Causal Analysis in Happiness Research

Abstract

There are some misunderstandings in the way we interpret estimated coefficients in happiness equation regressions, especially when the words “effect”, “cause”, or “impact” are used to describe a relationship between self-rated happiness and some personal characteristics which may or may not be causal. This paper describes the potential damages in the way we interpret estimated coefficients on some of the observable characteristics in happiness equations as having causal impacts on well-being, and reviews a number of methods used to make effective causal inferences about what determine our happiness in the economic literature.