliness, health, purpose and more. Future projects could also collect primary data tailored specifically to explore these mechanisms, allowing us to ask the right questions, measure the right things, and understand wellbeing change in greater depth.
The wellbeing economics research so far tells us what matters most for wellbeing, but mediation analysis could get us one step closer to understanding how this happens, and that insight holds real potential for better policy and practice.