How a quarterly wave of our signature index will be reported, from the component questions to the composite score.
By Integrity Polling research staff
Sample data. This release is a demonstration of our reporting template. No survey was fielded. Every number on this page is illustrative.
Share with “a great deal” or “quite a lot” of trust, by institution
U.S. adults, n = 2,011. Institutions rotated. Gold marks our own industry.
Integrity Polling, Sep 17–24, 2026. n = 2,011 u.s. adults. Question Q1a–Q1i.
Which comes closest to your view?
U.S. adults. The Representation Index is the share who feel represented by either major party, rescaled with intensity.
Whiskers show ±2.5 percentage points, the margin of sampling error.
Integrity Polling, Sep 17–24, 2026. n = 2,011 u.s. adults. Question Q2.
Index trajectory (illustrative back-series)
Demonstration only. Real waves begin in Q1 2027; this series shows how the trend chart will read after eight quarters.
Integrity Polling, Sep 17–24, 2026. n = 2,011 u.s. adults.
Trust is local. Fifty-eight percent of adults trust the officials who run elections in their own community a great deal or quite a lot. That is twenty points higher than trust in the Supreme Court and forty points higher than trust in Congress. People trust what they can see.
Thirty-eight percent say neither party represents them. That is the plurality position. It is also the plurality among adults under 50, and it reaches three in four among independents. In most horse-race polling these people are visible only as “undecided” or “other.” Measuring them directly is the point of this index.
Polls are trusted slightly more than the press and far less than local government. We include our own industry in the battery on purpose. At 29 percent, public opinion polling has a trust problem that this firm exists to address, and we will publish this number every quarter whether it rises or falls.
The index moves slowly, by design. A composite of nine institutions damps the quarter-to-quarter noise that any one item shows. In the illustrative series above, no wave-to-wave change exceeds two points. When a real wave moves the index by more than its standard error of about 0.9 points, we will say so, and when it does not, we will say that too.
The Institutional Trust Index is the unweighted mean of nine 0–100 institution scores. Each institution is scored from the five-point trust question (none at all = 0, very little = 25, some = 50, quite a lot = 75, a great deal = 100). We publish every component so you can rebuild the index yourself, or build a different one, from the crosstab file. The Representation Index combines the party-representation question with its intensity follow-up, so that someone who “somewhat strongly” feels represented counts for less than someone who feels it “very strongly.” Full construction notes are on the Tracker page.
The headline question broken out by group. The CSV includes every question and the unweighted n for every column.
| Response | Democrat | Republican | Independent |
|---|---|---|---|
| The Democratic Party | 71% | 2% | 9% |
| The Republican Party | 1% | 69% | 8% |
| Neither | 23% | 25% | 76% |
| Not sure | 5% | 4% | 7% |
| Unweighted n | 684 | 612 | 715 |
Bold marks the largest share in each row. Columns may not sum to 100 because of rounding. The overall margin of error is ±2.5 points; subgroup margins are larger and grow as the unweighted n shrinks.
Exactly as respondents heard or read it. Bracketed text is interviewer instruction. Rotations are noted where they occurred.
How much trust do you have in each of the following to act in the public interest: a great deal, quite a lot, some, very little, or none at all? [ROTATE: Congress; the Supreme Court; the presidency; federal agencies such as the FDA and the IRS; your state government; your local government; the officials who run elections in your community; the news media; public opinion polls]
Each institution is scored 0–100 (none at all = 0, a great deal = 100). The Institutional Trust Index is the unweighted mean of the nine institution scores.
Which of the following comes closest to your view, even if none is exactly right? The Democratic Party represents people like me. The Republican Party represents people like me. Neither party represents people like me.
The two party statements were rotated. 'Neither' was always read last so it was not advantaged by position.
How strongly do you feel that way: very strongly, somewhat strongly, or not that strongly?
The Representation Index weights Q2 by Q3 intensity: very strongly = 100, somewhat = 67, not that strongly = 33 for respondents who feel represented; 0 for those who chose Neither or Not sure.
Thinking about the last election you voted in, how confident are you that your vote was counted as you intended: very confident, somewhat confident, not too confident, or not at all confident?
In the past twelve months, have you contacted an elected official, attended a public meeting, or volunteered for a campaign or civic organization?
Used as a weighting target to correct for the civic-engagement skew common to survey panels.
Demographic items: age, gender, race and ethnicity, education, ZIP code, party identification with leaners, and 2024 vote. Exact wording is in the questionnaire download.
Percentages may not sum to 100 because of rounding. The margin of error for subgroups is larger than the overall margin. In addition to sampling error, question wording, order and practical difficulties in fielding surveys can introduce error or bias; we publish every question in the order asked so you can judge for yourself.
Paste this into your page. The frame stays current if we issue a correction, and always links back to the full release.
<iframe src="https://integritypolling.com/embed/sample-trust-tracker-pilot-wave/" title="Demonstration release: Trust Tracker pilot wave — Integrity Polling" width="100%" height="440" style="border:1px solid #E4E0EA;border-radius:8px" loading="lazy"></iframe>Among likely voters in a hypothetical statewide race, Maria Delgado (D) leads Thomas Reed (R) 44–40, with Libertarian Alicia Park at 7 and Green Devon Okafor at 2. The lead sits just inside the margin of error.
After hearing a neutral description, 58 percent of U.S. adults support letting voters rank candidates in order of preference. Nearly two-thirds say independent and third-party candidates should face the same signature requirements as major-party nominees.