Key findings

This research examined characteristics associated with transitions into and out of employment between 2011 and 2021, with a particular focus on disability status, socio-economic conditions and household circumstances.

Entering employment (2011-2021) (Figure 2) and (Table 7).

Overall, the main factors associated with moving into employment were:

  • Being male.
  • Age 20-39 years.
  • Holding qualifications, in particular a degree.
  • Being unemployed or in the other economically inactive category in 2011.
  • Having reported a disability in 2011 but not in 2021.
  • Providing lower levels of unpaid care (1-19 hours per week).

Among females, being separated or divorced and living in northern and eastern areas of Belfast Local Government District were associated with a move into employment.

Leaving employment (2011-2021) (Figure 4) and (Table 8).

Overall, the main factors associated with leaving employment were:

  • Being female.
  • Reporting disability, particularly persistent disability (reported in both 2011 and 2021) or disability acquired during this period.
  • Self-reported poor mental health.
  • Providing high levels of unpaid care (35 or more hours per week).
  • Socio-economic disadvantage, including living in social rented accommodation and in lone-parent households.
  • Age 40-49 years.

Among males, leaving employment was associated with:

  • Being single.
  • Being separated or divorced.
  • Reporting neurodiversity.
  • Having a physical health condition.
  • Living in urban areas of Derry City and Strabane Local Government District or in western and northern areas of Belfast Local Government District.

Among females, having a child with a disability in the household or having three or more dependent children was associated with leaving employment.

Variable definitions

Age and labour market status refer to the baseline characteristics in 2011. Disability status captures changes between 2011 and 2021. All other characteristics reflect status in 2021 and circumstances associated with employment outcomes at the end of this period. The leaving employment analysis excludes people who left due to retirement. Definitions of health, socio-economic, household and area variables are provided in Tables 1 and 2.


Background

Improving labour market participation and reducing economic inactivity are longstanding policy priorities in Northern Ireland (see reference 1). Effective policy development can be enhanced by gaining a greater understanding of the characteristics associated with transitions into and out of work over time.

People with disabilities and those with long-term health conditions continue to experience poor labour market outcomes (see reference 2). Having a disability remains one of the strongest predictors of not being in employment. In 2021, only 38% of people with disabilities were in employment in Northern Ireland, compared with 86% of people without disabilities, resulting in a disability employment gap of 48 percentage points (see reference 3). People with disabilities are also likely to leave employment once in work, highlighting barriers to both labour market entry and retention (see reference 4).

This analysis examined employment transitions between 2011 and 2021, focusing on those aged 20-49 in 2011.

This work builds on an earlier NISRA policy brief (see reference 5), which presented high-level findings on disability and employment transitions using the same linked data. This report extends that work by:

  • providing a more detailed profile of those moving into and out of employment including the role of specific health conditions (such as poor mental health and neurodiversity) and changes in disability status over time; and
  • applying modelling techniques to identify socio-economic, household and area-level characteristics associated with labour market transitions.
    Two policy briefs summarising the report’s findings, on transitions into and out of employment, are available on the NISRA-led research website.

Note 1

Percentages and percentage changes have been rounded to the nearest whole number to aid readability. As a result, percentages shown in tables and charts may not always sum to exactly 100 per cent. For the modelling analysis (commentary relating to Figures 2 and 4 and the data in Tables 7 and 8), results are presented to two decimal places to maintain precision, while associated percentage differences are rounded to the nearest whole number (for example, an odds ratio (modelling result) of 1.75 indicates being 75% more likely to move into employment between 2011 and 2021).

Note 2
Figures are not directly comparable with the previous policy brief, as the age ranges differ. This report focuses on individuals aged 20-49 years in 2011, whereas the policy brief examined those aged 25-54 years in 2011.


Study overview

The analysis used the Northern Ireland Longitudinal study (NILS), a high-quality linked dataset combining Census data from 2011 and 2021, to examine employment transitions. In the linked 2011-2021 dataset, the sample of individuals aged 20-49 in 2011 represented 22% of the 2011 Northern Ireland population (167,784 out of 755,704) (see reference 6).

Key strengths of NILS include:

  1. Longitudinal design enabling analysis of changes in employment, and self-reported disability and health across ten-year intervals.
  2. Large representative sample (around 28% of the Northern Ireland population per Census year) allowing detailed subgroup analysis, including by disability and other Section 75 equality groups.
  3. Household-level linkage, supporting analyses incorporating area context, living arrangements and shared socio-economic characteristics.

Study population

The study focused on people aged 20-49 years in 2011. This age range was chosen to exclude most individuals still in full-time education and to minimise the impact of planned retirement on labour market exits.

  • Students in 2011 were excluded from the baseline population.
  • Students in 2021 were included in the economically active (employed) group, reflecting engagement in work or study.
  • Retired individuals (those retired in both 2011 and 2021, or who entered retirement between 2011-2021) were excluded, with the exception of a small number who were retired in 2011 but returned to employment or reported a different form of economic inactivity in 2021.

Transitions into employment (2011-2021)

(i) Descriptive characteristics: who moves?

Full results are presented in Annex 2, Tables 3 and 4, with separate breakdowns by sex provided in Annex 3, Tables 5 and 6. See Tables 1 and 2 for definitions of health, socio-economic, household and area variables.

In the commentary below, the comparison group is those who were not employed in both 2011 and 2021.

Sex, age and marital status

  • Those who moved into employment were more likely to be male (44% vs 37%) and married (47% vs 33%).
  • Among males aged 20–29, 45% entered employment between 2011 and 2021, compared with 24% who were not employed in both years.
  • Among females, higher proportions entered employment than remained out of employment in the 20–29 (33% vs 23%) and 30–39 (37% vs 30%) age groups.

