Accounting for Unobserved Country Heterogeneity in Happiness Research : Country Fixed Effects Versus Region Fixed Effects

Many empirical studies are ambiguous about whether good formal institutions are conducive to subjective well-being or not. Possibly, this ambiguity is caused by cross-section models that do not account for unobserved cultural and institutional effects. Using the World Value Survey 1980-2005, this paper supports a positive relation in a country panel framework that accounts for unobserved, time-invariant country heterogeneity. This study also shows that using supranational region dummies (by geography or language) in a country-random effects model appears to be a sufficient substitution for omitted country fixed effects