Representative Institutions Lose Trust; Implementing Institutions Hold or Rise

Why This Question Matters

Many scholars and commentators argue that democracies face a broad 'crisis of political trust.' Empirical studies, however, disagree: some report clear declines while others find only fluctuations. Viktor Valgarðsson shows that these divergent results reflect analytical confusion—whether trust is falling depends on which institutions are measured, which countries are considered, and which time period is studied.

What Valgarðsson Did

Valgarðsson assembles an unprecedentedly broad dataset on institutional trust, pooling 3,377 surveys from 50 survey projects covering 143 countries between 1958 and 2019. He groups trust responses for six institutions into conceptually distinct types and uses Bayesian dynamic latent trait models to recover underlying long-term trends while accounting for measurement differences across surveys and countries.

How Trust Was Measured

• Cross-national survey items about trust in six institutions were modeled as noisy indicators of latent trust over time.
• Bayesian dynamic latent trait models trace how those latent levels change year by year for each country and globally, allowing the author to separate short-term volatility from persistent trends.

Key Findings

• There are important regional and country-level differences in trust trends; some places show sharp swings while others remain stable.
• Globally, trust in representative institutions has generally declined in recent decades.
• By contrast, trust in so-called 'implementing' institutions has been stable or has increased over the same period.
• These patterns help explain why previous studies reached different conclusions: aggregating across institutions or periods can mask divergent trends.

What This Means for Debates About Democratic Crisis

Valgarðsson's analysis suggests that claims of a universal, uniform collapse of political trust are oversimplified. The political relevance of trust trends depends on which institutions are changing and where; declines concentrated in representative bodies pose different democratic questions than stable or rising trust in implementing agencies. The paper provides a clearer empirical basis for debates about institutional legitimacy and for scholars seeking to link trust trends to electoral behavior, policy capacity, and democratic resilience.