Ageing Strand Abstracts

Strand organiser: Gloria Chepngeno-Langat, University of Southampton

Ageing populations: Monday 8 September 1.30pm

Moving towards a gerontocracy? The influence of demographic changes and legal reforms on the age structure of the German electorate
Nora Sánchez Gassen, ESRC Centre for Population Change, University of Southampton

With population ageing well on its way, politicians, media and interest groups across many European countries have suggested that older people will become a pivotal electoral group in the near future and that inter-generational conflicts about policies and public spending may arise (“war of generations”). Nonetheless, the exact degree to which electorates will age is not clear, since demographic projections of electorates have so far not been calculated. This study provides such calculations for the first time by focusing on Germany. This country already has the largest proportion of older persons among all EU countries and is expected to age further in the future. I calculate a set of scenario-based multistate population projections and quantify the extent to which the electorate in Germany will age in the coming decades. In addition, I incorporate policy scenarios into the projections to estimate whether changes in electoral law, e.g. a lowering of the minimum voting age, could slow down this process. The projection results suggest that the median age of the electorate in Germany will indeed increase by 6 to 7 years during the next two decades unless radical policy changes are implemented. Understanding the extent to which electorates will age in the future is important, since changes in the composition of the electorate may influence political debates, election results and policy outcomes. The study closes by discussing to what extent the ageing of the German electorate may influence politics and policy making in the years to come.


To downsize or not? Housing adjustment at older ages in Scotland since 1991
Francesca Fiori, Elspeth Graham, Zhiqiang Feng, Centre for Population Change, University of St Andrews

In the past decades, the proportion of older households in Scotland has increased. At the same time, social and economic changes have prompted greater diversity in individual life courses and housing careers. This paper investigates the housing consumption of older individuals and couples, and the extent to which this is adjusted in response to changes in household composition. In particular, it examines who is and who is not downsizing and whether moves to smaller housing units have increased since the economic downturn of 2008. Data from the Scottish Longitudinal Study are used to examine the housing adjustments of older adults who change residential address between censuses, as well as the characteristics of those who do not move. Repeated cross-sectional analyses compare two decades (1991-2001 and 2001-2011) to observe decadal changes in residential mobility and immobility, and whether housing adjustment – especially downsizing – has become more or less common. The samples consist of older adults aged 55-69 at the beginning of each decade. Covariates includes individual socio-economic characteristics and housing characteristics at the beginning of the decade and changes in health, work and family composition across the decade. Preliminary findings indicate significant differences by socio-economic status in the likelihood of a residential move and in the direction of housing adjustment, with some older individuals upsizing while others downsize. Changes in individual and family conditions across the decade also play an important role.


Paying for care costs in later life using the value of the home
Les Mayhew Cass Business School David Smith, Cass Business School Faculty of Actuarial Science

Paying for care costs in later life using the value in people’s homes Even though the upward trend in life expectancy was anticipated in the 1980s, the rate of growth in the number of older people has consistently exceeded forecasts. Life expectancy continues to increase and one of the clear trends is that males are catching up with females. With an ageing population that includes a higher proportion of people with manageable illnesses, health and care costs will rise. It is expected that there will be a massive increase in the demand for social care which unlike health care must be paid for out of pocket and is subject to a means test. Recent reforms resulting from the Dilnot Commission propose a cap on care costs at £72k. Income and wealth distribution in the 65+ population is skewed towards housing wealth to such an extent that many could be forced to sell their homes to pay for care. At the same time there is considerable interest in using the capital value in people’s home to protect their assets through insurance or to draw down those assets gradually to pay care costs. This paper will explore the distribution of income and assets in the older population and propose different funding solutions for people in different circumstances, in particular those who are asset rich and income poor. It will point out that the home values have outstripped care costs such that the multiple between house prices and care costs in a nursing home have increased from about 3-fold to 10-fold over two decades. It will include discussions of ways people could use equity in their home to pay for care and also address possible funding solutions for non-home owners. The paper will draw on our recent research and engagement with policy makers and joint research with the think tank, DEMOS. It will also touch on behavioural insights into why people are so reluctant to plan or save for possible care costs in later life and policy solutions that might alter this tendency.


Ageing: transition in health – Tuesday 9 September, 1.30pm

The health of grandparents caring for their grandchildren: The role of early and mid-life conditions
Giorgio Di Gessa, Karen Glaser, Anthea Tinker, King’s College London

Background: Grandparents are an important source of childcare. However, concerns have been raised that caring for grandchildren may come at the expense of grandparents’ own wellbeing. Our study examines the cross-sectional and longitudinal associations between various types of grandparental childcare and grandparents’ own health, and focuses on the extent to which such associations are directly and indirectly affected by cumulative advantage/disadvantage across the life course.

