Yuehong Cassandra TaiData ▸ TrustGov

Trust in Government (TrustGov)

Comparable estimates of public trust in national government for 115 countries and territories, 1973–2020, harmonizing 1,545 country-year observations from 189 national and cross-national surveys using a Bayesian latent variable model. Select cases to compare levels and trends; shaded bands show the 80% credible interval.

Year range
to
Display
Download full dataset (CSV) Code & replication Read the paper

About these estimates

TrustGov scores are posterior means from a dynamic comparative public opinion (DCPO) latent-variable model, on a 0–1 scale where higher values indicate greater public trust in the national government. Because they are model estimates rather than direct survey percentages, each country-year carries measurement uncertainty — the sd, q10, and q90 columns. Analyses that use these scores should propagate that uncertainty rather than treating the means as known quantities.

Coverage is uneven: cases surveyed more often have more precise estimates and longer series. Wide bands mean sparse underlying survey data, not volatile opinion.

Citation

Tai, Yuehong Cassandra. 2026. “Introducing the trust in government (TrustGov) dataset: A new resource for cross-national time-series trust research.” Research & Politics.

Full posterior draws

This page ships the country-year summaries. The full posterior draws, which you need to propagate measurement uncertainty properly, are in the replication repository.