Trust and Legitimacy
The slow-moving story that explains much of the fast-moving news. Trust in governments, parliaments and parties has fallen in many democracies, and in the United States since the 1960s; trust in courts and police has held up better.
(73% in 1958, 77% in 1964)
(46% in 1972, General Social Survey)
(the lowest since it began measuring in 2015)
A note on framing. Trust is hard to measure precisely, and trust surveys are themselves imperfect tools. The numbers below are drawn from long-running surveys (Pew, Gallup, the General Social Survey, the World Values Survey, Eurobarometer, the OECD Trust Survey, Edelman) that have asked similar questions for years or decades. The slow direction is clear; the precise magnitude is debatable. The page tries to walk through what is well-established about declining institutional trust, what is contested, and what the practical consequences are for political life and individual decisions.
What has declined
Americans have been asked the same question since 1958, first by the American National Election Studies and later by Pew Research and others: how much of the time can you trust the government in Washington to do what is right? In 1958, 73% said "just about always" or "most of the time"; in 1964, 77%; in September 2025, 17%. The fall was not smooth. Trust dropped to 25% by 1980 amid Vietnam, Watergate and economic troubles, recovered in the mid-1980s, rose in the late-1990s boom and reached 60% just after the September 11 attacks, then fell with the Iraq war and the financial crisis. It has not been above 30% since 2007.
Trust in other major institutions has followed similar patterns. In Gallup's June 2026 poll, average confidence in 14 core institutions was 27%, a point above the 2023 record low, with 12 of the 14 at or near record lows. Trust in the mass media fell to 28% in 2025, Gallup's first reading below 30% in over 50 years. Trust in big business, organized religion and higher education is low; trust in the medical system fell during and after the pandemic. The military, the police and the Supreme Court now divide the parties, with Republicans 40 to 46 points more confident than Democrats.
Interpersonal trust - the share of people who say that "most people can be trusted" rather than "you can't be too careful" - has also fallen in the United States. It was 46% in 1972 (General Social Survey) and 34% in Pew's 2023-24 surveys, with 26% among adults under 30. This may matter even more than trust in institutions. A society where most adults expect to be cheated is a different society from one where most adults expect basic fairness, and most measures of social functioning depend on that baseline.
Outside the United States the pattern is more mixed. The largest study, of 3,377 surveys in 143 countries from 1958 to 2019 (British Journal of Political Science, 2025), found trust in parliaments, governments and parties generally falling, but trust in courts, police and civil services stable or rising. Europe shows the same split: in spring 2026, 37% of EU citizens trusted their national government and 74% the police (Eurobarometer). The European Social Survey shows the variation is not random - countries with stronger welfare states, less inequality, and more inclusive political institutions have generally kept higher trust.
Why the decline happened
No single cause explains the decline of institutional trust, and the leading explanations are hard to separate. Each has evidence behind it and serious critics.
Specific institutional failures. Vietnam, Watergate, the Iraq war, the 2008 financial crisis, the Catholic Church abuse scandals, the Boeing safety failures, repeated intelligence failures, the pandemic missteps, and other specific events have given citizens real reasons to lower their trust, and Pew's series moves with them. A 2018 review by Jack Citrin and Laura Stoker finds the sources of change in trust mainly political, from government performance to process. Critics reply that performance is judged through party: since the 1970s Americans have trusted the government more when their own party holds the White House (26% of Republicans, 9% of Democrats in 2025).
Information abundance. In the 1970s, three television networks set the news agenda for most Americans; today millions of sources compete, and more institutional failures are visible. In a study of 840,000 people in 116 countries, Sergei Guriev, Nikita Melnikov and Ekaterina Zhuravskaya (2021) found that approval of government fell as 3G mobile internet spread where the internet was not censored, partly because it exposed real corruption. Yet most of the US fall, from 77% in 1964 to 25% in 1980, came before cable news and the internet.
Polarization. When the same university, newspaper, or government agency is seen as an ally by one side and an enemy by the other, neither side fully trusts it. Marc Hetherington and Thomas Rudolph (2015) found that Americans whose party is out of power now have almost no trust in a government run by the other side. But the rise in polarization is largely American: of 12 rich democracies studied by Levi Boxell, Matthew Gentzkow and Jesse Shapiro, the US had the largest increase over four decades, and six had decreases.
Inequality and economic insecurity. Citizens who feel institutions serve a different class than their own have less reason to trust them (see the Wealth and Inequality page). More unequal countries have lower social trust, and Yann Algan, Sergei Guriev and colleagues found that European regions where unemployment rose in the Great Recession lost trust in national and European institutions. Critics dispute the direction of cause: Andreas Bergh and Christian Bjørnskov (2011) found that historical trust levels predict the size of today's welfare states, suggesting that trust may come first.
