The Question
Two prominent researchers had published contradictory findings on whether higher income improves emotional wellbeing, and the contradiction had generated genuine debate. Kahneman and Deaton’s 2010 study reported that happiness increased with income up to a threshold of around $75,000 per year and then plateaued. Killingsworth’s 2021 study, using experience sampling rather than retrospective reporting, found that happiness continued to rise with income with no plateau. Killingsworth, Kahneman, and Mellers (2023) engaged in an adversarial collaboration to find a coherent interpretation of both.
The Resolution
The key finding is that both studies were partly right, but about different people. A reanalysis of the experience sampling data confirmed the flattening pattern, but only for the least happy people. For happier people, happiness continued to rise steadily with income, and in the happiest group the relationship actually accelerated (Killingsworth et al., 2023).
The two earlier studies had each missed this because both assumed that the relationship between income and happiness was the same for everyone. When that assumption of homogeneity is relaxed and the distribution of happiness is examined rather than just its average, the apparent contradiction dissolves: money matters more for those who are already relatively happy, and the plateau that Kahneman and Deaton observed was driven primarily by the experience of the least happy, for whom additional income above a threshold brings diminishing returns (Killingsworth et al., 2023).
A Lesson About Method
The authors also identify specific methodological reasons why each earlier study reached the conclusion it did. Kahneman and Deaton’s dichotomous questions had a ceiling effect that prevented discrimination among degrees of happiness, causing them to overstate the flattening pattern. Killingsworth’s experience sampling captured finer-grained variation but averaged across the distribution in a way that obscured the heterogeneity (Killingsworth et al., 2023). Both errors reflect practices that are standard in social science but, the authors argue, should be questioned more often.
Reference
Killingsworth, M. A., Kahneman, D., & Mellers, B. (2023). Income and emotional well-being: A conflict resolved. Proceedings of the National Academy of Sciences, 120(10), e2208661120. https://doi.org/10.1073/pnas.2208661120
