Unfortunately, your browser is too old to work on this website. Please upgrade your browser
Analysis

Socioeconomic disadvantage and self-reported health
Changes among the working-age population over time

Published 9 March 2026
Time to read clock icon About 15 mins
Authors
Topics

Key points

  • Good health underpins our ability to work and lead fulfilling lives. Yet the UK faces a growing working-age health challenge and widening health inequalities.
  • This analysis, underpinned by modelling from the Office for National Statistics, is the largest study of its kind to show the links between self-reported health and socioeconomic deprivation over time (between 2011 and 2021).
  • We found where people live affected the rate of decline in good health. 1 in 5 people (18%) of working age who lived in the most deprived areas, no longer reported good health a decade later (compared with 13% overall).
  • Working-age people (16–69 years) living in the most deprived areas faced an at least 43% higher risk of no longer reporting good health, compared with those in the least deprived areas after accounting for demographic and socioeconomic differences.
  • Females aged 20–24 years in the most deprived areas in 2011 had a 72% higher risk of no longer reporting good health a decade later compared with those in the least deprived areas (70% higher for males).
  • Our analysis shows the main socioeconomic factors linked to people no longer reporting good health are unemployment, being economically inactive and living in the private or social-rented sector. These factors are also associated with reduced likelihood of returning to good health. For example, people who were unemployed but looking for work in 2011 had a 67% (female) and 82% (male) higher likelihood of no longer being in good health a decade later, compared with their peers in employment.  
  • Our analysis adds to evidence on the need for a preventative approach that helps keep people healthy, so they can fulfil their individual and economic potential. Policy action should focus on preventing people from leaving the labour market and on improving housing quality and security.
 

Introduction

Good health is a critical asset for society, underlying our ability to work, lead fulfilling lives and participate in family and community life. Yet the UK faces a growing challenge: life expectancy improvements have stalled, health inequalities are widening and working-age health has deteriorated significantly compared with peer countries.

It is well understood that our health generally deteriorates with age. But this tends to happen faster for people who experience socioeconomic disadvantage over their lifetimes. This novel analysis links the latest census data from 2021 with that from 2011, allowing us to look at how people’s health changed over the decade. 

This is important given growing evidence of the consequences of poor working-age health. In the UK, 4 million people have a work-limiting health condition and are not employed. Since the COVID-19 pandemic, this has increased across all age groups, with older age groups reporting higher rates of work-limiting health conditions. Our Commission for Healthier Working Lives looked at how health and employment influence one another and included recommendations for action to improve working-age health. 

A safe, independent and fulfilling life is shaped by different factors, including our health and how we perceive it. Our own perception can sometimes give a better picture of how our health shapes our lives than how medical professionals define our health. The impact of waiting for a diagnosis or struggling to manage a health condition is not easy to define or capture in health records. 

In this analysis, we look at changes from 2011 to 2021 to understand which socioeconomic factors were associated with the working-age population no longer reporting good health. The findings help us to understand factors related to self-reported health and inequalities within this. This could inform policy interventions to improve health and where to target these, given the established link between self-reported health and population-wide morbidity and mortality measures. This is particularly important given health in the UK ranks poorly against other high-income countries. 

 

Trends in the population’s self-reported health

Self-reported health is an important measure of the public’s health. It is a subjective measure based on individuals evaluating their own health status. It can be used to predict morbidity and mortality at a population level, as well as levels of economic inactivity. Self-reported health data are not routinely collected as part of NHS health records. But self-reported health is often a feature of surveys looking at changes in population health, which go beyond defining health by the diagnosis of health conditions. Health conditions can affect people in different ways, including how well they feel able to manage their conditions and the support they can access. 

Self-reported health is used to calculate healthy life expectancy – a measure that indicates how many years people might spend in poor health towards the end of life. Healthy life expectancy is often used to look at how the health of populations change over time, as well as how they compare to each other. Health can also change throughout our lifetimes, meaning people can spend periods of life in poor health and then return to good health. However, once people report having transitioned to poorer health, their chances of reporting a return to good health are low. Therefore, it is important policymakers understand the factors behind this transition so that it might be prevented. The census gives us a unique chance to see how everyone in England views their health, something we usually estimate from surveys with much lower participation. 

People’s perceptions of their health change throughout their lifetimes. Older adults are more likely to have health conditions and/or be disabled, which can change how people view their health. In the 2021 Census, 91% of 20 to 39-year-olds reported good or very good health, compared with 56% of those aged 70 to 89. Overall, 1% more people reported good health than in 2011 – a surprising result against the backdrop of the COVID-19 pandemic and an older population across England (Figure 1). 

