Case Study 1: The Long Decline of Trust in Institutions, 1958–2026

"In Government we trust" never appeared on American currency. The closest thing was "In God We Trust," which Congress added in 1956 — at the height of an era when Americans trusted both. The trust in God has held up better than the trust in government.

The data series

The American National Election Studies (ANES) has asked the same question, in the same wording, every two years since 1958: "How much of the time do you think you can trust the government in Washington to do what is right — just about always, most of the time, or only some of the time?"

The combined "just about always" plus "most of the time" response — the trust-in-government number — is one of the longest-running political-attitude time series in any democracy. Its trajectory:

• 1958: 73% trusted government most of the time or always.
• 1964: 77%, the highest reading on record.
• 1972: down to 53%, with Vietnam and the early Watergate revelations.
• 1976: 33%, the post-Watergate trough.
• 1984: modest recovery to 44% during the Reagan economic expansion.
• 1992: back down to 29%.
• 2000: rebound to 44% during the late-Clinton-era prosperity.
• September 2001: spike to 60% in the immediate post-9/11 polling — the only major upward shock in the entire 65-year series.
• 2004: declining back to 47%.
• 2008: further decline, around 30%.
• 2014: historic lows, around 24%.
• 2020: brief rise during pandemic to around 28%.
• 2022: decline to 20%.
• 2024: around 22%, statistically indistinguishable from 2022.

The pattern, stripped of detail: Americans of the 1960s trusted their federal government roughly three times as much as Americans of the 2020s do. The decline has been long, mostly steady, and has resisted every administration's effort to reverse it.

What did Americans trust in 1964 that they don't trust now?

The trust readings of 1964 are striking partly because they came at what most Americans now think of as a moment of high political division — the Civil Rights movement, Vietnam beginning to escalate, the Goldwater-Johnson election. The country was divided on substance, but Americans nonetheless said, in large numbers, that they trusted government to do what was right.

Several explanations have been advanced.

The era's peer institutions were also trusted. Confidence in churches, the military, large corporations, banks, and the press were all in the 60–80% range in the late 1950s and early 1960s. The high-trust posture toward government was part of a high-trust posture toward American institutions generally. The decline in trust in government has been mirrored by decline in trust in nearly every other major American institution, with the partial exception of the military (which remains relatively trusted, around 60% in recent Gallup data) and small business (around 65%).

The electorate was less politically engaged with daily news. Households watched the evening news on one of three networks; political coverage was thinner; partisan media was rare. The trust-in-government number partially reflected unfamiliarity with the day-to-day texture of governance.

The mid-century federal government had recent triumphs in living memory. Americans interviewed in 1958 had personal experience of the New Deal, World War II, the Marshall Plan, the Berlin Airlift, the Interstate Highway System, the GI Bill. The federal government had won a war, defeated a depression, and built a middle class. The trust readings reflected the prestige earned in that era.

The first inflection point came with Vietnam, then Watergate. The decade between 1965 and 1975 saw escalating distrust as the federal government's failures became visible. The Tet Offensive (1968) revealed that official optimism about Vietnam was contradicted by reality. The Pentagon Papers (1971) showed that the government had been systematically misleading the public for years. Watergate (1972–74) showed the President himself participating in cover-up. By 1974, the trust readings had collapsed from 77% (1964) to 36%.

This collapse has never been fully reversed. Subsequent inflection points — the Reagan recovery, the Clinton boom, the post-9/11 rally — produced temporary rebounds but did not return trust to its pre-Vietnam levels.

The post-9/11 moment

September 2001 deserves separate attention because it is the only major positive shock in the entire 65-year series. ANES polling and Gallup polling both showed trust-in-government numbers leaping from around 30% in mid-2001 to around 60% in the weeks after the attacks. Confidence in the military jumped to over 80%. Approval of the President went from 51% (early September 2001) to 92% (late September 2001) — the highest ever recorded.

The "rally around the flag" effect lasted longer than the typical week-or-two crisis spike. Trust-in-government remained elevated through 2002. Then it began to decline as the Iraq War's justifications were contested, as the military operation extended longer than promised, as the Department of Homeland Security's color-coded threat-level system became a target of mockery, and as the post-Hurricane-Katrina (2005) federal response was widely judged a failure. By 2008, trust was back below 30%, and the financial crisis pushed it lower.

The 9/11 episode shows what does and does not move trust. A unifying external threat, combined with effective leadership and a sense of national solidarity, can substantially raise trust in government. But the rise is fragile; perceived government failures dissipate it within years.

The financial-crisis decline

The 2008 financial crisis produced a different kind of trust shock. Unlike Vietnam or Watergate, the crisis was not primarily a story of government deception or moral failure. It was a story of government — and elite-financial-sector — failure to prevent a foreseeable disaster, followed by an emergency response that bailed out the institutions whose mistakes had caused the crisis while leaving ordinary Americans to absorb the costs.

The trust readings reflect this: not just decline in trust in government, but decline in trust in expertise — in the people who claimed to know what they were doing. The post-2008 era produced both the Tea Party movement on the right (skeptical of bank bailouts, deficit spending, expanding federal power) and Occupy Wall Street on the left (skeptical of financial-sector capture, oligarchic inequality, regulatory failure). The two movements disagreed on most things but shared a foundational distrust of the consensus elite, including the federal government's role in supporting that consensus.

