Health and socio-economic status over the life course

First results from SHARE Waves 6 and 7

Axel Börsch-Supan, Johanna Bristle, Karen Andersen-Ranberg, Agar Brugiavini, Florence Jusot, Howard Litwin, Guglielmo Weber, Axel Börsch-Supan, Johanna Bristle, Karen Andersen-Ranberg, Agar Brugiavini, Florence Jusot, Howard Litwin, Guglielmo Weber

• Publisher: De Gruyter Oldenbourg
• Year: 2019
• ISBN:
  9783110617450
• Edition: 1
• Topic: Economics
• Subtopic: Microeconomics
• Index: Economics

Health in later life is shaped by behavior and policies over the life course and reflects the differences between the societies in which we are ageing. This multidisciplinary book answers questions from all life course phases and its interconnections from a European perspective based on the most recent SHARE data, such as: How is our health related to personality traits and influenced by our childhood conditions and careers? Which role does our social network play? Which impacts of the different health care and societal regimes can we trace at older ages? Which are the differences and similarities across European countries?

Axel Börsch-Supan, Johanna Bristle, Karen Andersen-Ranberg, Agar Brugiavini, Florence Jusot, Howard Litwin and Guglielmo Weber

1A spotlight on health and life courses in Europe using SHARE Waves 6 and 7

1.1Health and socio-economic status over the life course

The Survey of Health, Ageing and Retirement in Europe (SHARE) puts special emphasis on the interplay among the triangular connections of health, social embeddedness and the socio-economic status of older individuals. Waves 6 and 7 add three important innovations to this triangle and make SHARE a highly powerful tool for investigating ageing societies in Europe. First, Wave 6 deepens the objective measurement of health via biomarkers obtained from dried blood spot samples (DBSS). Second, Wave 7 finally achieves the cross-nationality that was demanded in the SHARE-ERIC statutes and covers all 26 continental EU member states plus Switzerland and Israel. Third, Wave 7 strengthens longitudinality reaching far back into childhood by collecting life-history data in all 28 countries.

These three innovations substantially enrich the multidisciplinary SHARE data and belong together because health, economic and social status in later life emerge from complex interactions over the entire life course (see Figure 1.1).

Departing from a person’s biological make-up, parental conditions and early education (indicated by the left box in Figure 1.1), the trajectories of health, economic status and social embeddedness are not determined in isolation but through mutual interactions over the entire life course (as indicated by the many two-sided arrows between the three trajectories). Health, for instance, influences economic status because healthier bodies are likely to support higher learning capacities at younger ages and higher workloads at older ages (e.g. Deaton 2002). In turn, income inequalities are likely also to cause inequalities in health because richer individuals can afford higher out-of-pocket healthcare costs and may have easier access to healthcare, especially in certain healthcare systems (e.g., Smith 2003). Health behaviours, lifestyle and environmental and occupational conditions add to these mutual interactions between health and economic status and simultaneously introduce interactions with the social environment in which individuals live. For example, ample evidence exists that embeddedness in a good family background is beneficial for the health of the family members (Fagundes et al. 2011). An important insight of recent research is that these interactions manifest their effects starting very early in life and then accumulate during positive and negative feedback cycles over the entire life course (Heckman and Conti 2013) before they determine later-life health, economic and social outcomes at older ages (right box in Figure 1.1).

Figure 1.1: Conceptual background.
Source: Own illustration.

Many of these interactions can be modified by policies, such as education, workplace regulations, poverty prevention or healthcare (indicated by the boxes at the top and bottom of 1. 1). Some welfare state interventions directly affect health and employment. Early retirement, for example, is directly and often immediately influenced by the rules of the pension, disability and unemployment systems (Börsch-Supan and Coile 2018). Health is directly affected by healthcare systems (Sirven and Or 2011). In addition, long-run interventions of the welfare state exist, such as education, preventive healthcare and workplace regulations, which have complex indirect and interrelated effects over the life course on both health and employment. Preventive healthcare, for instance, not only increases health but also makes meaningful occupation feasible at older ages (Jusot et al. 2012). High workplace standards not only improve employment at older ages by reducing early retirement but also tend to improve physical and mental health (Reinhardt et al. 2013).