Education and health

Those who moved into employment were:

  • More likely to hold intermediate-level (54% vs 42%) or degree-level (28% vs 11%) qualifications.
  • More likely to not report a disability in either 2011 or 2021 (73% vs 29%).
  • Less likely to report a health condition in 2021 including poor mental health (9% vs 41%), neurodiversity (3% vs 8%), and a physical health condition (21% vs 56%).

Other
Those who moved into employment were more likely to:

  • Live in owner-occupied housing (61% vs 37%).
  • Have no unpaid caring responsibilities (83% vs 77%).
  • Live outside Belfast and Derry City and Strabane Local Government Districts (73% vs 65%).
  • (For females) live in households where no children had a disability (87% vs 76%).

(ii) Where do they move from?

People who were not employed in 2011 were outside of the labour market for different reasons. Figure 1 shows the proportion of males and females moving into employment between 2011 and 2021, by economic activity group in 2011.

Figure 1: Proportion of males and females moving into employment (2011-2021), by economic inactivity group in 2011.

Figure 1 shows the proportion of males and females moving into employment between 2011 and 2021 broken down byr economic activity group in 2011

Figure 1: Interpretation and note

Figure 1 shows, for example, of the 6,346 unemployed males in 2011 63% were in employment ten years later.

Note: Figure 1 is based on a 22% population sample, see study overview.


Among those who were not employed in 2011 but were employed in 2021, distinct patterns emerged by sex:

  • Among males, the largest group was unemployed (63%), followed by other forms of economic inactivity (49%), looking after home or family (34%) and long-term sick or disabled (19%).
  • Among females, unemployment was again the largest group (56%) but looking after home or family (41%) had similar levels to other economically inactive (42%). Long-term sick or disabled (17%) was again the smallest group.

The results show clear differences in employment trajectories depending on individuals’ economic activity status in 2011, with notable variations between males and females. In particular, the comparatively higher number of females who were looking after the home or family in 2011 highlights the important influence of caring responsibilities on female’s subsequent employment transitions.

(iii) Modelling moving into employment, 2011 to 2021

Modelling analysis was undertaken to estimate the characteristics associated with moving into employment between 2011 and 2021, while accounting for all other characteristics included in the model.

Models adjusted for:

  • Census 2011 characteristics: age; economic inactivity category (e.g, unemployed or looking after home or family).
  • Changes in disability status from 2011 to 2021.
  • Census 2021 characteristics: sex, educational attainment, caring responsibilities, living arrangements, number of dependent children, housing tenure, Local Government District, and health conditions.

Figure 2 presents the results as odds ratios, highlighting the characteristics most strongly associated with entering employment between 2011 and 2021, each compared with a defined reference group. Full modelling results are presented in Annex 4, Table 7, with further details on the modelling approach provided in Annex 1.

Figure 2: Groups most likely to enter employment, 2011-2021 (aged 20-49 years in 2011)

Figure 2 shows which groups had the highest likelihood of entering employment between 2011 and 2021.

Note: The ‘rest of Northern Ireland’ refers to areas outside of Belfast and Derry City and Strabane Local Government Districts.

Figure 2: Note and interpretation

Odds ratios (ORs) above show how much more likely a group is to enter employment between 2011 and 2021, compared to a reference group, after adjusting for other factors. For example, those with a degree
(OR = 3.95) were 3.95 times more likely to move into employment between 2011 and 2021, compared to those with no qualifications.


The commentary below refers to results in Figure 2 above and sex-specific results presented in Annex 4, Table 7.

Factors strongly associated with moving into employment between 2011 and 2021

Qualifications

Compared to individuals with no qualifications:

  • Those with a degree were 3.95 times more likely to enter employment (OR = 3.95).
  • Those with intermediate level qualifications were 2.34 times more likely (OR = 2.34).

Sex and age

  • Males were 33% more likely to enter employment than females (OR = 1.33).
  • Individuals aged 20-29 years were 2.11 (OR = 2.11) times more likely, and those aged 30-39 years were 1.60 times more likely (OR = 1.60) to move into employment than those aged 40-49 years.

Economic activity status in 2011

Compared with individuals who were long-term sick or disabled in 2011:

  • Those unemployed in 2011 were over two times more likely to be in employment in 2021 (OR = 2.04).
  • Those classified as ‘other economically inactive’ in 2011 were 1.54 times more likely to be in employment in 2021 (OR = 1.54).

Caring responsibilities

  • Providing 1-19 hours of unpaid care per week in 2021 (compared with having no unpaid care responsibilities) was associated with a 37% greater likelihood (OR = 1.37) of moving into employment between 2011 and 2021.

Disability change

  • Having a disability in 2011 but not 2021 (compared with having no disability in either year) was associated with a 31% greater likelihood of entering employment (OR = 1.31).

Female-specific factors
When the models were run separately for males and females, notable differences emerged (see Annex 4, Table 7):

  • Females who were divorced or separated (compared with married), were 40% more likely to enter employment (OR = 1.40).
  • Females living in northern (31%: OR 1.31) and eastern areas (20%: OR = 1.20) of Belfast Local Government District had a higher likelihood of entering employment compared to the rest of Northern Ireland.


The findings should be interpreted as associations rather than causal relationships and are intended to inform understanding of broader patterns in movements into employment over time.

Transitions out of employment (2011-2021)

(i) Descriptive characteristics: who moves?

Full results are presented in Annex 2, Tables 3 and 4, with separate breakdowns by sex provided in Annex 3, Tables 5 and 6. See (Tables 1 and 2) for definitions of health, socio-economic, household and area variables.

