Hegseth military personnel changes and Army secretary search
4 outlets covered the same story. Here's how their headlines landed on the −5 (Far Left) to +5 (Far Right) spectrum — a 1.5-point gap between the most-left and most-right framing.
How the coverage framed it
The bias spread here is modest at 1.5 points, but one headline does stand out from an otherwise largely neutral pack. The New York Times framed the story most dramatically, describing "The Battle to Save General Donahue From Hegseth's Army Purge" — the word "purge" carries strong negative connotations, while "Save" casts Donahue as a victim requiring rescue, positioning Hegseth as an aggressor. This stands in contrast to The Hill's multiple headlines on the same story, which used measured, procedural language: "faces competition in search for new Army secretary," "eyes Driscoll replacement," and "gave polygraph tests to high-ranking officials to find leakers." These phrases describe institutional actions without assigning moral weight. The shared neutral framing across most outlets treated the personnel changes as administrative processes, making the Times' charged vocabulary — particularly "purge" — the primary driver of the spread's left-leaning edge.
- -1.5NY Times · LeftThe Battle to Save General Donahue From Hegseth’s Army Purge
- ▸ 'Save General Donahue From Hegseth's Army Purge' — 'purge' is left-charged
- ▸ 'Save' frames Donahue as victim of Hegseth
- +0.0The Hill · CenterPentagon gave polygraph tests to high-ranking officials to find leakers: Reports
- ▸ neutral 'gave polygraph tests', factual 'to find leakers'
- ▸ no loaded adjectives or framing
- +0.0The Hill · CenterLive updates: Hegseth eyes Driscoll replacement, Missouri asks to restore GOP-friendly congressional map
- ▸ neutral 'eyes Driscoll replacement', factual live update framing
- ▸ no ideological loading
- +0.0The Hill · CenterHegseth’s pick faces competition in search for new Army secretary
- ▸ neutral 'faces competition', factual search reporting
- ▸ no loaded adjectives detected
Scores are AI estimates of headline language, not factual ratings. Each headline is scored 0 (neutral) outward to Far Left / Far Right by Claude and Grok.