Health Effects of Risky Lifestyles and Adverse Working Conditions: Are Older Individuals More Penalized?
| Author | Elena Cottini,Paolo Ghinetti |
| Published date | 01 December 2020 |
| DOI | http://doi.org/10.1111/bjir.12502 |
| Date | 01 December 2020 |
British Journal of Industrial Relations doi: 10.1111/bjir.12502
58:4 December 2020 0007–1080 pp. 973–1003
Health Effects of Risky Lifestyles and
Adverse Working Conditions: Are Older
Individuals More Penalized?
Elena Cottini and Paolo Ghinetti
Abstract
Using unusually rich panel data from Denmark, we investigate dierences by
age in the health implications of risky lifestyles and adverse working conditions.
Accounting for time-invariant unobserved heterogeneity, overall, we find no
health penalties for older workers(55 and over) compared to younger ones (18–
34; 35–54). However, the former suer more from the health consequences of
risky lifestyles — especially the lack of consumption of fruit and vegetables and
physical inactivity. Working conditions negatively relate with health, but fewer
dierences across age groups exist. Selection bias, namely the healthy worker
eect, does not alter our results.
1. Introduction
A growing body of evidencestudies how workers’ health depends on a variety
of physical, biological, behavioural and social dimensions. Some studies show
that a job’sphysical hazards and psychosocial stressors play a big role (Green
and Mostafa 2012). Others suggest that riskylifestyles, such as smoking, heavy
drinking, bad food habitsor physical inactivity, are keydeterminants of major
preventable diseases (Patja et al. 2005).
In this context, the age of workers plays an important role. On the one
hand, the ageing process incorporates several health risks. On the other hand,
health implications of work and non-work activities may depend on age,
limiting the work potential of older workers compared to younger ones.
The empirical literature on these issues is limited. Some studies focus on
age, but without specific reference to risky lifestyles and working conditions.
Others analyse only the population of older workers. Overall, detailed
longitudinal information on key variables is rarely available for the entire
Elena Cottini is at the Universit`
a Cattolica del Sacro Cuore – Milano. Paolo Ghinetti is at the
Universit`
a degli Studi del Piemonte Orientale.
C
2019 John Wiley& Sons Ltd.
974 British Journal of Industrial Relations
working-age population. This makes it dicult to control for fixed individual
unobservable health components correlated with both key variables and the
ageing process.
Wecomplement the literature in several ways. First, we present new evidence
on health dierentials by age, controlling for lifestyles, workingconditions as
well as individual fixed eects. Second, we explore if lifestyles and working
conditions do play a role for such dierentials. To this purpose, we allow
for the health gradient of key explanatory variables to be age-specific. Third,
we use a broader range of health measures and a richer set of lifestyles and
working conditions. This may help in developing a better understanding of
how health gradients are generated, which may have implications for policies
aimed at limiting the consequences of major risks at dierent ages (see,
e.g., Cai and Kalb 2006). To these purposes, we exploit a unique dataset,
the Danish Work Environment Cohort Study (DWECS), which includes
longitudinal information on lifestyles, working conditions and health at all
ages.
Dierently frommany other countries, in Denmark, the health implications
of risky behaviours and working conditions have long been central in the
policy debate. Danish authorities introduced several regulations to adopt
best practices for the improvement of working conditions and screening of
enterprises in terms of health and safety at work, as well as policies aimed at
promoting healthylifestyles and extending the working life. An example is the
‘Flex-Seniority’ initiative,which encourages older workers to remain active in
the labour market longer (see Barslund 2015). For these reasons, Denmark is
then an interesting and relevant country for our analysis, although its specific
institutional setting suggests taking particular care in extending the policy
implications to other countries.
We define three age groups (18–34; 35–54; 55 and over), five indicators for
working conditions and a comprehensive set of measures for four lifestyle
dimensions. We study their association with four health measures: self-
assessed health (SAH), mental health (MH), vitality (VT) and musculoskeletal
health (MSH). Our estimates describe robustassociations that in our preferred
specification account for time-invariant individual heterogeneity.
Random eects (RE) estimates show that, in general, older cohorts of
Danish workers experience worse health than younger ones. However, fixed
unobservable individual traits explain a large portion of these dierences.
Once controlling for that, perceived health tends to be similar across age
groups,except for VT.Overall, there are no systematic dierences in the health
eects of risky lifestyles and working conditions between age groups. This
applies in particular to the negative association between working conditions
and health, which is usually statistically significant but similar across age
groups.Still, a number of age gradients do exist, although they vary across the
health dimensions considered, and concern only specific lifestyles and working
conditions. For example, older Danish individuals (55 and over) suer the
health consequences of inappropriate food habits and physical inactivity,
especially for MH and VT.
C
2019 John Wiley& Sons Ltd.
To continue reading
Request your trial