In the commentary below, the comparison group is those who were employed in both 2011 and 2021.

Sex, age and marital status

  • Compared with those who were in employment in both 2011 and 2021, those who left employment by 2021 were more likely to be female (62% vs 50%) with a higher proportion aged 40-49 years (39% vs 34%).
  • Those leaving employment were more often single (37% vs 24%) or divorced/separated (19% vs 11%).

Education and health Those who left employment were more likely to:

  • Have no formal qualifications (27% vs 9%).
  • Report a disability in both 2011 and 2021 (13% vs 2%).
  • Have acquired a disability between 2011 and 2021 (36% vs 9%).
  • Report poor mental health either alone or in combination with other health conditions (29% vs 5%).
  • Have a physical health condition (44% vs 17%).

Other
Those who left employment were more likely to:

  • Live in private-rented (21% vs 11%) or social-rented housing (23% vs 5%).
  • Live in a lone-parent household (22% vs 10%).
  • Provide a high level of unpaid care (35+ hours per week: 14% vs 5%).
  • Live in the 20% most deprived areas within Northern Ireland (23% vs 16%).
  • Live in urban areas (69% vs 63%).

Sex-specific patterns (Annex 3, Tables 6) were also evident:

  • Males who left employment were more likely to have no dependent children in the household (41% vs 27%), live in a lone-parent household (14% vs 5%) or be living alone or living other, e.g. unrelated individuals living together, such as young professionals (36% vs 16%).
  • Females who left employment were more likely to have three or more dependent children in the household (16% vs 11%) or have a child with a disability (18% vs 7%).

(ii) Where do they move to?

Figure 3 shows the economic activity status by sex for those leaving employment between 2011 and 2021 (excluding retirees).

Figure 3: Economic activity status in 2021 (%), among those who left employment (2011-2021), by sex

Figure 3 shows the 2021 Economic activity status for males and females among those who left employment between 2011-2021.

Note: Figure 3 is based on a 22% population sample, see study overview.

Among those who were employed in 2011 but were not employed in 2021 (after excluding retirees):

  • Among males, the largest group was long-term sick or disabled (40%), followed by other forms of economic inactivity (26%) and unemployed (19%).
  • Among females, the largest group was looking after home or family (45%), followed by long-term sick or disabled (32%) and other economically inactive (15%).

(iii) Modelling leaving employment, 2011-2021.

Modelling analysis was undertaken to estimate the characteristics associated with leaving employment between 2011 and 2021. The key advantage of the modelling approach is its ability to assess the effect of multiple characteristics simultaneously on the likelihood of leaving employment.

The models adjusted for:

  • 2011 characteristics: age.
  • Changes in disability status from 2011 to 2021.
  • 2021 characteristics: sex, educational attainment, caring responsibilities, living arrangements, number of dependent children, housing tenure, Local Government District, and health conditions.

Figure 4 presents the results as odds ratios, highlighting the characteristics most strongly associated with leaving employment between 2011 and 2021, each compared to a defined reference group. Full modelling results are presented in Annex 4, Table 8, with further details on the modelling approach provided in Annex 1.


Figure 4: Groups most likely to leave employment, 2011-2021 (aged 20-49 years in 2011)

Figure 4 shows which groups had the highest likelihood of leaving employment between 2011 and 2021.

Figure 4: Note and interpretation

Odds ratios (ORs) in Figure 4 show how much more likely a group is to leave employment between 2011 and 2021 compared to a reference group, after adjusting for other factors. For example, individuals who acquired a disability between 2011 and 2021 (OR = 4.08) were 4.08 times more likely to leave employment between 2011 and 2021 compared to those with no disability in either year.


The commentary below refers to results in Figure 4 above and sex-specific results presented in Table 8.

Factors strongly associated with leaving employment between 2011 and 2021

Overall, females were more likely than males to leave employment (OR: 1.69, +69%).

Health conditions in 2021 and disability change, 2011-2021

The likelihood of leaving employment was higher among those who reported poor mental health alongside other conditions (OR: 3.01, 3 times higher) and reported poor mental health alone (OR: 2.36, 2.36 times higher).

Compared with having no disability in either year, the likelihood of leaving employment was higher among those who:

  • Had a disability in both 2011 and 2021 (OR: 5.49, 5.49 times higher).
  • Acquired a disability between 2011 and 2021 (OR: 4.08, over 4-times higher).
  • Reported a disability in 2011 but not 2021 (OR: 1.46, +46%).

Other

  • Providing 35 or more hours of unpaid care per week (vs. none) was strongly associated with leaving employment (OR = 2.27, 2.27 times).
  • Those living in social-rented (OR = 2.86, 2.86 times) and in private-rented accommodation (OR = 1.88, 1.88 times), had a higher likelihood of leaving employment compared to those living in owner-occupied accommodation.

Sex-specific associations
When the models were run separately for males and females, notable differences emerged see Annex 4, Table 8.

For males, higher associations of leaving employment were found among those who:

  • Were single vs married (OR = 2.22, 2.22 times).
  • Were separated or divorced (OR = 1.62, +62%).
  • Reported neurodiversity (OR = 1.49, +49%).
  • Reported a physical health condition (OR = 1.25, +25%).
  • Lived in a lone-parent household (OR = 1.67, +67%).
  • Resided in urban areas of Derry City and Strabane Local Government District (OR = 1.29, +29%) or in the western (OR = 1.26, +26%) and northern areas (OR = 1.25, +25%) of Belfast Local Government District.

