Longitudinal studies & the life course abstracts

Strand organiser: Stephen Jivraj, University College London

Longitudinal studies and the life courses: Linked 2011 census data on social change in the UK - Monday 8 September 4.45pm

Inter-cohort Trends in Intergenerational Mobility in England and Wales: income, status, and class (InTIME)
Franz Buscha, University of Westminster, Patrick Sturgis, University of Southampton

The level of intergenerational mobility in a society is widely taken as a key barometer of its fairness and equality, outwardly signalling whether citizens achieve social and economic status through hard work and ability, or as a result of advantages bestowed upon them by their parents. As a concept social mobility has become one of the key motifs of our political epoch, with politicians of both left and right now championing it as a core policy objective (Saunders 2010). In 2011, for example, the coalition government announced its ‘social mobility strategy’ in which improving relative intergenerational mobility was specified as the government’s most important social policy objective for the parliament (Cabinet Office, 2011). In this paper we make use of the Office for National Statistics Longitudinal Study (LS). The LS is a one per cent sample of the population of England and Wales, with individual records linked across successive censuses. The LS thus provides representative cross-sectional and longitudinal information about the population of England and Wales for the years 1971, 1981, 1991, 2001 and 2011 and is thus well suited to investigate the development of social mobility over time. Our core objectives in this research project are to bring clarity and resolution to the debate over the nature of recent trends in social mobility in the UK by providing robust estimates of relative intergenerational social mobility for cohorts, in England and Wales defined by year of birth, born between 1955 and 1993 and covering the period 1981 to 2011. Our aim is to estimate and compare relative mobility rates using three different measures of socio-economic position: social class; social status; and income.

Email: buschaf@westminster.ac.uk

The stability of ethnic group and religion in the Censuses of England and Wales 2001-2011
Ludi Simpson, University of Manchester, Stephen Jivraj, Institute of Education, UCL, James Warren, Office for National Statistics

When a change in a population is recorded, for example a decrease in long-term limiting illness of the Caribbean group, or an increase in the employment of Bangladeshi men, these comparisons are fair and accurate as descriptions of how the group is different at two points in time. This paper digs a little deeper and asks the extent to which such change can occur because an ethnic group or religious group is made up of different people, rather than the same people having changed their characteristics. This paper intends to answer the most urgent practical questions for analysts of ethnicity and of religion from the censuses. After a summary of findings and a recap of population change 1991-2001-2011, for each ethnic group and religion category this paper identifies for 2001-2011: • How much population change between two censuses is due to people joining or leaving the population through being born, dying, moving into or out of England and Wales, or being not counted ? • For those recorded at both 2001 and 2011 Censuses, the extent to which people did not change their ethnic group, and whether amalgamating ethnic groups increases the stability of responses. • The comparisons between ethnic groups across 1991, 2001 and 2011 which are most valid because least subject to instability. • The implications for analysts of ethnic group from census datasets . • The extent to which people did and did not change their religion. The paper uses the ONS Longitudinal Study, containing individual records for about 1% of the population of England and Wales, linked across censuses since 1971.

Email: james.warren@ons.gsi.gov.uk

Sibship size and educational attainment in the UK
Tak Wing Chan, University of Oxford, Morag Henderson, Cardiff University, Rachel Stuchbury, CeLSIUS

We use ONS LS data from 1971 through 2011 to explore whether the sibship size of LS members has a causal effect on their educational outcome. Our strategy is to extract household information of LS members (number of siblings, whether they lived with both parents up to mid-teens, parental class and education, housing tenure, etc) when they were aged 4-7 and 14-17, and link this to their educational outcome when they were aged 24-27. By using twin births as an instrumental variable, we seek to determine whether number of siblings really leads to lower educational outcome. And by comparing LS members who were 4-7 in 1971, 1981 and 1991 respectively, we seek to determine whether the causal effects of family size has changed over time. linking LS members household composition information when they were of 4-7 and 14-17.

