Unhappily Single? Towards a More Accurately Estimated Association Between Romantic Relationship Status and Life Satisfaction

Partnered individuals tend to report higher life satisfaction (LS) than singles (weighted Cohen’s d = .43). However, many studies have suffered from either small sample sizes, unaddressed moderators or confounding variables, and/or single-method biases. To simultaneously address these issues, we used the Estonian Biobank ( N = 77,378) to more accurately estimate the association ( d ) between romantic relationship status and LS. We examined potential moderators (age, gender, income, parental status) and confounders (personality domains and nuances, genetic/familial influences, single-method biases). Individuals in relationships had higher average LS (β = .34), but in a sibling discordant analysis controlling for shared familial backgrounds, polygenic scores for LS, and personality traits the effect was smaller (β = .23; only controlling for personality traits resulted in a similar association, suggesting they captured the other potential confounders). However, combining self-report with informant report to reduce measurement error, the within-family, personality traits-controlled association was d = .32. We believe this represents a more rigorous estimate to date, suggesting that previous studies may have both over- and underestimated the link due to confounders and measurement issues that can only be addressed with multiple methods. Finally, the single-partnered difference was larger among men and those without children, and only those highly or very highly satisfied with their relationship had higher LS than single individuals.