For females, higher associations of leaving employment were found among those who:

  • Had three or more dependent children in the household (OR = 2.23, 2.23 times).
  • Had a child with a disability in the household (OR: 1.28, +28%).

These associations do not imply causation but highlight key groups where targeted employment and support interventions may be effective.

Summary

Between 2011 and 2021, transitions into employment was more common among males, those aged 20-39 years, and those with higher levels of educational attainment. Improvements in disability status were particularly important, while lower levels of unpaid care and being unemployed or ‘other economically inactive’ in 2011 were also associated with entering work. Among females, employment entry was more likely for those who were separated or divorced, and for those living in northern and eastern areas of Belfast Local Government District.

Transitions out of employment, by contrast, were more common among females, and those with lower levels of educational attainment. Health and disability were key drivers: those exiting employment were more likely to report a disability at both time points or to have developed a disability over the decade. They were also more likely to report poor overall, mental, and physical health in 2021. It should be noted that poor mental health is likely to be underreported, particularly among males (see reference 7).

Higher unpaid caring responsibilities, renting, and living alone or in lone-parent households were also associated with leaving employment. Distinct sex-specific patterns were evident. For females, leaving employment was strongly associated with having three or more children or having a child with a disability. For males, leaving employment was associated with being single, reporting neurodiversity or physical health conditions, and living in northern and western areas of Belfast Local Government District, as well as urban areas of Derry City and Strabane Local Government District.

Overall, disability, poor mental and physical health, educational attainment and caring responsibilities play a central role in employment transitions. The findings highlight different underlying drivers for males and females. For females, family size and caring-related constraints appear to be particularly important, pointing to the potential value of policies that support childcare provision, caring responsibilities and flexible working arrangements. For males, the results suggest a stronger role for health, disability and skills-related factors in sustaining employment, indicating the importance of targeted health interventions, skills development and employability support. More broadly, this points to the need for coordinated policies to support those with health conditions, disabilities or caring responsibilities to remain in work and reduce the risk of long-term economic inactivity.

Policy context

This analysis supports the Department for Communities’ Disability and Work Strategy for Northern Ireland (see reference 8) which aims to increase employment among people with disabilities. The findings also support the Department for the Economy’s Economic Vision (see reference 9) which emphasises delivering good jobs including for people with disabilities and area-based economic interventions to support regional balance.

The findings also align with wider cross-departmental strategies in Northern Ireland, including:

  • The Department for Health-led Mental Health Strategy (2021-2031) (see reference 10) which highlights mental-ill health as a key barrier to employment and social participation.
  • The Department for Health-led Autism Strategy (2023-2028) (see reference 11) which focuses on creating more inclusive environments for neurodiverse individuals across education and employment.
  • The Department for Communities-led Disability strategy (see reference 12) which will aim to tackle inequalities and the removal of barriers for people with disabilities.
  • The Programme for Government (2024-2027) (see reference 1) which addresses economic inactivity and improving wellbeing outcomes for people out of work due to ill/health or disability.

The NISRA research team continues to work with Queen’s University Belfast and policy leads at DfC and DfE to maximise the policy impact of this research.

Limitations

  • The ten-year interval between Census years limits the ability to pinpoint the timing of disability onset.
  • Economic activity is observed at two points in time ten years apart and therefore does not capture periods of employment, unemployment or economic inactivity between 2011 and 2021.
  • The 2021 endpoint in the analysis coincided with the final phase of Northern Ireland’s Covid-19 lockdown, with restrictions only beginning to ease from early March 2021. During this period, furlough schemes and public‑health measures impacted labour‑market behaviour. As a result, some individuals who may otherwise have become unemployed remained classified as employed, slightly understating employment exits. It is also possible that some furloughed individuals reported themselves as economically inactive rather than temporarily away from work, which may contribute to a small degree of misclassification in economic‑activity transitions.

Next steps

Building on this analysis, possible future research could:

  • Examine whether some areas in Northern Ireland experience higher or lower levels of economic inactivity than would be expected based on their population characteristics, to determine if additional neighbourhood or area‑based factors contribute to increased likelihood of economic inactivity.
  • Examine younger people (aged under 30 years) with disabilities, neurodiversity or Special Education Needs (SEN) to better understand education-to-employment transitions.
  • Undertake condition-specific analysis (e.g. poor mental health, neurodiversity and physical health conditions) to inform targeted interventions.

Further information

Acknowledgements

The help provided by the staff of the Northern Ireland Longitudinal Study (NILS) and the NILS Research Support Unit is acknowledged. The NILS is funded by the Health and Social Care Research and Development Division of the Public Health Agency (HSC R&D Division) and NISRA. The NILS-RSU is funded by the Economic & Social Research Council and the Northern Ireland Government. We also thank colleagues in NISRA, the Department for the Economy (DfE) and the Department for Communities (DfC) who contributed to the development of this report.

About ADR UK

ADR UK (Administrative Data Research UK) is a partnership transforming the way researchers access the UK’s wealth of public sector data, to enable better informed policy decisions that improve people’s lives. By linking together data held by different parts of government and facilitating safe and secure access for accredited researchers to these newly joined-up and de-identified data sets, ADR UK is creating a sustainable body of knowledge about how our society and economy function – tailored to give decision makers the answers they need to solve important policy questions. ADR UK is funded by the Economic and Social Research Council (ESRC), part of UK Research and Innovation.

