National daily
Where does Le Figaro sit?
On the FrIdéo scale, Le Figaro scores +1.02, which places it in the "right" band and 26th of 30. The 95% confidence interval runs from +0.50 to +1.55. That interval excludes zero: placing this outlet on its side of the panel is supported by the data, not just by its point position. 7 of the 9 evidence families cover it directly.
Among the best-covered outlets in the panel, with seven of nine families reporting directly, and its interval excludes zero. Statistically indistinguishable from CNews.
- +1.02
- FrIdéo score
- +0.50 … +1.55
- 95% confidence interval
- right
- Band
- 26 / 30
- Rank
What each source of evidence says
FrIdéo does not measure ideology once. It fuses nine independent families of evidence, each standardised across the same thirty-outlet panel. A blank row means that family reports nothing for this outlet, which is not the same as reporting zero.
| Evidence family | Position |
|---|---|
| D0Ownership ideology | +0.99 |
| D3Hyperlink and follower network position | +0.21 |
| D2Survey perceived ideology | +1.13 |
| D6aStated editorial orientation | no data |
| D6bThird-party encyclopedic profiling | +0.98 |
| D1Broadcast guest ideological bias | no data |
| D4Media Bias Fact Check rating | +0.84 |
| D8Ad Fontes Media bias rating | +0.84 |
| D7Distancing-aware lexical score | +0.82 |
What placed it there
The signed contribution of each evidence group to the final score. These sum, up to rounding, to the score above.
- Ownership+0.14
- Academic and network evidence+0.19
- Third-party ratings+0.24
- Editorial content+0.26
- Founding-tradition prior+0.21
Ownership
- Final owner
- Dassault family
Ownership attributions follow the FrIdéo ownership family, which draws on the Acrimed / Le Monde diplomatique "Médias français : qui possède quoi ?" mapping distributed as a community dataset on data.gouv.fr, cross-checked against Cagé (2017), with press coverage used only for changes of control after 2023.
Nearest outlets on the scale
The outlets immediately either side of this one. Where the intervals overlap, the ordering between them is not something this scale can resolve.
All thirty outlets →How the score is built
Nine evidence families are standardised across this panel, weighted by how much independent information each carries, and fused into one continuous position, with outlets covered by few families shrunk toward a founding-tradition prior. The scale is zero-mean and unit-variance by construction across these thirty outlets, so a score is a statement about where an outlet sits relative to this panel and not an absolute ideological coordinate.
How to read this responsibly
- Zero on this scale is the mean of these thirty outlets, not a neutral point in French politics. An outlet at zero is average for this panel, which is a different claim.
- Scores are comparable within this panel only. They cannot be compared against a scale built from a different set of outlets without re-standardising.
Common questions
Is Le Figaro left wing or right wing?
On the FrIdéo scale, Le Figaro scores +1.02, placing it in the "right" band and 26th of 30 French news outlets. Its 95% confidence interval (+0.50 to +1.55) excludes zero, so placing it on that side of the panel is supported by the data.
Who owns Le Figaro?
The final owner used by the model is: Dassault family. The attribution follows the Acrimed / Le Monde diplomatique "Médias français : qui possède quoi ?" mapping, distributed as a community dataset on data.gouv.fr and cross-checked against Julia Cagé (2017).
How is Le Figaro's orientation measured?
FrIdéo fuses nine independent families of evidence, among them ownership, position in the hyperlink network, survey-perceived orientation, editorial charters, two rating panels and a lexical score computed from the text. 7 of those nine cover Le Figaro directly. Each family is standardised across the same thirty-outlet panel, then the positions are fused into a single continuous score.
Is Le Figaro further right than Atlantico?
Atlantico is the outlet immediately to the right of Le Figaro on the scale. The two positions are close and their confidence intervals overlap substantially: the ordering between these two is not something this scale can resolve. Comparing outlets far apart on the scale is reliable; separating two neighbours is not.
How to cite
Cite the paper. The scale, the method and this outlet’s position all come from it, and a citation to a web page is not something a reviewer can check or a bibliography can resolve.
@inproceedings{sobhy2026frideo,
title = {{FrIdéo}: French News Outlet Ideology Scoring via Multi-Source
Evidence Fusion and Distancing-Aware Lexical Scoring},
author = {Sobhy, Amr},
booktitle = {Proceedings of ICNLSP 2026},
year = {2026}
}If you are citing the numbers as data rather than the method, cite the released dataset as well. It carries the scores, intervals and per-family evidence for all thirty outlets under CC BY 4.0.
Scores from FrIdéo v1.0