Social media dynamics. Engagement-driven feeds tend to reward outrage over agreement, and experiments by Diana Mutz and Byron Reeves found that televised incivility, rather than disagreement itself, lowers political trust. But in the largest field test, moving Facebook and Instagram users to chronological feeds for three months around the 2020 US election cut the uncivil content they saw on Facebook without measurably changing their polarization or political attitudes (Andrew Guess and colleagues, 2023).
Evidence stops authenticating itself. The newest shift is technological, and the least tested. For most of the last century a photograph, a recording or a piece of video carried a presumption of having happened, because faking one took skill, time and money. Generative tools have removed all three barriers at once. The obvious worry is people believing fabrications, but the slower effect may be larger: once anything can be fabricated, anything real can be dismissed as fabricated, and the person dismissing it no longer has to explain how the fake was made. That gives every party to every dispute an exit from inconvenient evidence.
This matters because "check it yourself" has been the fallback whenever institutional trust failed: a reader who did not believe the newspaper could look at the footage. Removing that fallback pushes verification back onto institutions - provenance standards, signed camera output, chains of custody - at exactly the moment when trust in institutions is in short supply. Whether those standards arrive fast enough to matter is an open question.
How countries compare
Trust varies dramatically across countries, and the variation matters for almost every other social outcome. Interpersonal figures below come from the World Values Survey and European Values Study "most people can be trusted" question (latest round, 2017-2022); institutional ones from Edelman (2026), the OECD Trust Survey (2025) and Eurobarometer (spring 2026). All are imperfect; together they give a useful rough picture.
The takeaway: high-trust societies are rare, valuable, and hard to replicate. They tend to share several features (relative equality, inclusive institutions, strong rule of law, low corruption, high social mobility) that take generations to build. The trend is not uniformly down: interpersonal trust rose in Denmark (from 46% in 1984), Germany and the Netherlands by the latest survey round. The Nordic-Swiss high-trust pattern is geographically narrow but is also the closest thing to an existence proof that high-trust modern societies are possible.
Why this matters
Trust is the invisible infrastructure of every other thing societies do. When trust is high, people sign contracts confidently, cooperate with strangers, accept losing votes, pay taxes without coercion, listen to medical advice, follow building codes, hand over their children to teachers, and engage with institutions that have power over their lives. When trust is low, every one of these activities gets harder, more expensive, and more contentious.
Concretely: low-trust societies have higher transaction costs for everything that requires cooperation across strangers. Loans are harder to get. Lawyers do more work. Elections are more contested. Public-health measures fail more often. Tax compliance falls. Vaccination rates drop. Trust in the courts, in elections, in news, in science, in basic safety claims about food or medicine, all interact. A low-trust society is not just a less pleasant place to live - it is measurably less productive, less able to handle crises, and more vulnerable to political manipulation.
The most sobering implication is that some forms of trust loss are very hard to reverse. Specific scandals can be repaired by accountability. Slow erosion across generations is much harder to fix - it requires rebuilding the conditions that produced the trust in the first place, which is a multi-decade project that no political system has fully figured out how to deliver. Whether the Nordic countries got where they are mainly through policy choices over generations or through an older inheritance of trust is one of the disputes below.
The paths from here
Trust is a slow variable. Reversal is possible but the path is harder than it sounds. Each path below is one realistic shape the next two decades could take. Dated September 2026, with US trust in the federal government at 17% (Pew, 2025).
Continued slow erosion
Institutional trust keeps drifting down at roughly the current pace, with periodic small shocks (scandals, failed elections, crises) producing visible drops that do not fully reverse. The information environment keeps amplifying distrust. By 2040 trust in most major institutions is lower still, and political life is more volatile as a result.
Will it happen? For the United States this is the base case: Pew's measure has not been above 30% since 2007. Elsewhere the picture is flatter: the OECD found trust in national governments stable in 2025, after a fall in 2021-23.
Crisis-driven institutional reform
A serious enough crisis (a financial collapse, a major war, a pandemic worse than 2020, an election disputed beyond recovery) forces a serious institutional housecleaning. Reforms similar to the post-Watergate or post-1929 period reset the trust baseline upward. The pattern in history is that some such resets do happen, even though they are unevenly successful.
