Government Trust Survey
- Introduced by
- OECD (Trust in Government programme)
- Year
- 2017
- Type
- Population survey instrument for institutional trust
- Family
- Public trust and governance measurement
What does it do?
A government trust survey is a population-based survey instrument for measuring how much citizens trust their public institutions and identifying the drivers of that trust.
Building on the OECD's Trust in Government work, modern instruments treat trust not as a single mood but as a set of measurable expectations: that government is competent and reliable in delivering services, and that it acts on values such as integrity, openness, fairness and responsiveness. By surveying a representative sample of the population, weighting responses to the population, and analyzing trust alongside its drivers, the method produces comparable indicators that diagnose where and why public trust is high or low and what policy levers might raise it.
The intuition behind the method
Trust in government is like trust in any relationship — it rests both on competence (does this institution do what it says, reliably and well?) and on values (does it treat me fairly, honestly and openly?). You cannot improve a relationship by simply asking 'do you trust me?' once a year; you have to understand which specific behaviors build or erode trust. A government trust survey unpacks the overall feeling into its components — reliability of services, responsiveness to needs, integrity of officials, openness of decisions, fairness of treatment — so that a falling trust score points to a concrete cause, such as a perception of corruption or of being ignored, rather than leaving leaders to speculate about a vague malaise.
History
Survey measurement of political and institutional trust dates to mid-twentieth-century studies such as the American National Election Studies' trust-in-government items, and large cross-national programs like the World Values Survey, the European Social Survey and Eurobarometer have tracked institutional confidence for decades. From the 2010s the OECD developed a dedicated Trust in Government work stream, culminating in its 2017 report Trust and Public Policy and subsequent OECD Surveys on Drivers of Trust in Public Institutions, which established a standardized driver-based framework and guidelines for measuring trust comparably across countries. Scholars such as Steven Van de Walle and Geert Bouckaert have contributed to the conceptual and methodological foundations of public-sector trust measurement.
When to use it
Use a government trust survey when policymakers need a representative, comparable measure of public trust and an evidence-based account of what is driving it, whether for national monitoring, international benchmarking, or evaluating the trust effects of reforms. It is well suited to repeated cross-sectional monitoring and to linking trust to specific institutional behaviors. The method assumes a sound probability sample, that respondents can report trust meaningfully and reasonably candidly, and that the driver model captures the relevant determinants. It is less appropriate when only a non-representative convenience sample is available, when sample sizes are too small for reliable subgroup estimates, or when the question is about a specific service encounter rather than systemic trust, where a service-quality instrument fits better.
Limitations and considerations
- Trust questions are sensitive to wording, survey mode and the political climate, so apparent changes can reflect measurement artifacts rather than real shifts.
- Cross-country comparison depends on measurement invariance, which is hard to guarantee given cultural and linguistic differences in how trust is understood.
- Survey measures capture stated trust, which may diverge from behavior such as compliance, and are vulnerable to social-desirability and nonresponse bias.
- Identifying drivers through correlational analysis cannot by itself establish that changing a driver will causally raise trust.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Decomposes trust into measurable competence and values drivers, turning a diffuse sentiment into actionable diagnostic information.
- Population-weighted probability sampling yields estimates that generalize to the public and support rigorous subgroup and trend analysis.
- Alignment with OECD trust frameworks enables credible international benchmarking and comparison across jurisdictions.
- Repeated waves create a governance-feedback loop, allowing the trust effects of reforms to be tracked over time.
Applications
- National governments monitoring public trust in institutions as a headline governance indicator.
- International benchmarking of trust across countries using the OECD's standardized trust survey.
- Evaluating whether public-management or anti-corruption reforms move trust and its underlying drivers.
- Diagnosing disparities in trust across regions and social groups to target responsiveness and inclusion efforts.
- Informing communication and service-improvement strategies by identifying which drivers most shape trust in a given context.