Contact details

Research team John Hughes, Brian Foley, Jana Ross, Carmel Colohan, and Deborah Lyness (all Administrative Research Unit, NISRA)
Professor Duncan McVicar and Dr Aideen Maguire (both Queen’s University Belfast)

For further information on this research, please contact

Accessibility contact

Please contact Dissemination Branch for assistance with accessibility requirements or alternative formats. Contact details are:

Email:

Telephone: +44 (0)300 200 7836

Dissemination Branch
NISRA
Colby House
Stranmillis Court
BELFAST
BT9 5RR

Annexes

Annex 1 - Definitions and methods used

Table 1: Health definitions

Category Definition
Disability Aligned with the Disability Discrimination Act (DDA). Respondents were asked: “Are your day-to-day activities limited because of a health problem or disability which has lasted, or is expected to last, at least 12 months?” (No; Yes, limited a little; Yes, limited a lot).
Disability change
(2011–2021)
  1. No disability in 2011 or 2021.
  2. Acquired a disability between 2011 and 2021.
  3. Disability in 2011 but not 2021.
  4. Disability in both 2011 and 2021.
Neurodiversity The definition of neurodiversity used is based on respondents self-reporting Autism, Asperger syndrome, or a learning difficulty (e.g. dyslexia) on the 2021 Census. This represents a bespoke classification created for this analysis and should not be interpreted as an official or exhaustive definition. An official definition of neurodiversity is provided by the Royal College of Nursing (see reference 13).
Physical health condition Long-term pain or discomfort, breathing difficulty, mobility/dexterity difficulty, or other long-term condition (e.g. cancer, diabetes, heart disease).
Poor mental health Self-reported emotional, psychological, or mental health condition (e.g. depression, schizophrenia).


Table 2: Other definitions

Category Definition
Economic activity group
  1. Economically active: employed.
  2. Economically active: unemployed.
  3. Economically inactive: looking after home/family.
  4. Economically inactive: long-term sick/disabled.
  5. Other economically inactive: individuals not seeking work for reasons other than looking after home/family or long-term sickness/disability. This group included people discouraged from seeking employment, those who gave no specific reason for their inactivity and individuals temporarily sick/disabled.
Intermediate level qualification School-level qualification, other vocational qualification or apprenticeship.
Living other Unrelated individuals living together, such as young professionals.
Urban/rural residence Based on the 2021 Census, settlements (see reference 14) with a population of greater than or equal to 5,000 were classified as urban while settlements with a population of less than 5,000 were classified as rural.
Area deprivation The Northern Ireland Multiple Deprivation Measure (NIMDM) (see reference 15) assessed deprivation across seven domains. In this study, the 2017 Income NIMDM domain grouped individuals into deprivation quintiles based on their place of residence. Analysis compared those in the most deprived quintile, ‘the top 20% of deprived areas’ with those in the remaining four quintiles, ‘the bottom 80% of deprived areas’.
Local Government District (LGD) A person’s area of residence within one of 11 LGD’s introduced in 2015 (see reference 16). Belfast LGD was subdivided into four areas (northern, eastern, southern and western) using lower level geographical areas called District Electoral Areas (DEA) (see reference 17). In addition, Derry City and Strabane LGD was separated into urban and rural areas.

Methods

To assess the likelihood of moving into (Figure 2) and out of (Figure 4) employment, we used a statistical method called logistic regression (see reference 18). This helps us understand how different factors are linked to an outcome while taking other influences into account. For each factor, one group is chosen as the reference group, which is given an odds ratio (OR). of 1. All other groups are compared against this reference group. OR shows whether a group is more likely, less likely, or equally likely to have the outcome:

  • OR = 1 → the comparison group is just as likely as the reference group
  • OR > 1 → the comparison group is more likely to have the outcome
  • OR < 1 → the comparison group is less likely to have the outcome

Example Those with a degree (OR = 3.94) were 3.94 times more likely to move into employment than those with no qualifications.

Understanding the charts The odds ratios in Figures 2 and 4 are displayed on a log scale, which spaces the numbers in a way that makes them easier to compare. For example, an odds ratio of 0.5 (half as likely) and 2 (twice as likely) appear the same distance from 1 on this type of chart. Each odds ratio also includes a 95% confidence interval (CI). This shows the range of values we are reasonably confident contains the true result.

Choosing which characteristics to include Our analysis included everyone aged 20–49 in 2011. Each possible factor (for example, having a degree compared to having no qualifications) was first tested individually, while adjusting only for age and sex. Factors showing very weak evidence of an effect (p > 0.2) on entering or leaving employment were removed. The remaining factors were then tested together in one model to see the effect of each factor on its own, after taking the others into account.

Variable definitions Age and labour market status refer to baseline characteristics in 2011, while disability status captures changes between 2011 and 2021. However, due to data availability and improvements in data quality, all other characteristics reflect status in 2021 and circumstances associated with transitions into and out of employment over the period. As a result, the role of disability onset in leaving employment (e.g. see Figure 4) may be underestimated, as disability onset may also influence other characteristics (e.g. area of residence or marital status) measured at the end of the period.

Annex 2 - Characteristics by employment change

Table 3: Census 2021 Individual characteristics (%) by employment transition (2011-2021)

