Kahneman · Proceedings of the National Academy of Sciences of the United States of America 2010 · cross-sectional survey analysis · n=>450,000

High income improves evaluation of life but not emotional well-being.

Level 4 - case-series / case-control

Cross-sectional survey without prospective longitudinal follow-up

PubMed 20823223 · doi:10.1073/pnas.1011492107 · record verified 2026-08-26

What was done

Researchers analyzed more than 450,000 survey responses from US residents in the Gallup-Healthways Well-Being Index, a daily telephone survey of approximately 1,000 individuals. They examined the relationship between household income and two distinct dimensions of subjective well-being: emotional well-being (measured via self-reported experiences of joy, stress, sadness, anger, and affection yesterday) and life evaluation (measured via Cantril's Self-Anchoring Scale).

What was found

Life evaluation and emotional well-being showed distinct relationships with income and life circumstances. Life evaluation rose continuously with log income. Emotional well-being also improved with log income but plateaued beyond an annual income of ~$75,000. Income and education were stronger correlates of life evaluation, whereas health, caregiving, loneliness, and smoking were stronger predictors of daily emotional quality. Low income was also found to exacerbate the emotional pain associated with divorce, poor health, and loneliness.

Why it matters

This study demonstrates that income has divergent effects on different aspects of well-being, clarifying that higher earnings continue to improve overarching life evaluation but do not meaningfully enhance day-to-day emotional happiness beyond a moderate threshold.

Limits

The study relies on cross-sectional survey data, precluding causal conclusions. All measures are self-reported and subject to recall bias for the prior day's emotions. The findings are restricted to US residents surveyed at a specific point in time and do not track individual financial or emotional trajectories longitudinally.

Cited by