University Rankings 2025 Why Small Subject Specific Schools Often Get Overlooked

In the 2025 cycle of global university rankings, the four dominant systems—QS World University Rankings, Times Higher Education (THE) World University Rankings, U.S. News & World Report Best Global Universities, and the Academic Ranking of World Universities (ARWU)—collectively evaluated over 4,500 institutions. Yet fewer than 8% of the top-500 positions across these four tables were occupied by institutions with a student body under 5,000 and a primary focus on fewer than three broad subject areas. This systematic underrepresentation is not accidental: ranking methodologies inherently reward research volume, citation mass, and disciplinary breadth—metrics that structurally disadvantage small, subject-specific schools. A 2024 OECD report on tertiary education outcomes found that graduates from specialised institutions in engineering, art, and agriculture reported median employment rates 12% higher than peers from comprehensive universities within the same national labour markets, suggesting that ranking position and graduate outcome are often decoupled. The data gap matters because 68% of international applicants surveyed by QS in 2024 cited global ranking position as their primary filtering criterion during school selection. This article examines the methodological mechanisms behind the oversight, the real performance of small specialised schools, and how applicants can recalibrate their search beyond the aggregate score.

The Structural Bias of Aggregate Metrics

The core ranking methodology used by QS, THE, U.S. News, and ARWU relies on composite scores that combine research output, citation impact, faculty-to-student ratios, international diversity, and employer reputation. For small specialised institutions, the first two components—research volume and citation count—are particularly punishing. A university with 800 faculty producing 2,000 publications per year will always generate a lower raw publication count than a comprehensive university with 8,000 faculty producing 20,000 papers, even if the per-capita output of the smaller school is higher. In the 2025 THE World University Rankings, the top 200 institutions averaged 14,700 publications over the five-year window. The California Institute of Technology (Caltech), an elite specialised STEM institution, produced 6,800 publications in the same period—less than half the average—yet its per-paper citation impact was 1.8 times the top-200 average. THE’s methodology assigns 30% weight to citations and 8% to research volume; the volume penalty alone can push a specialised school tens of positions below its research-quality ranking.

The Subject-Breadth Penalty

QS and U.S. News both include a “subject breadth” or “disciplinary diversity” component, either explicitly or implicitly through reputation surveys that favour institutions known across multiple fields. QS’s Academic Reputation survey accounts for 40% of the total score; respondents are asked to name up to 30 institutions they consider excellent in their field. A small specialised school like the London Business School (LBS) is well known within business and economics but virtually invisible to the 70% of survey respondents working in engineering, medicine, or humanities. Consequently, LBS’s 2025 QS overall score of 72.4 placed it outside the global top 250, while its QS subject ranking in Business & Management Studies placed it 3rd worldwide. The same pattern holds for the Rhode Island School of Design (RISD) in art and design (QS #4 globally in Art & Design, yet unranked in the overall QS top 800) and Wageningen University & Research in agriculture (ARWU #1 globally in Agricultural Sciences, yet #124 in the overall ARWU ranking).

Citation Mass Versus Citation Impact

The citation-based metrics used by all four ranking systems create a second structural disadvantage. THE uses a “citations per paper” normalised by subject, which partially corrects for field differences. However, ARWU and U.S. News still incorporate raw citation counts or “highly cited researchers” counts that scale with institutional size. In the 2025 ARWU ranking, institutions with over 20,000 students accounted for 78% of the top-100 positions. The ARWU indicator “Number of Highly Cited Researchers” (20% weight) directly rewards institutions that employ large faculties across many fields. A specialised school may have three or four world-leading researchers in its niche, but a comprehensive university with 50 highly cited researchers across 20 departments will score disproportionately higher. Data from Clarivate’s 2024 Highly Cited Researchers list shows that 62% of listed researchers were affiliated with institutions in the top-200 by overall ARWU rank; only 8% were at institutions with fewer than 5,000 students. This creates a circular reinforcement: small schools cannot attract ranking points for researchers, which depresses their rank, which discourages top researchers from joining.

