Abstract
While the relationship between own health and subjective well-being is well documented, evidence is limited regarding the relationship of the latter with respect to the health of a family member. Recent models of human capital formation link human development to the stability of the home environment and to parental health. Using a unique longitudinal data from four developing countries, we extend this strand of the literature by investigating the role of family health on the well-being of adolescents. Our results show that family illness negatively affects subjective well-being. This effect is stronger for health problems of siblings than parents. We explore potential mechanisms and find that the most significant transmission channel for parental health is the shared social environment. The effect of the illness of siblings appears to be mostly driven by psychological factors.
The data used in this publication come from Young Lives, a 20-year study of childhood poverty and transitions to adulthood in Ethiopia, India, Peru and Vietnam (www.younglives.org.uk). Young Lives is funded by UK aid from the Foreign, Commonwealth & Development Office and a number of further funders. The views expressed here are those of the author(s). They are not necessarily those of Young Lives, the University of Oxford, FCDO or other funders.
You have full access to this open access chapter, Download chapter PDF
Similar content being viewed by others
Keywords
1 Introduction
Health is one of the most important predictors of subjective well-being (Clark et al. 2019).Footnote 1 A large literature shows that both physical health and mental health matter (Ferrer-i Carbonell and Van Praag 2002; Shields and Price 2005). While adverse health outcomes immediately affect well-being, they also have long-lasting effects in terms of happiness (Oswald and Powdthavee 2008).
Subjective well-being is also known to be relative. Clark and Oswald (1996) use the British Panel Household Survey to show that job satisfaction increases with own wage but diminishes with the wage level of colleagues. Using a quasi-natural experiment, Card et al. (2012) demonstrate that disclosing information on peers’ salaries lowers job satisfaction of workers below the median of their pay unit. However, there is also evidence that the relationship between one’s own well-being and other’s incomes can be positive. Using the Russian Longitudinal Monitoring Survey, Senik (2004) shows that the reference group’s income exerts a positive influence on individual life satisfaction. Similarly, D’Ambrosio et al. (2020) find that relative wealth increases life satisfaction in German data. This latter phenomenon has come to be known in the literature as the information effect: the presence of richer or wealthier individuals signals that there is a possibility for oneself to get richer in the future, which increases own well-being even before any actual enrichment takes place (Hirschman and Rothschild 1973).
The aforementioned articles often consider individuals with common characteristics (for instance age, gender, and education) as the reference group at the basis of these relative comparisons. However, one can also postulate that subjective well-being may be influenced by family members’ outcomes and behaviours. This is the case in Clark (2003) where life satisfaction of the unemployed is higher when they live with an unemployed partner. Powdthavee (2009) and Wunder and Heineck (2013) demonstrate that empathetic (or altruistic) reactions affect well-being to the extent that one’s life satisfaction is a direct function of the partner’s life satisfaction.
Very little is known about the link between the health of relatives and life satisfaction. Much of the related literature is limited to investigating partners’ health and find negative effects on subjective well-being (Van den Berg et al. 2014; Clark et al. 2019). Using the 1970 British Cohort Study and the Malaise score to measure mental health, Layard et al. (2014) and Clark and Lepinteur (2019) show that mothers’ mental health status in 1980 still affects subjective well-being of their children in 2000. To the best of our knowledge, only Powdthavee and Vignoles (2008) specifically address the question of whether parents’ mental distress influence their child’s life satisfaction. Using British data, they show that lower levels of paternal mental health reduces life satisfaction of children. Maternal mental health only affects that of daughters.
We here take advantage of a longitudinal birth-cohort data from four developing countries (i.e. Ethiopia, India, Peru, and Vietnam) and investigate the consequences of relatives’ health on subjective well-being of children. Our study’s contribution to the literature is twofold. First, we account simultaneously for the effect of the health of the father, the mother and the siblings on own life satisfaction. To the best of our knowledge our work is the first contribution addressing this point. Our approach paints a better picture of the key pathways through which parental and sibling health translates into child well-being. Siblings spend considerable time together, and siblings’ characteristics and sibling dynamics substantially influence developmental trajectories and outcomes. Second, most of the extant literature focused on developed countries and we here provide what we believe to be the first evidence from developing countries. This is particularly important since families in developing countries are likely to be further affected by potential health disparities, persistent poverty, and poor access to health services. The dataset we use allows us to explore how patterns are similar or different across those countries and make comparisons that are relevant for other countries with similar circumstances.
We first show that the health of the parents and the siblings are strongly correlated with the life satisfaction of the children of our estimation sample. We find that the effect of the health of siblings is more important than the effect of parental health. We then explore whether the underlying transmission mechanism of the observed correlation is due to objective causes, namely the shared environment and the parental investment decisions affecting the intra-household time allocation, or to psychological factors such as the information effect or empathy. We find that the health of the parents have little impact beyond the shared environment. This is particularly true for mothers. While having ill siblings reduces the probability of attending school and increases the time spent on paid and housework, the drop in life satisfaction due to the sibling’s illness seems to be mainly driven by psychological factors.
