The classic Muller and Seligson’s model of relationships between civic culture and democracy presented in the article Civic Culture and Democracy: A Question of Causal Relationships (1994) shows that civic culture, along with income inequality and ethnolinguistic fractionalization, influences democracy. However, when the model is replicated for more recent time periods and extended to more countries, the data do not support their hypothesis. Further analysis shows that Muller and Seligson used statistical methods (OLS regression) that are not appropriate for the data and this caused problems with ceiling effect. For this reason ordered logit regression is used and model is tested on data collected from 85 countries (in 1981-2010 time period). Results show that the new analysis of contemporary and more elaborate data seems to uphold the original Muller and Seligson's hypothesis; not only economical factors, but also cultural factors influence the change in democracy.