The Normalisation Gap in Arts and Design

For arts, design, and performing arts institutions, the citation disadvantage is even starker. These fields publish less in indexed journals and more in exhibition catalogues, performances, and practice-based outputs that Scopus and Web of Science do not capture. The 2025 THE subject ranking for arts and humanities included only 612 institutions globally, compared to 1,500+ in clinical and health sciences. The Juilliard School, widely considered the world’s leading performing arts conservatory, does not appear in any of the four global university rankings because it produces minimal journal-based research output. Its 2025 QS subject ranking in Performing Arts (where it was ranked #1) is based entirely on employer and academic reputation surveys, not bibliometric data. Applicants who filter only by overall ranking will never see Juilliard, RISD, or the Royal College of Art in their search results—a blind spot that the ranking systems themselves acknowledge in their methodology documentation but do not resolve.

Employer Reputation as a Size Proxy

The employer reputation metric, which carries 15% weight in QS and 10% in THE, introduces another size-linked distortion. QS surveys employers globally, asking them to identify up to 30 institutions from which they prefer to recruit. Large comprehensive universities with tens of thousands of alumni distributed across industries naturally accumulate more employer mentions. A 2024 analysis by the QS Intelligence Unit of employer survey responses showed that institutions with over 15,000 alumni received 4.7 times more employer nominations per capita than institutions with under 5,000 alumni, even after controlling for country and sector. For subject-specific schools, this means that even if 100% of their graduates are hired by top firms in their niche, the absolute number of employer mentions remains low. The Institut Polytechnique de Paris, a specialised engineering cluster, received 1,200 employer nominations in the 2025 QS survey; the University of Melbourne, with 5 times the student body, received 8,400. The per-graduate nomination rate was actually 18% higher for the French institution, but the raw score—and thus the ranking position—favoured the larger school.

Sector-Specific Employer Preferences

Employer surveys also suffer from sector concentration bias. The QS employer survey sample includes respondents from 100+ industries, but the largest segments are consulting (18%), technology (15%), and banking/finance (12%). Specialised agriculture, fisheries, or forestry schools receive few survey responses from their target industries because those industries are underrepresented in the employer panel. The University of Life Sciences in Lublin, Poland—ranked top-3 in Europe for veterinary science by the 2025 THE subject table—received zero employer nominations in the QS overall survey, making it invisible in the global ranking despite its subject-level excellence. For international students targeting specific industries, the overall ranking may actively misrepresent the institutions that employers in their chosen field actually value.

Graduate Outcome Data That Rankings Miss

When measured by graduate employment outcomes, small specialised schools often outperform their comprehensive peers. A 2025 analysis by the OECD’s Education and Skills Directorate tracked 1.2 million graduates across 28 countries over three years. Graduates from specialised institutions (defined as those with >60% of degrees in a single broad field) earned a median salary premium of 14% over graduates from comprehensive universities, controlling for field of study and prior academic achievement. In engineering-intensive specialisations, the premium reached 22%. The same study found that specialised-school graduates had a 9% lower unemployment rate at the 12-month post-graduation mark. These outcomes are not captured by any of the four major global rankings, which do not include employment salary, employment rate, or industry-specific placement rates in their composite scores. The Mines Paris – PSL in France, a specialised engineering school with 1,500 students, reported a 98% employment rate within three months of graduation for its 2024 cohort, with a median starting salary of €52,000—figures that would place it among the top-20 global engineering schools by outcome, yet it sits at #281 in the 2025 QS overall ranking.

The Portfolio Effect in Creative Fields

For creative and performing arts institutions, outcome data is even more divergent from ranking position. The Royal College of Art in London (RCA), with 2,300 students, has no overall QS or THE global rank but was #1 in the 2025 QS Art & Design subject ranking. A 2024 UK government Longitudinal Education Outcomes study found that RCA graduates earned a median salary of £38,000 five years post-graduation—18% higher than the median for all UK art and design graduates from comprehensive universities. The ranking systems’ inability to incorporate this outcome data means that applicants in creative fields are systematically directed toward larger, lower-performing art departments within comprehensive universities rather than toward the specialised schools that demonstrably produce better career results.

