The Perils of Survey Measures of Culture in Historical Economics

Measures of values and beliefs derived from answers to survey questions are often used to investigate the cultural roots of long-term economic development, based on the idea that cultural attitudes are persistent traits shaped by historical factors. We show that these survey proxies are not persistent and result into fragile historical correlations. These points are illustrated by analyzing proxies for cultural attitudes derived from the World Values Survey and the Afrobarometer, and by revisiting analyses carried out in two early, influential studies in historical economics that use these data sources. Diagnostic tools and possible solutions are discussed