Characteristic
Category
Entered employment
(%)
(N = 14,163)
Not employed in 2011 or 2021
(%)
(N = 21,141)
Left employment
(%)
(N = 15,649)
Employed in 2011 and 2021
(%)
(N = 116,831)
Sex Male 44 37 38 50
Sex Female 56 63 62 50
Age 20-29 years 38 23 31 31
Age 30-39 years 33 30 30 34
Age 40-49 years 29 47 39 34
Marital status Single 38 46 37 24
Marital status Married 47 33 43 65
Marital status Divorced or separated 15 21 19 11
Educational
attainment 		No qualifications 18 47 27 9
Educational
attainment 		Intermediate
qualifications 		54 42 50 45
Educational
attainment 		Degree
qualifications 		28 11 23 46
Unpaid care Provides none 83 77 79 82
Unpaid care Provides 1–19 hours 8 4 5 11
Unpaid care Provides 20-34 hours 2 2 2 2
Unpaid care Provides 35 or more hours 7 17 14 5
Disability change
(2011–2021) 		No disability
in either year 		73 29 48 87
Disability change
(2011–2021) 		Disability in both
2011 and 2021 		7 44 13 2
Disability change
(2011–2021) 		Disability in 2021
but not 2011 		9 22 36 9
Disability change
(2011–2021) 		Disability in 2011
but not 2021 		11 5 2 2
Poor mental health No 91 59 71 95
Poor mental health Yes without other condition(s) 4 10 9 3
Poor mental health Yes with other condition(s) 5 31 20 2
Neurodiversity No 97 92 95 98
Neurodiversity Yes 3 8 5 2
Physical health condition No 79 44 56 83
Physical health condition Yes 21 56 44 17

Table 4: Census 2021 Household and area characteristics (%) by employment transition (2011-2021)

Characteristic
Category
Entered employment
(%)
(N = 14,044)
Not employed in 2011 or 2021
(%)
(N = 20,669)
Left employment
(%)
(N = 15,442)
Employed in 2011 and 2021
(%)
(N = 116,127)
Housing tenure Owner occupied 61 37 56 85
Housing tenure Social rental 18 40 23 5
Housing tenure Private rental 22 22 21 11
Living arrangements Couple household 60 41 52 75
Living arrangements Lone-parent household 21 27 22 10
Living arrangements Living alone/
living other 		20 32 26 16
Number of dependent
children in household 		None 23 33 31 26
Number of dependent
children in household 		One or two 63 55 57 62
Number of dependent
children in household 		Three or more 13 12 13 12
Has child with disability
in household 		No 90 90 86 93
Has child with disability
in household 		Yes 10 10 14 7
Local Government
District 		North Belfast 7 9 6 4
Local Government
District 		East Belfast 5 5 5 6
Local Government
District 		South Belfast 2 3 3 3
Local Government
District 		West Belfast 4 7 4 3
Local Government
District 		Urban Derry and
Strabane 		5 5 5 3
Local Government
District 		Rural Derry and
Strabane 		4 5 3 3
Local Government
District 		Rest of
Northern Ireland 		73 65 73 79
Area deprivation Top 20%
of deprived areas 		25 31 23 16
Area deprivation Bottom 80%
of deprived areas 		75 69 77 84
Area of residence Rural 34 27 31 37
Area of residence Urban 66 73 69 63

Annex 3 - Characteristics by employment change by sex

Table 5: 2021 Individual characteristics (%) by employment transition (2011-2021)

Males
Females
Characteristic
Category
Entered employment (%)
Not employed both years (%)
Left employment (%)
Employed both years (%)
Entered employment (%)
Not employed both years (%)
Left employment (%)
Employed both years (%)
Age 20-29 45 24 26 30 33 23 34 32
Age 30-39 27 29 30 34 37 30 30 34
Age 40-49 27 48 44 35 29 47 36 34
Marital status Single 47 61 45 24 31 37 33 24
Marital status Married 42 23 37 67 51 39 47 63
Marital status Divorced/separated 11 16 17 8 19 24 20 13
Educational attainment No qualifications 21 54 34 12 16 43 23 6
Educational attainment Intermediate qualifications 56 37 48 49 52 45 51 42
Educational attainment Degree-qualifications 23 9 18 39 32 12 26 52
Unpaid care Provides none 88 83 82 86 79 74 76 79
Unpaid care Provides 1-19 hours 5 3 4 9 10 4 6 13
Unpaid care Provides 20-34 hours 2 2 2 1 3 2 3 2
Unpaid care Provides 35 or more hours 5 12 11 4 8 19 15 6
Disability change (2011-2021) No disability in either year 72 23 44 88 73 33 51 86
Disability change (2011-2021) Disability in both 2011 and 2021 7 50 14 2 7 40 12 2
Disability change (2011-2021) Disability in 2021 but not 2011 8 21 39 8 10 22 35 9
Disability change (2011-2021) Disability in 2011 but not in 2021 13 6 3 2 10 5 2 2
Poor mental health No 93 58 71 96 90 60 71 94
Poor mental health Yes without other condition(s) 3 10 9 2 5 9 8 4
Poor mental health Yes with other condition(s) 4 32 20 2 5 30 20 3
Neurodiversity No 96 87 92 98 98 95 97 99
Neurodiversity Yes 4 13 8 2 2 5 3 1
Physical disability No 80 42 54 83 78 45 58 82
Physical disability Yes 20 58 46 17 22 55 42 18

Table 6: 2021 Household and area characteristics (%) by employment transition (2011-2021)