The remainder of the chapter is organised as follows. In Sect. 2 we outline a brief conceptual framework. Section 3 describes the data, and the empirical strategy appears in Sect. 4. Results are presented in Sect. 5. We conclude in Sect. 6.
2 Conceptual Framework
Several studies in Developmental Psychology have found that child well-being is predicted by family well-being and parenting quality (see Newland 2015, for an extended review). Studies in Economics that focuses on the effect of parental health, however, have a narrow scope and are limited mostly to school enrolment outcomes (Gertler et al. 2004; Sun and Yao 2010; Bratti and Mendola 2014). Recent models of human capital formation (Heckman 2007) link human development to the stability of the home and to parental (mental) health. The home environment has notable impacts on child well-being that may affect the stock of human capital in adulthood (Knudsen et al. 2006; Heckman 2007). The extant literature heavily focused on measuring child well-being by cognitive/non-cognitive standardized test scores. Conti and Heckman (2014) stress that well-being indicators must incorporate dynamic processes such as self-realization and the degree to which a person is fully functioning in society. We contribute to these strands of literature by asking if there is a robust relationship between the health of family members (e.g. parents, siblings) and subjective well-being (SWB) of adolescents.
Our conceptual framework is set-up in the spirit of Todd and Wolpin (2003) and Cunha and Heckman (2008), but we focus on the role of family circumstances in producing a child’s subjective well-being. The framework has several important characteristics. It models how environments and investments joined with own characteristics affect the evolution of well-being; it accounts for the preferences of the parents which help shape the investments in children; and it factors in the constraints that the families face (see Conti and Heckman 2014, for a comprehensive review of a human capital model).
In Fig. 1 we outline the basic conceptual framework for how we specify the empirical model of the relationship between family health shocks and child well-being outcomes. We consider key pathways from family health to child well-being: both the direct pathways, and the indirect pathways through parental investment decisions and the shared environment.
2.1 The Effects of Parental Investment and the Shared Environment on Child’s Life Satisfaction
The literature in the Economics of Happiness mostly explored the determinants of adult life satisfaction in developed countries. Little is comparatively known about the causes of life satisfaction of children and adolescents living in developing countries. However, several hypotheses can be formulated based on the extant literature.
Children with parents investing in their education are expected to have higher levels of life satisfaction for at least two reasons. According to Kahneman et al. (2004), housework and working time are the activities associated with the lowest levels of pleasures. In developing countries, children might be asked to quit school and spend more time in housework or on the labour market. Consequently, parental investments in education make children spending relatively less time in activities that would reduce their life satisfaction. Beyond its impact on current time allocation, education is expected to give children better expectations regarding their future life; and positive expectations about future outcomes are associated with higher contemporaneous life satisfaction (Senik 2008).
Gaps in human capital development of children across different socio-economic groups have counterparts in gaps in family environments. Francesconi and Heckman (2016) review the related empirical literature and conclude that disadvantaged children have compromised early environments as measured on a variety of dimensions. For example, children from disadvantaged environments are exposed to a substantially less rich vocabulary than children from more advantaged families. Cunha et al. (2013) also documents the lack of parenting knowledge among disadvantaged parents. Parenting styles are important determinants of early child development (Fiorini and Keane 2014; Del Bono et al. 2016). Consequently, a shared environment where resources (monetary and time) are scarce is expected to inhibit child development and happiness.
2.2 The Effect of Family Health on Child’s Life Satisfaction
2.2.1 Through Parental Investment
Health shocks to parents might reduce the time they invest in the education of their children; for instance, parental involvement in the child’s education and care-giving may reduce when one or both parents face serious illness (Dhanaraj 2016). Parents may also force children to participate in home/market work, or they may decide to take children out of school as the opportunity cost of children’s time increases. Bratti and Mendola (2014) study the effect of parental morbidity on child school enrolment and find that a young adult (age 15–24) with an ill mother but healthy father is significantly less likely to be enrolled in school. Alam (2015) use longitudinal data from Tanzania and shows that father’s illness reduces children’s education by decreasing their attendance.
2.2.2 Through the Shared Environment
Parental health shocks (in the absence of insurance) affect elements shared by the household by constraining its income (and consumption).Footnote 2 The usual mechanisms of consumption smoothing are often limited for resource constrained households in low and middle income countries due to the absence of well-developed credit and insurance markets (Jensen 2000). In this situation, households optimize by investing less in children (Becker and Tomes 1986). Luca and Bloom (2018) show that, regardless of gender, each parent’s health is significantly associated with his or her own labour market activity and earnings, as well as intrinsically linked to their spouse’s labour market status and earnings. They find paternal health to be associated with child schooling outcomes, especially for girls. However, child schooling outcomes are substantially and negatively affected when both parents suffer health shocks. The latter in the case of the father negatively affect aggregate household consumption, particularly food and education expenditures.
