Eudaimonic happiness as a leading health indicator: cross-country European evidence

ABSTRACT

Eudaimonic happiness (measured in terms of sense of life) is a relatively unexplored subjective wellbeing indicator. The empirical findings presented in this paper show that it has a significant and quantitatively remarkable correlation with the future insurgence of some chronic diseases and the reduction of most functionalities in the ageing population. These results document that eudaimonic happiness is a relevant leading indicator of future health outcomes and expenditure and that its correlation is independent from that of the traditional life satisfaction measure.

KEYWORDS:

• Eudaimonic satisfaction
• health outcomes
• functionalities
• life satisfaction

JEL CLASSIFICATION:

• I12
• I31

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

¹ For a survey on subjective wellbeing measures see among others Helliwell and Barrington‐Leigh (2010) and Dolan, Layard, and Metcalfe (2011).

² ‘The evidence base for eudaimonic measures is less clear. While some specific measures – such as those relating to “meaning” and “purpose” clearly capture unique and meaningful information, the picture with respect to eudaimonia as a whole is more ambiguous. This suggests that further work is needed before a definitive position can be taken on the validity of these measures’ (OECD 2013,13).

³ With reference to extreme life events, Victor Frankl (1946, 16) observed that life can be meaningful even under conditions of extreme adversity and that having a sense of purpose is essential to maintaining psychological health and wellness. In these extreme cases life satisfaction and eudaimonic happiness may clearly diverge.

⁴ Relevant examples are contributions showing how job satisfaction affects objective outcomes such as employment status, productivity, likelihood of job change and job quit (see Judge 1992; Staw and Barsade 1993; Judge et al. 2001).

⁵ Note that other diseases were included in different versions of the questionnaire. However, since we base our analysis only on waves 4 and 5, the final diseases considered for the computation of the number of chronic diseases are only those listed in the text.

⁶ The rationale for using the last two waves is that SHARE waves are irregularly spaced and the database contains a discontinuity between the second and fourth wave (the third wave has a completely different structure with respondents being asked to record experiences of the past related to their health). In addition to it, the question we use as dependent variable has been slightly changed after the second wave. While individuals in the first wave were asked – ‘Has a doctor ever told you that you had any of the conditions on this card?’ from the second wave on the question becomes ‘[Has a doctor ever told you that you had/Do you currently have any] of the conditions on this card? [With this we mean that a doctor has told you that you have this condition, and that you are either currently being treated for or bothered by this condition.]’. The limit of the first question is that respondents may report also illnesses from which patients recovered in the past (for those pathologies for which recovery is possible).

⁷ The negative binomial model can be considered as a generalization of the Poisson model having the same mean structure as Poisson plus an extra parameter – α – to model the over-dispersion (we, therefore, opt for the former after rejecting the null of α = 0).

⁸ We alternatively use 1997 ISCED (International Standard Classification of Education) standards. Results are not substantially different and do not exhibit particular nonlinearities in the relationship between education degrees and number of pathologies. The more parsimonious specification with number of education years is therefore preferred.

⁹ As is well known there are different measures of equivalised income that can be alternatively used to divide household income for the number of its members (Schwarze 2003). Our findings are substantially unaffected by such changes. Evidence is omitted for reasons of space and available upon request.

¹⁰ Following the standard international classification the underweight class starts below a body mass index of 18.5, the overweight class above 24.99 and the obese class above 30. The normal weight class is, therefore, our omitted benchmark.

¹¹ For a similar approach on the attrition weighting procedure in the literature see, among others, Raab, Purdon, and Buckner (2005), Nicoletti and Peracchi (2005) and Vandecasteele and Debels (2007).

¹² Findings on education confirm the well-known positive nexus between education and health in the literature (for a survey on this literature see Grossman 2006).

¹³ Individuals with high ideals and high gap between aspirations and realizations may have low life satisfaction and high sense of life where the latter captures their strong willed pursuit of such aspirations and the former the frustration of not seeing them realized. On the opposite individuals may have low levels of aspirations that are fully realized and be satisfied with their life even though they see not much sense on it. Even though in most cases eudaimonic and cognitive subjective wellbeing should coincide these cases cannot be neglected and explain why the two variables do not overlap.

¹⁴ Let $p_j$ be the predicted probability of a positive outcome and $y_j$ be the actual outcome, a prediction is classified as positive if $p_j\ge 0,25$ and otherwise is classified as negative. The classification is correct if the prediction is positive and $y_j=1$ or if it is negative and $y_j=0.$.