Males
Females
Characteristic
Category
Entered employment (%)
Not employed both years (%)
Left employment (%)
Employed both years (%)
Entered employment (%)
Not employed both years (%)
Left employment (%)
Employed both years (%)
Housing tenure Owner occupied 64 35 57 86 58 39 56 84
Housing tenure Social rental 15 43 23 4 20 39 23 5
Housing tenure Private rental 21 22 20 11 22 22 21 11
Living arrangements Couple household 62 35 51 79 58 44 53 71
Living arrangements Lone-parent household 11 16 14 5 28 33 27 14
Living arrangements Living alone/living other 27 48 36 16 14 23 20 15
Number of dependent children in household None 33 48 41 27 16 25 25 24
Number of dependent children in household One or two 59 48 53 60 67 60 59 64
Number of dependent children in household Three or more 8 5 7 12 17 16 16 11
Has disabled child in household No 94 91 91 93 87 76 82 93
Has disabled child in household Yes 6 9 9 7 13 24 18 7
Local Government District North Belfast 7 10 7 3 7 8 6 4
Local Government District East Belfast 5 5 5 5 5 4 5 6
Local Government District South Belfast 2 3 3 3 2 2 2 3
Local Government District West Belfast 4 6 4 3 5 6 4 3
Local Government District Urban Derry and Strabane 6 8 5 3 5 8 5 4
Local Government District Rural Derry and Strabane 4 5 3 3 4 4 4 3
Local Government District Rest of Northern Ireland 72 63 73 80 73 67 74 77
Area deprivation Top 20% of deprived areas 26 32 23 16 25 30 23 16
Area deprivation Bottom 80% of deprived areas 74 68 77 84 75 70 77 84
Area of residence Rural 35 26 30 39 33 28 32 36
Area of residence Urban 65 74 70 61 67 72 68 64

Annex 4 - Full modelling results

Table 7: Odds ratios with 95% confidence intervals (CIS) for moving into employment between 2011 and 2021, overall and by sex

Category (i) Overall
(95% CI)		 (ii) Male
(95% CI)		 (iii) Female
(95% CI)
Male (vs Female) 1.33 (1.25,1.42) NA NA
Unemployed (vs Long-term sick or disabled) 2.04 (1.84,2.26) 2.05 (1.74,2.40) 1.98 (1.73,2.28)
Looking after home or family (vs Long-term sick or disabled) 1.04 (0.94,1.16) 0.88 (0.71,1.1) 1.11 (0.98,1.26)
Other economically inactive (vs Long-term sick or disabled) 1.54 (1.38,1.71) 1.79 (1.51,2.12) 1.37 (1.19,1.58)
20-29 (vs 40-49) 2.11 (1.95,2.29) 2.18 (1.91,2.48) 2.09 (1.88,2.32)
30-39 (vs 40-49) 1.60 (1.49,1.72) 1.42 (1.26,1.61) 1.72 (1.58,1.88)
Single (vs married) 0.72 (0.66,0.79) 0.46 (0.4,0.53) 0.94 (0.83,1.06)
Divorced/separated (vs married) 1.22 (1.10,1.35) 0.84 (0.70,1.00) 1.40 (1.22,1.59)
Intermediate qualifications (vs no qualifications) 2.34 (2.19,2.49) 2.45 (2.21,2.72) 2.30 (2.12,2.5)
Degree-qualified (vs no qualifications) 3.95 (3.63,4.29) 3.57 (3.08,4.14) 4.29 (3.87,4.76)
Provides 1-19 hours unpaid care (vs no care) 1.37 (1.22,1.54) 1.38 (1.09,1.74) 1.43 (1.25,1.63)
Provides 20-34 hours unpaid care (vs no care) 0.68 (0.57,0.80) 0.57 (0.42,0.77) 0.76 (0.62,0.92)
Provides 35 or more hours unpaid care (vs no care) 0.30 (0.28,0.33) 0.25 (0.21,0.29) 0.34 (0.31,0.38)
Disability in both 2011 and 2021 (vs no disability in either year) 0.16 (0.14,0.18) 0.14 (0.12,0.17) 0.16 (0.14,0.19)
Acquired a disability between 2011 and 2021 (vs no disability in either year) 0.29 (0.26,0.32) 0.23 (0.19,0.27) 0.33 (0.29,0.37)
Disability in 2011 but not in 2021 (vs no disability in either year) 1.31 (1.17,1.45) 1.34 (1.12,1.59) 1.26 (1.10,1.45)
Poor mental health alone (vs none) 0.44 (0.39,0.50) 0.32 (0.26,0.39) 0.53 (0.46,0.61)
Poor mental health + other conditions (vs none) 0.37 (0.33,0.41) 0.32 (0.27,0.38) 0.41 (0.36,0.47)
Neurodiversity (vs not) 0.67 (0.58,0.78) 0.70 (0.58,0.86) 0.70 (0.56,0.88)
Physical health condition (vs not) 0.95 (0.87,1.03) 0.94 (0.81,1.08) 0.93 (0.84,1.04)
Social-rented (vs owner-occupied) 0.53 (0.49,0.57) 0.39 (0.35,0.44) 0.58 (0.53,0.64)
Private-rented (vs owner occupied) 0.68 (0.64,0.73) 0.60 (0.53,0.68) 0.69 (0.63,0.76)
Lone-parent household (vs living in a couple household) 0.86 (0.78,0.94) 0.66 (0.56,0.78) 0.86 (0.76,0.97)
Living alone/other (vs living in a couple household) 0.75 (0.68,0.82) 0.65 (0.56,0.74) 0.91 (0.80,1.03)
One or two dependent children in household (vs none) 1.00 (0.92,1.08) 0.85 (0.74,0.97) 1.07 (0.96,1.20)
Three or more dependent children in household (vs none) 0.63 (0.56,0.70) 0.74 (0.59,0.93) 0.61 (0.53,0.70)
Has child with disability in household (vs not) 0.89 (0.82,0.97) 0.73 (0.60,0.88) 0.84 (0.76,0.93)
North Belfast (vs rest of NI) 1.19 (1.06,1.32) 0.98 (0.82,1.17) 1.31 (1.14,1.50)
East Belfast (vs rest of NI) 1.07 (0.94,1.22) 0.92 (0.74,1.14) 1.20 (1.02,1.43)
South Belfast (vs rest of NI) 1.13 (0.94,1.35) 1.07 (0.80,1.44) 1.21 (0.96,1.54)
West Belfast (vs rest of NI) 1.02 (0.89,1.15) 0.79 (0.64,0.98) 1.14 (0.97,1.34)
Derry and Strabane - urban (vs rest of NI) 0.78 (0.70,0.87) 0.64 (0.53,0.77) 0.85 (0.74,0.98)
Derry and Strabane - rural (vs rest of NI) 0.83 (0.73,0.95) 0.74 (0.59,0.92) 0.87 (0.73,1.03)
Note: 95% confidence intervals (CIs) are a range of likely values around the odds ratio. CIs that do not cross 1 are statistically significant, while CIs that do cross 1 are not.

The ‘rest of NI’ refers to areas outside of Belfast and Derry City and Strabane Local Government Districts.

Table 8: Odds ratios with 95% confidence intervals (CIS) for leaving employment between 2011 and 2021, overall and by sex

Category (i) Overall
(95% CI)		 (ii) Male
(95% CI)		 (iii) Female
(95% CI)
Female (vs male) 1.69 (1.62,1.76) NA NA
30-39 (vs 20-29) 0.89 (0.85,0.94) 1.13 (1.03,1.23) 0.82 (0.77,0.87)
40-49 (vs 20-29) 1.06 (1.01,1.12) 1.43 (1.31,1.56) 0.94 (0.87,1.00)
Single (vs married) 1.38 (1.31,1.47) 2.22 (2.03,2.44) 0.99 (0.92,1.08)
Divorced/separated (vs married) 1.08 (1.00,1.16) 1.62 (1.44,1.82) 0.84 (0.77,0.92)
Intermediate qualifications (vs no qualifications) 0.43 (0.41,0.46) 0.46 (0.42,0.49) 0.42 (0.39,0.45)
Degree-qualified (vs no qualifications) 0.24 (0.23,0.26) 0.28 (0.25,0.31) 0.22 (0.20,0.24)
Provides 1-19 hours unpaid care (vs no care) 0.58 (0.54,0.63) 0.57 (0.50,0.66) 0.58 (0.52,0.63)
Provides 20-34 hours unpaid care (vs no care) 1.08 (0.95,1.22) 1.39 (1.12,1.73) 0.94 (0.81,1.10)
Provides 35 or more hours unpaid care (vs no care) 2.27 (2.13,2.42) 2.38 (2.13,2.66) 2.14 (1.97,2.32)
Disability in both 2011 and 2021 (vs no disability in either year) 5.49 (5.06,5.96) 5.64 (4.97,6.40) 5.30 (4.76,5.91)
Acquired a disability between 2011 and 2021 (vs no disability in either year) 4.08 (3.84,4.32) 4.60 (4.20,5.05) 3.74 (3.47,4.04)
Disability in 2011 but not in 2021 (vs no disability in either year) 1.46 (1.29,1.66) 1.62 (1.35,1.94) 1.35 (1.14,1.60)
Poor mental health alone (vs none) 2.36 (2.18,2.56) 3.16 (2.76,3.61) 2.00 (1.80,2.21)
Poor mental health + other conditions (vs none) 3.01 (2.81,3.23) 3.69 (3.29,4.14) 2.76 (2.52,3.02)
Neurodiversity (vs not) 1.35 (1.21,1.51) 1.49 (1.28,1.73) 1.10 (0.93,1.31)
Physical health condition (vs not) 1.09 (1.03,1.15) 1.25 (1.14,1.36) 1.03 (0.95,1.11)
Social-rented (vs owner-occupied) 2.86 (2.69,3.03) 3.28 (2.98,3.62) 2.73 (2.52,2.95)
Private-rented (vs owner occupied) 1.88 (1.78,1.98) 1.82 (1.67,1.99) 1.96 (1.83,2.10)
Lone-parent household (vs living in a couple household) 1.30 (1.21,1.40) 1.67 (1.48,1.89) 1.27 (1.16,1.39)
Living alone/other (vs living in a couple household) 1.30 (1.23,1.39) 1.34 (1.22,1.48) 1.22 (1.12,1.33)
One or two dependent children in household (vs none) 1.01 (0.95,1.06) 1.02 (0.94,1.11) 1.05 (0.98,1.13)
Three or more dependent children in household (vs none) 1.75 (1.62,1.88) 1.13 (0.98,1.30) 2.23 (2.03,2.45)
Has child with disability in household (vs not) 1.15 (1.08,1.23) 1.08 (0.95,1.22) 1.28 (1.18,1.39)
North Belfast (vs rest of NI) 0.95 (0.87,1.04) 1.25 (1.09,1.44) 0.81 (0.73,0.91)
East Belfast (vs rest of NI) 0.84 (0.77,0.92) 0.88 (0.76,1.02) 0.82 (0.74,0.92)
South Belfast (vs rest of NI) 0.92 (0.81,1.04) 1.09 (0.90,1.32) 0.80 (0.68,0.95)
West Belfast (vs rest of NI) 0.95 (0.85,1.05) 1.26 (1.06,1.50) 0.84 (0.74,0.95)
Derry and Strabane - urban (vs rest of NI) 1.14 (1.04,1.24) 1.29 (1.11,1.49) 1.08 (0.96,1.21)
Derry and Strabane - rural (vs rest of NI) 1.12 (1.01,1.25) 1.14 (0.95,1.35) 1.13 (0.99,1.30)
Note: 95% confidence intervals (CIs) are a range of likely values around the odds ratio. CIs that do not cross 1 are statistically significant, while CIs that do cross 1 are not.

The ‘rest of NI’ refers to areas outside of Belfast and Derry City and Strabane Local Government Districts.