Will it happen? Possible but not predictable. Crises tend to bring rallies that fade, as after the September 11 attacks, and the US-Israeli war with Iran since February 2026 has brought no visible rally. Whether a crisis produces reform or further breakdown depends on choices that have not yet been made.
Information environment reform
Some combination of platform regulation, media literacy at scale, technological tools for verifying claims, and the rise of new credible institutions slows or reverses the erosion of news and information trust. Citizens begin to be able to tell reliable from unreliable sources more consistently than today.
Will it happen? Pieces are happening: the EU's Digital Services Act has produced its first fines (on X, Temu and AliExpress in 2025-26); media-literacy programs and AI-aided verification tools are growing. Finland comes closest among rich countries, with 63% trusting most news, but trust fell in 29 of 48 markets in 2026 (Reuters Institute). The political will is uneven and the technological challenges are real.
Local trust replaces national trust
As trust in national institutions falls, citizens increasingly invest their trust in local institutions, communities, and personal networks instead. The trust deficit at the national level produces stronger local-level engagement, civic life, and identity. Politics decentralizes de facto even where the formal structure remains national.
Will it happen? Already partly happening. Local government is trusted more than national government in most OECD countries; in the EU, 63% trust their regional or local authorities against 37% their national government (Eurobarometer). Several political movements (in the US, Europe, and elsewhere) explicitly emphasize local sovereignty and devolved decision-making. The trade-off is that some problems require national or international cooperation that purely local institutions cannot provide.
Strongman politics fills the gap
Some societies respond to declining institutional trust by embracing leaders who promise to restore order and credibility through personal authority rather than through institutional reform. The short-run effect can restore a sense of direction and decisiveness - and in some cases delivers real improvements in specific governance areas. Under Nayib Bukele, El Salvador's government reports that the homicide rate fell from 53 per 100,000 in 2018 to 1.9 in 2024, after more than 80,000 arrests under a state of exception that human-rights groups criticize. Singapore's People's Action Party is often cited on public services. The longer-run track record is more mixed: personalized authority that bypasses institutions can produce effective governance under a competent leader and arbitrary governance under a less competent successor, because the same institutional weakness that enabled the first also enables the second.
Will it happen? Already happening in multiple countries, though voters can reverse it, as Hungary's did by voting Viktor Orbán out in April 2026 after 16 years. Turkey and, in some respects, India are other commonly cited cases. Critics of Donald Trump's second administration say the United States fits the pattern; his supporters say he is carrying out an electoral mandate against unaccountable institutions. Whether these surges produce durable governance improvements or long-run institutional erosion is one of the central political questions of the period.
Continued slow erosion is the base case in the United States; across the OECD, the latest survey points to a plateau. Some institutional reforms in response to specific events are likely. Authoritarian temptations will be visible in some countries. The Nordic high-trust societies will probably keep being cited as models without being widely replicated. The most useful question for any specific country is not "will trust be restored" but "what specific institutions can be made trustworthy by what specific changes" - the question that countries that have actually made progress on this have answered well.
Where serious analysts disagree
Trust is one of those topics where careful researchers using similar data reach different conclusions about cause and consequence. The named voices below are worth reading directly.
Inequality is the deeper driver
Across countries and across time, the level of economic inequality correlates strongly with the level of social trust. Countries that have allowed inequality to widen have seen trust fall; countries that have kept inequality moderate have maintained trust. The road back to higher trust runs through tax-and-transfer policy, affordable housing, accessible education, and the other supports that produce broad shared prosperity.
Held by: Richard Wilkinson and Kate Pickett ("The Spirit Level"), Eric Uslaner (Maryland), Bo Rothstein, and many welfare-state researchers. The cross-country correlation is strong; the direction of cause is disputed (see above), and the policy implication is more contested.
Diverse societies have a harder time maintaining trust
Robert Putnam's controversial 2007 work found that in more diverse US communities, residents of all races tend to "hunker down": trust, even of their own group, is lower, even after controlling for income, education, and many other factors. Putnam notes that successful immigrant societies have overcome this in the long run. Holders argue the short-run finding is uncomfortable but should be faced, not dismissed.
Held by: Robert Putnam (Harvard) himself; a 2020 meta-analysis of 87 studies by Peter Thisted Dinesen and colleagues found a negative association, strongest for trust in neighbors and changed only slightly by controls. The finding is contested on the empirics, not only on policy implications. Maria Abascal and Delia Baldassarri (2015) re-analyzed Putnam's own data and called the association a "compositional artifact": groups that report lower trust for other reasons are overrepresented in diverse communities. A 2014 review of 90 studies by Tom van der Meer and Jochem Tolsma found consistent effects only on neighborhood cohesion, more often in the US than elsewhere.
