Government and rights, the evidence

The evidence on government and rights in the United States comes mainly from long-running surveys of trust and confidence, comparative democracy indices, and research on whether public opinion predicts policy. The surveys agree that trust in the federal government is low, 17 percent in Pew’s September 2025 reading against 73 percent in 1958, but researchers disagree about why, about whether ordinary citizens influence policy, and about how to measure rights at all. This page sets out what is measured, what is disputed, and what is unknown; the subject overview and who decides pages give context.

In plain words

Researchers have asked Americans the same trust question for almost 70 years. Trust has fallen a long way. That much is solid.

Why it fell, and whether it matters, is less clear. Trust goes up and down with which party is in power. Some studies say ordinary people have almost no influence on what government does. Other researchers say those studies overstate the problem. Rankings that score countries on democracy are useful, but they are built on judgment calls.

The honest summary: we know a lot about what people say, less about what drives it, and least about how well rights are actually protected day to day.

Defined terms

Trust in government (survey measure). Agreement that one can trust “the government in Washington to do what is right” just about always or most of the time. The question began in the American National Election Studies and is continued by Pew Research Center.

Confidence in institutions. Gallup’s measure of the share expressing “a great deal” or “quite a lot” of confidence in named institutions.

Policy responsiveness. The degree to which government policy changes in line with public preferences.

Democracy index. A composite score that rates countries on elections, civil liberties, and related dimensions using expert judgment and documentary evidence.

Partisan asymmetry. The pattern in which supporters of the party holding power report more trust than supporters of the party out of power.

What the data show: trust and confidence

Government and rights data are most robust for stated trust. Pew’s series, built on American National Election Studies readings from 1958 and continued in its own surveys, recorded 73 percent trust in December 1958, a peak of 77 percent in October 1964, 22 percent in May 2024, and 17 percent in a poll dated September 28, 2025.⁠¹

Confidence in specific institutions varies widely. Gallup’s June 2025 survey found confidence in Congress around 10 percent, the presidency at 30 percent, the military at 62 percent, and an average of 28 percent across the nine institutions it has tracked since 1979.⁠² Gallup’s June 2026 survey of 1,001 adults found the average across 14 core institutions at 27 percent, one point above the 2023 record low, and reported partisan gaps of 70 points on the presidency and 40 points on the Supreme Court.⁠³

Perceived voice is low too. In Pew’s July 2023 survey, 70 percent said people in their own congressional districts have too little influence on Congress. Fifty-seven percent said voting by people like them can affect the country’s future direction, but only 20 percent said it can do so “a lot”.⁠⁴

What the data show: proposals for structural change

Government and rights reform proposals draw different levels of support, and the differences are informative. In Pew’s July 2023 survey, 79 percent favored a maximum age for federal elected officials and 74 percent favored age limits for Supreme Court justices, while a proposal to increase the number of Supreme Court justices drew 46 percent in favor and 51 percent opposed.⁠⁴ The pattern suggests that proposals framed around general fairness or turnover attract broad agreement, while proposals understood as shifting power between current coalitions divide the public closely. That is an interpretation of the pattern, not a finding the survey itself tested.

How the trust measure is built, and its limits

Government and rights trust readings form a long series, but the series is stitched together. Pew’s trend combines American National Election Studies readings for the early decades (for example December 1958 and October 1964) with Pew’s own polls in recent years, such as May 2024 and September 2025.⁠¹ Interpretation: when a series spans different survey organizations, modes of interviewing, and sampling methods, small year-to-year changes should be read cautiously, while a fall of more than 50 points over six decades is far larger than any plausible method effect. The question also asks about “the government in Washington,” so it says nothing directly about state or local government, and it does not specify which branch or which officials the respondent has in mind. A respondent who distrusts Congress but trusts the courts gives one answer for both. See question design.

