Self-Rated Health: Is Happiness the Missing Link?

Andrew Noymer, University of California, Irvine
Leah Ruppanner, University of California, Irvine

We analyze the relationship between happiness and self-rated health, using the World Values Survey. We look at the joint distribution of health and happiness in a sample of over 200,000 observations from 82 countries. Both in the case of happy predicting healthy, and vice versa, the relationship is highly statistically significant and extremely robust to the inclusion of seemingly every possible appropriate confounding variable. Applications are in areas like the study of marital status and health. Other coefficients in the models can be very fragile to the inclusion of happiness as a control, requiring more attention to modeling issues in the future. If happiness should be included, then many models are misspecified. If happiness should not be included, then misspecification is not a problem, but justification for exclusion of such an important variable needs to be developed. The demonstrated fragility of the models makes it more than a theoretical concern.