Comparative Public Trust in Courts

This paper develops a measurement framework to estimate public trust in judiciaries across more than 40 countries over several decades. We compile and harmonize surveys that ask about respondents’ trust or confidence in courts, addressing differences in scales, question wording, and sampling designs. Our approach centers on a Bayesian multilevel model that links ordinal responses to a continuous latent measure of trust and models its evolution through time using flexible hierarchical splines. The model incorporates individual-level characteristics and survey-specific effects, and we use multilevel regression with poststratification to generate population-representative estimates. By pooling many surveys fielded at distinct moments, we obtain precise, continuous trajectories of judicial trust along with credible intervals, enabling researchers to examine how institutional reforms or democratic backsliding correspond to shifts in public confidence. We conclude with case studies of Poland and Turkey, two countries where courts have played central roles in recent episodes of institutional erosion, to illustrate how the resulting comparable time series can deepen understanding of the political dynamics surrounding judicial institutions.