Figure 1

When we compared age groups, we found fewer younger working-age people (16 to 40-years old) reported being in good health in 2021 compared with 2011. In contrast, older age groups showed the opposite trend. While we might expect the latter in a society where longevity is improving for successive cohorts, fewer younger working-age people reporting good health is a significant concern. This decline could be related to the crisis in young people’s mental health. There are several possible explanations for the relative improvement among older age groups, with research suggesting the observed changes may be related to how people viewed their health in the context of the pandemic. 

In this analysis, we looked at changes from 2011 to 2021 to understand which socioeconomic factors were associated with the working-age population no longer reporting good health. The findings help us to understand factors related to self-reported health, and inequalities within this. This could inform policy interventions to improve health, and where to target these given the close link between self-reported health and population-wide morbidity and mortality measures. This is particularly important given health in the UK ranks poorly against other high-income countries. 

Box 1:

The data underpinning this analysis were produced by the Office for National Statistics (ONS) on behalf of the Health Foundation. Relative risk estimates were produced by the authors using these data. 

This research looks at data from 2011 Census results and links them to 2021 results and death records to understand how people’s health changed over that time period. The analysis includes people living in England who were usual residents, aged 16 to 90 years in 2011 and completed both the 2011 and 2021 censuses; or completed the 2011 Census and had died before Census 2021. People living in communal establishments (care homes, hospitals, prisons, student halls of residence) were excluded. Full details on linkage and methods can be found in the ONS publication. 

In each census, people were asked to give an assessment of the overall state of their health (‘How is your health in general?’), from ‘very good’ to ‘very bad’ across five categories. Here, these are summarised as ‘good health’ (‘very good’, ‘good’) and ‘not good health’ (‘fair’, ‘bad’ and ‘very bad’). The term ‘poorer health’ is used to describe ‘not good health’ in discussion of results. The term ‘no longer reporting good health’ refers to the transition from self-reported good health in 2011 to self-reported not good health in 2021. This is distinct from declining health, which could occur within the two categories (eg moving from ‘fair’ to ‘very bad’ health would see someone remain within the ‘not good health’ category). 

The analysis combined data from the 2011 Census about each person’s demographic characteristics (age, sex, ethnicity, region) and socioeconomic characteristics (household tenure and composition, employment type, highest qualification, area-level deprivation). It also included how they viewed their health in 2011 and 2021, along with death records for those who had died before March 2021. These data were used to build statistical models which adjusted for differences in these demographic and socioeconomic factors. This led to estimates of the relative likelihood of reporting good health, not good health or having died by March 2021. This was compared against reference groups for each of those who were in good or not good health in 2011. Reference groups were selected either as the most frequent category or to provide a clear baseline for comparison.  

This analysis focuses on the transition from good to not good health in the working-age population and includes additional analysis building on the findings from the ONS. The full dataset is available on the ONS website

The analysis adjusts for several demographic and socioeconomic factors related to self-reported health, but the estimates might have been different if it was possible to account for a larger number of factors. For example, exposure to common risk factors – such as smoking, alcohol consumption or sedentary behaviours – or access and experience of health care services were not included due to data availability. These factors affect someone’s health and how they perceive it. 

The data used to conduct this study included all people living in England counted in the 2011 Census. However, some groups are still more likely to be underrepresented, such as those experiencing homelessness. In addition, people who moved out of England between 2011 and 2021 will not be captured here, as well as people who moved to England after 2011. The self-reported health trends among these groups may differ from the wider population. 

The 2021 Census was completed during the gradual lifting of restrictions following the second national lockdown of the COVID-19 pandemic. Some groups were more likely to have been displaced on census day, meaning some information collected for them was not representative of their ‘normal’ circumstances. While the pandemic may have affected how people perceived their health, exploration of these self-reported health data compared to reporting of disability show that completion in 2011 and 2021 were similar. 

 

How different factors affect the transition to poorer health

Age

As people age, fewer of us remain in good health – as shown by the blue bar on the left-hand side of each panel in Figure 2. This aligns with what we know about the development of long-term conditions being related to ageing, with earlier onset for people living in more deprived areas. Therefore, we would expect to see some of those reporting good health in 2011 to report being in poorer (not good) health a decade later. This is shown by the flow from the blue bar on the left changing to the purple bar on the right. We see this across all age groups, but to a much larger extent for the group who were 50 to 69 years old in 2011. Around 1 in 20 (5%) of those aged 16 to 29 years old in 2011 no longer reported good health by 2021. This was higher (16%) for those who were age 50 to 69 years in 2011. 

