Elsevier

Social Science & Medicine

Volume 70, Issue 3, February 2010, Pages 428-438
Social Science & Medicine

Socioeconomic differences in health over the life cycle in an Egalitarian country

https://doi.org/10.1016/j.socscimed.2009.10.020Get rights and content

Abstract

A strong cross-sectional relationship between health and socioeconomic status is firmly established. This paper adopts a life cycle perspective to investigate whether the socioeconomically disadvantaged, on top of a lower health level, experience a sharper deterioration of health over time. Data are drawn from the Dutch Central Bureau of Statistics (CBS) Health Interview Surveys covering the period 1983–2000. The analysis focuses on the self-rated health and disability of persons aged 16–80. We show that in the Netherlands, as in the US, the socioeconomic gradient in health widens until late-middle age and narrows thereafter. The analysis and the available evidence suggests that the widening gradient is attributable both to health-related withdrawal from the labor force, resulting in lower incomes, and the cumulative protective effect of education on health outcomes. The less educated appear to suffer a double health penalty in that they begin adult life with a slightly lower health level, which subsequently declines at a faster rate.

Introduction

Health differs by socioeconomic status. The socially and economically advantaged enjoy better health, irrespective of whether it is measured by morbidity, disability or mortality. This strong socioeconomic gradient in health is firmly established in evidence from both the developed and the developing world (CSDH, 2008, Mackenbach et al., 2002, Marmot, 1999, Smith, 1999). Despite an abundance of literature, the causal mechanisms and pathways responsible for the association are still poorly understood. In part, this stems from the static nature of much of the analysis. Both health and socioeconomic status (SES) are multidimensional and dynamic, and the relationship between them may reflect different effects and feedbacks over the course of life. Does low economic status in early life lead to the development of health problems in middle and later life? Or, does poor health interfere with the acquisition of education and, subsequently, the chances of securing, or holding onto, a well-paid job? The difficulty of answering such questions led Cutler, Lleras-Muney, and Vogl (2008) in a recent review to remark that “(…) differential patterns of causality make a single theory of socioeconomic gradients in health difficult to imagine. We suspect, though, that the right theory will emphasize the life cycle.”

The advantage of bringing a life cycle perspective on the gradient has proved extremely useful already in the United States, where Smith, 1999, Smith, 2005a and Case and Deaton (2005) have challenged the common view that the socioeconomic gradient in health reflects the effect of socioeconomic status, in particular income, on health, arguing instead that a large part of the gradient derives from a feedback effect of health on income through labor force participation. Banks, Marmot, Oldfield, and Smith (2007) made a first attempt to unravel these mechanisms in the UK and their preliminary conclusion is that the same mechanism is able to explain an important part of the gradient there as well.

This paper does not aim to present evidence on the causal impact of SES on health, or vice versa, but rather to motivate such analysis by describing how socioeconomic differences in health evolve over the life cycle. More specifically, we investigate whether the stylized facts emerging from the US literature are also apparent in Dutch data. The Netherlands differs markedly from the US in relation to social structure, income inequality, health and disability insurance, social protection and health care organization (see e.g. Hurd & Kapteyn, 2003). For example, the gross replacement rate for disability insurance benefits in 1993 was 63 percent in the Netherlands, as opposed to only 30 percent in the US (MacFarlan & Oxley, 1996). Similar large differences exist for unemployment benefits and social assistance, where also the maximum period of entitlement is much longer in the Netherlands than it is in the US (MacFarlan & Oxley, 1996). No doubt as a consequence of these differences, socioeconomic inequality in health is much lower in the Netherlands than it is in the US (van Doorslaer et al., 1997). But little or nothing is currently known about whether and how the countries differ in the way in which the socioeconomic gradient in health changes over the life course. If there were no such differences, it would suggest that the observed patterns result from the fundamental relationships between education, occupation, work and health over the life course, and are not responsive to the social, health and economic policy environment.

A life cycle perspective additionally provides information on how much more rapidly health declines for some groups than others. While it is clear that, at any given age, the socially disadvantaged experience a lower level of health, there is no consensus over whether they can also expect their health to deteriorate more rapidly. On the one hand, proponents of the cumulative-advantage hypothesis maintain that differences in health by SES are established early in life and subsequently widen as the economic and health disadvantages of the less privileged interact and accumulate (House et al., 1994, Lynch, 2003, Ross and Wu, 1996, Willson et al., 2007). The competing view— the age-as-leveler hypothesis— maintains that deterioration in health is an inevitable part of the process of ageing irrespective of economic means or social position, with the result that the SES-health gradient narrows at advanced ages (Beckett, 2000, Herd, 2006). If there is cumulative advantage, then interventions that were effective in breaking this process by expanding opportunities to socially disadvantaged groups earlier in life would have large pay-offs in terms of improved health and labor-market prospects over the life cycle.

As in the US, we find that socioeconomic differences in health first diverge, reach a peak around late middle-age, and then converge in old age. It appears that a large part of the socioeconomic gradient in health is governed by labor force participation. Once one restrict attention to working individuals, health differences across income groups are greatly reduced and do not widen with age up to middle-age. The convergence in old age may partly stem from selective mortality—only the most robust of the lower income groups survive—although it appears also to reflect the reduced dependence of income on health after retirement.