Email: tw.chan@sociology.ox.ac.uk

Does mixed marriage by religion increase the risk of marital dissolution in Northern Ireland?
Dr David M Wright, Dr Dermot O’Reilly, Centre for Public Health, Queen’s University Belfast, Dr Michael Rosato, Bamford Centre for Mental Health and Wellbeing, University of Ulster

Risks of marital dissolution are often elevated when partners are dissimilar (e.g. by age, ethnicity or socioeconomic position) but may also depend on the social context of the marriage. Northern Ireland has a history of conflict and residential segregation along religio-cultural (Protestant-Catholic) lines and we investigated whether marriages crossing this social boundary were at increased risk of dissolution and whether risk was further elevated in the most segregated areas. We identified 115,465 married couples in the 2001 Census and estimated the risk of marital dissolution in the subsequent decade (indicators of dissolution were living apart or no longer married at the 2011 Census). Multilevel models were adjusted for age, education, economic activity, housing tenure, country of birth, history of previous marriages, presence of dependent children, housing tenure, urban/rural residence and degree of segregation. Risk of dissolution for non-mixed marriages was lowest within the most conservative religious groups and highest for people with no affiliation. Some types of mixed marriages had a slightly higher risk of dissolution than corresponding non-mixed marriages but this effect was weak for Protestant-Catholic marriages. Risk of dissolution for Protestant-Catholic marriages was independent of residential segregation but such marriages were rare and these couples were more likely to cohabit than other types of mixed couple (e.g. among Protestant denominations). Mixed relationships (both marriage and cohabitation) were more common among the young. Social barriers to interdenominational relationships remain in Northern Ireland, particularly along Protestant-Catholic lines but there are signs that cross-community relationships may become more common in future.

Email: d.wright@qub.ac.uk

Longitudinal studies and the life course: Health and employment - Tuesday 9 September 11.00am

Women’s economic activity trajectories over the life course: implications for the self-rated health of women aged 60+ in England
Juliet Stone, Maria Evandrou, Jane Falkingham and Athina Vlachantoni, ESRC Centre for Population Change, University of Southampton

Women’s role in the labour market changed substantially during the second half of the twentieth century, potentially necessitating an increased need to balance work and family life. In turn, previous research has highlighted the importance of accumulated life course labour market status for understanding inequalities in health in later life. This study uses optimal matching analysis and cluster analysis to produce a taxonomy of women’s life-course economic activity trajectories based on their experiences between ages 16 and 60 years. This classification is then used in logistic regression analysis to investigate associations with self-rated health in later life. A relatively limited set of five trajectories emerge as the dominant patterns of women’s economic activity over the life course. However, the more ‘flexible’ patterns including temporary breaks from paid employment to look after a home and family are more common among younger cohorts. Regression analyses further show that for women, being in full-time paid employment across the life course is not necessarily good for their later health, while women who combine full-time work with family life appear to have the most favourable outcomes. The findings are discussed with reference to the accumulation of social and economic resources over the life course and the balancing of multiple roles in work and family domains. In conclusion, development of policies that facilitate women, if they wish, to successfully combine paid employment with family life could have a positive impact on their health in later life.

Email: j.stone@soton.ac.uk

Diverging experiences of transition to the labour market in Germany: The effect of social background and labour market context
Wouter Zwysen, Institute for Social and Economic Research (ISER), University of Essex

Family background has long-lasting effects on labour market outcomes. A crucial period is the transition from the educational system to the labour market. It is here that inequality can get reinforced through an unsuccessful transition involving low work attachment and low pay. A bad first experience has been shown to have long-lasting effects, making this an important point for any policy intervention. While previous studies found an effect of family background on different adverse outcomes, the estimates differ strongly over time and between countries. The main empirical approaches focused on individual characteristics to explain the transmission of disadvantage. This paper studies the transition from school to work in the context of labour markets to explain variation. Family disadvantage is associated with lower human and social capital which makes the disadvantaged less desirable on the labour market. In tighter labour markets employers are expected to prefer more advantaged young adults. This results in a larger gap between the disadvantaged young adults and the rest in times of high unemployment. Using the German socio-economic panel study (SOEP) from 1984 to 2011 this paper shows that disadvantaged young adults (16-35) are more affected by the labour market context than the others. Random coefficient models are used to account for unobserved characteristics. A disadvantaged young adult is estimated to earn €0.50 or €1.50 less per hour than a similar more advantaged young adult, depending on whether the unemployment rate was 6% or 18% at the time of job entry.

