I can’t smile without you: Spousal correlation in life satisfaction

https://doi.org/10.1016/j.joep.2009.06.005Get rights and content

Abstract

This paper tests whether one partner’s happiness significantly influences the happiness of the other partner. Using 10 waves of the British Household Panel Survey, it utilizes a panel-based GMM methodology to estimate a dynamic model of life satisfaction. The use of the GMM-system estimator corrects for correlated effects of partner’s life satisfaction and solves the problem of measurement error bias. The results show that, for both genders, there is a positive and statistically significant spillover effect of life satisfaction that runs from one partner to the other partner in a couple. The positive bias on the estimated spillover effect coming from assortative mating and shared social environment at cross-section is almost offset by the negative bias coming from systematic measurement errors in the way people report their life satisfaction. Moreover, consistent with the spillover effect model, couple dissolution at t + 1 is negatively correlated with partners’ life satisfaction at t.

Introduction

The idea that married people care a great deal about the well-being of their partner is not new to economists (Becker, 1973, Becker, 1974, Friedman, 1986). The past three decades have seen a significant increase in the number of studies showing that people in marriage tend to behave altruistically towards their partner (see, for example, Ermisch, 2003, Foster and Rosenzweig, 2001). However, while it may be possible to make some inferences about the degree of caring between partners from their behaviour, the idea that there may be a direct spillover of well-being from one partner to the other has rarely been tested empirically.

This paper aims to do just that. Using a long-run panel of nationally representative randomly sampled married and cohabiting individuals, it examines the extent of spousal correlation in subjective well-being data, particularly self-rated life satisfaction (LS). It proposes that a positive correlation between partners’ LS may reflect three distinct processes. First, individuals who are born happy or are born with innate predispositions that make them happy may tend to marry those who are similar to them. In addition to this, people of the same family background or life styles – in other words, same unobserved social factors – may also tend to marry each other. This matching of fixed personal characteristics on the marriage market is analogous to the concept of assortative mating (Becker, 1974). Manski (1995) refers to such phenomena as correlated effects of social interactions.

Second, given that marriage allows individuals to share with their partner the kind of physical and emotional resources that may not be available for each person to obtain outside marriage (Waite & Gallagher, 2000), correlated effects may also arise from the shared social environment (which can either be time-invariant or time-variant) that is simultaneously affecting LS for both spouses.

Lastly, the observed correlation may be the result of a direct spillover of LS within the couple. The idea is that, if a husband cares about his wife, then her LS becomes one of the main determinants of his own LS, and vice versa. This generates a possibility that a husband will be ceteris paribus happier when his wife is happier for whatever reasons that make her happy but not him directly. Hence, we would expect an increase in one partner’s LS to be positively correlated with the other partner’s LS even after allowing for all the factors that can affect both partners’ LS at the same time. This phenomenon is likened to the endogenous effects in Manski’s terminology, whereby the individual outcome is a function of group achievement.

In addition to the above confounding influences which make it difficult for the true relationship between partners’ well-being to be identified, the estimates of spousal correlation in LS may also suffer from the negative measurement error bias. There may be, for example, a tendency for individuals to misreport their true LS in surveys. The low signal-to-noise ratio caused by misreporting can result in an estimated coefficient on partner LS that is biased towards zero in a large sample. In short, because there are both positive (correlated effects) and negative (measurement error) biases involved, the direction of bias is unclear on a priori ground.

This paper uses 10 waves of the British Household Panel Survey (BHPS) data to examine the extent of spousal correlation in LS. In particular, it uses the “system GMM estimator” proposed by Arellano and Bover (1995) and Blundell and Bond (1998) to estimate the causal spillover effect that runs from one partner’s life satisfaction to the other partner’s life satisfaction. The use of the GMM-system estimator, which is a unique approach in the study of happiness, control for the correlated effects and solve the problem of measurement error bias in self-rated life satisfaction through instrumentations and first-differencing. The results show that there is strong evidence of a spillover effect of LS, which suggests that well-being is transferable from one partner to the other. Consistent with the spillover effect model, partners’ LS today are also associated with lower probabilities of partners separating or divorcing one period into the future.

There are similarities in terms of research questions and analytic strategy between this paper and previous studies that examined similarities in a husband’s and wife’s behaviour such as smoking (Clark & Etile, 2006), their political preferences (Kan & Heath, 2006), and their sporting activities (Farrell & Shields, 2002).

This article is organised as follows: Section 2 reviews relevant past research on marriage and well-being. Section 3 addresses theoretical issues revolving around the various interpretations of the correlation between partners’ LS. Section 4 describes how to implement a test and the data set. Section 5 discusses the results, and Section 6 concludes.

Section snippets

Marriage, subjective well-being, and spillovers

Previous research on marital status and emotional well-being is clear on one point: married persons are significantly happier and more satisfied with life than those who are divorced, separated, widowed, or single, across various countries and time periods (Gove et al., 1983, Marks and Lambert, 1998, Mastekaasa, 1994). The large psychological benefits of marriage persist even when the selection of happy people into marriage is controlled for in the analysis (Frey and Stutzer, 2006, Mastekaasa,

Theory

In this section, I will briefly discuss the three underlying mechanisms that may account for the raw correlation between a husband’s and wife’s LS levels: assortative mating, shared social environment, and spillover effect.

