The Transparency Test: What Six Agencies' 2026 FOIA Records Actually Show
The Freedom of Information Act sets a clear baseline: agencies should generally respond to requests within 20 working days. The reality is more complicated. Requests vary enormously in scope, agencies use different tracking systems, and a single median can hide thousands of difficult cases. But the records agencies publish still offer one of the most practical ways to judge whether a transparency claim is supported by actual performance.
This scorecard examines six major federal agencies in 2026: the Department of Justice, the Department of Homeland Security, the Department of Health and Human Services, the Environmental Protection Agency, the Department of State, and the Department of Veterans Affairs. The figures below are rounded and drawn from agency FOIA dashboards, annual reports covering fiscal 2025, and quarterly data through September 30, 2026. They are not a perfect ranking. They are a starting point for readers who want to test the rhetoric.
The 2026 Snapshot
| Agency | Median simple-track days | Backlog | Full release | Partial release | Full denial | No records |
|---|---|---|---|---|---|---|
| DOJ | 34 | 12,400 | 22% | 41% | 18% | 19% |
| DHS | 48 | 31,900 | 17% | 39% | 21% | 23% |
| HHS | 29 | 9,800 | 28% | 35% | 14% | 23% |
| EPA | 41 | 6,700 | 25% | 44% | 16% | 15% |
| State | 96 | 21,500 | 11% | 46% | 24% | 19% |
| VA | 22 | 15,300 | 31% | 37% | 12% | 20% |
The first column is the median number of working days for requests placed on the simple track. The backlog column shows pending requests at the end of the period. Release and denial figures are shares of processed requests, rounded to the nearest percentage point. 'No records' means the agency said it could not locate responsive records; it is not the same as a denial, though the distinction matters.
What the Numbers Can Tell You
The most obvious measure is the gap between the statutory 20-day standard and the median simple-track time. Every agency in this snapshot exceeds 20 days. The Department of Veterans Affairs comes closest at 22 days. The Department of State is far off the pace at 96 days, more than four times the statutory target. That does not automatically mean State is acting in bad faith. It does mean that a request sent to State should be expected to take months, not weeks.
Backlog size tells a different story. DHS has the largest backlog in this group, at roughly 31,900 requests. State has fewer pending requests but a much slower median. A large backlog can reflect a surge in requests, staffing shortages, or both. A smaller backlog with a long median can indicate that the agency is triaging aggressively, moving easy requests quickly while hard ones sit.
Release rates are useful only when read together. HHS reports the highest full-release rate in this group at 28%, along with a relatively low full-denial rate of 14%. VA reports the highest full-release rate at 31% and the lowest full-denial rate at 12%, but its backlog remains substantial. State has the lowest full-release rate at 11% and the highest full-denial rate at 24%. That combination deserves scrutiny, though it may also reflect the volume of classified and diplomatic material State handles.
Trends matter more than a single snapshot. DHS's backlog has grown compared with the prior year, while EPA's has declined modestly. State's median has improved from earlier in the decade but remains far above the statutory standard. VA's median is comparatively strong, but its pending caseload suggests that speed on simple requests has not solved the larger queue.
What the Numbers Cannot Tell You
FOIA statistics are not an x-ray. They are a self-reported summary, and they can mislead in several ways.
First, median times hide outliers. An agency can process most simple requests quickly while a small number of complex requests take years. The median will look respectable even when the hardest cases are stuck.
Second, agencies define tracks differently. A 'simple' request at one agency may be broader than a 'complex' request at another. Backlog counts may include requests that are awaiting clarification from the requester, referrals to other agencies, or consultations that are outside the agency's control.
Third, denial rates do not explain why a denial happened. A full denial may rest on a legitimate exemption, or it may rest on an overly broad reading of one. A 'no records' response may be accurate, or it may reflect an inadequate search. The aggregate number cannot distinguish between those possibilities.
Fourth, release rates can be shaped by request patterns. An agency that receives many narrow, routine requests may post high release rates without being especially transparent on contested matters. An agency that receives broad requests on sensitive topics may post lower release rates even when it is complying with the law.
Finally, the numbers do not measure whether the public got useful information. A requester can receive a stack of heavily redacted pages and still count as a partial release. A timely response can be technically compliant and substantively empty.
A Practical Scorecard for Readers
A transparency claim should be tested with a few concrete questions.
- Compare the median simple-track time with the 20-day standard, but do not stop there. Ask how the agency defines 'simple' and how many requests fall outside that track.
- Look at the three-year backlog trend. A single year can be distorted by a surge or a staffing change. A consistent rise is harder to explain away.
- Read the denial reasons, not just the denial rate. Agencies publish exemption codes. If one exemption accounts for most denials, that is worth knowing.
- Check the 'no records' rate alongside the wording of the request. A high no-records rate may signal a search problem, or it may signal that requesters are asking for records that do not exist.
- Search the Office of Government Information Services and federal court dockets for disputes involving the agency. Litigation is expensive, but it is also a public signal of where the administrative process broke down.
- Use the agency's own reading room and proactive disclosure pages before filing a request. If the material is already posted, the response time is immediate.
- File a narrow test request. A request for a specific contract, memo, or date range is easier to evaluate than a broad demand for 'all documents' on a topic.
The Bottom Line
No single number settles the transparency question. DHS's backlog is a warning sign. State's median is a practical caution. VA's simple-track speed is a genuine strength, but its pending caseload shows the limits of that measure. HHS and EPA look stronger on several indicators, though neither is exempt from the problem of redactions and slow complex cases.
The useful lesson for 2026 is that transparency is not a press release. It is a set of records about records. Readers who want to judge an agency's claims should look at the trend, the backlog, the denial reasons, and the paper trail. The scorecard does not declare a winner. It tells you where to look next.
