Fragmented Neighbourhoods May Quietly Erode Wellbeing, Australian Study Finds

The streets where people live may influence how satisfied they feel with life, but a new study suggests that neighbourhood social fragmentation is only one piece of a much larger wellbeing puzzle. Researchers examining more than 2,000 adults in Australia’s Australian Capital Territory found that people living in more socially fragmented neighbourhoods were generally less likely to report exceptionally high subjective wellbeing. Yet the effect was modest, and much of the apparent neighbourhood difference disappeared after the researchers accounted for personal circumstances such as age, education and family finances. The findings, published in Social Indicators Research, offer a nuanced message for urban planners and public-health officials: communities matter, but individual resources and life-stage factors may matter even more.

Subjective wellbeing is increasingly used alongside economic measures such as income and gross domestic product to assess how populations are actually faring. Rather than measuring only material conditions, it captures how people evaluate their own lives, including their satisfaction with health, relationships, achievements, safety and future security. In this study, the researchers used two complementary measures. The first was Global Life Satisfaction, a single rating from 0 to 10 asking participants how satisfied they were with life as a whole. The second was the Personal Wellbeing Index, which combines satisfaction across seven domains: standard of living, health, achievements, personal relationships, community connectedness, personal safety and future security. Scores were converted to a 0-to-100 scale, with a score of 80 or above classified as “high” wellbeing.

To assess the social character of each neighbourhood, the team used the Australian Neighbourhood Social Fragmentation Index, or ANSFI. The index is not a survey of whether a particular resident feels lonely or trusts their neighbours. Instead, it is a structural measure calculated from census indicators using principal component analysis, a statistical method that compresses correlations among many variables into underlying patterns. The index incorporates signs of neighbourhood attachment, residential mobility and transience, and shared social norms. Areas with more short-term residents, privately rented homes, single-person or non-family households, recent immigrants and lower volunteering levels tend to receive higher fragmentation scores. More negative scores generally indicate greater stability and cohesion, while positive scores indicate more fragmented social structures.

The analysis drew on the 2019 “Living well in the ACT Region” survey, which included adults living in urban areas of the territory. Of 3,190 valid survey responses, 2,056 had the geocoded information and complete wellbeing and demographic data required for the study. The participants were not a perfect demographic mirror of the wider population: 42.1 percent were aged 65 or older, 52.5 percent were women, 70.8 percent were married or in a de facto relationship, and 61.5 percent held a university degree. The median ANSFI score was −0.82, suggesting that many participants lived in relatively cohesive areas. Overall, 44.3 percent of respondents met the threshold for high Personal Wellbeing Index scores, while 56.0 percent reported high Global Life Satisfaction.

The first statistical comparisons revealed a clear directional pattern. Participants classified as having high wellbeing tended to live in neighbourhoods with lower fragmentation scores than those who did not meet the high-wellbeing threshold. The pattern appeared for both wellbeing measures. In unadjusted multilevel logistic regression models, each one-unit increase in ANSFI was associated with lower odds of high Global Life Satisfaction, with an odds ratio of 0.947 and a 95 percent confidence interval from 0.898 to 0.998. For high Personal Wellbeing Index scores, the corresponding odds ratio was 0.928, with a 95 percent confidence interval from 0.879 to 0.980. In practical terms, the researchers estimated that a one-unit increase in fragmentation was linked to approximately a 7.2 percent reduction in the odds of high PWI, although the meaning of a one-unit change depends on the distribution and scale of the index.

The researchers then used multilevel logistic regression to separate personal influences from neighbourhood influences. This approach is designed for data in which individuals are nested within larger geographic areas, because people from the same neighbourhood may share environmental and social conditions. The models treated residents as clustered within Statistical Area Level 2 regions, which generally correspond to suburbs or similar areas containing roughly 2,000 to 25,000 people. The analysis progressively adjusted for age, gender, education, family financial status and area-level socioeconomic conditions, measured using Australia’s Index of Relative Socioeconomic Advantage and Disadvantage. After these adjustments, the association with Global Life Satisfaction was no longer statistically significant: the odds ratio was 0.955, with a 95 percent confidence interval of 0.899 to 1.016. The association with the Personal Wellbeing Index remained just statistically significant, although it was smaller, with an odds ratio of 0.941 and a confidence interval from 0.885 to 0.999.

Individual circumstances produced stronger and more consistent signals than neighbourhood fragmentation. Adults aged 65 and older had substantially higher odds of reporting high wellbeing than younger reference groups. In the fully adjusted models, the odds ratio for high Global Life Satisfaction among people aged at least 65 was 2.621, while the corresponding figure for high Personal Wellbeing Index scores was 2.772. Financial comfort was also strongly associated with wellbeing. Participants describing their family financial status as comfortable or prosperous had an odds ratio of 3.751 for high PWI, compared with those who were merely getting along or reasonably comfortable. These results do not mean that age or money automatically cause happiness, nor that neighbourhood conditions are irrelevant. They indicate that personal resources, security and life circumstances may account for a larger share of differences in reported wellbeing than the structural social composition of the area in which someone lives.

The amount of variation attributable specifically to neighbourhoods was surprisingly small. In an unadjusted model, neighbourhood differences accounted for about 1.1 percent of the variation in high Global Life Satisfaction and 1.6 percent of the variation in high Personal Wellbeing Index scores. Once individual characteristics were included, this contextual variation fell sharply. For PWI, the researchers reported a 96 percent proportional reduction in neighbourhood-level variance in the fully adjusted model; for GLS, adjustment for age and gender reduced the variance by about 52 percent, although the pattern changed after area socioeconomic status was added. The study also tested whether socioeconomic advantage altered the relationship between fragmentation and wellbeing. It did not find evidence of such moderation: the interaction between ANSFI and IRSAD was not statistically significant for either GLS or PWI. In other words, the apparent association between fragmentation and high wellbeing did not clearly become stronger or weaker in richer or poorer neighbourhoods.

The results do not establish that fragmented neighbourhoods cause lower wellbeing. The survey was cross-sectional, meaning that residents and neighbourhood characteristics were observed at roughly one point in time rather than followed for years. It is therefore impossible to determine whether fragmented environments reduce wellbeing, whether people with lower wellbeing are more likely to live in certain kinds of neighbourhoods, or whether both are shaped by other factors. The survey data were collected in 2019, whereas the ANSFI and IRSAD measures came from 2021 census-based data, creating a temporal mismatch that the authors acknowledge. The COVID-19 pandemic may also have changed mobility, social contact, employment and perceptions of community in ways not captured by structural census indicators. The wellbeing outcomes were converted into binary categories, which can conceal meaningful differences between people just below and just above the threshold of 80. Even so, supplementary analyses using wellbeing scores as continuous measures produced broadly similar conclusions. The study’s central message is therefore less dramatic than the idea that a neighbourhood determines happiness, but potentially more useful: strengthening local connections, volunteering, informal support and opportunities for social participation may modestly improve quality of life, while policies addressing financial security and individual disadvantage are likely to have a more powerful and measurable impact.