What Does Dynamic Panel Analysis Tell Us About Life Satisfaction?

Abstract

Static fixed effects estimation of life satisfaction often gives rise to serial correlation in the residuals, indicating omitted dynamics. Rather than correcting the standard errors, “to make them robust,” this study explores dynamic panel analysis to incorporate the otherwise omitted dynamics into the estimation of life satisfaction models. The analysis utilizes the System General Method of Moments approach to dynamic analysis. The results indicate that overall life satisfaction is largely (though not wholly) contemporaneous; correspondingly, the persistence of satisfaction from previous periods is limited. This finding is robust to a variety of different specification choices. Accordingly, the issue of omitted dynamics in a life satisfaction context is a minor one. This use of dynamic panel analysis, therefore, offers support for the ongoing use of static models (like fixed effects), even though they omit dynamics. Given the highlighted challenges of dynamic panel analysis, static fixed effects analyses will often be preferred.