History of the Shanghai Ranking
The Academic Ranking of World Universities (ARWU) — formally the Academic Ranking of World Universities — was first published in 2003 by Shanghai Jiao Tong University. Its original purpose was not to produce a consumer product for students, but to benchmark Chinese universities against global peers as part of China's effort to build world-class institutions. It was the first systematic global university ranking, and its publication triggered the entire modern rankings industry.
In 2009, the ranking separated from Shanghai Jiao Tong University and is now published by the independent Shanghai Ranking Consultancy. It is updated annually, typically released in August, and now covers 1,000 institutions with a top 500 published in full. Subject rankings (ShanghaiRanking's Global Ranking of Academic Subjects) cover 55 academic disciplines separately.
ARWU is distinctive among the major rankings for one critical reason: it uses no surveys. Every indicator is derived from verifiable, objective data from public records and bibliometric databases. This makes ARWU both the most immune to gaming through reputation campaigns and the most narrow in what it measures — essentially, the rankings are a proxy for research prestige at the very highest level.
Alumni Winning Nobel Prizes and Fields Medals (10%)
The Alumni indicator (weight: 10%) counts the total number of alumni who have won Nobel Prizes in Physics, Chemistry, Medicine, Economics, and Peace, plus Fields Medals in Mathematics. "Alumni" is defined as those who received their bachelor's, master's, or doctoral degree from the institution.
Nobel Prize records extend back to 1901, which means Harvard, Columbia, Chicago, and other long-established American and European universities have accumulated decades of alumni laureates that newer institutions — regardless of their current quality — cannot match. The indicator is temporally weighted, giving more weight to recent laureates, but the cumulative advantage of old elite universities is substantial. An institution founded in 1980 could be producing world-class research today and still score near zero on this indicator.
Staff Winning Nobel Prizes and Fields Medals (20%)
The Staff indicator (weight: 20%) is ARWU's single largest component. It counts Nobel Prizes and Fields Medals won by people who were, at the time of the award, working at the institution as academic staff. This is temporally weighted and captures current, not historical, faculty excellence.
Because recent laureates carry more weight, universities that actively recruit star scientists — including potential Nobel candidates — can improve their scores. However, Nobel Prizes are typically awarded 20–30 years after the seminal research, meaning a laureate at Institution A today may have done the prize-winning work at Institution B decades ago. Attribution therefore depends as much on where a scientist ended their career as where they did their most important work.
Research Output at the Nobel level reflects decades of sustained investment, national research infrastructure, and historical accident as much as current institutional quality.
Highly Cited Researchers (20%)
ARWU's Highly Cited Researchers (HiCi) indicator (weight: 20%) uses Clarivate Analytics' annual list of Highly Cited Researchers — scientists and social scientists who rank in the top 1% by citations in their field over a given decade. Each institution receives a score based on the number of HiCi researchers on its staff.
This indicator is more current than Nobel Prize counts and more sensitive to institutional momentum. A university that strategically recruits prolific, highly cited scholars in key fields can improve its ARWU score within a few years. The indicator also slightly favours large institutions (more faculty = more chances to include HiCi researchers), though ARWU applies a per-capita correction at the end of the calculation.
The H-Index of individual researchers indirectly feeds into this indicator, as the selection of Highly Cited Researchers by Clarivate is based on sustained citation impact across a researcher's career — essentially a career-level h-index.
Papers in Nature and Science (20%)
ARWU's Nature and Science (N&S) indicator (weight: 20%) counts papers published in the journals Nature and Science over a rolling five-year window. These two journals are among the most selective in academic publishing, with acceptance rates below 10% and global readership across all scientific disciplines.
The N&S indicator is deliberately narrow: it covers only two journals, even though thousands of outstanding research papers appear in other high-impact venues. This concentrates scoring advantage on institutions with strong traditions in multidisciplinary natural sciences — biology, chemistry, physics, materials science — and disadvantages institutions where excellence is concentrated in mathematics, social sciences, humanities, or professional fields where publication in Nature and Science is unusual.
Citation Impact from these publications contributes indirectly, since Nature and Science papers are typically among the most-cited in their fields.
Papers in SCIE/SSCI (20%)
The Publications indicator (weight: 20%) counts papers indexed in the Science Citation Index Expanded (SCIE) and Social Sciences Citation Index (SSCI), both components of Clarivate's Web of Science platform. This captures total research output across all indexed disciplines over a rolling period, not just the prestigious two journals counted in N&S.
This indicator rewards volume of publication alongside quality (since only peer-reviewed journals are indexed). Large, comprehensive research universities — particularly those with large medical schools and life science faculties — benefit from the sheer volume of indexed output they produce annually.
Per Capita Performance (10%)
The Per Capita (PCP) indicator (weight: 10%) divides each institution's weighted score across the five objective indicators by the number of full-time equivalent academic staff. This normalisation is ARWU's attempt to prevent extremely large universities from dominating purely through volume.
Per-capita normalisation significantly reshuffles the ranking relative to the raw totals. Small specialist institutes — such as the École Normale Supérieure in Paris or Rockefeller University in New York — that might rank in the 200s on raw indicators can rise into the top 50 when their output is divided by a very small faculty. For students, this is a meaningful signal: a high per-capita score indicates that almost every member of faculty is operating at elite research levels, suggesting a highly demanding and intellectually charged environment.
Why ARWU Favors Large Research Universities
Despite the per-capita adjustment, ARWU systematically advantages large, comprehensive, English-language research universities for several interconnected reasons:
- Time horizon: Nobel Prize indicators reward institutions with 100+ years of history accumulating laureates. KAIST, founded in 1971, cannot compete with Harvard (founded 1636) on this indicator regardless of current quality.
- Discipline bias: SCIE/SSCI indexing and Nature/Science publication norms heavily favour natural sciences over arts, humanities, and social sciences. A university specialising in law, fine arts, or philosophy will score near zero on all research indicators despite potentially being world-leading in its domains.
- Language bias: Web of Science and Nature/Science both have strong English-language biases. Excellent research published in Chinese, German, French, or Arabic journals may be underrepresented or absent from the citation databases ARWU relies upon.
- National funding differences: Institutions in countries with lower research funding levels — regardless of how efficiently they deploy that funding — are structurally disadvantaged against institutions in countries that invest heavily in basic research.
For students, ARWU is most useful as a measure of a university's standing in the global research community, particularly in natural sciences and medicine. It should be used alongside other rankings for a complete picture, especially if your intended field is not a traditional STEM discipline.