Email: wzwyse@essex.ac.uk

Inequality in the Transition to Adulthood: Individualization or Social Allocation
Ross Macmillan, Bocconi University, Frank Furstenberg, University of Pennsylvania, Francesco Billari, Oxford University

The paper uses longitudinal data from three distinct generation of American adolescents to consider the role of the transition to adulthood for the production and re-production of inequality. To frame the research, we describe a life course capitalization perspective that allows us to define more and less efficacious pathways with respect to economic viability in later life. We the apply latent class techniques to longitudinal data drawn from the US National Longitudinal Series to formally model unique and distinct pathways into adulthood and then consider the whether inequality is generated from increasing disorder and de-standardization of role transitions within pathways or the selection and allocation of individuals within culturally resonant, pan historic pathways that generalize across race and sex. Results suggest a remarkably strong social structure to the transition to adulthood over time and across social groups and indicate the importance of selection and allocation mechanisms, rather than reconfiguration, in the generation of heightened inequalities in the contemporary era.

Email: ross.macmillan@unibocconi.it

Assessing the potential impact of markers of social support on levels of ‘excess’ mortality in Scotland and Glasgow compared to elsewhere in the UK
Kevin Ralston, David Walsh, Zhiqiang Feng, Chris Dibben, University of Edinburgh

Scotland has higher rates of mortality compared to the rest of the UK. Also premature mortality in Glasgow is 30% higher than in Liverpool and Manchester, with deaths at all ages around 15% higher. This excess is observed across almost all age groups, both males and females. In recent analyses reciprocity, trust, volunteering and religious affiliation including proxies for social capital of religious participation has been shown to be an important catalyst for social connectivity among some populations, and has been shown to be associated with lower mortality. This research combines the ONS Longitudinal Study of England and Wales with the Scottish Longitudinal Study to examine whether levels of ‘excess’ mortality in Scotland (compared to E&W) and Glasgow (compared to Liverpool/Manchester) are modified by the existence of social supports in peoples lives such as through practicing religion and living arrangements. We look at all-cause mortality by various age ranges including all age and 35 to 74. Poisson regressions are used along with a pioneering application of eDataSHIELD to undertake analysis on two restricted access datasets. The findings show that indicators of social support moderate mortality but that mortality in Scotland remains above that of England and Wales.

Email: kev.ralston@ed.ac.uk

Longitudinal studies and the life course: Methods and combining data - Tuesday 9 September 1.30pm

What does the ONS Longitudinal Study tell us about the quality of NHS registration data?
Steve Smallwood, James Warren, Office for National Statistics

What does the ONS Longitudinal Study tell us about the quality of NHS registration data? The aim of this research project was to assess the quality of NHS registration data by comparing location of residence between the census with that of the NHS registration system. By using the Longitudinal Study (LS) data matched to the NHS data we have assessed: the number of people that have a different location of residence in the census compared to the NHS data; the time lag between the location of residence being amended on the NHS data to match the location of residence on the census; migration indicators from the census and, the characteristics of these people. Data based on NHS registration activity is a key source feeding into current and potential new population statistics systems. Within the current population statistics system, these data are used: to estimate the migration of people within the UK; to allocate international migrants to local authority level; and to produce population estimates for lower level geographies. The Beyond 2011 Programme has been established to carry out research on the options for a new population statistics system. One likely option is greater use of administrative data, and NHS registration data would play a key part in this. This analysis has the potential to identify certain population groups that may be under represented in the NHS data to estimate the population, so that statistical methods and research can be undertaken to address this. This is a continuation of earlier work produced using data from the 2001 Census and published in Population Trends 141.