The paper is organized as follows. Section 2 summarizes existing evidence on the SES health gradient over the life cycle. Although we will often refer rather loosely to the ‘life cycle’, since we are particularly interested in the interactions between health, work and income, we restrict attention to the years of adultood. For this reason, we do not survey either the ever-expanding literature on socioeconomic differences in the evolution of child health (e.g. Case et al., 2002, Currie and Stabile, 2003, Currie et al., 2007, Murasko, 2008), or that on the impact of early-life conditions on health in adulthood (Barker, 1995, van den Berg and Lindeboom, 2007, van den Berg et al., 2006). In Section 3, we introduce the data and methods, and in Section 4 evidence on the SES-health gradient over the adult life course in the Netherlands is presented. In Section 5 we consider the consequences of the evidence presented for health inequalities and policies to tackle them.

Section snippets

Related literature

Case and Deaton (2005) report large differences in average self-reported health (SRH) in the US by income quartile that increase up to age 50–55, before narrowing particularly after age 60 until they disappear by age 80. A similar life cycle pattern has been observed for other indicators of SES and health (Smith, 2005a, Case and Deaton, 2005) and in one European (Belgian) study (Deboosere & Neels, 2008).

Interpretations of the observed widening and then narrowing of the SES-health gradient with

Data and methodology

We examine socioeconomic differences over the adult life cycle in self-reported health, disability and mortality.

The life cycle profile of self-reported health by income

Fig. 1 shows a clear income gradient in reported bad health at all but the youngest ages. For example, there is a 10% prevalence of morbidity in the bottom income quartile of females already at age 35; while in the top quartile this prevalence is only reached at 65. The age profile of the gradient displays striking consistency with the evidence reviewed in Section 2. For both genders, but particularly for males, income differences in health diverge until 55 or so, at which point almost 30% of

Discussion

Our analysis demonstrates that in the Netherlands socioeconomic differences in health widen until late middle-age before narrowing in old-age. This life cycle profile in the socioeconomic-health gradient is remarkably similar to that observed in the US despite the stark differences that exist between the countries in characteristics such as health system finance and organisation, income inequality, social protection and disability insurance, which would be expected to influence socioeconomic

References (47)

  • J. Banks et al.

    The SES health gradient on both sides of the Atlantic

    (2007)
  • D.J.P. Barker

    Fetal origins of coronary heart disease

    British Medical Journal

    (1995)
  • M. Beckett

    Converging health inequalities in later life: an artifact of mortality selection?

    Journal of Health and Social Behavior

    (2000)
  • G. van den Berg et al.

    Birth is the messenger of death – But policy may help to postpone the bad news; new evidence on the importance of conditions early in life for health and mortality at advanced ages

    (2007)
  • G. van den Berg et al.

    Economic conditions early in life and individual mortality

    American Economic Review

    (2006)
  • J. Bound

    Self reported versus objective measures of health in retirement models

    Journal of Human Resources

    (1991)
  • A. Case et al.

    Broken down by work and sex: how our health declines

  • A. Case et al.

    Economic status and health in childhood: the origins of the gradient

    American Economic Review

    (2002)
  • P. Contoyannis et al.

    The dynamics of health in the British household panel survey

    Journal of Applied Econometrics

    (2004)
  • E. Crimmins

    Trends in the health of the elderly

    Annual Review of Public Health

    (2004)
  • CSDH

    Closing the Gap in a Generation: Health Equity through Action on the Social Determinants of Health

    (2008)
  • J. Currie et al.

    Socioeconomic status and child health: why is the relationship stronger for older children?

    American Economic Review

    (2003)
  • D.M. Cutler et al.

    Socioeconomic status and health: Dimensions and mechanisms

    (2008)
  • Cited by (65)

    • Lifetime socioeconomic determinants of health trajectories among older adults

      2021, Advances in Life Course Research
      Citation Excerpt :

      There is also some evidence supporting the age as-leveler hypothesis. This hypothesis posits that health inequalities may diminish as a consequence of older age if health declines surpass cumulative life-course disadvantages (Dupre, 2007; Siegel, Luengen, & Stock, 2013; van Kippersluis et al., 2010). The higher mortality of poorer individuals at younger ages or the reduction in inequality with older age may explain this effect (Dupre, 2007; Siegel et al., 2013; van Kippersluis et al., 2010).

    • Biomarkers, disability and health care demand

      2020, Economics and Human Biology
    View all citing articles on Scopus

    This paper derives from the NETSPAR funded project “Income, health and work across the life cycle”. The authors acknowledge access to linked data resources (RIO 1998–2001, 2002–2004, and DO 1998–2005) by the Netherlands Central Bureau of Statistics (CBS). We are grateful to James Smith, Federica Teppa, Anton Kunst, Arthur van Soest, Bart Boon, Ivana Gomes Durão, Peter Kooreman, Bernard ter Haar, and two anonymous referees for comments on an earlier version of this paper. Tom Van Ourti is a Postdoctoral Fellow of the Netherlands Organisation for Scientific Research – Innovational Research Incentives Scheme – Veni.

    View full text