TY - GEN
T1 - Tracing the societal impact of variations in health: Three empirical contributions
AU - Aaskoven, Maiken Skovrider
PY - 2021/8
Y1 - 2021/8
N2 - Health is one of the most important dimensions of welfare. Variations in health will therefore have
great societal implications. Adverse health events may impact individual’s and others’ opportunities
and outcomes in life such as education, labour market participation, and overall well-being. Personal
characteristics and social circumstances influence individuals’ ability to invest in health, which may
make a significant difference when faced with adverse health events. In order to improve societal
welfare, there is a need for a better understanding of adverse health events, to what extent they may
impact the welfare of the individual and others, and whether there is heterogeneity in ability to cope
with such events. These are the aims of this PhD thesis.To fulfil the aims of the thesis, I utilise available observational data which enables me to approach
causal inference in the absence of exogenous shocks. I exploit three comprehensive datasets including
both survey data and register data which enables me to overcome some of the shortcomings of the
individual data sources. Combining the versatility of the survey and the comprehensiveness of the
registers generates highly enriched data which broadens the scope and quality of the research.This thesis encompasses three empirical contributions that examine the societal impact of variations
in health focusing on: 1) negative spillover effects of health shocks on the next generation’s human
capital, 2) the impact of long-lasting health detriments on individual utility, and 3) the impact of
health on individual productivity.Paper I studies the intergenerational transmission of human capital. A rich longitudinal registerbased dataset of Danish children is utilised to investigate whether severe parental health shocks affect
children’s school achievements. Coarsened exact matching is used in combination with parametric
models to control for potential endogeneity. Strength of the shock is measured using cancer specific
survival rates independent of stage at diagnosis. There are significant negative (although small)
effects of a parental health shock on children’s basic school grades as well as their likelihood of
commencing and completing secondary education. The risk of attenuated educational attainment
appears stronger in the case of severe parental cancer (i.e., poor prognosis) and parental cancer death.
More detailed analyses indicate a presence of mainly mortality, but also morbidity effects. There is
heterogeneity with respect to the children’s gender and age and gender of the ill parent, but no effects
v
of family-related resilience factors such as child’s ethnicity or parental education level. The effects
appear to be driven by non-pecuniary costs such as time and emotional investments rather than by
pecuniary costs.Paper II studies the impact of long-lasting health detriments on individual utility. A cross-sectional
survey of Danish citizens aged 50-80 is used in combination with administrative register data to
investigate the association between chronic illness and subjective wellbeing (SWB) among the
elderly. Six common chronic illnesses are identified using an algorithm: asthma, COPD, rheumatoid
arthritis, osteoporosis, type 1 and type 2 diabetes. The Personal Well-being Index (PWI) is used to
measure SWB in seven domains of life: standard of living, personal health, achieving in life, personal
relationships, personal safety, community connectedness, and future security. To mitigate
endogeneity, OLS is employed and adjusted for numerous confounders such as socio-economics and
personality traits. A range of heterogeneity analyses explore the role of 16 personal characteristics as
moderators of the relationship between SWB and chronic illness. Furthermore, two possible sources
of variation in SWB are examined. There is a significant negative association between chronic illness
and elderly’s SWB. All the included domains of life are affected by chronic illness. There is little
evidence of personal characteristics moderating the association between SWB and chronic illness,
although some tendencies are seen for ethnicity, extraversion, emotional stability, openness to
experiences, and internal health locus of control. Heterogeneity in SWB across personal
characteristics may be explained by poorer health state and in some cases by differential marginal
utility of health. Paper III studies how variations in health affect labour market participation. Several countries,
including Denmark, face demographics changes that threatens the financial sustainability of the
pension systems. Therefore, they have linked the statutory retirement age with increases in life
expectancy. The implicit assumption is that older workers have the health capacity to work longer.
The validity of the assumption is examined as well as the fairness of pension systems that link the
statutory retirement age with increases in average life expectancy. The combination of the Survey of
Health, Ageing and Retirement in Europe (SHARE) and Danish register data allows for composite
health indices that an extensive range of health indicators. Health capacity to work is estimated using
the Cutler et al. (2013) method. Health capacity is decomposed into physical and mental health as
well as subjective and objective health. Health capacity to work is investigated across educational
vi
and occupational attainment. Substantial additional work capacity is found for older Danes aged 55
and above, indicating that individuals retire for reasons beyond health. Depending on the type of
health index applied, the health capacity to work varies, especially at the individual level. There is
evidence of an educational and an occupation gradient in ability to work.Overall, this thesis shows that variations in health impact societal welfare, even in a society with an
extended welfare system. Individual utility and productivity are reduced when faced with health
detriments and there are negative spillover effects on the next generation’s human capital. Some
personal characteristics drive individuals’ ability cope with adverse health events. Both mental and
physical health affect welfare, and strength and trajectory of adverse health events modify the impact
of such events. These findings may guide policy makers. Finally, this thesis demonstrates how rich
observational register data and survey data can be used for policy-relevant research and for inferring
plausible causal effects.
