A considerable body of work explores the relationship between the economy and governmental popularity. These ‘popularity functions’ exhibit a good deal of instability in the economics coefficient, leading some to question its very existence. It is argued in this article that this instability is apparent, rather than inherent. Improvements in model specification, measurement, sample size and estimation reveal a strong and stable economic effect. In particular, fixed and random effects models on pooled time-series (from France, Germany, Italy, Spain, the United Kingdom and the United States) are estimated here. The impact of national economic perception on popularity emerges as statistically and substantively significant, across this sample of countries.