Trust can be a double-edged sword for democracy. On the one hand, social capital theory highlights the importance of social trust in societies for the functioning of democracy. Similarly, democratic governments require some level of political trust to function effectively, and declining levels of political trust in consolidated democracies over the past decades have been stylized as a “crisis of democracy”. On the other hand, democracy also relies on a healthy scepticism to hold governments accountable and “critical citizens” may fuel pressure for democratic reforms. Despite a plethora of arguments on the relationship between trust and the quality of democracy, only few studies have tested this relationship empirically, reaching different conclusions. Two crucial limitations of these studies are a) that they rely on single survey projects with a restricted number of observations and b) that they do not distinguish between different forms and objects of trust. In this paper, I aim to overcome these limitations and put the relationship between trust and the quality of democracy to a rigorous empirical test. Using a Bayesian latent variable model, I combine data from various international surveys to estimate smooth country-year panels of different forms of social and political trust. I employ dynamic panel data models to test whether they are related to subsequent changes in V-Dem democracy indices. The findings have crucial implications for understanding the importance of social and political trust for the emergence and consolidation of democracies.
