Summary

  • The NPR/PBS News/Marist poll of 1,280 registered voters documents a 13-percentage-point decline in confidence in USPS timely delivery of mail-in ballots, with partisan rhetoric, documented service-quality problems, and heightened institutional salience following the Sept. 14 Supreme Court ruling all identified as plausible contributing factors.
  • USPS reports a 97.94% on-time delivery rate for more than 233 million election-mail pieces through August, while Government Accountability Office and USPS Office of Inspector General findings record unmet service-performance goals and disproportionate rural impacts.
  • The partisan perception gap is substantial, with 70% of Republican respondents versus 18% of Democratic respondents reporting belief that there is a great deal or good amount of mail-voting fraud.
  • The Sept. 14 Supreme Court ruling constrained the Trump administration’s documented position of using USPS to restrict voting by mail for the midterms.
  • Washington state data shows 19 counties saw late-postmark ballot rejections more than double in this year’s primary compared with 2024, providing jurisdiction-specific evidence of postmark changes tied to USPS restructuring.

Confidence in the U.S. Postal Service’s timely delivery of mail-in ballots has fallen to 58% from 71% in 2024, according to the NPR/PBS News/Marist poll of 1,280 registered voters conducted Sept. 14–15 in English and Spanish with a reported 3.8-point margin of error. The same survey recorded 75% confidence in state and local election administration — up from 69% in March — indicating the confidence decline is concentrated on USPS rather than on election administration generally. The poll was released shortly after a Sept. 14 Supreme Court ruling that blocked the Trump administration from using USPS to restrict voting by mail for the midterm elections. The article reports that USPS, in its own statement, attributed sentiment changes to “a lot of misinformation recently about the Postal Service’s role in the electoral process.” The poll, USPS’s performance claims, and the GAO and OIG findings together support a multi-causal picture rather than a single explanation.

What USPS reports and what its oversight bodies record

USPS stated it delivered more than 233 million pieces of election mail this fiscal year through the end of August with an on-time score of 97.94%. “The Postal Service’s actual performance metrics speak for themselves,” the USPS statement said, according to the article. “Postal employees take great pride in their responsibility to support the democratic process.”

Against that figure, the article cites a recent Government Accountability Office report finding that USPS’s 10-year restructuring has produced longer expected delivery times and lowered service-performance goals that USPS has been unable to fully reach. The USPS Office of Inspector General, in a report issued this month, concluded that major changes to USPS’s delivery network have not improved its financial position and have disproportionately affected rural communities. The inspector general warned of the risk of losing “customers who perceive the service as inequitable, which could harm its brand and reputation, and adversely affect revenue.”

The USPS performance metric and the GAO and OIG findings describe different dimensions of service. A national on-time score does not speak directly to whether service-performance goals are being met, and unmet goals do not by themselves establish voter-level experiences. The article places these claims in tension without reconciling them, and the supplied record does not provide jurisdiction-level performance data comparable to the national figure.

Postmark changes and Washington state evidence

The article reports that USPS reorganization has affected when mail receives a postmark, which is often used to determine whether a mail-in ballot was sent on time. A report released by Democratic U.S. Sen. Maria Cantwell and Washington Secretary of State Steve Hobbs found that 19 Washington counties saw their rate of ballots rejected due to late postmarks more than double for this year’s primary election compared with 2024.

The Washington evidence documents a jurisdiction-specific postmark problem but does not establish a national general-election pattern. The article notes that the Cantwell–Hobbs release supplied jurisdiction-specific information that supplements USPS’s national metric. The supplied record does not establish whether USPS has published equivalent jurisdiction-level information for other states.

The Supreme Court ruling and USPS’s planned operations

The Sept. 14 Supreme Court ruling blocked the Trump administration’s effort to use USPS to restrict voting by mail for the midterms. Following the ruling, USPS was expected to keep delivering election mail as originally planned, including putting in place its “extraordinary measures” policy for speeding the delivery of general-election ballots beginning a week before Election Day.

The article does not document a follow-on filing or alternative legal effort by the administration. The ruling supplies a legal criterion concerning USPS’s role in mail-voting restrictions, but the article does not establish how the ruling’s timing may have affected the poll’s recorded confidence figures.

Partisan perception and mail-voting participation

The poll found a continuing partisan divide in views about mail-in voting. Democratic voters (32%) were more likely than Republican voters (12%) and independent voters (26%) to say they would vote by mail in this year’s general election. Republicans (70%) were more likely than Democrats (18%) and independents (32%) to say they thought there was a great deal or good amount of fraud when people are given the option to vote by mail.

The article attributes to President Trump the unsupported claim that mail voting is “corrupt” and reports that he votes by mail. The article cites a 2025 Brookings Institution review as having found mail-voting fraud to be “exceptionally rare”; that claim remains dependent on the article’s attribution because the supplied material does not independently provide the review. According to the new poll, 55% of registered voters overall said they believed there was little or no fraud when mail-in voting is available.

The 52-point gap between Republican and Democratic responses on perceived mail-voting fraud is large in directional terms, but its precise statistical significance cannot be determined from the supplied material because subgroup-specific margins are not provided. Partisan crosstabs for confidence in USPS delivery by political affiliation or geography are also not supplied, leaving unresolved whether the confidence decline has the distribution that a political-rhetoric account would predict.