Email: james.warren@ons.gsi.gov.uk

Generating synthetic microdata to widen access to sensitive data sets: method, software and empirical examples
Beata Nowok 1, Gillian Raab 2, Chris Dibben 1,2, 1 University of Edinburgh, ADRC-S, 2 University of Edinburgh, LSCS

In many contexts, confidentiality constraints severely restrict access to unique and valuable microdata. The UK Longitudinal Studies linking census and other health and administrative data for individuals and their immediate families across several decades provide a good example which also motivated this study. In order to allow academics and other users to carry out their research more freely, synthetic version of a bespoke data set can be generated and provided to users with fewer access restrictions. Synthetic data mimic the real data and preserve the relationships between variables and transitions over time, but they do not include any real individuals. The basic idea of data synthesis is to replace genuine data with values sampled from conditional probability distributions. We develop a sequential algorithm for producing synthetic data set and implement it in R software (freely available R package called ‘synthpop’). The users have a choice between different parametric and non-parametric synthesising models. The latter includes classification and regression trees (CART) models. As a validation of the method we compare statistical inference based on real and synthetic data for research projects using the SLS data (including 2011 Census data). In addition, we compare the relative performance of parametric and non-parametric synthesising models.

Email: beata.nowok@ed.ac.uk

Augmenting MCS with external data on neighbourhood poverty rates and change
Ludovica Gambaro, Heather Joshi, Institute of Education, Ruth Lupton, University of Manchester, Alex Fenton, Leibniz Universität, Hannover, Mary Clare Lennon, City University, New York

This paper asks whether survey respondents’ views of their neighbourhood are in line with an objective measure of local poverty rates. We link cases from the first three sweeps of the Millennium Cohort Study (MCS) to a new indicator of small-area poverty, the Unadjusted Means-tested Benefits Rate (UMBR) (Fenton 2013). It is the ratio of claimants of means-tested benefits in a small area to the number of households and is available annually across Great Britain, making it preferable for this purpose to existing Indices of Multiple Deprivation (IMDs). We use UMBR to outline the patterns of change in poverty, and show how small areas in “disadvantaged urban” and “multicultural city life” communities have changed considerably between 2001 and 2006, the first five years of the cohort’s life. By combining UMBR with data from the MCS, we explore the association between residents’ perceptions of their neighbourhood and the poverty level of their area. We first find that respondents who lived in areas of higher poverty tended to express more negative views of their neighbourhood in general and in relation to raising children. We then distinguish between families who moved home and those who remained in the same area. Our findings reveal that those who retrospectively named the search for a “better neighbourhood” as a reason for moving have moved into areas with markedly lower poverty rates. However, among those who did not move, small changes in local poverty over time were not reflected in changes in residents’ views.

Email: heather.joshi@ioe.ac.uk

Assessing the association of retrospective and contemporaneous life course variables on later life outcomes
Alissa Goodman, Stephen Jivraj, Centre for Longitudinal Studies, Institute of Education, University of London

The growth of the older age population in the UK has sparked increased research interest on socioeconomic and health inequalities in later life and how they have accumulated during the life course. This is evident through the creation of longitudinal studies of older samples and the renewed investment in the birth cohort studies that have older age samples. This paper explores the potential in bringing together a prospective birth cohort study, the 1958 National Child Development Study (NCDS), and a panel study of older people that includes a retrospective life history module, the English Longitudinal Study of Ageing (ELSA). We present a series of harmonised variables available in the two datasets that measure experience during the life course and in later life. These data are used to predict a range of health, economic and social outcomes in mid to late age that have been found to be related to earlier life circumstances using retrospectively and contemporaneously collected data. We evaluate the limitations of both data collection techniques and offer suggestions of how each can inform the other in future research.

Email: s.jivraj@ioe.ac.uk

Longitudinal studies and the life course: Life events and families - Tuesday 9 September 4.45pm

Pathways to First Birth and the Changing Role of Education in Europe and the United States
Julia Mikolai, Brienna Perelli-Harris, Ann M. Berrington, University of Southampton