Figure 2

Some people who reported being in good health in 2011 had died by 2021 (the flow from the blue bar on the left to the red bar on the right). This analysis does not look at inequalities in transitions to death. We have looked at variation in premature mortality between local areas in a previous analysis. 

Levels of deprivation

The proportion of people in 2011 who no longer reported being in good health a decade later increased with age. Age remains a major factor even after accounting for socioeconomic differences. While the trends across the life course are similar across all levels of deprivation, there are pronounced inequalities in the risk of no longer reporting good health.  

Figure 3 shows the percentage of people who no longer reported good health a decade later across each age group in each level of deprivation. Starting with females aged 16 to 19 in 2011, 12% of those living in the most deprived areas of England no longer reported good health, compared with 7% of those living in the least deprived areas. This equates to a 69% higher risk for those in the most deprived areas, compared with those in the least deprived areas (Figure 4). The higher likelihood of those living in the most deprived areas no longer reporting good health is seen across all working-age groups, with at least a 43% higher risk of no longer reporting good health than those in the least deprived areas. These results account for differences in ethnicity, household tenure and composition, and employment. This helps to make sure the reported variation reflects true differences in these transitions, rather than group characteristics. This inequality is more pronounced for females than males in older working-age groups. 

Figure 3

We see a larger absolute difference in the proportion of people who no longer report good health at older age groups, compared with younger age groups. For example, 10% of females aged 20–24 years in the most deprived areas no longer reported good health a decade later, 4 percentage points lower than their peers living in the least deprived areas. The percentage point difference widens in older age groups (eg a 9 percentage-point difference for females aged 65–69 years). This highlights the cumulative impact of the experiences people have in more deprived areas. The difference is especially large among those aged 50 to 69 years, meaning that people living in the most deprived areas are at much greater risk of being unable to maintain good health to state pension age.

Relative differences in the risk of no longer reporting good health are highest among younger age groups, who generally have higher levels of good health. While the percentage who make this transition is smaller for younger age groups, the differences are more marked (Figure 4): 

  • Females aged 20 to 24 years living in the most deprived areas had a 72% higher risk of no longer reporting good health. Males in the same group had a 70% higher risk compared with counterparts in the least deprived areas.
  • For 40 to 44 year olds, the risk was 63% higher for females and 61% higher for males, falling to 54% (females) and 48% (males) at ages 60 to 64. 

Figure 4

At older ages, the relative difference in the risk of no longer reporting good health gets smaller between those living in the most and least deprived areas. This is because the overall proportion of people no longer reporting good health makes up a larger share of these groups. Inequalities remain to the same extent when including people who either transitioned to poorer health or who died between 2011 and 2021. 

Females aged 55 to 59 years in 2011 who lived in the most deprived areas had a 57% higher risk of no longer reporting good health a decade later, compared with those in the least deprived areas. This age group were 65 to 69 years in 2021, the period during which people reach state pension age. The risk was slightly lower for males, who had a 51% higher risk. This shows that people living in the most deprived areas are at much greater risk of being unable to maintain good health to state pension age.

This inequality is a concern for the planned rise in state pension age (which will increase to 67 from April 2026) and future potential rises. Any further increase risks widening existing inequalities if people are unable to work due to ill health, relying on the less generous working-age welfare system.

Building blocks of health

The building blocks of health are the social and economic factors that influence health in our day-to-day lives. This analysis compares the role of different socioeconomic factors in people’s risk of no longer reporting good health. The most significant, explored below, are employment status and housing tenure.

Economic activity

We have looked at different types of economic activity: whether or not someone is in employment and, if not, for what reason. Here, we have focused on the health changes seen by people who were not in employment in 2011. These are categorised as ‘unemployed’ (looking for work and could start in the next 2 weeks or waiting for a job to start) or ‘economically inactive’ (not in employment and either not looking for work or not able to start work within 2 weeks). Economic inactivity is further categorised by reason: ‘long-term sick or disabled’; studying; retired; and ‘looking after home or family, or other reason’. Results for those studying or retired are not discussed but are available in the ONS data release. 

People in employment were chosen as the reference group because they are most likely to maintain good health given the positive role employment can play for health. We accounted for demographic factors (age group, sex, ethnicity and region) and socioeconomic factors (area-level deprivation, highest qualifications and household tenure and composition). Figure 5 shows:

  • Females and males who were unemployed but seeking work in 2011 had a 67% and 82% higher respective likelihood of no longer reporting good health a decade later, compared with those in employment, and after accounting for demographic and socioeconomic factors.
  • Females and males who were economically inactive due to looking after family or home or for other reasons in 2011 had a 50% and 82% higher respective likelihood of no longer reporting good health a decade later, compared with those in employment.
  • Females who were not working due to long-term sickness or being disabled in 2011 had a 272% higher likelihood of no longer reporting good health a decade later compared with females in employment. The equivalent group of males had a 279% higher likelihood of no longer reporting good health than their counterparts. 