What the data show: comparative context

Government and rights dissatisfaction is not uniquely American. Pew’s 2024 surveys in 31 countries found a median of 54 percent dissatisfied with how democracy is working and 45 percent satisfied; the U.S. survey was conducted April 1 to 7, 2024, and Pew reported a double-digit fall in U.S. satisfaction between 2021 and 2024.⁠⁵

Freedom House’s Freedom in the World 2025 rates the United States “Free,” with 84 of 100 points (34 of 40 for political rights and 50 of 60 for civil liberties), up one point from the prior year. Its narrative cites polarization, gerrymandering, threats against election workers and officials, partisan judicial appointments, press freedom incidents, and an immigration court backlog of about 3.6 million cases.⁠⁶

Where the evidence is disputed

Does the average citizen influence policy? Gilens and Page analyzed 1,779 policy issues and concluded that economic elites and business-oriented groups have substantial independent effects on policy, while average citizens and mass-based groups have “little or no independent influence”.⁠⁷ The study is widely cited by people across the spectrum who argue that the system is captured. Peter Enns responded that even on issues where middle-income and affluent preferences diverge, “policy ends up about where we would expect if policymakers represented the middle class and ignored the affluent,” because the two groups usually rank policies in the same order; he concluded that “the opportunity of unequal representation of the ‘average citizen’ is much less than previously thought,” while finding “substantial opportunity” for unequal representation of strong partisan preferences.⁠⁸ Gilens replied in the same journal issue disputing the extent of that coincidence. The dispute is about method (how to treat correlated preferences) and remains open.

Is low trust a verdict on performance or on partisanship? Pew’s series shows the parties trading places: in May 2024, 35 percent of Democrats and 11 percent of Republicans expressed trust; in September 2025, 9 percent of Democrats and 26 percent of Republicans.⁠¹ That pattern supports reading trust partly as partisan identity. It does not rule out a performance component, since the overall level has stayed low under both parties.

Do democracy indices measure the same thing? Freedom House’s score depends on its analysts’ judgments about which events matter and how much. Critics on different sides contest specific items in its U.S. narrative, for instance whether particular state laws on abortion, transgender policy, or protest are restrictions of civil liberties or legitimate democratic choices.⁠⁶ An index is evidence of expert assessment, not a neutral measurement like a vote count.

Which rights are under threat? Survey evidence shows Americans weigh the same right differently depending on the threat in view. Between 2018 and 2021, support for government steps to restrict false information online, even at a cost to freedom of information, rose from 39 to 48 percent, with Democrats (65 percent in 2021) and Republicans (70 percent preferring freedom of information) on opposite sides.⁠⁹ Neither side’s answer is a measurement of how well speech is actually protected.

What is unknown

Government and rights research leaves several important gaps.

• Causes of declining trust. Candidates include government performance, scandals, polarization, media change, and economic insecurity. The survey series records the decline; it does not identify the cause.
• Local versus national trust. Most of the data above concern national institutions. The United States has 90,837 local governments,⁠¹⁰ and how trust in them moves, and whether it moves with national trust, is far less studied in public data.
• Rights in practice. Constitutional text and court opinions say what rights are; they do not say how often officials respect them in ordinary encounters. Counting violations depends on reporting, which is itself uneven.
• Whether participation restores trust. Proposals to rebuild trust through more participation rest on plausible reasoning, but evidence that new participation tools raise trust over time is limited, and early adopters are self-selected. See evaluation.

Evidence

Government and rights evidence on this page is drawn from Pew Research Center and Gallup survey releases with stated dates and samples, Freedom House’s 2025 country report, two peer-reviewed articles in Perspectives on Politics, and the Census Bureau’s count of governments. Each number above carries its date; readers should expect newer readings to differ.

Competing interpretations

Government and rights evidence supports several competing readings, each held in good faith.

• Crisis reading. Low trust, low perceived voice, and contested elite influence add up to a legitimacy problem that requires structural reform.
• Polarization reading. Much of the measured distrust is the out-party’s distrust of the in-party; reduce polarization and trust recovers without structural change.
• Skeptical citizen reading. A free people should not trust government “most of the time”; low stated trust can coexist with continued compliance, participation, and peaceful transfers of power, and may even be healthy.
• Measurement reading. Trust questions are vague, indices embed judgment, and responsiveness studies disagree; strong conclusions in any direction outrun the data.

Application to this framework

Government and rights evidence shapes what My Vote Forward claims about itself. The framework is built on the idea that people want ongoing, structured ways to state priorities between elections, but the research on measurement and evaluation means it treats that idea as a hypothesis, not a finding. Any published release of priorities in this subject should report partisan composition where consent allows and privacy rules permit, because the trust data show how strongly partisanship drives answers. See published releases.

Unresolved questions

• What survey design would separate “distrust of the party in power” from “distrust of the institution”?
• Could priority data, gathered repeatedly from the same people, show whether trust changes when institutions respond, and what would count as responding?
• How should a participation tool report evidence that its own participants are unrepresentative of the jurisdiction?