This paper applies multistate event history models to study the changing role of educational attainment on five pathways to first birth for women born between 1950 and 1969 using harmonised retrospective union and fertility histories (“Harmonized Histories”) from 13 European countries and the United States. Controlling for educational enrollment and birth cohort, we find a persistent negative educational gradient of first birth within cohabitation which remains negative even in countries where the transition into cohabitation has a positive educational gradient. Similarly, having a first birth while being never partnered is associated with low education in all countries. Moreover, on the pathway to first birth within marriage that was preceded by cohabitation, what seems to matter is that the more educated women have a higher risk to marry their cohabiting partner. Once they do so, they are, however, more likely to delay first birth than their lower educated counterparts. Although the educational gradient of direct marriage shows less consistent results, the timing pattern of the transition to first birth within direct marriage resembles that of the transition to first birth within marriage that was preceded by cohabitation. All in all, these results suggest that education plays an important role in the transition to first birth within cohabitation or while being never partnered and in the transition to marriage following cohabitation.

Email: jm1e11@soton.ac.uk

The role of unemployment and job instability in mediating the relationship between parental resources and the tempo of marriage and cohabitation in Europe
Elena Mariani, London School of Economics, Andras Gabos, TARKI Social Research Institute

In this study we aim to clarify the role of employment status (job instability and unemployment) in mediating the relationship between parental socioeconomic resources and the timing of transitions to marriage and cohabitation across Europe. High parental resources are associated with a delay in marriage and cohabitation formation. While parental resources play a strong role in the decision to form a new family, own resources are probably a necessary condition for transition to adulthood. On the other hand, recent economic changes in Europe, especially the increase in job instability and the high rates of youth unemployment, have raised the concern that young adults are not able to reach the desired economic independence for forming a new family. In particular, the recent economic crisis has led many young adults to return to their family homes, so that the legacy and importance of family of origin is now stronger. We answer our research question applying discrete time hazard models via multinomial logistic models with country fixed and random effects to European Union Statistics on Income and Living Conditions data, 2005-2011.

Email: e.mariani@lse.ac.uk

Do parents have a happier life than non-parents? The role of preferences
Nicoletta Balbo, Bocconi University, Dondena Centre, Bruno Arpino, Pompeu Fabra Universitat

This paper aims to investigate whether there is a causal relationship between parenthood and subjective well-being, studying the effect of first childbirth on an individual’s life satisfaction. Specifically, we aim at answering to the following research question: Is an individual who has a child more satisfied than his/her childless counterpart who has the same socio-economic characteristics and preferences? Existing longitudinal studies only look at parents, not comparing them with their non-parent counterparts. At the same time, previous studies do not take into account that people might have different values, preferences and expectations about childbearing. These cultural differences might shape how life satisfaction is affected by parenthood. We therefore adopt a longitudinal approach, within which we compare parents with their childless counterparts, who share not only the same socio-economic characteristics but also the same preferences. Using the British Household Panels Survey and engaging in a propensity score matching, we show that parents are significantly more satisfied than non-parents. This effect is found to be stronger among men than among women. However, the positive effect of childbearing seems to be mostly limited to the year before childbirth, which can be interpreted as a strong anticipation effect. Almost no long-lasting positive effects on life satisfaction after childbirth are found. We moreover find that family-oriented women are those who benefit more from childbearing in terms of increased life satisfaction. Career-oriented mothers show a decrease in life satisfaction in the long-run compared to career-oriented non-mothers. Childbearing seems to barely affect life satisfaction of adaptive women.

Email: balbo.nicoletta@unibocconi.it

Cohort and gender differences in work-family life courses in Great Britain: evidence of increasing individualisation?
Rebecca Lacey, Amanda Sacker, Anne McMunn, UCL, Meena Kumari, Cara Booker, University of Essex

The United Kingdom has seen a decline in traditional social institutions, such as patriarchal institutions of marriage and gender divisions of labour as encapsulated in the male breadwinner model – as described and predicted by the individualisation thesis. Participation in paid work is now a fact of life for the majority of women, including those with family responsibilities; fathers are more involved in childcare; and family forms have become increasingly diverse. In addition, these processes of change are occurring within the context of, and may be partially driving, increasing socioeconomic inequality between households and between women. This study uses the wealth of longitudinal data now available in Britain to describe gender & cohort differences in work-family life course sequences across the three national birth cohort studies in the UK: the National Survey of Health and Development 1946 birth cohort, the National Child Development Study 1958 birth cohort, and the 1970 Birth Cohort Study. Sequence analysis is used to derive life course work-family typologies for each cohort between ages 16-42. The potentially changing role of early life predictors of work-family patterns is also investigated. Early results suggest that work-family life courses are strongly gendered. While the majority of men in all cohorts occupy stable life courses with strong labour market ties, the proportion of women in these typologies has increased, with a resultant decrease in the proportion of long-term homemaking women. We find expected increases in long-term cohabiting couples but increases in lone parent typologies are not as great as expected.