Figure 5

Being unemployed or economically inactive was associated with changes in how people perceived their health over time. People who viewed themselves to be in good health and were out of work due to long-term sickness or being disabled, had a much higher relative likelihood of no longer reporting good health a decade later than those who were out of work to look after home/family, or for another reason. In 2011, 159,000 people reported being in good health and were out of work due to long-term sickness or being disabled, with 64,000 of this group no longer in good health by 2021. For comparison, among those who were unemployed in 2011 962,000 reported being in good health in 2011, with 199,000 no longer reporting a decade later. 

This is a significant concern given the around 35% increase in health-related economic inactivity between 2019 and 2024. Furthermore, future growth in health-related economic inactivity could be three times greater than growth in the working-age population as a result of an ageing population and increasing multimorbidity. As set out by the Commission for Healthier Working Lives, policy efforts must take a preventative approach to maintain people’s health and employment – particularly as people’s chances of returning to work diminish the longer they are out of work, and further limited by a subsequent decline in health. 

This analysis only considers employment status in 2011 and does not track changes over the decade. Our findings, however, highlight the need to act early to keep people in employment as a preventative measure to protect their health. 

Housing tenure

Living in social or private-rented housing in 2011 was associated with a higher likelihood of no longer reporting good health by 2021 (Figure 6). This was even after accounting for the same demographic and socioeconomic factors as before, although this time accounting for economic activity status instead of household tenure. We compared the likelihood of no longer reporting good health among people who lived in private or social-rented housing with those who lived in owned housing as this is the most frequent type.

Figure 6

Females and males who lived in private rented housing had a 42% and 34% higher respective likelihood of no longer reporting good health, compared with those in owned housing. This difference increases substantially when we compare people living in social-rented housing to those in owned housing. For females, the difference was 92% while for males it was 75%. 

Differences may reflect the contrasting experiences of rented versus owned housing. Insecurity of tenure in the private-rented sector and higher risks of overcrowding in both social and private-rented housing may contribute to the higher likelihood observed. 

Other wealth and income-related differences not fully captured by deprivation or employment type may also affect these findings. Affordability of housing, measured as the share of income spent on housing, varies across tenures. This analysis only includes people who reported being in good health in 2011. As a result, relatively few disabled people are included since being disabled is closely linked to reporting poorer health. However, some disabled people may have reported good health at the time. Disabled people are also more likely to live in social-rented housing. This suggests that the higher likelihood in no longer reporting good health seen for people living in social rented housing might come from the effects of being disabled, rather than housing tenure itself.

 

Conclusion

This analysis provides new insight into how working-age people transition out of good health over time. The findings show stark heath inequalities between the least and most deprived areas that persist and widen across the life course. Our analysis also compares how the role of different socioeconomic factors affect people’s risk of no longer reporting good health – the most significant of those we looked at being employment status and housing tenure.

Differences in self-reported health across deprivation levels raise concerns for state-pension age policy. Among females aged 55 to 59 years in 2011, the risk of no longer reporting good health a decade later is 57% higher for those living in the most deprived areas compared with the least deprived areas and 51% higher for males. This suggests many people in disadvantaged areas are at greater risk of being unable to maintain good health to a higher retirement age. This may limit people’s ability to work, which could lead to living in poverty in the years before reaching state-pension age. This could be addressed by a higher rate of Universal Credit payments to people unable to work to a higher state-pension age.

There is growing evidence on the impact of unemployment and economic inactivity on working-age health. Health inequalities among those not in work are often assumed to reflect people’s state of health having already worsened. Our research shows that people who were out of work and in good health in 2011 had a substantially higher likelihood of no longer reporting good health a decade later, compared with those in employment. This reinforces the need to keep people in work, where possible, through preventative measures, including more proactive sickness absence management and the introduction of vocational rehabilitation. This can have positive impacts on both people’s health and the economy. 

Housing plays an important role in health over time. The Renters’ Rights Act 2025 takes some important steps to improve the security of private-rented housing, but the government has a much harder task in delivering on its housebuilding programme to improve housing affordability. 

This research adds to existing evidence on employment, housing and area-level inequalities. It also highlights areas for further research on how the building blocks of health affect self-reported health. 

Acknowledgements 

The data underpinning this analytical work was produced by the ONS health research group on behalf of the Health Foundation. All interpretation of those data reported here are the work of the Health Foundation. 

Further reading

Related content

View all related content