International Student Diversity: Why Enrollment Numbers Are Not Enough

Share

A high share of international students tells you how many enrolled students crossed a border, or hold a foreign passport, at one point in time. It does not tell you how many countries they come from, whether one source country dominates, whether they are degree students or short-term visitors, at which level they study, whether they progress and graduate at rates similar to other students, who could afford to come, or whether they take part in campus life. Official statistics also count "international" in different ways, so even the headline share is not always comparable. A credible assessment of international student diversity therefore reads the enrollment number together with at least six other kinds of evidence, which this article sets out.

Who counts as an international student?

The answer depends on the statistical rule, and the rules differ in ways that change the headline number.

The UNESCO Institute for Statistics (UIS) defines internationally mobile students as individuals who have physically crossed an international border to take part in education in a destination country that differs from their country of origin. For tertiary students, the UIS glossary defines the country of origin as the country where the student gained the upper secondary qualification, also called the country of prior education. Where that cannot be measured, the glossary recommends country of usual or permanent residence, and it allows citizenship to be used "only as a last resort". The same entry excludes distance learning from a foreign institution and foreign programs delivered in the student's own country.

The OECD's Education at a Glance 2025 keeps two concepts apart. Foreign students are those who are not citizens of the country where they are enrolled; the OECD notes that they may be long-term residents or even born in the host country, and that naturalization policies differ between countries, so comparisons based on citizenship should be made with caution. International students are those who left their country of origin to study, with origin defined by upper secondary education, prior education or usual residence, depending on national legislation and data availability. The OECD calls students who are either international or foreign "mobile students".

National statistics add further variants. Statistics Canada's persistence and graduation tables classify students by status in Canada: "international students" include study-permit holders, non-Canadians on other visas, non-Canadians whose status is unknown and students studying from outside Canada, for example online. That last group would not count as internationally mobile under the UIS definition. France's statistical service (SIES) splits foreign students into two groups with different profiles: of nearly 50,000 foreign students who first enrolled in a university bachelor's (licence) or master's program in 2018-2019, about three quarters were degree-mobile and one quarter were foreign residents of France.

How common definitions of "international student" differ (sources opened 21 November 2026)
DefinitionMain criterionWho may be included or excluded
Internationally mobile student (UNESCO UIS)Physical border crossing; origin by country of prior (upper secondary) educationExcludes cross-border distance learning and foreign programs taken at home; citizenship only as a last resort
Foreign student (OECD)Citizenship different from the country of studyCan include long-term residents and people born in the host country
International student (OECD)Moved to study; origin by upper secondary education, prior education or usual residenceGenerally a subset of foreign students, according to the OECD
International student (Statistics Canada, persistence tables)Status in Canada (not a citizen or permanent resident)Includes students studying from outside Canada, for example online
Foreign student (France, SIES Note 2025-02)Nationality, then split by mobilitySeparates degree-mobile students from foreign residents

The practical consequence is simple: before comparing two institutions or two years, check which rule produced each number. The OECD's April 2026 report on international students in six host countries notes that national and internationally harmonized figures differ partly because of time lags, definitions and the levels of education covered.

Share versus spread: country-of-origin diversity

Two universities can report the same international share and have very different student bodies. One may enroll students from dozens of countries in roughly balanced numbers; another may draw most of its international students from a single country. The enrollment share cannot distinguish the two.

Concentration is a documented system-level pattern. The OECD's 2026 report describes the origins of international students in Australia, Canada, France, Germany, the Netherlands and the United Kingdom as "increasingly concentrated". In 2023, students from Asia made up 89% of international students with a known country of origin in Australia, 73% in Canada and 68% in the United Kingdom, while students from other European countries made up 74% in the Netherlands and students from Africa 53% in France. The same report states that reliance on a single country or region raises risks for systems and institutions, because enrollments can fall suddenly with geopolitical tensions, pandemics or immigration policy changes. Across OECD countries, Education at a Glance 2025 reports that students from Asia accounted for 58% of all internationally mobile students in 2023.

Governments have also treated concentration as a policy question. In February 2022, Australia's then Department of Education, Skills and Employment released a discussion paper inviting views on the benefits and challenges of diversifying international student cohorts at public universities. Education at a Glance 2025 introduced a separate system-level indicator that asks a different question: how far a country's mobile students are clustered in its most international institutions. If mobile students were spread evenly across institutions in proportion to their size, the indicator would equal 10%; if all were in the selected institutions, it would equal 100%. The OECD notes that this indicator does not describe concentration inside any single institution.

