Yale School of Management professor Edieal Pinker published the analysis in Studies in Conflict & Terrorism. He examined articles that appeared from October 7, 2023, through June 7, 2024, and concluded that the dominant frame assigned “near-exclusive agency and responsibility to Israel while diminishing Hamas’s role in sustaining the conflict.”
Ninety-one percent of the articles did not mention Israeli casualties after October 7. A total of 1,423 pieces omitted both those later Israeli losses and deaths among Hamas combatants. Half the articles made no reference to Israeli hostages held by Hamas, and 41 percent did not mention the Israeli dead from the initial October 7 attack.
Personal accounts of Palestinian and Lebanese suffering appeared on average two out of every three days. Comparable stories of Israeli suffering not tied directly to October 7 were rare.
Across the dataset, the word “Israel” appeared 27,205 times and “Hamas” 8,499 times—more than a three-to-one gap that, the study argued, placed responsibility almost entirely on Israel.
Pinker described a recurring narrative: Hamas’s October 7 attack, followed by Israeli military operations and rising Palestinian casualties, with little space for Israeli military losses, Hamas fighter deaths, or later Palestinian attacks. Israel, the paper found, was often presented as the sole aggressor that “seemingly bears no costs from conducting the war except for a loss of international support,” while Palestinians appeared as passive victims.
The study used ProQuest archives of Times articles together with casualty and incident data from Israeli government sources and the Armed Conflict Location & Event Data project.
During the period after October 7 that Pinker examined, those sources recorded hundreds of Israeli soldiers killed, additional civilian deaths inside Israel and the West Bank, and repeated attacks. Most Times articles in the sample did not report those figures or distinguish Hamas combatants from civilians among Palestinian dead. An earlier, non-peer-reviewed version of the paper circulated in 2025; the journal version, accepted in 2026, is described by the author as substantially the same analysis with formal review.
The findings concern frequency of mentions, omissions, and narrative patterns in one newspaper’s output over eight months, not a claim about every reporter or every story.

