Gaza's numbers, Gaza's sourcing problem

Three numbers crossed the wire at 05:44, 05:48, and 06:00 UTC on 2 July 2026, all carrying the same byline — the Government Media Office in Gaza, as relayed by Al-Alam Arabic's Telegram channel. Thirty-eight hospitals and ninety-six healthcare centres bombed, destroyed, or put out of service. More than a thousand children under one year of age killed, five hundred and twenty of them born and killed during the war. Nine thousand five hundred people missing, presumed under the rubble or fate unknown. The figures are staggering on their face. They are also, structurally, the kind of figures that Western newsrooms receive with one eyebrow raised — and then, too often, print anyway.

This publication has a duty to report Gaza's toll. It also has a duty to be honest about the chain of custody those numbers travel through before they reach a reader's screen.

The wire is not the story

The Government Media Office in Gaza operates under the Hamas-run administration that has governed the strip since 2007. It is a political body, not an independent statistical agency. Its casualty figures for civilians killed and wounded are widely cited — by Al Jazeera English, Middle East Eye, The Cradle, the wires — because there is no Western, Israeli, or UN body producing day-by-day counts that reliably cover the same ground. UNICEF and OCHA-Occupied Palestinian Territory report aggregate figures, not running tallies. The IDF issues its own running count, which is smaller in some categories and larger in others, and which uses a different methodology. WHO hospital-availability counts come from a third logic again.

The result is that the same events are being measured by at least three different rulers. The figure you cite depends on whose tape measure you trust.

When the Government Media Office says thirty-eight hospitals were bombed, destroyed, or put out of service, that is a real claim with a real evidentiary base: satellite imagery, on-the-ground videographers, WHO sit-reps, hospital directors who post their own outages. The number is not invented. But it is also a number the issuer has every institutional incentive to maximise, on the assumption — often correct — that Western editors will bury it on page ten if it is too low.

Reading the framing

The Telegram posts use the language of "war of extermination" and "occupation." That is the issuer's framing. Whether a Western wire republishes those words unfiltered is a separate editorial decision, and usually the answer is no — Reuters and the BBC will paraphrase, attributing the count to "Hamas-run authorities" or "the health ministry in Gaza." The framing shifts; the number travels.

That travel is the problem. The number is real and the framing is partisan. Both can be true. A reader who sees "38 hospitals bombed" in a Reuters graf and never sees who counted, and by what method, is being told the truth incompletely. A reader who sees "war of extermination" unfiltered is being told the framing without the source. Either error is an error.

There is a third option. Report the number. Name the issuer. Quote the framing. Add an UN- or WHO-cross-checked figure where one exists. Then let the reader weigh.

The structural context

Casualty reporting in active conflict is an information war as much as a kinetic one. Both sides know the first day's count frames every subsequent day's coverage. Hamas authorities issue hourly figures knowing full well that an Al Jazeera live blog will republish. The IDF issues daily figures knowing full well that Reuters will word them with an IDF caveat. Western wire editors sit between the two, fact-checking time they do not have against independent corroboration they cannot find.

This is not a Hamas problem or an IDF problem. It is a war-reporting problem, and Gaza is its most acute case. The structural failure is the absence of a neutral, in-theatre body with the access and the standing to produce a running count. WHO has access but is not a casualty registrar. OCHA has standing but is not a frontline journalist. The ICRC is bound by confidentiality. The UN Human Rights Office produces retrospective counts, not real-time ones.

Stakes

If the reading public learns to discount casualty numbers from the strip because the issuer has a politics, the cost is borne by the dead — Palestinian first, Israeli at sites such as the 7 October aftermath, but Palestinian overwhelmingly so in this war. That is the human stakes. The editorial stakes are not smaller: a media ecology that cannot distinguish between a partisan count, a sloppy count, and a forged count has lost the ability to do its job at the moment the job matters most.

What we do not know

The Telegram posts do not specify methodology, do not give a time window, do not break down civilian versus combatant for the missing-persons figure, and do not state how the under-one-year-old count was verified. A responsible desk treats the numbers as the highest credible floor, not the precise ceiling, and waits for cross-source verification before publishing any of them in a headline. This publication has read the same three posts as every other outlet on the wire; it has chosen to make the sourcing problem the story, because the numbers themselves deserve a better receipt than a Telegram forward.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/s/alalamarabic
• https://t.me/s/alalamarabic
• https://t.me/s/alalamarabic
• https://en.wikipedia.org/wiki/Government_Media_Office_(Gaza)