The decline is partly an information artifact
Citizens today see more institutional failures because more of them are visible. The actual performance of institutions may not have declined; the information environment has gotten harsher about what it shows. If this view is correct, "fixing" the trust decline requires addressing the information environment more than the institutions themselves.
Held by: Martin Gurri, for 28 years an analyst of open media at the CIA ("The Revolt of the Public"), and media-effects researchers such as Diana Mutz. Whether performance has actually changed or just become more visible is hard to disentangle: the 3G study above found the internet lowered approval partly by exposing real corruption.
The decline reflects real institutional problems
The standard reading: institutions have actually performed less well in recent decades than in the postwar consensus. Failed wars, financial crises, scandal cover-ups, regulatory capture, and elite self-dealing have given citizens valid reasons to lower their trust. Restoring trust requires actually fixing the institutions, not better communications or better media.
Held by: a wide range of political scientists, journalists, and reform advocates on the left and the right, drawing on reviews such as Citrin and Stoker's. Critics answer that perceived performance is filtered through partisanship. The open question is which reforms actually move the trust needle and which leave it largely unchanged.
High-trust countries are not exceptional - they are explainable
The Nordic countries did not get to high trust through cultural magic. They got there through specific decisions about welfare, equality, education, public service quality, and institutional design that were made over generations. Other countries can replicate the conditions; they have not chosen to. On this view, treating Nordic exceptionalism as a mystery is a way of avoiding the uncomfortable lessons.
Held by: Bo Rothstein (Gothenburg) and the broader Quality of Government research tradition, which ties trust to impartial, uncorrupt public institutions. Critics argue that trust is partly inherited and slow to move: Yann Algan and Pierre Cahuc (2010) found that the trust of descendants of US immigrants still reflects their ancestors' country of origin.
Where the evidence is firmest: trust in representative institutions has fallen in many democracies, steeply in the United States, while trust in courts and police has held up better. Each explanation above has real support and serious critics. The diversity effect, where found, is strongest in neighborhoods, and whether high-trust societies were built by policy or inherited their trust is still argued. Readers may weight these factors differently depending on their own political and cultural framework.
What this means for you
Trust is one of those topics that touches everything indirectly and is hard to act on directly. A few practical observations:
If you are a citizen
Notice the difference between specific and generalized distrust. Specific institutional failures deserve specific anger and demand specific accountability. Generalized distrust ("they're all corrupt," "you can't trust anyone in government") is rarely useful and is often the entry point for cynicism that is paralyzing. Engaged citizenship in low-trust times requires the harder work of distinguishing which specific institutions and practices need reform from the broader cultural mood that dismisses everyone equally.
If you read news
The information environment rewards content that reinforces distrust. Almost every algorithmic feed will surface the most dramatic and most outrage-inducing claims. Active selection of news sources - including some you disagree with, and some that focus on what is working rather than only on what is failing - is a real exercise in maintaining personal calibration. The cost of consuming only outrage-driven content is mood, judgment, and energy that better-curated reading saves.
If you work in or with institutions
The trust deficit lands disproportionately on the people inside institutions doing the daily work. Whether you are a public-school teacher, a city employee, a journalist, a healthcare worker, or someone running a small business, the public mood toward your institution affects your daily life. The most useful response is the one that actually rebuilds trust through specific consistent action - showing up reliably, being clear about what you can and can't deliver, and being honest about mistakes when they happen. Trust rebuilding happens at the unit-of-trust scale, not at the institutional scale.
If you are raising children
Children pick up trust calibrations from the adults around them. Children raised by adults who model relentless cynicism often grow into adults who struggle to trust anyone, including the people who would benefit them most. Children raised by adults who model thoughtful trust - "trust by default, verify when consequences are large, learn from specific betrayals without globalizing them" - tend to navigate complex social environments better. The home is the first and most durable trust-formation environment.
If you are thinking about emigration or immigration
The high-trust countries are real and the difference is large in everyday life. Countries with high institutional trust feel different to live in - public services work, basic transactions are honest, expectations of fairness are met more often than not. If you are choosing where to spend a long career or raise children, the trust environment of a country deserves serious weight alongside more visible variables like taxes, weather, and language. Most measures of national quality of life correlate strongly with trust, even when the official statistics don't capture it directly.
The mechanics behind this
The trust story sits on top of three deeper mechanisms covered elsewhere on this site. If the analysis above depends on ideas you want to understand first, these fundamentals make the conversation more legible:
