Deconstructing the hedonic treadmill: is happiness autoregressive?

Abstract

Affective habituation is well-documented in social sciences: people seem to adapt to many life events, ranging from lottery windfalls to terminal illnesses. We propose a subtle but critical difference: current happiness may depend directly on past happiness. We test our hypothesis running dynamic happiness regressions using panel data from the German Socio-Economic Panel Study, the British Household Panel Survey and the Swiss Household Panel. Contrary to the widespread prior among economists and non-economists, the coefficient on lagged happiness is positive and statistically significant. We discuss some explanations for the puzzle. Our favorite is that reported happiness is timeinconsistent, even within individuals. We test this conjecture by using a 52-days study. As expected, the coefficient on lagged happiness is negative and statistically significant. We find that changes in hedonic states bounce back 30% in only 5 days.