Email: rebecca.lacey@ucl.ac.uk

How does childhood socioeconomic hardship affect reproductive strategy? Pathways of development
Paula Sheppard, Rebecca Sear, Department of Population Health, London School of Hygiene and Tropical Medicine, Mark Pearce, Institute of Health and Society, Newcastle University

In high income populations, there is strong evidence to suggest socioeconomic disadvantage early in life is correlated with reproductive strategy. Children growing up in unfavourable rearing environments tend to experience earlier sexual maturity and first births. Earlier first births may then be associated with larger family sizes. Less is known about the causal pathways linking childhood disadvantage to reproductive outcomes in later life. Using data from the Newcastle Thousand Families Study, a 1947 British birth cohort, we developed a path model to identify possible physiological factors linking early life conditions (childhood socioeconomic status (SES) and poor housing standards) to two reproductive outcomes: age at first birth and total surviving children. We explored age at menarche, and adult height as possible mediators. Further, we posited that other markers of child health such as birth weight, weight gain after birth, childhood illnesses, and body mass index at age nine may connect early adversity to both of these mediators. We found direct effects of SES and housing grade on age at first birth, and of housing on total fertility. Although we found links between childhood disadvantage and markers of health, neither of our proposed mediators was significantly correlated with either reproductive outcome, although age at first birth is a strong predictor of total fertility in this population. We conclude that while there are clear links between early childhood adversity and child health, the pathways to reproduction are less clear later on in life.

Email: paula.sheppard@lshtm.ac.uk

Longitudinal studies and the life course: Linked 2011 census data on health and caring in the UK - Wednesday 10 September 9.00am

The long-term impacts of NEET experiences on health: evidence from the Scottish Longitudinal Study
Zhiqiang Feng, Elspeth Graham, University of St Andrews, Kevin Ralston, Gillian Raab, Chris Dibben, University of Edinburgh

Young people who are not in employment, education or training (NEET) have drawn consistent attention from policy makers in the last decades. A spell of NEET experiences may impact on socioeconomic and health outcomes in later life. Theoretically there have been debates on the consequences of NEET experiences and so far a number of empirical studies have yielded mixed results. This paper aims to investigate whether experiences of being NEET have long term adverse effects on health outcomes in the Scottish context between 1991 and 2011. We used the Scottish Longitudinal Study (SLS), which is a 5.3% representative sample of the Scottish population. We followed young people who were aged 16-19 in 1991 to 2001 and then to 2011 when they were aged 26-29 and 36-39 respectively. We explored whether young people who had NEET experiences in 1991 displayed higher risks of poor physical and mental health ten and twenty years later. The outcomes include the self-reported from the censuses and objective ones from NHS patient records such as hospitalisation and prescription. We used descriptive and modelling approaches in our analysis. Covariates include a number of individual socioeconomic characteristics and local area characteristics in the models. Our research found that the NEET status in 1991 appears to be associated with negative health outcomes in 2001 and 2011. However the association varies with outcomes and by gender.

Email: zf2@st-andrews.ac.uk

Does the Month of Birth of Children in Northern Ireland Affect their Educational Success and self-reported Health later in Life?
Stefanie Doebler, Ian Shuttleworth, Queen's University Belfast