2.2.3 Direct Psychological Effects
Family illness could cause stress that may affect the well-being of children. There is some evidence about emotional and psychological transmissions of negative emotions and mental distress from one family member to the other (Powdthavee and Vignoles 2008). Consequently, a lower level of family health might directly translate into a lower level of life satisfaction. The information effect (Hirschman and Rothschild 1973) can have a similar effect: the presence of ill individuals in the same household signals that there is a possibility for oneself to get ill in the future, which decreases own happiness even before any actual change in health takes place. Along the same line, one might postulate that children have altruistic preferences and might value having a family in good health.
Siblings’ rivalry might also play an ambiguous role in the relationship between life satisfaction and siblings’ health. Knowing that parents make decisions on how to allocate investments across different siblings and might compensate for or reinforce initial differences among them (Behrman et al. 1982), children may become rivals in the face of labour and capital constraints. Hence, a sibling in relatively poor health might be a source of lower or higher happiness if parents invest their resources to respectively compensate or reinforce differences in health between siblings.
3 Data
The data for this study are from the Young Lives survey, a study tracking the lives of children in four countries: Ethiopia, India (in the states of Andhra Pradesh and Telangana), Peru and Vietnam over 15 years. In each study country, the Young Lives surveys involve tracking 3000 children in two cohorts. The younger cohort consists of 2000 children who were born between January 2001 and May 2002. The older cohort consists of approximately 1000 children from each country born in 1994–1995.
Currently, five survey waves are available: the baseline round in 2002 and four follow-up waves in 2006, 2009, 2013 and 2016. The data are clustered and cover 20 sites in each country across rural and urban areas. The objectives of Young Lives project is to study the causes and consequences of childhood poverty. Hence, the sample was designed to include a high proportion of poor children, while at the same time including other children with whom their experiences could be compared. Study sites in each country were selected non-randomly, with rich areas excluded from the sample and poor areas over-sampled. Children in the right age group in the selected sites were then sampled randomly. For the younger cohort the children were aged between 6 and 18 months, while they were aged 7 to 8 years for the older cohort sample.Footnote 3
Even though the study sites were chosen purposely to reflect the diverse socioeconomic conditions within the study countries and therefore are not statistically representative for the country, comparisons with representative datasets like the Demographic and Health Survey (DHS) samples do show that the data contain a similar range of variation as nationally representative datasets in each of the countries.
The Young Lives survey consists of three main sources of data: a child questionnaire, a household questionnaire, and a community questionnaire. Our variables are from the household and child questionnaires. From the household questionnaire we obtain data on parental background, household and child education, the number of household members by sex and age groups, the size of the household, urban/rural location, and indicators of household socio-economic status.
From the child questionnaire, we obtain data on a wide range of child well-being indicators such as schooling, time-use, health, social networks, feelings and attitudes, anthropometry, as well as cognitive and non-cognitive tests. Subjective well-being is measured with a Cantrill ladder question that seeks to capture a child’s view on where she places herself on a ladder ranging from 1 to 9 reflecting worst to best possible life. A picture of the ladder is shown to the child by the survey enumerator as the question is asked.
The survey records detailed information about the history of illness and injury suffered by members of the household. Health shocks are defined such that the respondent perceives the illness of the household member to have affected the welfare of the household negatively.
Last, the time use diary documents all of the activities of the children over a 24 h period on one randomly chosen weekday. These activities are then grouped into the following eight major categories: Caring for others (younger siblings, the elderly, ill household members); domestic chores (fetching water, firewood, cleaning, cooking, shopping); tasks at the family business (farm, cattle herding, other family business); activities for pay outside of household; at school; studying outside of school (including extra tutorship, and studying at home); play time/general leisure, and sleep.
4 Empirical Strategy
We first estimate how the health of relatives affects the life satisfaction of Young Lives respondents via the following OLS regression:
where \(L{S}_{ijt}\) is the life satisfaction reported by the Young Lives respondent \(i\) from household \(j\) in wave \(t\) on a 9-point scale. The other variables are described below.Footnote 4
The validity and reliability of responses to well-being questions can be a source of concern. We have discussed in details some facts in favour of its use in Chapter 1 (Borga, D’Ambrosio & Lepinteur). We know, for example, that responses to subjective well-being questions and physiological expressions of emotions or brain activity are strongly correlated (Urry et al. 2004). In addition, life satisfaction is a good predictor of future behaviours, such as marital break-up (Guven et al. 2012) or job quits (Clark 2001). This evidence shows that subjective measures of well-being reveal useful information about individual preferences and thus behaviour, even after controlling for a wide range of objective variables. As these findings are also based on pooled regressions, it seems that survey respondents share a common understanding of life-satisfaction scales and it supports its use for interpersonal comparisons.