This anti-establishment moment shaped the 2016 primaries. Bernie Sanders's left-populism and Donald Trump's right-populism both drew from the same well of post-2008 elite distrust, and both significantly outperformed expectations. The trust data pointed at the political possibility before most analysts saw it.

Polarization and the partisan structure of trust

A defining feature of post-2010 trust-in-government data is that the level among the in-party (the party currently holding the White House) has not recovered to mid-century norms. In the 1950s, both Republican and Democratic respondents expressed high trust regardless of which party was in power; the question seemed to be about confidence in the federal apparatus generally rather than about evaluation of the current administration.

By the 2010s, trust-in-government had become almost entirely a partisan-evaluation question. When Democrats held the White House, Democratic respondents expressed elevated trust (typically 35–45%) and Republican respondents very low trust (typically 15–25%). When Republicans held the White House, the pattern reversed. The aggregate number — averaged across both parties — has stayed in the 20–30% range because the in-party gain is offset by the out-party loss.

This is itself a sign of polarization. The question "do you trust the government?" used to be answered partly on the basis of how respondents evaluated the institution. It is now answered substantially on the basis of which party they belong to. Institutional confidence has been absorbed into partisan identity.

The implication for democratic theory: a polity in which institutional trust is conditional on partisan victory is a polity at risk. When trust is identity-driven rather than evidence-driven, no amount of good government performance under a given administration can rebuild the trust of the out-party. The institution becomes a hostage to electoral outcomes rather than a stable backdrop for them.

Cross-national context

The American trust-in-government collapse looks even more extreme when compared to peer democracies. Cross-national data from the OECD's Trust in Government index, the World Values Survey, and the European Social Survey show, as of 2023:

• Switzerland: around 65% trust in national government.
• Norway: 65%.
• Sweden: 60%.
• Germany: 55%.
• Canada: 50%.
• Australia: 48%.
• United Kingdom: 35% (also having declined sharply, especially after Brexit).
• France: 35%.
• Italy: 30%.
• United States: 22%.

The United States is the outlier on the low-trust end among advanced democracies. Switzerland's 65% is roughly where the United States was in 1964. Most Northern European democracies are at trust levels the U.S. has not seen in 40 years.

This comparison is humbling for several reasons. First, it suggests the American decline is not inevitable: other countries with comparable problems (immigration debates, economic inequality, populist political movements) have maintained much higher institutional trust. Second, it shifts the analytic question: not "why has trust declined?" but "why has American trust declined more than peer democracies?" The candidate answers — extreme polarization, the structure of American media, the depth of the post-Vietnam-Watergate trauma, the design of American electoral institutions, the asymmetric impact of inequality, the religious and racial dimensions of American political conflict — are still being debated.

What it means for governance

A federal government operating at 20% public trust faces severe constraints.

• Compliance. Citizens are less likely to comply voluntarily with laws and tax obligations when they distrust the institutions imposing them. Tax-compliance research suggests trust correlates with voluntary compliance independent of enforcement intensity.
• Public-health response. The COVID-19 pandemic demonstrated this in real time. Public-health guidance was ignored by substantial portions of the population, especially the portions with low institutional trust. The pandemic's death toll was higher in low-trust regions and demographics, controlling for other variables.
• Coalition-building. Major reforms — to Social Security, to immigration, to climate policy — historically required building cross-partisan coalitions. With trust collapsed and polarized, coalition-building is harder; the political capital required to push major change exceeds what most administrations possess.
• Constitutional resilience. A democracy with 20% institutional trust is more vulnerable to erosion of constitutional norms. When citizens already believe institutions are failing, accusations that institutions have been captured (by deep-state Democrats; by oligarchic Republicans; by either party's elites) land more easily.

None of this is irrecoverable. The cross-national data show that trust can be high in modern democracies. The post-9/11 American data show that trust can rise rapidly under the right conditions. But the recent American trajectory is downward, and reversing it would require something different from current political dynamics.

What this case study tells us about public-opinion measurement

A few methodological observations. First, the ANES question's stability is itself a methodological achievement: 65 years of identically worded survey items make trend analysis possible. Most survey questions are not asked with this consistency, and trend claims about other variables (e.g., racial attitudes, religious belief, gender norms) require careful harmonization across differently worded items.

Second, the gap between ANES and Pew on this question is small but real: Pew typically reports trust readings 2–4 points lower than ANES, partly because Pew's online panel produces more candid responses on questions where social-desirability bias affects answers.

Third, the question "trust government to do what is right" is itself contestable. "Right" is open to interpretation; "the government in Washington" elides the federal-state-local distinction (state and local government have higher trust readings); "most of the time" is a frequency judgment with substantial respondent variability. The trend is meaningful precisely because the question wording has been held constant; it would be hazardous to compare absolute levels across questions worded differently.

For a Chapter 17 reader: this single time series illustrates many of the chapter's themes. Public opinion is measurable; the measurement is partial; the trends are real; the partisan structure of opinion has changed; the institutional stakes are large. The trust-in-government decline is not just one fact about America. It is the backdrop against which most of what we discuss in subsequent chapters — polarization, social movements, the media environment, constitutional erosion — must be understood.

Sources: ANES Time Series Cumulative Data File, variable VCF0604, accessed 2026 via electionstudies.org. Gallup historical confidence-in-institutions trend, news.gallup.com. OECD Trust in Government 2023 indicators. Pew Research Center, "Public Trust in Government, 1958–2024," pewresearch.org.