British and international research in education consistently found a positive effect of children’s month of birth on their educational success and health later in life. Children born soon before the school-enrolment date have been found to be disadvantaged, as they are almost a year younger than their classmates. For example research in England and Wales, where school-enrolment starts on the 1 September found children born in July and August to be significantly more likely to perform worse: they obtain lower grades and are less likely to obtain a degree later in life than their older class-mates. The birth-month disadvantage was found across social strata. This paper tests whether the birth-month effect observed in England and Wales is also found in Northern Ireland. Northern Ireland is a well-suited setting for this comparison, because in contrast to England and Wales, the school-enrolment date is 1 July. Northern Ireland may thus (with caveats) function as a quasi-natural experiment. This paper presents findings from statistical analyses of the Northern Ireland Longitudinal Study (NILS), a representative sample of the Northern Irish population based on 2001 and 2011 Census data linked to administrative data including health records. OLS and binary logistic regression models of relationships between the month of birth, socio-structural, and childhood-deprivation indicators and two outcomes: educational success (having obtained a degree, having no academic qualification later in life) and self-reported health show a null-result. In Northern Ireland the month of birth does not statistically significantly influence educational success and self-reported health later in life.

Email: s.doebler@qub.ac.uk

Health, housing tenure, and entrapment 2001-2011: Does changing tenure and address improve health?
Myles Gould, School of Geography, University Leeds, Ian Shuttleworth, School of Geography, Archaeology and Palaeoecology, Queen's University Belfast

There is considerable academic literature on the inter-relationships between: housing tenure, health, and wider dimensions of social well-being measured at both the individual and area level (e.g. Marmot, 2010; Macintyre et al, 2002). Some academics have also considered the concepts of housing entrapment and selective placement (Smith & Easterlow, 2005). Indeed there is considerable political and policy debates about how health, housing tenure and economic status reduce/(re)produce social-spatial mobility patterns. This paper seeks to contribute to this research area by explicitly exploring the relationships between changing health and housing tenure status, and also spatial mobility between 2001 and 2011 using the Northern Ireland Longitudinal Study. A particular focus will be on the extent to which different tenure trajectories (e.g. movements from social rented to owner occupied housing) are associated with changes in health status, and how these are linked to different types of place. The analysis presented will describe different observed transitions (e.g. particularly changes in individuals’ limiting long-term illness and general health, 2001-2011). Following this, multilevel analysis (Gould & Jones, 1996) will be undertaken to determine individual and areal-level factors associated with changing housing tenure and health status. In doing so the paper will examine a selection of other key variables such as gender, age, social economic status (SES), education, and community background, 2001-2011. The paper will also explore relationships between the new chronic illness question included in the 2001 Northern Ireland Census and its interrelationships with housing tenure change (2001-2011).

Email: m.i.gould@leeds.ac.uk

Characteristics of and living arrangements amongst informal carers in England and Wales at the 2011 and 2001 Censuses: stability, change and transition
James Robards 1, Maria Evandrou 1,2,3, Jane Falkingham 1,2, Athina Vlachantoni 1,2,3, 1 EPSRC Care Life Cycle, Social Sciences, University of Southampton, 2 ESRC Centre for Population Change, Social Sciences, University of Southampton, 3 Centre for Research on Ageing, Social Sciences, University of Southampton

Informal caring in England and Wales has become a key social policy issue in relation to population ageing and expenditure cuts in local services of adult social care. At the 2011 and 2001 Censuses a question on the provision of informal (unpaid) care was included. Headline results from the 2011 Census showed that more people are likely to become informal carers at some point in their lives and informal caring of 20 hours or more per week had increased in prevalence from 2001. Using a 1% sample of England and Wales 2011 Census records matched to the 2001 Census responses from the same individuals, the Office for National Statistics Longitudinal Study, this paper investigates the characteristics of informal carers at 2001 and 2011, identifies transitions between caring intensities and for carers at 2001 identifies characteristics associated with repetition of informal caring at 2011. This is the first study to present results for informal caring transitions between 2001 and 2011. Results suggest that a greater number of people started caring at some point between 2001 and 2011 than stopped caring. Characteristics associated with informal caring at 2001 or 2011 show stability. Around a third of those caring at 2001 were also caring ten years later. Multivariate analyses to predict informal caring at 2011 among the carers at 2001 show that those providing 50 hours or more care in 2001 were the most likely to be caring at 2011, suggesting that past provision of care is crucial in predicting future caring.

Email: james.robards@soton.ac.uk