\(FatherIl{l}_{ijt}\), \(MotherIl{l}_{ijt}\) and \(SiblingIl{l}_{ijt}\) are dummies equal one when respectively the father, the mother or any of the child’s siblings were ill between wave \(t\) and \(t-1\). The households were asked to identify which among a predefined list of events/shocks had a negative consequence on them. Illness of a family member (father, mother, and siblings) is among these predefined events asked to every household. These shock-related variables are binary (the variable equals 1 when shock was reported during the period in between rounds, and 0 otherwise). Note that answers are based on perceptions; that is, they do not show whether a negative event has occurred, rather they show whether the respondent considers the event has affected the welfare of the household negatively.
\(OwnHealt{h}_{ijt}\) is self-reported health ranging from 1 to 5 (i.e., “very poor health” to “very good health”). It is reported by the main caregiver of the child. Regarding the different concerns about the interpretation and reporting biases of self-assessed health, Doiron et al. (2015) illustrate its predictive power on mortality, risk of coronary heart diseases and chronic diseases using an Australian survey data linked to administrative individual medical records. They find that self-assessed health predicts future health such as hospitalizations and prescription drugs. According to Doiron et al. (2015), the predictive power of self-assessed health is even more precise in cases of serious and chronic diseases.
\({\delta }_{i}\) and \({\lambda }_{t}\) represent individual fixed effects and survey-round fixed effects. The former captures the influence of time-invariant characteristics of the respondents, such as personality traits or genes. Survey-round fixed effects neutralize the influence of elements that are shared by all respondents in a given survey round, e.g. macroeconomic environment, design of the questionnaire, etc.
\(S{E}_{ijt}\) is a set of control variables capturing the impact of the shared environment (the parental education, the household size, a wealth index and an urban dummy). We will present in our main results two versions of Eq. 1 where we will respectively not control for and then control for \(S{E}_{ijt}\). In the first version, \({\alpha }_{1},\) \({\alpha }_{2},\) \({\alpha }_{3}\) can be considered as the net effects of the health of the different family members on the life satisfaction of the Young Lives respondents. In the second version of Eq. 1, \({\alpha }_{1}\), \({\alpha }_{2}\), \({\alpha }_{3}\) will measure the effect of the various measures of health net of the shared environment. The comparison of the different vectors of estimates will give us the opportunity to evaluate the relative importance of the shared environment in explaining the relationship between family health and own life satisfaction.
However, this exercise does not allow to quantify to what extent the effect of family members’ health is due to objective detrimental changes due to parental investment decisions (i.e. higher probability to quit school and to work) or psychological reasons. This is why we estimate this second regression via OLS:
\(P{I}_{ijt}\) is a vector of parental investments that may mediate the relationship between family members’ health and life satisfaction as outlined in Sect. 2 (and summarized in Fig. 1). We include the probability to be enrolled at school, the share of time spent on paid work outside of the home as well as the share of time spent in housework. We believe that these variables capture the influence of the most important objective channels that may explain a potential objective relationship between family members’ health and life satisfaction. If this is true, \({\beta }_{1}\), \({\beta }_{2}\) and \({\beta }_{3}\) should respectively be significantly lower than \({\alpha }_{1}\), \({\alpha }_{2}\) and \({\alpha }_{3}\). If \({\beta }_{1}\), \({\beta }_{2}\) and \({\beta }_{3}\) remain significantly different from zero, it would mean that family illness affects life satisfaction beyond objective reasons. This would imply that a part of the relationship between family illness and happiness is based on psychological factors such as the information effect or empathy.
Each regression we report here is carried out using all of the survey members who have non-missing values for the dependent variable and independent variables. This produces an estimation sample of 22,352 observations. The children from our estimation sample are on average 14 years old and their average life satisfaction is 5.71 (on a scale from 1 to 9). They report a level of health of 3.71 on average (on a scale from 1 to 5). The probability of having a father ill is 9 %. The same figure applies to mother and siblings. The complete descriptive statistics appear in Table 1.
5 Results
5.1 Family Health and Subjective Well-being
Table 2 shows how life satisfaction is correlated with the measures of health status of the different household members. The first column presents the estimation results that only include dummies for illness of the father, mother and siblings. The second column adds own health and the third includes all of the other family outcomes. The last column introduces individual fixed-effects. All specifications control for the gender and age of the respondent as well wave and region fixed effects.
Column (1) of Table 2 reports the association between life satisfaction and family health. We find a significant negative correlation between household members’ health and the child’s life satisfaction. Pairwise Wald tests confirm that the effect of father and mother illness are not significantly different while the negative impact of siblings’illness is significantly larger.
The estimates from column (1) might confound the effect of unobserved variables that may simultaneously affect life satisfaction and household members’ health. To account for this, we first control for own health in Column (2). Consistent with the literature on health and subjective well-being (Clark et al. 2019), we find that health is an important predictor of happiness. Own health attracts a positive and precise estimate. Note that one cannot directly compare the magnitude of the effect of own health to the effect family illness as they do not share a common scale. However, the most precisely estimated coefficient is unsurprisingly the one associated with own health. Controlling for own health also lowers the magnitudes of family health estimates (but they remain significantly different from zero at 1 % level).
We control for the effect of the shared environment in column (3). The parents’ health variables still attract positive estimates but their significance is reduced. Standard errors remain the same implying that the loss in significance is only due to the lower magnitude of the estimates. The effect of siblings’ health remains unchanged and still highly significant. The effect of own health is still positive and significantly different from zero at 1 % level albeit its magnitude is slightly reduced. The introduction of household characteristics aims at capturing the influence of the common shared environment. Regarding the evolution of our estimates, it seems that it partially explains the effect of parental health.
Last, we introduce individual fixed-effects in column (4). This allows controlling for the influence of all time invariant factors (such as genetics and personality traits) that may affect the relationship between life satisfaction and family illness. Maternal health is now no longer significantly different from zero. Paternal health remains a negative predictor of life satisfaction but only at the 10 % significance level. This result suggests that most of the effect of parental health observed in column (1) transits via the objective living conditions, i.e. the common shared environment, and the time invariant characteristics we controlled for. However, the effect of siblings’ health is remarkably stable across specifications and, as such, relatively insensitive to the influence of the effects of aforementioned objective living conditions. We explore more deeply the mechanisms that might explain our results in the next sub-section.
5.2 Parental Investment or Psychological Factors?
We now ask whether the negative effects of the father’s and siblings health are explained by objective and detrimental changes in parental investments or by psychological factors. To do so, we first check whether lower levels of health of the relatives reduce the probability of the main respondent attending school. We also verify how the time a child spends on the market work and on household chores reacts to changes in the health of the relatives.
We replicate our main model but use the probability of school enrolment, the time spent on the market work, and the time spent in housework as dependent variables. Results are shown in Table 3. The probability of school enrolment is unsurprisingly lower when the father is ill but also when the siblings are ill. This might be explained by the results in column (2). A sick father is also less likely to work and it might create the need for the other household members to spend a larger share of their time on the labour market. This is consistent with a mechanism of intra-household compensation in working time. Furthermore, we find that when the mother or the siblings are sick, the time spent in housework also increases (see the last column of Table 3). As expected from countries where the adherence to conservative gender norms is still high, we do not find any substitution mechanism between the cohort members and the fathers’ health when we consider the time spent in housework.
As suggested by Table 3, children who grew up with sick family members in our estimation sample are less likely to go to school and more likely to devote time in the labour market and on housework. Such changes in the time allocation of the cohort members would explain the estimates we found in column (4) of Table 2 if well-being is positively correlated with school enrolment and negatively correlated with the time spent on the labour market and housework.
We formally test this hypothesis in Table 4. The first column replicates our baseline estimates. The results in the next three columns are in line with our expectations: schooling is positively associated with well-being and the time spent on the labour market and in housework reduces life satisfaction. We may then expect the detrimental effect of father and sibling’s sickness to be potentially driven by the reduction in the probability to be enrolled and the increase in the time spent on the labour market and in housework. However, none of the health estimates appears to be affected by the inclusion of these different variables. The effect of a father’s illness is no longer significantly different from zero in column (2) but it is not significantly lower than the baseline estimate in column (1). We finally follow a “horse race” approach in the last column, i.e. we control for all the potential channels in a single regression. Both school enrolment and the time spent doing housework remain significant predictors of well-being. The effect of the siblings’ health remains negative and highly significant while the effect of the father illness is no longer different from zero.
These results suggest that the well-being effect of the health of the parents is mostly explained by the control variables we used in the columns (3) and (4) of Table 2, i.e. the shared environment. However, the reduction in well-being caused by sibling’s illness is neither explained by the shared environment nor objective changes in parental investment (school enrolment, time on the labour market, time in housework). As none of the objective factors we controlled for in our empirical analysis influenced the well-being effect of the health of the siblings, we suspect the relationship between life satisfaction and sibling’s illness to be mostly driven by psychological factors such as the information effect or empathy.
5.3 Heterogeneity
Table 2 shows the average effect of family illness on the whole estimation sample. One may then argue that certain individuals are more affected by the health of their relatives than others. For instance, it is shown in the related literature that the determinants of happiness differ across gender (Fugl-Meyer et al. 2002). We account for this concern by splitting the sample by gender and re-estimating our main model. These results are shown in the first two columns of Table 5 and we do not find significant differences between boys and girls.
We then split our sample between “poor” and “non-poor” household. As there is no measure of income in our dataset, we use the median of the wealth index computed by the data provider as the poverty threshold. Results are displayed in columns (3) and (4). The effect of own health in both cases is positive and significant. The effects of the illness of the siblings are always negative and comparable in magnitude across samples. Note that the effect of having an ill father is only negative and significant for children from poor households. However, the estimate is only significant at 10 % level.
We next consider rural and urban households separately in columns (5) and (6). While most of the estimates are comparable across samples, a noticeable element here is that the health of mothers is significantly different from zero in rural households only.
One might finally wonder whether the effect of the health of relatives depends on the age of the respondents. We here take advantage of the particular setting of the Young Lives Study as two different cohorts are surveyed at the same time: the Older cohort and the Younger cohort. The former is on average 19 years old in our estimation sample while the latter is on average 12.6 years old. Once again, we find few differences between the two categories as revealed by columns (7) and (8). The most important factors are own health and the health of siblings while the health of the parents attract estimates that are hardly significantly different from zero.Footnote 5
6 Conclusion
The role of the health of parents and siblings have so far received scant attention in the literature. Our contribution aimed to fill this gap.
Family health is a strong predictor of child well-being through its potential impact on parental investment decisions and family interactions. Health shocks entail both economic and psychological costs to children and may alter their time allocation patterns. Resource constrained families living in underdeveloped communities with limited access to health services are further affected by this. Hence, understanding the impact of health shocks and their transmission mechanisms helps inform public policy.
Using a dataset covering four developing countries (namely Ethiopia, India, Peru and Vietnam), we show that having parents and siblings in poor health significantly reduces own life satisfaction. We then find that the loss in well-being produced by the poor health of parents is mostly explained by the shared environment (education level of the parents, level of wealth). The health of the parent’s matters more in terms of well-being for children who grew up in poor and rural households.
Having at least one sibling in poor health has a stronger effect on life satisfaction than a sick parent. Keeping constant the shared environment and accounting for time-invariant characteristics does not explain the effect. While we show that a sick sibling decreases the probability to be enrolled at school and increases the time on market and domestic work, none of these channels explain the negative effect of a sick sibling.
As none of the objective factors we considered could explain the loss in wellbeing induced by the poor health of siblings, we postulate that it may be caused by psychological processes in line with the information effect or empathic behaviour. We know from the Economics of Happiness literature that individuals tend to compare to persons they consider as peers to have a better understanding of what the future might bring to them (i.e. the information effect). But we also know that individuals genuinely care about their relatives (i.e. empathic behaviours). Future research should determine whether our estimates reflect a “pure” empathy or if it reflects a “tunnel effect”. Are the children of our estimation sample simply worrying for the health of their siblings or are they concerned for themselves? Note also that our conclusion regarding the psychological nature of the association between sibling’s health and own life satisfaction holds under the assumption that the influence of all objective channels is kept constant.
There are some limitations and potential weaknesses in our study. The empirical implementation of our research questions is problematic for a number of reasons. First, not all the determinants of subjective well-being (and their histories) are observable; second, there may be some endogeneity effects with respect to unobserved endowments and prior realizations of outcome (for example, indirect effects caused by parental behavioural responses or other confounding factors); and third, our outcome and main explanatory variables could be measured with error. Our study overcomes some of these limitations. We control for time invariant omitted variables through the inclusion of the child fixed effect, and account for correlated omitted variables that are shared by all respondents in a given survey round (such as macroeconomic environment) thanks to the survey-round-fixed effect. Despite our attempt to demonstrate as close to a causal relationship between family health and child life satisfaction, we are aware of the fact that our results cannot be interpreted as completely causal. There is still ample room for future research to find an appropriate identification strategy that purges the issues of omitted variables and endogeneity and establish the causal impact of family health on child’s subjective well-being.
Our results are important for policy makers. The benefits of health policies would be underestimated if one only accounts for the better health of the targeted individuals. This means that cost–benefit analyses would be more accurate if they were likely to consider the potential positive spillover effects of any health policy.
Notes
- 1.
As standard practice in Economics, we will use the terms “subjective well-being” and “happiness” interchangeably in this chapter. Subjective well-being is generally measured by variables such as self-assessed satisfaction with own job, or own life. See Chap. 1 (Borga, D’Ambrosio & Lepinteur) for more details about those concepts in Economics.
- 2.
While it is not the focus of our analysis, it is important to stress that elements composing the shared environment (e.g. income, parental education) are also known to affect health. There is a well documented evidence that across countries, within countries at a point in time, and over time with economic growth, wealthy people live longer and have lower morbidity, on average, than poor people (Case et al. 2002). This disparity in health across socio-economic status (SES) groups is often known as the SES-health gradient. A number of SES measures have been proposed, other than income, such as wealth, education, employment status, and race/ethnicity.The SES-health gradient is also remarkably similar between countries with relatively low levels of social protection from loss of work and health risks, such as the US, and those with stronger welfare systems, such as the Netherlands (Galama and Van Kippersluis 2019).
- 3.
See the survey documentation at https://www.younglives.org.uk/ for further details.
- 4.
While we lack statistical power to convincingly correct for the Nickell bias, a dynamic panel model yields similar estimates. Results are available upon requests.
- 5.
We considered here individual sources of heterogeneity. One may argue that heterogeneity may arise from the institutional context and that we should analyse separately the four countries from our estimation sample. When we do so, we find very similar results in Ethiopia, Peru and Vietnam. We do not find any effect of health (own and relative) on life satisfaction in India. This is not surprising as we have less observations and less within-variability in this country: we then attribute the absence of significant estimates to the limited statistical power.
References
Alam, S. A. (2015). “Parental health shocks, child labor and educational outcomes: Evidence from Tanzania.” Journal of Health Economics, 44, 161–175.
Becker, G. S., & Tomes, N. (1986). “Human capital and the rise and fall of families.” Journal of Labor Economics, 4, S1–S39.
Behrman, J. R., Pollak, R. A., & Taubman, P. (1982). “Parental preferences and provision for progeny.” Journal of Political Economy, 90, 52–73.
Bratti, M., & Mendola, M. (2014). “Parental health and child schooling.” Journal of Health Economics, 35, 94–108.
Card, D., Mas, A., Moretti, E., & Saez, E. (2012). “Inequality at work: The effect of peer salaries on job satisfaction.” American Economic Review, 102, 2981–3003.
Case, A., Lubotsky, D., & Paxson, C. (2002). “Economic status and health in childhood: The origins of the gradient.” American Economic Review, 92, 1308-1334.
Clark, A. E. (2001). “What really matters in a job? Hedonic measurement using quit data.” Labour Economics, 8, 223–242.
Clark, A. E. (2003). “Unemployment as a social norm: Psychological evidence from panel data.” Journal of Labor Economics, 21, 323–351.
Clark, A. E., Flèche, S., Layard, R., Powdthavee, N., & Ward, G. (2019). The origins of happiness: The science of well-being over the life course. Princeton University Press.
Clark, A. E., & Lepinteur, A. (2019). “The causes and consequences of early-adult unemployment: Evidence from cohort data.” Journal of Economic Behavior and Organization, 166, 107–124.
Clark, A. E., & Oswald, A. J. (1996). “Satisfaction and comparison income.” Journal of Public Economics, 61, 359–381.
Conti, G., & Heckman, J. (2014). “The economics of child well-being.” In A. Ben-Arieh, F. Casas, I. Frønes, & J. Korbin (Eds.), Handbook of child well-being, 363–401.
Cunha, F., Elo, I., & Culhane, J. (2013). Eliciting maternal expectations about the technology of cognitive skill formation. NBER Working Paper N. 19144.
Cunha, F., & Heckman, J. J. (2008). “Formulating, identifying and estimating the technology of cognitive and non-cognitive skill formation.” Journal of Human Resources, 43, 738-782.
D’Ambrosio, C., Jäntti, M., & Lepinteur, A. (2020). “Money and happiness: Income, wealth and subjective well-being.” Social Indicators Research, 148, 47–66.
Del Bono, E., Francesconi, M., Kelly, Y., & Sacker, A. (2016). “Early maternal time investment and early child outcomes.” Economic Journal, 126, F96–F135.
Dhanaraj, S. (2016). “Effects of parental health shocks on children’s schooling: Evidence from Andhra Pradesh, India.” International Journal of Educational Development, 49, 115–125.
Doiron, D., Fiebig, D. G., Johar, M., & Suziedelyte, A. (2015). “Does self-assessed health measure health?” Applied Economics, 47, 180–194.
Ferrer-i Carbonell, A., & Van Praag, B. M. (2002). “The subjective costs of health losses due to chronic diseases: An alternative model for monetary appraisal.” Health Economics, 11, 709–722.
Fiorini, M., & Keane, M. P. (2014). “How the allocation of children’s time affects cognitive and noncognitive development.” Journal of Labor Economics, 32, 787–836.
Francesconi, M., & Heckman, J. J. (2016). “Child development and parental investment: Introduction.” Economic Journal, 126, F1–F27.
Fugl-Meyer, A. R., Melin, R., & Fugl-Meyer, K. S. (2002). “Life satisfaction in 18-to 64- year-old swedes: In relation to gender, age, partner and immigrant status.” Journal of Rehabilitation Medicine, 34, 239–246.
Galama, T. J., & Van Kippersluis, H. (2019). “A theory of socio-economic disparities in health over the life cycle.” Economic Journal, 129, 338–374.
Gertler, P., Levine, D. I., & Ames, M. (2004). “Schooling and parental death.” Review of Economics and Statistics, 86, 211–225.
Guven, C., Senik, C., & Stichnoth, H. (2012). “You can’t be happier than your wife. Happiness gaps and divorce.” Journal of Economic Behavior & Organization, 82, 110–130.
Heckman, J. J. (2007). “The economics, technology, and neuroscience of human capability formation.” Proceedings of the National Academy of Sciences, 104, 13250–13255.
Hirschman, A. O., & Rothschild, M. (1973). “The changing tolerance for income inequality in the course of economic development: With a mathematical appendix.” Quarterly Journal of Economics, 87, 544–566.
Jensen, R. (2000). “Agricultural volatility and investments in children.” American Economic Review, 90, 399–404.
Kahneman, D., Krueger, A. B., Schkade, D. A., Schwarz, N., & Stone, A. A. (2004). “A survey method for characterizing daily life experience: The day reconstruction method.” Science, 306, 1776–1780.
Knudsen, E. I., Heckman, J. J., Cameron, J. L., & Shonkoff, J. P. (2006). “Economic, neurobiological, and behavioral perspectives on building America’s future workforce.” Proceedings of the National Academy of Sciences, 103, 10155–10162.
Layard, R., Clark, A. E., Cornaglia, F., Powdthavee, N., & Vernoit, J. (2014). “What predicts a successful life? A life-course model of well-being.” Economic Journal, 124, F720–F738.
Luca, D. L., & Bloom, D. E. (2018). The returns to parental health: Evidence from Indonesia. NBER Working Paper N. 25304.
Newland, L. A. (2015). “Family well-being, parenting, and child well-being: Pathways to healthy adjustment.” Clinical Psychologist, 19, 3–14.
Oswald, A. J., & Powdthavee, N. (2008). “Does happiness adapt? A longitudinal study of disability with implications for economists and judges.” Journal of Public Economics, 92, 1061–1077.
Powdthavee, N. (2009). “I can’t smile without you: Spousal correlation in life satisfaction.” Journal of Economic Psychology, 30, 675–689.
Powdthavee, N., & Vignoles, A. (2008). “Mental health of parents and life satisfaction of children: A within-family analysis of intergenerational transmission of well-being.” Social Indicators Research, 88, 397–422.
Senik, C. (2004). “When information dominates comparison: Learning from Russian subjective panel data.” Journal of Public Economics, 88, 2099–2123.
Senik, C. (2008). “Is man doomed to progress?” Journal of Economic Behavior & Organization, 68, 140–152.
Shields, M. A., & Price, S. W. (2005). “Exploring the economic and social determinants of psychological well-being and perceived social support in England.” Journal of the Royal Statistical Society: Series A (Statistics in Society), 168, 513–537.
Sun, A., & Yao, Y. (2010). “Health shocks and children’s school attainments in rural China.” Economics of Education Review, 29, 375–382.
Todd, P. E., & Wolpin, K. I. (2003). “On the specification and estimation of the production function for cognitive achievement.” Economic Journal, 113, F3-F33.
Urry, H. L., Nitschke, J. B., Dolski, I., Jackson, D. C., Dalton, K. M., Mueller, C. J., … Davidson, R. J. (2004). “Making a life worth living: Neural correlates of well-being.” Psychological Science, 15, 367–372.
Van den Berg, B., Fiebig, D. G., & Hall, J. (2014). “Well-being losses due to care-giving.” Journal of Health Economics, 35, 123–131.
Wunder, C., & Heineck, G. (2013). “Working time preferences, hours mismatch and well-being of couples: Are there spillovers?” Labour Economics, 24, 244–252.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Open Access Dieses Kapitel wird unter der Creative Commons Namensnennung 4.0 International Lizenz (http://creativecommons.org/licenses/by/4.0/deed.de) veröffentlicht, welche die Nutzung, Vervielfältigung, Bearbeitung, Verbreitung und Wiedergabe in jeglichem Medium und Format erlaubt, sofern Sie den/die ursprünglichen Autor(en) und die Quelle ordnungsgemäß nennen, einen Link zur Creative Commons Lizenz beifügen und angeben, ob Änderungen vorgenommen wurden.
Die in diesem Kapitel enthaltenen Bilder und sonstiges Drittmaterial unterliegen ebenfalls der genannten Creative Commons Lizenz, sofern sich aus der Abbildungslegende nichts anderes ergibt. Sofern das betreffende Material nicht unter der genannten Creative Commons Lizenz steht und die betreffende Handlung nicht nach gesetzlichen Vorschriften erlaubt ist, ist für die oben aufgeführten Weiterverwendungen des Materials die Einwilligung des jeweiligen Rechteinhabers einzuholen.
Copyright information
© 2022 Der/die Autor(en)
About this chapter
Cite this chapter
Borga, L., D’Ambrosio, C., Lepinteur, A. (2022). It’s a Family Affair: Family Health and Child Well-being. In: Heinen, A., Samuel, R., Vögele, C., Willems, H. (eds) Wohlbefinden und Gesundheit im Jugendalter. Springer VS, Wiesbaden. https://doi.org/10.1007/978-3-658-35744-3_8
Download citation
DOI: https://doi.org/10.1007/978-3-658-35744-3_8
Published:
Publisher Name: Springer VS, Wiesbaden
Print ISBN: 978-3-658-35743-6
Online ISBN: 978-3-658-35744-3
eBook Packages: Social Science and Law (German Language)