The utility model of couples

Consider first Gary Becker’s (1974) simple utility function of an individual i in a marriage to individual j at any given time, which can be written asUi=U(X,Uj(X)),where X is a vector of the consumption of commodities within the household, and Uj is the individual i’s partner’s utility. The individual’s utility is assumed to be increasing with X, which is divisible and can be shared between the couple. An increase in X therefore raises the individual’s utility both through a direct effect upon

Life satisfaction spillovers

Table 2 reports results from the GMM-system estimator described in the previous section. The dependent variable is the respondent’s self-rated life satisfaction measured cardinally (on a scale of 1 to 7).

Conclusion

This paper has used 10 waves of BHPS data to study intra-spousal correlations in self-reported life satisfaction data. Its primary objective was to determine whether the observed correlation is due largely to partners’ fixed traits are similar through assortative mating by personality traits on the marriage market, partners sharing the same social environment that simultaneously affects their well-being, or a spillover effect of life satisfaction from one partner to the other.

A simple OLS model

Acknowledgements

I am grateful to Andrew Oswald, Andrew Clark, Paul Dolan, Ian Walker, Robin Naylor, Geeta Kingdon, Rainer Winkelmann, Ulrich Schimmack, Nateecha Ratanadilok Na Bhuket, Anthony Fielding, participants at the Royal Economic Society at Nottingham and the Economics of Happiness Symposium at the University of Southern California in March, 2006, and two anonymous referees for their valuable comments. The British Household Panel Survey data were made available through the UK Data Archive. The data were

References (65)

  • M. Arellano et al.

    Some tests of specification for panel data: Monte Carlo evidence and application to employment equations

    Review of Economic Studies

    (1991)
  • M. Arellano et al.

    Another look at the instrumental-variable estimation of error-components models

    Journal of Econometrics

    (1995)
  • M. Argyle

    Causes and correlates of happiness

  • G.S. Becker

    A theory of marriage: Part I

    Journal of Political Economy

    (1973)
  • G.S. Becker

    A theory of marriage: Part II

    Journal of Political Economy

    (1974)
  • P.L. Benson

    Intrapersonal correlates of nonspontaneous helping behaviour

    Journal of Social Psychology

    (1980)
  • P. Berger et al.

    Marriage and the construction of reality: An exercise in the microsociology of knowledge

    Diogenes

    (1964)
  • L.F. Berkman

    The changing and heterogeneous nature of aging and longevity: A social and biomedical perspective

    Annual Reviews in Gerontology and Geriatrics

    (1988)
  • N. Bolger et al.

    The contagion of stress across multiple roles

    Journal of Marriage and the Family

    (1989)
  • N.M. Bradburn

    The structure of psychological well-being

    (1969)
  • A. Bruhin et al.

    Happiness function with preference interdependence and heterogeneity: The case of altruism within family

    (2007)
  • K. Chamberlain

    On the structure of well-being

    Social Indicators Research

    (1988)
  • Clark, A. E., & Georgellis, Y. (2004). Kahneman meets the quitters: Peak-end behaviour in the labour market. CNRS and...
  • J.F. Ermisch

    An economic analysis of the family

    (2003)
  • L. Farrell et al.

    Investigating the economic and demographic determinants of sporting participation in England

    Journal of Royal Statistical Society A

    (2002)
  • A. Ferrer-i-Carbonell et al.

    How important is methodology for the estimates of the determinants of happiness?

    Economic Journal

    (2004)
  • A.D. Foster et al.

    Imperfect commitment, altruism, and the family: Evidence from transfer behaviour in low-income rural areas

    Review of Economics and Statistics

    (2001)
  • B. Frey et al.

    Introducing procedural utility: not only what, but also how matters

    Journal of Theoretical Economics

    (2004)
  • B. Frey et al.

    Does marriage make people happy, or do happy people get married?

    Journal of Socio-Economics

    (2006)
  • D.D. Friedman

    Price theory: An intermediate text

    (1986)
  • W.R. Gove et al.

    Does marriage have positive effects on the psychological well-being of the individual?

    Journal of Health and Social Behavior

    (1983)
  • Graham, L., & Oswald, A.J. (2006). Hedonic capital. IZA Discussion Paper No....
  • Cited by (65)

    • Untangling causal beliefs: A lay theory of happiness determinants using a factorial survey

      2020, Poetics
      Citation Excerpt :

      Chileans also believe in a stronger effect of income on happiness when persons are in a recent relationship (F). Scholars in economic psychology (e.g. Powdthavee, 2009) have pointed out that partner relationships provide higher real income per partner. The literature has also shown that spending money in another person increases the level of happiness, and even more than in oneself (Dunn et al., 2008).

    • Working time mismatches and self-assessed health of married couples: Evidence from Germany

      2019, Social Science and Medicine
      Citation Excerpt :

      Spousal working time mismatches may also have an indirect effect on health. Powdthavee (2009) show that the wellbeing of an individual is an increasing function of her partner’s wellbeing. Similarly, we may expect health to be interdependent between couple members.

    • The welfare implications of addictive substances: A longitudinal study of life satisfaction of drug users

      2018, Journal of Economic Behavior and Organization
      Citation Excerpt :

      To understand the potential channels through which the consumption of addictive substances may relate to individual’s overall life satisfaction, Table 5 re-estimates the full specification using seven different standardised domain-specific satisfaction measures (health satisfaction, housing satisfaction, neighbourhood satisfaction, employment satisfaction, financial satisfaction, safety satisfaction, and relationship satisfaction) as outcomes. This approach is similar to the approach adopted by Powdthavee (2009a, 2012) in the study of how disability and job loss indirectly shapes one’s life satisfaction through their effects on domain satisfactions. The most notable results are as follows.

    View all citing articles on Scopus
    View full text