Is the ‘world’s top 2% scientists’ list measuring merit or geography?

You have seen the badge. Once or twice a year your LinkedIn feed fills with the same post: someone you know, usually a professor, announces they made the “World’s Top 2% Scientists” list. There is often a certificate. Sometimes a screenshot of a spreadsheet with one row highlighted in yellow. The comments say congratulations, and they mean it.

The list is real, and it is not junk. It comes from John Ioannidis and colleagues at Stanford, who built a set of science-wide author databases of standardized citation indicators out of Elsevier’s Scopus. The logic is reasonable. Raw citation counts are unfair to compare across fields, so they normalize them, fold six measures (total citations, three author-position citation counts, the h-index, and a co-authorship-adjusted h-index) into one composite, and publish the top slice inside each field. That slice is what everyone now calls the top 2%. It comes in two flavors, a single-year snapshot and a career-long one.

Before you print the certificate: in at least one field, the list tracks a US postal code about as closely as it tracks merit.

What the paper actually checked

A 2025 paper in the Journal of Informetrics took the 2024 edition of the list, picked one field, and counted. Péter Sasvári and Gergely Ferenc Lendvai, both at Ludovika University of Public Service in Hungary, chose legal studies, partly because “Law” is one of the fastest-growing fields on the list: since the first edition it has grown 811% in the career-long ranking and 491% in the single-year one. They pulled every law scholar from both rankings, 766 people in all (393 in the single-year list, 373 in the career-long), then scraped roughly 50,000 of their publications from Scopus to see who was citing whom.

The United States is heavily overrepresented. In the career-long ranking, 213 of 373 legal scholars were based in the US, about 57%. In the single-year ranking it was just under half, around 46%. The same person tops both lists: Cass R. Sunstein, whom the authors call “leagues above his peers.” The best-ranked non-US scholar, David Weisburd, sits far behind. In the single-year data Sunstein’s rank is 242 and Weisburd’s is 4,787. Across the board the US median rank beats the non-US one (131,187 against 163,484 in the single-year set, where a lower number is better).

Where they publish is most of the story

The gap is not really about who writes better law. It is about where the writing lands. The top 10 journals for US scholars are, all ten of them, American law reviews, and not one is published by a “Big 5” commercial house (Elsevier, Sage, Wiley, Taylor and Francis, Springer). They are the in-house reviews of elite law schools: Harvard, Chicago, Yale, Stanford, Penn. The Harvard Law Review is the single most popular venue, and articles there pull a mean of about 101 citations each.

Non-US scholars publish somewhere else entirely. Their top journals are European and international ones, with far thinner citation means. Their most common venue by article count is actually Nature, but those articles average just 5.39 citations. Not a single Global South journal appears in any of the top-journal tables. And when Sasvári and Lendvai mapped the citation network, US scholars mostly cited other US scholars, a dense, inward-looking web. The authors’ description of the ideal candidate for the list in law is worth quoting in full: “one has to fit three criteria: publish frequently in Ivy League universities’ journals, cite U.S. scholars, and it definitely does not hurt to be American either.”

A ranking that measures the neighborhood

Put the pieces together and, in this field, the ranking is not really measuring the scientist. It is measuring the neighborhood the scientist publishes in. Citations pool where the papers already cluster, the composite standardizes that pool, and the ranking hands the trophy back to where it started. Call it the Matthew effect with a passport: citations accrue to those who already have them, and here “those” tend to carry a US affiliation. The authors describe an American “citation ecosystem that reinforces its central position,” frequently at the expense of Global South scholarship, which stays on the periphery of both the publishing and the citing.

What this is, and what it is not

I want to be careful, because it’s easy to over-read one paper. This is a single field, and law may be the strangest field you could pick for citation counting. A large share of US legal scholarship lives in student-edited law reviews that behave nothing like a physics or a medical journal, many of them without clean metadata for the databases to even read. Law is also split by legal system, common law against continental, so a scholar’s work is often local by nature. Sasvári and Lendvai point out that researchers working in Islamic, customary, or Jewish law are “visibly non-existent” on the list, and that Scopus itself under-indexes non-English work. None of this is claimed to generalize. I wouldn’t assume the same US share turns up in molecular biology or applied mathematics.

The authors are also not telling you to burn the list. Their words: “Completely disregarding ranking such as this one is, of course, not necessary.” What they are saying is narrower, and more useful. In legal studies the list carries a loud geographic accent, so it should not be read as the measure of who the best legal scholars are.

So what do you do with the badge?

Use it for less than the certificate invites. If you’re on it, enjoy it. It’s a fair sign your work gets read and cited, which isn’t nothing. If you’re ranking other people with it, for hiring, promotion, or a tenure file, remember that in at least one field it rewards a postal code about as much as a mind. “This person made the top 2%” honestly means “this person publishes in well-cited venues.” That is worth something. It is not the same sentence as “this person is among the best scientists alive,” even though the yellow highlight invites you to read it that way.

And if the world’s top 2% is this sensitive to where you happen to publish, what is it we’re actually ranking?