A standing study · updated monthly
Whose press actually gets heard?
We set national coverage of the same story side by side every day. Across 434 comparisons we have shown 2,436 sources from 120 countries — and the distribution is nothing like the world.
The whole of sub-Saharan Africa has supplied fewer sources to our comparisons than Mexico alone — 44 against 47.
That is not a claim about which press is worth reading. It is a measurement of which press a system actively looking for global coverage actually reaches — and some of the reason is us.
The shape of the distribution
Every country that has supplied at least ten sources. The tail is long and thin: 27 countries have appeared exactly once, and the median country has appeared 7 times in 434 comparisons.
Gini coefficient of attention: 0.70, where 0 would mean every country appears equally often and 1 would mean one country supplies everything.
By region
The same figures grouped. Regional labels are blunt instruments, but the gap they show is not subtle.
Is it just the stories we choose?
Largely, yes — and the rest we cannot honestly separate.
7 of our 434 comparisons were about sub-Saharan Africa at all — about 2% of them. If we rarely choose the story, we will rarely reach the press that covers it most closely.
The obvious follow-up is whether we also under-reach African outlets when the story is theirs. With 7 such stories we cannot answer that. Any figure we quoted from a sample that size would look like evidence and would not be.
So the honest account is: the dominant driver we can actually see is story selection, which is entirely ours. Beyond that sit causes we cannot cleanly separate — how much of the world’s journalism is indexed and reachable at all, how much sits behind paywalls or in formats machine discovery handles badly, and how our source selection behaves once a story is chosen. We are not going to guess at the split.
Why we publish this
A publication whose entire premise is comparing national coverage should be able to say how national its comparisons actually are. We would rather put the number in public than let readers assume it is better than it is.
One thing the figures do not show is straightforward Western dominance. Only 5 of 434 comparisons were sourced entirely from Western outlets, and 20% contained no Western source at all. The concentration is real, but it runs along different lines than the usual story about global media — toward whichever countries are party to the conflict of the moment, and away from everywhere that is not.
Method, and what this does not show
Every source shown in a Bulletin carries the country of the outlet that published it. This counts those attributions. It measures which national presses appear in our comparisons — not which countries have media, and not how good that media is.
- Window: 2026-05-11 to 2026-09-02. 2,436 sources across 434 comparisons. The window opens when we moved to comparing national coverage; before that the Bulletin compared outlets by political lean instead, and carried no country data.
- Not a measure of the world’s press. It is a measure of the press our process reached. Indexing coverage, paywalls, language and our own selection all sit between the two, and this study cannot fully separate them.
- Sources per comparison changed. Earlier editions carried more sources each than current ones, so raw monthly totals are not directly comparable; the shares here are computed across the whole window.
- Regional grouping is ours and, like all such groupings, arguable. Turkey and the North African states are counted with the Middle East, which is why the sub-Saharan figure is quoted separately.
This study is rebuilt automatically at the start of each month, as new comparisons are published — every figure above moves as the record grows. Last updated 2026-09-02.
Every comparison names the country behind every source.
Several times a day, we take one story moving through the global press and set out how different countries are reporting it.