Congress Ordered an Annual Accounting of Terrorism. The Counting Has Stopped.

On May 1, 2026, the State Department released the 2024 edition of the Country Reports on Terrorism — a full year past the deadline in effect when the report came due, and months past even the more forgiving date Congress has since written into law. No Annex of Statistical Information, the report’s statistical backbone in years past, has been published with it. And the report’s own pages cite terrorism-incident rankings from the Global Terrorism Trends and Analysis Center, or GTTAC — the same data program whose funding the department allowed to lapse in March.

For two decades, professionals met that statistical requirement: the National Counterterrorism Center from 2004 to 2011, the University of Maryland’s Global Terrorism Database from 2012 to 2017, and GTTAC’s Record of Incident Database — GRID — from 2018 on. GRID was purpose-built for the statutory task. It applied the government’s own definition of terrorism, named its sources, published its codebook, and put its methodology through peer review this year in the journal Democracy and Security.

Note the statutory asymmetry. The report is mandated unconditionally; only the statistics carry the qualifier “to the extent practicable.” Defund the sole mechanism that made the statistics practicable, and the qualifier becomes an exit. The 2024 report’s own compliance chapter makes the gap explicit: it states that the statistical requirement “is satisfied through the inclusion of a statistical annex to the report” — an annex that has not been published. The 2025 edition will be worse still, because the incidents of 2025 are not being systematically collected by anyone charged with the task.

The reflex answer — use the alternative datasets — does not survive contact with the data. I spent my doctorate inside those datasets, and their custodians have been admirably candid about the limits. By its own founding director’s account, the Global Terrorism Database’s source-validity rule excluded 28 percent of otherwise-eligible cases worldwide in an early-2014 sample — and 76 percent in Syria, precisely where the fighting was.

The Armed Conflict Location & Event Data project, the other presumed substitute, measures political violence broadly and does some things well; it was never designed to be the statutory instrument. My dissertation documented what that costs. ACLED codes a suicide bombing only as a stand-alone event, never as part of a battle. Re-coding its Iraq narratives for 2016–2018 by hand, I found 860 suicide bombings where ACLED had coded 344 — and my recount landed within 1 percent of the Global Terrorism Database’s independent tally, so the missing events were real. The same re-coding captured 48 percent more incidents than ACLED had coded in Afghanistan, and 111 percent more in Nigeria. ACLED’s data are not without merit, as I wrote then. But extracting reliable answers from them requires heavy engineering that most consumers — congressional staff included — will never perform.

Methodology is not a technicality. Across 2018–2020, the years both instruments covered, GRID recorded 30 percent more incidents than the Global Terrorism Database and nearly twice the fatalities — and the two disagreed about whether global terrorism was rising or falling. The choice of instrument decides the trend line the government sees. Choose none, and there is no trend line at all.

There is precedent for the way back. When the Global Terrorism Database lost its State Department funding in 2018, allied governments — first Germany’s foreign office, then Britain’s — stepped in to keep collection alive, because the capability was too valuable to lose. GRID’s case is stronger: the mandate is still on the books; the archive — public once again, and current through mid-May — and the trained pipeline still exist; and the gap to close is measured in weeks, not years. The cost of restarting collection is a rounding error against what the government spends to fight the groups it is no longer counting. The combatant commands consume this data. So do the insurers and reinsurers who price political-violence risk. So do the scholars and students who will train the next generation of analysts.

The 2024 report proves the mandate outlived the capability. The question Congress should be asking — and that any serious funder, public or private, should be answering — is who restores it. The threats were never in doubt. What remains in doubt is whether we intend to see them.

David E. Firester, Ph.D., is a former U.S. Army intelligence analyst and the founder and CEO of TRAC Intelligence, LLC. He is the author of “Failure to Adapt: How Strategic Blindness Undermines Intelligence, Warfare, and Perception” (2025).

image: United States Airforce B1 bombers are seen at RAF Fairford, a US Airbase, after an incident in Fairford, England, Sunday, Sept. 27, 2026. (AP Photo/Alastair Grant) - Five men are suspected of preparing a terrorist act near an English air base the US uses in its war against Iran.