Methods: We used a sample of grandparents aged 50+ from waves 1-4 of two nationally representative longitudinal studies of older people from selected European countries, i.e. the English Longitudinal Study of Ageing, and the Survey of Health, Ageing and Retirement in Europe. Both datasets provide detailed retrospective life histories, including childhood characteristics (such as health and socio-economic position), and adulthood conditions (marital history, unemployment spells, adverse life events such as bereavement). The health outcome considered was self-rated health. Both logistic regressions and structural equation models were used in order to disentangle how and whether childhood and adulthood factors interact –both directly and indirectly– with grandparental childcare to affect grandparents’ own health.

Results: We found little evidence to suggest that intensive grandparental childcare provision has negative effects on grandparents’ health. We also found some evidence to suggest health benefits to grandparents who provide occasional childcare. Grandparents who co-reside with grandchildren (either in skipped- or multi-generation households) are more likely to report poor self-rated health. However, such health differences are largely associated with grandparents’ prior characteristics – particularly childhood circumstances and work history. Discussion: Findings suggest that better health among grandparents who provide grandchild care do not appear to be a consequence of care provision per se, but rather of their initial health and socio-economic advantage.


Cohort differences in the levels and trajectories of frailty among older people in England
Alan Marshall, James Nazroo, University of Manchester

In this paper, we use the English Longitudinal Study of Ageing to model cohort-specific trajectories (change) in levels of frailty among the older (age 50+) community-dwelling population (2002 to 2010) in England. We stratify our analysis by wealth and sex and use a frailty index, based on accumulation of ‘deficits’. For the older population and for males and females, frailty trajectories for adjacent trajectories converge between the ages of 50 and 70. However, levels of frailty are higher in more recent compared to later cohorts at the older ages (70+). These cohort differences are observed most strongly in the poorest quintile, whilst for the most affluent, frailty trajectories overlap across all adjacent cohorts suggesting no change in levels of frailty over time. A number of factors are likely to explain the cohort differences we observe including increased survival of frail individuals as well as increases in the frailty of more recent compared to older cohorts. Importantly, this paper illustrates that the social conditions experienced across the wealth distribution have important impacts on the rate of deficit accumulation in older populations. Our results on trajectories of frailty between 2002 and 2010 are pessimistic and, in the context of rising life expectancies, suggest that older people spend these additional years of life in a frail state. Whilst the gender frailty gap appears to have remained constant over the study period, our results suggest a widening of wealth inequalities.


Is healthy life expectancy increasing in terms of cognitive functioning? 
Valeria Bordone, Nadia Steiber, Sergei Scherbov, Wittgenstein Centre for Demography and Global Human Capital (IIASA, VID/ÖAW, WU)

Higher chronological age tends to be associated with lower cognitive functioning, especially regarding cognitive skills that are based on the brain’s capacity to process and memorize new information. In line with increases in healthy life expectancy, the question is whether or not people of a certain age today (e.g. age 60-75) perform better than people of the same age in the past. With this study, we aim to answer this question using data from the English Longitudinal study of Ageing (ELSA), the Survey of Health, Ageing and Retirement in Europe (SHARE), and the German Socio-Economic Panel (SOEP). These data allow us to investigate cognitive functioning in a cross-country comparable framework along three dimensions (memory, verbal fluency, and the speed of processing). Preliminary results based on methods that overcome retest effects, which tend to produce upward bias in longitudinal studies of cognition, suggest that cognitive functioning has generally improved over time (i.e. across survey waves). This result holds for women and men, all age groups, and educational strata. Moreover, this result holds for all of the three surveys analysed. Several explanations will be proposed.


Predicting the start of care in later life: Exploring the relationship between the development of needs and the commencement of care using the English Longitudinal Study of Ageing (Waves 1 to 4).
Robert Sanders, University of Stirling

Current population projections suggest that ‘by 2034, 23 per cent of the population is projected to be aged 65 and over compared to 18 per cent aged under 16’ (ONS, 2010). Current debate on the impact of current and future increases in the older population, particularly the oldest old who potentially represent an expanding and increasingly dependent section of society with the greatest risk of requiring long-term care, have focused on management of expected rising costs associated with providing formal care services. In this context, the only source of support for many older people remains unpaid care provided by family or friends. Within the wider continuum of care, the interrelationship between informal and formal care is central in understanding the continuing care arrangements of older people. The research uses data from four waves of the English Longitudinal Study of Ageing (ELSA) to investigate the effect of a number of key factors (including gender, age, socioeconomic status, and potential availability of informal care) on the probability of receiving care from informal, formal and privately-paid sources. In order to disentangle the way in which different needs affect the types of care that people receive in later life the research uses the longitudinal nature of ELSA to explore how the onset of six activities of daily living (ADLs) may predict initial commencement of each of these types of care.


Ageing: Determinants of health and wellbeing in later life – Wednesday 10 September 11.00am

A gendered analysis of protective factors for old age mortality in the Nairobi slums, Kenya
Rachel Bennett, Gloria Chepngeno-Langat, Maria Evandrou and Jane Falkingham, University of Southampton

Urbanisation and population ageing are occurring simultaneously in many sub-Saharan African countries, contributing to an increasingly pressing need to understand which factors enable older people in this region to navigate the challenges associated with city life. Research has shown that slum dwellers face disproportionate health disadvantages relative to the rest of the population yet very few studies have offered gendered explanations for the health inequities within informal settlements. This paper uses unique longitudinal data collected in two Nairobi slums to examine the role of individual characteristics, social networks, health status and socio-economic status as protective factors for mortality amongst older men and women. The results indicate that women have poorer health and socio-economic status and contrasting social networks to men, and female gender is only a significant protective factor for mortality after adjusting for social networks, health and socio-economic status. The distribution of protective factors is more dispersed amongst men than women and the gender-specific mortality analyses reveals more factors differentiate mortality risk amongst men than women. Therefore the finding that gender is only significant after controlling for other factors is indicative of both the cumulative disadvantage of women: masking the female survival advantage observed across world regions, and large within-gender differences amongst men: failure to account for which exaggerates a male advantage in survival. Consequently, there is a need for gender sensitive policies which seek to both redress the disadvantages women face across the life course and to account for the special needs of both genders as they grow old.


What can demography reveal about the association between social networks and self-rated health at older ages?
Heather Booth, Pilar Rioseco, Heather Crawford, Australian Demographic and Social Research Institute, Australian National University

Though studies abound on social networks (SN) and self-rated health (SRH), few address the role of demography in determining their association. Demography differentiates family networks from friend networks, determines family network structure, and influences friend and group network structure through homophily. Demography also defines the socio-structural context of the SN-SRH association. This study seeks to illuminate the role of demography in the SN–SRH association among older people by focusing on the intersection of three social sub-networks (family, friend and group) and seven demographic groups defined by age, sex and partnership status. The aim is to identify and understand differences in the SN-SRH association among demographic groups while also comparing the roles of different sub-networks. Data are from an Australian survey of seniors (n=2122). SRH is a binary variable contrasting Good and Poor (14%). Logistic regression is used to identify the predictors of Poor SRH in terms of network structure (size) and ties (frequency of contact) and the network social-support functions of instrumental support and emotional support. Covariates include a measure of disability. Though no effects are found in the total sample, the disaggregated analysis reveals differences among demographic groups in terms of the relative importance of different network variables within family, friend and group sub-networks. By using this three-dimensional analytical framework (network variables within sub-networks for each demographic group), we identify important interactions between specific sub-network ties and disability. It is concluded that disability is important to understanding the SN-SRH association. Possible explanations include disability-associated personal agency in networks.


Social determinants of health disparities and function at older ages in Ghana
Paul Ayernor, Department of Sociology, University of Oxford

The paper focuses on social determinants of health disparities and function at older ages made possible by additional years of life. Specifically, the paper investigates the following research questions: (1) how does age and gender affects health disparities and function at older ages? (2) Does socioeconomic status predict health disparities and function at older ages? (3) To what extent do individual and household wealth influence health disparities and function at older ages? (4) How much an effect type place of residence has on health disparities and function at older ages? In this paper, I reviewed the relevant theoretical and the empirical scholarship accruing on Ghana. I utilized data from a nationally representative survey of the Study on global Ageing and adult health (SAGE) Wave 1 (2007). A total of 5000 respondents aged 50 years old and over were interviewed. Descriptive and multivariate regression models are used to assess the social determinants of health disparities and function at older ages. The paper is an ongoing work with incomplete analyses and findings therefore it should be considered for a poster session. However, the preliminary results show a gendered health disparity and functional impairment. Health disparity and functional impairment tend to mimic socioeconomic inequalities experienced at older ages. Both health disparities and functional impairment differ by place type of residence. Overall, health disparities and function at older ages have significant health and economic implications for the individual, family and country.