How it works
How it works
The survey is structured around an explicit model of what drives trust, distinguishing competence drivers (responsiveness and reliability of public services) from values drivers (integrity, openness and fairness), and separating trust in different institutions such as national government, local government, the civil service, the courts and the police. Defining these drivers turns a diffuse concept into a set of measurable dimensions, so that results explain trust rather than merely record its level.
A probability sample of the resident population is designed to be representative by age, gender, region, education and other relevant characteristics, with a sample size large enough to support reliable national estimates and key subgroup comparisons. Sampling frames, modes (face-to-face, telephone, web or mixed) and weighting targets are specified in advance so the eventual estimates can be generalized to the population and compared across waves and countries.
The questionnaire is fielded using validated trust items — overall trust in each institution plus batteries on the competence and values drivers — phrased neutrally and consistently to support comparison over time and across jurisdictions. Careful translation, pretesting and mode control are essential, since trust questions are sensitive to wording, social-desirability pressure and the political climate at the moment of fielding.
Responses are combined with survey weights to estimate population-level trust indicators as the weighted mean of individual trust ratings, with the weights correcting for unequal selection probabilities and nonresponse. Standard errors and confidence intervals are computed to convey precision, and indicators are produced both overall and by institution and population subgroup so that disparities in trust are visible.
Formula
\hat{T} = \frac{\sum_{i=1}^{n} w_i \, t_i}{\sum_{i=1}^{n} w_i}The internal consistency of multi-item driver scales is checked with Cronbach's alpha, and the measurement structure is validated through factor analysis to confirm that the competence and values items behave as intended dimensions. Reliability and cross-country measurement-invariance checks are important when results are compared internationally, because a trust scale must mean the same thing in each setting for comparisons to be valid.
Formula
\alpha = \frac{K}{K-1}\left(1 - \frac{\sum_{i=1}^{K}\sigma_i^2}{\sigma_T^2}\right)The estimated indicators and driver analyses are interpreted to guide policy: regressing trust on its drivers reveals which factors most strongly shape trust in a given context, pointing to actions — improving service reliability, strengthening integrity safeguards, opening up decision-making — likely to rebuild it. Tracking the indicators over successive waves shows whether reforms move the needle, turning the survey into a governance-feedback instrument rather than a one-off opinion poll.
Common pitfalls
- Comparing trust figures across surveys that used different question wording, modes or scales as if they were directly comparable.
- Reporting an unweighted convenience-sample result as if it represented the national population.
- Treating a single overall trust number as the headline while ignoring the driver analysis that explains it and guides action.
- Interpreting a driver-trust correlation as proof that improving that driver will raise trust, without considering confounding or reverse causation.
Frequently asked questions
What is the difference between trust in government and satisfaction with services?
Satisfaction is typically about a specific service experience, whereas trust is a broader, more durable expectation about whether institutions are competent and act on the right values. The OECD framework treats service reliability and responsiveness as competence drivers of trust, so satisfaction feeds into trust, but trust also depends on values drivers like integrity, openness and fairness that a service-satisfaction survey does not capture. A government can deliver decent services yet still suffer low trust if it is seen as corrupt or unresponsive to citizens' voice.
Why is the sampling design so important for a trust survey?
Because the goal is to describe the whole population's trust, the sample must be drawn by probability methods and weighted to the population; otherwise the result reflects whoever happened to respond rather than citizens at large. Trust often varies sharply by age, education, region and political alignment, so a non-representative sample can badly mislead. A rigorous design also ensures that comparisons across time and countries reflect real differences rather than differences in who was surveyed.
Can trust surveys be compared across countries?
They can, but only with care. Valid comparison requires measurement invariance — the trust items must mean the same thing across languages and cultures — plus consistent question wording, scales and modes. The OECD's standardized framework and guidelines were designed precisely to make cross-country comparison credible, and analysts test for invariance before drawing comparative conclusions. Without these safeguards, apparent international differences may be artifacts of measurement rather than genuine differences in trust.
Related methods
Sources
OECD (2017). Trust and Public Policy: How Better Governance Can Help Rebuild Public Trust. OECD Public Governance Reviews. Paris: OECD Publishing. Trust in Government.
Open publication referenceCopied.