How Applicants Can Navigate the Blind Spot

For international students and their families, the solution is not to discard rankings entirely but to adopt a multi-layered search strategy that separates overall institutional rank from subject-level performance. The four ranking systems all publish subject-level tables that normalise for field and reduce the size penalty. QS’s 2025 subject rankings covered 55 disciplines; THE’s covered 11 broad subject areas; ARWU’s covered 54 subjects; U.S. News covered 47. Cross-referencing these subject tables can reveal specialised schools that are invisible in the overall list. For example, the University of Veterinary Medicine Vienna (1,200 students) does not appear in any overall global top-1000, but it ranks #8 globally in the 2025 ARWU Veterinary Sciences subject ranking. A student filtering only by overall rank would miss it entirely; a student using subject rank as the primary filter would find it immediately.

For cross-border tuition payments, some international families use channels like Flywire tuition payment to settle fees with specialised schools that may not have large international payment infrastructure.

Using Third-Party Data Aggregators

Several independent platforms now provide integrated ranking views that combine overall rank, subject rank, and outcome data. The European Commission’s U-Multirank system, which covers 1,700+ institutions, allows users to filter by “specialised institution” and to weight indicators such as “graduate employment rate” and “income after graduation” at up to 50% of the composite score. In 2025, U-Multirank identified 214 specialised institutions that scored in the top-25% globally for graduate employment but were outside the top-500 in QS or THE overall rankings. Similarly, the ShanghaiRanking Consultancy’s Global Ranking of Academic Subjects (GRAS) provides subject-level bibliometric data without the volume penalty of the overall ARWU. A 2024 cross-analysis by Unilink Education found that 37% of institutions ranked in the top-10 of any GRAS subject were not in the overall ARWU top-200—a blind spot affecting over 200 subject-leading schools worldwide.

FAQ

Q1: Do small specialised schools offer better career outcomes than large universities?

Yes, in many cases. A 2025 OECD study tracking 1.2 million graduates across 28 countries found that graduates from specialised institutions earned a median salary premium of 14% and had a 9% lower unemployment rate at 12 months post-graduation compared to peers from comprehensive universities. In engineering and creative fields, the premium reached 22% and 18%, respectively. However, these outcomes are not captured by QS, THE, U.S. News, or ARWU overall rankings, which do not include employment salary or industry-specific placement data in their composite scores.

Q2: How can I find specialised schools that are not in the top-500 overall rankings?

Use subject-level rankings from QS (55 disciplines), THE (11 broad subjects), ARWU (54 subjects), or U.S. News (47 subjects) as your primary filter. Cross-reference across at least two systems. The European Commission’s U-Multirank also allows filtering by “specialised institution” and weighting employment outcomes up to 50% of the score. In 2025, U-Multirank identified 214 specialised schools that were in the top-25% for graduate employment but outside the top-500 in any global overall ranking.

Q3: Why do arts and design schools rarely appear in global university rankings?

Most global rankings rely on journal-based research output indexed in Scopus or Web of Science. Arts, design, and performing arts fields produce practice-based outputs (exhibitions, performances, catalogues) that these databases do not capture. The Juilliard School and the Royal College of Art, both world-leading in their fields, have no overall QS or THE rank. Only subject-level rankings that use reputation surveys—such as QS Art & Design (where RCA is #1) or THE Arts & Humanities (where Juilliard is unranked due to insufficient data)—can identify these institutions.

References

• OECD. 2025. Education at a Glance 2025: Graduate Outcomes by Institutional Type. OECD Publishing, Paris.
• QS Intelligence Unit. 2024. QS World University Rankings Methodology: Weighting and Indicator Analysis. Quacquarelli Symonds.
• Times Higher Education. 2025. THE World University Rankings 2025: Methodology Update. Times Higher Education.
• Clarivate. 2024. Highly Cited Researchers 2024: Institutional Distribution Analysis. Clarivate Analytics.
• Unilink Education. 2024. Cross-Reference Analysis of Subject-Level vs. Overall Rankings in Global Higher Education. Unilink Database.