Three live explanations and what the evidence does not resolve

Three causal accounts are consistent with the supplied record. A political-framing account holds that partisan claims about mail voting shifted perceptions independently of USPS operational performance; the partisan distribution of fraud perceptions and mail-voting intentions is consistent with that account. A service-quality account holds that changes in delivery performance, postmarking, or ballot processing affected voter confidence; the GAO and OIG findings and the Washington postmark data are consistent with that account. A salience account holds that the Sept. 14 Supreme Court ruling and the surrounding dispute increased public attention to USPS’s role and influenced perceptions; USPS’s statement attributing sentiment changes to “a lot of misinformation recently” is consistent with that account.

The three accounts are not mutually exclusive, and the supplied evidence does not isolate their relative contributions. Even with partisan crosstabs for confidence, distinguishing political rhetoric from service effects would require controlling for geographic exposure to delivery and postmark problems; the supplied material does not provide that test. The source therefore supports a mixed interpretation in which jurisdiction-specific service problems and partisan perception effects may reinforce each other, but it does not establish a single dominant cause.

Institutional roles and the absence of cross-partisan bridge-building

The William Ury framework for third-party roles, as characterized in the analytical literature, organizes responses to conflict into prevention, resolution, and containment functions. USPS performs a provider function for timely delivery and postmark integrity, but its performance and reputation are contested. NPR, PBS News, and Marist perform a reporting and information function by documenting public sentiment and partisan differences; the article’s editor’s-note disclosure that USPS is a financial supporter of NPR provides a transparency element relevant to evaluating the reporting institution. The Brookings attribution functions in the analysis as a source of counterevidence to claims of widespread mail-voting fraud, although its reach among Republican respondents is not established.

The Supreme Court performed an arbiter function by issuing a binding ruling on the administration’s documented position. The litigation and reporting institutions performed mediator-like functions by bringing the dispute into public and legal processes, although the source does not characterize them in those terms. The GAO, USPS Office of Inspector General, and the pollsters function as witnesses by documenting performance, finances, rural effects, and public sentiment. USPS’s extraordinary-measures policy for accelerating ballot delivery before Election Day is an operational containment mechanism described in the article.

The available record does not identify a bridge-building initiative that brings partisan constituencies together around shared election-administration concerns, nor does it document a developed healing function addressing the longer-term partisan dispute over election legitimacy. The record also leaves unclear how postmark standards are communicated to voters and how conflicting national and jurisdiction-specific performance measures are reconciled. Threats of exit or violence are not documented in the supplied article.

Editor’s note and reporting-institution transparency

The article discloses that the U.S. Postal Service is a financial supporter of NPR. That disclosure is relevant to evaluating the article’s presentation of both USPS performance claims and the public-trust findings. The article’s headline foregrounds the 13-point decline in USPS confidence, while the same poll records increased confidence in state and local election administration; that framing choice places institutional weight on the USPS-specific finding. The source does not establish whether the poll’s timing after the Supreme Court ruling caused the decline, merely that the poll was released shortly afterward.

Methodological limits of the supplied record

The 3.8-point margin of error applies to the poll’s topline result; the supplied article does not provide subgroup-specific margins for the partisan figures. The article’s attribution of mail-voting fraud as “exceptionally rare” is dependent on the cited Brookings review, which is not independently supplied in the package. The claims concerning the recent GAO report, the September USPS Office of Inspector General findings, and the Sept. 14 Supreme Court ruling are retained as source-attributed claims because the article is the operative source, while independent verification is unavailable in the supplied material.

A shared evidentiary basis would require comparable national and jurisdiction-level measures of delivery timeliness, postmarking, ballot rejection, and voter confidence; the supplied record contains these measures only unevenly. The source does not provide enough information to determine whether partisan rhetoric, operational experience, or institutional salience explains more of the decline, and the question of how voters would change voting methods if USPS confidence declined is not addressed.

Additional considerations

State and local election officials have an operational interest in timely ballot processing and in the relationship between postmarks, receipt deadlines, and ballot acceptance. The source documents alternatives available to voters, including in-person voting, early in-person voting, and drop-box use, but does not provide detailed evidence about the capacity of those systems. USPS has a documented institutional interest in maintaining public confidence, brand reputation, revenue, and compliance with its universal-service mandate while pursuing a 10-year restructuring. Voters’ interests described in the analyses divide between ballot access and confidence in election integrity; the source does not establish how voters would change voting methods if USPS confidence declined further. The Cantwell–Hobbs release demonstrates the value of publishing jurisdiction-specific ballot-acceptance information alongside national USPS performance data, and the absence of equivalent jurisdiction-level information elsewhere in the supplied record limits the basis for symmetric national inference.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Principled Negotiation
Works a negotiation from interests, options, and objective criteria rather than positions.
Process Tracing
Reconstructs the step-by-step causal pathway of a specific historical event.
The Third Side
Takes the vantage of the surrounding community that has a stake in resolving a conflict (Ury).