This volume presents 38 short studies that summarize SHARE-based research on these interactions over the lifecycle. These studies are showcases of the interdisciplinary and cross-nationally comparative research results obtained from Waves 6 and 7 of SHARE. Almost all contributions have a special focus on health. Collecting objective health data in Wave 6 was important because the health of the general population and, in particular, of the older aged population is very different across countries. Comparisons between SHARE, the English Longitudinal Study of Ageing (ELSA) and the US Health and Retirement Study (HRS) have documented that older people in continental Europe have better health than those in England. In turn, English people have better health than their American counterparts. For example, the percentages of individuals aged 50–74 years with at least one limitation in activities of daily living (ADL) is 12 per cent in the United States, 10 per cent in the United Kingdom and only 7 per cent in the European Union (Avendano et al. 2009).

Many reasons exist for these cross-national health differences. Healthcare systems are very different between the United States and Europe because almost all European countries have mandatory universal health insurance and the United States does not. Coverage, ease of access, co-payments, administrative rules and quality also differ across EU countries, as do historical life circumstances, income and wealth distributions, lifestyles and health behaviours. In addition, cross-Europe differences exist in the interactions between health-care systems and lifestyles, such as when healthcare systems attempt to influence health behaviours, and in social policies and programmes potentially affecting health across the life course. The latter includes differences in early education and childcare programmes, employment protection and support programmes during middle age and social security and pension systems affecting older individuals. Although the impact of many of these policies on social outcomes is well documented, the extent to which they influence health and contribute to differences in longevity among high-income countries has yet to be established.

Understanding the reasons for cross-national health differences requires that studies use comparable health measurements. The findings by Avendano et al. (2009) were based on comparable measures, but these were self-reports and may have suffered from reporting biases. Few studies use more objective measurements of health-related biomarkers. SHARE Wave 6 was designed to fill this gap: We collected dried blood spot samples (DBSS) from approximately 27,000 respondents in 13 countries, which is among the largest collection of DBSS from a representative adult population. The DBSS include a small calibration sample from Poland in which both DBSS and venous blood were collected. SHARE has also collected retrospective histories of participants’ life courses and health events. Although we are still awaiting the laboratory results, preliminary analyses and validations are described in Part 8 of this book.

The central innovation in SHARE Wave 7 was the collection of highly structured retrospect life histories that are fully harmonized in all continental EU countries, Switzerland and Israel. This collection was achieved by using electronic displays that show the timeline of events and risk factors in several dimensions, such as health, work, family and housing. This display permits the respondent to see related events in one domain (family) with events in another domain (health), which significantly facilitates recall and improves the accuracy of the retrospective data.

Although we are aware that the hindsight perspective may create reporting biases, Smith (2003) shows the power and usefulness of retrospective data in detecting associations between health and socio-economic variables. More specifically, Korbmacher (2014) demonstrated the accuracy of retrospectively collected employment histories in a large-scale validation study comparing SHARE with linked administrative data. Life-course data on the timing of the major social, health and economic events over long segments of the lifecycle have been shown to be extremely helpful in identifying the causal mechanisms in the dynamic and cumulative relationship among health, lifestyles and socioeconomic resources. Such data capture biological and socio-economic risk factors in early and mid-life, including health shocks, working conditions and behaviour during childhood and adulthood. Thus, these data allow us to not only quantify the long-lasting effects of early-life events (including interventions by health and social policies) on later-life health status but also to study potential behavioural channels causing associations between risk factors and health outcomes.

Many contributions in this book use the life-history data from Wave 7; some also use the earlier life histories collected in Wave 3. Therefore, this book is organized along the course of life. Part 1, edited by Agar Brugiavini and Guglielmo Weber, begins with the development of personality in early childhood and its influence until late in life. This part of the book features the inclusion of the widely used Big Five dimensions of personality in SHARE. Part 2, edited by Guglielmo Weber, describes how health inequalities are generated during the life course by inequalities in education and income. Part 3 is edited by Agar Brugiavini and uses another innovation of the SHARE data, namely, the ‘jobcoder’, to describe labour market careers, occupation and retirement. Part 4, edited by Axel Börsch-Supan, celebrates the inclusion of eight new countries in SHARE. Based on the life-history data, this part showcases social transitions in the new accession countries and the effects of the economic crisis. Part 5, edited by Howard Litwin, exploits the longitudinal social network data obtained in Wave 6, another innovation of the SHARE panel. Such data permit studies on the interactions between health and its social context. Healthcare and health behaviours are the subject of Part 6, which is edited by Florence Jusot. In Part 7, we focus on how objective health measures contribute to our understanding of the ageing process, which was edited by Karen Andersen-Ranberg. This topic is taken up in Part 8, edited by Axel Börsch-Supan, with a first glance at the dried blood spot analyses.

1.2Personality and childhood

In the seventh wave of SHARE, respondents were for the first time asked a set of questions aimed at eliciting their five most important personality traits (John and Srivastava, 1999). These traits, or factors, known as the Big Five are Openness to experience, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. For each of them, the respondents answered two – in one case, three – questions representing the high and the low poles (Rammstedt and John, 2007). Most respondents who participated in Wave 7 also answered a large number of questions on their life histories going back to their childhoods (with a special focus on their situation at age 10), and including the relationship they remember having with their parents.

In this part, edited by Agar Brugiavini and Guglielmo Weber, we present three chapters that analyse personality traits and their relationships with well-being, physical health and financial investment decisions late in life. We also present two chapters that relate the abuse respondents suffered from in their childhood with their current well-being and mental health.

A word of caution is necessary here. As Bertoni et al. note in Chapter 4, “estimating a model with adult-life outcomes as the dependent variable and personality traits as the explanatory variable is not without problems. First, personality cannot be measured directly: the measures used in our empirical analysis might be imperfect proxies of true non-cognitive abilities, introducing measurement error. Second, personality traits may themselves be the result of a dynamic process of investment in cognitive and non-cognitive skills.” We urge the reader to keep in mind these caveats when interpreting the evidence presented in this part.

The first chapter by Bracha Erlich and Howard Litwin describes in detail the Big Five personality traits and the manner in which they are elicited in SHARE. The chapter examines the way in which these traits vary with age – and finds that Agreeableness shows a moderate rise across age groups and Openness to experience and Extraversion show small declines. The authors are careful to point out that their evidence could alternatively be interpreted as cohort effects (younger respondents belong to cohorts born in more recent years) or might even reflect selective mortality (if more agreeable individuals live longer, for instance). Perhaps the most important take-home message is that some traits – Neuroticism and Conscientiousness – are age-invariant. This age-invariance is important because the authors also study how the personality attributes relate to life satisfaction and find that a major negative role is played by Neuroticism and positive roles are played by Extraversion and Conscientiousness.

The second chapter, written by Jonathan Shemesh, Ella Schwartz and Howard Litwin, concentrates on personality and physical health. The authors point out that the determinants of health include genetics and lifestyle, such as physical activity and diet. However, they cite recent evidence suggesting that personality also exerts significant effects on health throughout the lifespan (Murray and Booth 2015). The authors investigate the manner in which personality traits are associated with a number of health indicators in older European adults, controlling for age, gender, marital status, financial capacity, years of education and country of residence. The strongest and most consistent personality-level correlates of good health are shown to be high Conscientiousness and low Neuroticism. To the extent that these personality traits are stable over time, as we observed in the previous chapter, and are not themselves affected by health, one might be tempted to interpret this finding as indicating a possible cause of good health in old age. However, even if this is not the case, an interesting policy implication is that personality testing can be used to assess the health risks of older people.

The role of personality traits is also investigated in further research areas in addition to health. It is well known from the behavioural finance literature that an investor’s personality traits are significantly associated with his or her financial behaviour, even conditioning on his or her partner’s cognition and other observable characteristics. In the chapter ‘Personality traits and financial behaviour’, Marco Bertoni, Andrea Bonfatti, Martina Celidoni, Angela Crema and Chiara Dal Bianco show that the personalities of both partners matter in determining household financial decision making, although to different extents and through different traits. The authors find that some personality traits of the financial household head (that is, the person who volunteered to answer questions on assets and debts), namely, Consciousness and Neuroticism, are significantly associated with stock market participation. Also reported was that risk aversion plays a role, but its role is much less strong when the likelihood of having financial liabilities is considered. In this case, whenever the financial household head is a female, her degrees of Agreeableness and Neuroticism are positively associated with proneness to indebtedness, and the Openness and Consciousness of her (male) partner also attract significant coefficients.

The following chapter by Noam Damri and Howard Litwin, titled ‘Relationships with Parents in Childhood and Well-Being in Later Life’, shows that the familial environment that older Europeans experienced during childhood is associated with their well-being in later life. The chapter contributes to the vast socio-economic literature on the long-term effects of early childhood events by focusing on the broad quality of the parent–child relationship and more specifically on physical abuse by either parent. The authors show that the six survey questions can be combined into two factors: one for the quality of the relationship with the parents, and the other for the frequency of parental abuse. They then take two different measures of (current) well-being and relate them to the two indices of the childhood interpersonal environment in the home, controlling for a host of confounders that are generally associated with these same measures, such as age, gender, education, number of children, marital status, health status, financial status and social activity.

Their key finding is that people who had good relationships with their parents show higher well-being scores at older ages, whereas those who grew up in an abusive familial environment show a lower quality of life scores in old age. This result can partly be attributed to a form of recall or justification bias whereby unhappy people tend to blame others for the low quality of their lives, whereas happy people put to rest their memories of past negative events. Even if this is the case, the implication of these findings for policy and practice are that the childhood interpersonal environment of older people needs to be addressed when dealing with ways to maintain or promote well-being in late life.

The last chapter of this part is written by Raluca E. Buia, Matija Kovacic and Cristina E. Orso. Similar to the preceding chapter, the motivation of this chapter is the concern that adverse childhood experiences may exert a negative influence on emotional well-being later in life. However, and unlike the preceding chapter, this contribution puts a specific focus on mental health problems. The authors investigate the extent to which exposure to adverse early life experiences favours the onset of emotional disorders. The chapter addresses the potential relationship between emotional neglect and physical harm in childhood and adolescence, and the onset of emotional disorders later in life. The authors recognize a potential recall bias: if depressed individuals tend to remember negative episodes more than otherwise identical individuals, this situation may lead to an overestimate of the effect. However, to the extent that such bias is invariant across individuals, the reader may want to focus on some very interesting differences in the effects of adverse childhood circumstances by cohort and gender.

First, the authors find that the intensity of the effects of adverse childhood experiences on mental well-being displays important differences between the pre- and post-war cohorts. A poor relationship with parents has a stronger and more significant impact on the post-war cohort, whereas having experienced physical harm from parents is not significantly different from having experienced zero harm for the pre-war cohort. However, physical abuse from persons outside the family has a more important effect for the older respondents. Interestingly, most adverse childhood experiences have a stronger and more significant impact on women. In particular, the analysis shows that ...

Table of contents

• Cover
• Title Page
• Copyright
• Contents
• Editors
• Contributors
• 1 A spotlight on health and life courses in Europe using SHARE Waves 6 and 7
• Part I Personality and childhood
• Part II Health inequalities — Education and income
• Part III Labour market, occupation and retirement
• Part IV Social transitions and economic crises
• Part V Social context and health
• Part VI Healthcare and health behaviour
• Part VII Objective health
• Part VIII Dried blood spot samples