The volatility of survey measures of culture and its consequences

Measures of cultural attitudes derived from the World Values Survey are often used in economics to investigate the relation between individual culture and aggregate output. We show (i) that these measures are constructed from underlying variables that exhibit within-country volatility and time trends; and (ii) that such lack of persistence implies fragile correlations between cultural measures and output. These points are illustrated by revisiting the analyses of Tabellini (2010) and Gorodnichenko and Roland (2011).