AB - Health is one of the most important dimensions of welfare. Variations in health will therefore have
great societal implications. Adverse health events may impact individual’s and others’ opportunities
and outcomes in life such as education, labour market participation, and overall well-being. Personal
characteristics and social circumstances influence individuals’ ability to invest in health, which may
make a significant difference when faced with adverse health events. In order to improve societal
welfare, there is a need for a better understanding of adverse health events, to what extent they may
impact the welfare of the individual and others, and whether there is heterogeneity in ability to cope
with such events. These are the aims of this PhD thesis.To fulfil the aims of the thesis, I utilise available observational data which enables me to approach
causal inference in the absence of exogenous shocks. I exploit three comprehensive datasets including
both survey data and register data which enables me to overcome some of the shortcomings of the
individual data sources. Combining the versatility of the survey and the comprehensiveness of the
registers generates highly enriched data which broadens the scope and quality of the research.This thesis encompasses three empirical contributions that examine the societal impact of variations
in health focusing on: 1) negative spillover effects of health shocks on the next generation’s human
capital, 2) the impact of long-lasting health detriments on individual utility, and 3) the impact of
health on individual productivity.Paper I studies the intergenerational transmission of human capital. A rich longitudinal registerbased dataset of Danish children is utilised to investigate whether severe parental health shocks affect
children’s school achievements. Coarsened exact matching is used in combination with parametric
models to control for potential endogeneity. Strength of the shock is measured using cancer specific
survival rates independent of stage at diagnosis. There are significant negative (although small)
effects of a parental health shock on children’s basic school grades as well as their likelihood of
commencing and completing secondary education. The risk of attenuated educational attainment
appears stronger in the case of severe parental cancer (i.e., poor prognosis) and parental cancer death.
More detailed analyses indicate a presence of mainly mortality, but also morbidity effects. There is
heterogeneity with respect to the children’s gender and age and gender of the ill parent, but no effects
v
of family-related resilience factors such as child’s ethnicity or parental education level. The effects
appear to be driven by non-pecuniary costs such as time and emotional investments rather than by
pecuniary costs.Paper II studies the impact of long-lasting health detriments on individual utility. A cross-sectional
survey of Danish citizens aged 50-80 is used in combination with administrative register data to
investigate the association between chronic illness and subjective wellbeing (SWB) among the
elderly. Six common chronic illnesses are identified using an algorithm: asthma, COPD, rheumatoid
arthritis, osteoporosis, type 1 and type 2 diabetes. The Personal Well-being Index (PWI) is used to
measure SWB in seven domains of life: standard of living, personal health, achieving in life, personal
relationships, personal safety, community connectedness, and future security. To mitigate
endogeneity, OLS is employed and adjusted for numerous confounders such as socio-economics and
personality traits. A range of heterogeneity analyses explore the role of 16 personal characteristics as
moderators of the relationship between SWB and chronic illness. Furthermore, two possible sources
of variation in SWB are examined. There is a significant negative association between chronic illness
and elderly’s SWB. All the included domains of life are affected by chronic illness. There is little
evidence of personal characteristics moderating the association between SWB and chronic illness,
although some tendencies are seen for ethnicity, extraversion, emotional stability, openness to
experiences, and internal health locus of control. Heterogeneity in SWB across personal
characteristics may be explained by poorer health state and in some cases by differential marginal
utility of health. Paper III studies how variations in health affect labour market participation. Several countries,
including Denmark, face demographics changes that threatens the financial sustainability of the
pension systems. Therefore, they have linked the statutory retirement age with increases in life
expectancy. The implicit assumption is that older workers have the health capacity to work longer.
The validity of the assumption is examined as well as the fairness of pension systems that link the
statutory retirement age with increases in average life expectancy. The combination of the Survey of
Health, Ageing and Retirement in Europe (SHARE) and Danish register data allows for composite
health indices that an extensive range of health indicators. Health capacity to work is estimated using
the Cutler et al. (2013) method. Health capacity is decomposed into physical and mental health as
well as subjective and objective health. Health capacity to work is investigated across educational
vi
and occupational attainment. Substantial additional work capacity is found for older Danes aged 55
and above, indicating that individuals retire for reasons beyond health. Depending on the type of
health index applied, the health capacity to work varies, especially at the individual level. There is
evidence of an educational and an occupation gradient in ability to work.Overall, this thesis shows that variations in health impact societal welfare, even in a society with an
extended welfare system. Individual utility and productivity are reduced when faced with health
detriments and there are negative spillover effects on the next generation’s human capital. Some
personal characteristics drive individuals’ ability cope with adverse health events. Both mental and
physical health affect welfare, and strength and trajectory of adverse health events modify the impact
of such events. These findings may guide policy makers. Finally, this thesis demonstrates how rich
observational register data and survey data can be used for policy-relevant research and for inferring
plausible causal effects.
U2 - 10.21996/2y8p-j530
DO - 10.21996/2y8p-j530
M3 - Ph.D. thesis
PB - Syddansk Universitet. Det Sundhedsvidenskabelige Fakultet
ER -