The Fundamental Endogeneity of Survey-Based Cultural Dimension

Survey-based cultural dimensions are used extensively in empirical analysis as the primary source of quantitative cultural scores. However, survey-based cultural dimensions have a significant endogeneity problem. The individuals responding to the surveys are affected by their current socioeconomic condition and environment (social mood), meaning that the outcomes are a mixture of culture and current socioeconomic effects. The World Value Survey (WVS) waves provided empirical support for the endogeneity problem. Our empirical findings show that in-country scores frequently change between waves. In addition, the countries in the WVS change their ranking between waves; so, even the relative position held by a country is questionable when measured along survey-based cultural dimensions. This finding contrasts with the idea that cultural values are sticky. We provide an alternative method for capturing cultural dimensions based on the grammatical structures of languages that capture ancestral culture with no fear of the current socioeconomics pollution of the measure.