Two simple concentration measures (editorial illustration)

Institutions can describe the spread of their international students with two easily computed figures. The first is the top-source share: the percentage of international students who come from the largest source country. The second is a Herfindahl-type index: square each source country's share of international students and add the results. The index is close to 0 when students come from many countries in small numbers and equals 1 when all come from one country. Its inverse, 1 divided by the index, can be read as an "effective number of source countries". A related measure, based on the inverse of the sum of squared nationality shares and adjusted for group size, is used by Haelermans and colleagues (2026) to describe the nationality mix of small classroom groups.

The following worked example is hypothetical. The three universities do not exist, and the figures are invented only to show the arithmetic. Each enrolls 1,000 international students.

Hypothetical example: three universities with the same number of international students
Hypothetical universityDistribution of international students by source countryCountries representedTop-source shareHerfindahl-type indexEffective number of source countries
A60%, 20%, 10%, 10%460%0.4202.4
BTen countries at 10% each1010%0.10010.0
C30%, 30%, then 40 countries at 1% each4230%0.1845.4

The arithmetic for University A is 0.60² + 0.20² + 0.10² + 0.10² = 0.36 + 0.04 + 0.01 + 0.01 = 0.42, and 1 ÷ 0.42 ≈ 2.4. For University C it is 0.09 + 0.09 + 40 × 0.0001 = 0.184, and 1 ÷ 0.184 ≈ 5.4. The example shows why a simple count of nationalities can mislead: University C can say it has students from 42 countries, yet its intake is more concentrated than University B, which has 10. These measures are an editorial illustration, not a UNIRANKS indicator, and neither of them says anything about inclusion.

Why concentration may matter for students

Two recent studies examined nationality mix and academic outcomes in specific settings. Khoo (2026) used administrative data on doctoral students at public institutions in Ohio, United States, from 2009 to 2015, covering international students from six origin countries, and exploited year-to-year changes in cohort composition within programs. The author reports that a higher share or number of same-country peers reduced the likelihood of persisting in the program, particularly above certain thresholds, and had an adverse effect on completion that lost statistical significance in later years. The effect remained after accounting for the presence of other international peers. Khoo cautions that the mechanisms need further study before admissions policy is changed, and suggests that departments track cohort country-of-origin shares at matriculation and at the end of the first year.

Haelermans, Hendrikse, Smeets and Bijsmans (2026) studied 19,448 bachelor's students over six years at one international university in the Netherlands, where students are randomly assigned to small teaching groups. They report that a more varied nationality mix in the group largely improved grades, with differences by field, year of study and student nationality, and no evidence of harm to local students. Course evaluations from one field suggested that the effect was stronger where teachers encouraged English as the common language during classes and breaks. The authors note that the results come from one university, bachelor's programs only and a single outcome, grades.

Neither study shows that any particular level of concentration is "too high" in general. Together they suggest that the composition of a cohort, and not only its size, is worth monitoring.

Degree students, exchange students and short visits

Enrollment figures can mix students who came for a full degree with students who came for a semester. Degree mobility means moving abroad to obtain a full qualification; credit mobility means a temporary period abroad, often a semester or less, that counts toward a degree at the home institution. The OECD's 2026 report states that its primary focus is degree mobility, and Education at a Glance 2025 treats credit mobility separately. A university that counts every incoming exchange student, summer participant or visiting researcher in its "international students" figure is reporting a different quantity from one that counts degree-seeking students only. Published figures should state which groups are included and whether counts are headcounts at a census date or totals over a year. Incoming and outgoing exchange participation is examined in our companion article on measuring international student mobility.

Undergraduate and postgraduate distribution

An institution-wide share can hide large differences between levels. Bound, Braga, Khanna and Turner (2021), reviewing four decades of growth in international enrollment in the United States, state in their abstract that the impacts are "far from uniform", with significant differences by level of study and type of institution. In practice, an international share concentrated in one-year taught master's programs means something different, for teaching, belonging and graduate outcomes, from the same share spread across undergraduate and doctoral programs. A useful disclosure breaks the international share down by level (bachelor's, master's, doctoral) and, where numbers allow, by broad field, because international students are often clustered in particular programs. The OECD's 2026 report observes that such clustering, shaped by language of instruction and field of study, can limit everyday opportunities for international and domestic students to interact.

Do international students stay and graduate?

Retention and completion show whether a diverse intake becomes diverse graduates. The evidence from official statistics is mixed and depends on the measure.

  • Canada. Statistics Canada Table 37-10-0140 (undergraduate degree programs, full-time new entrants, persistence and graduation within the province of first enrollment) shows a one-year persistence rate of 82.9% for international students who entered in 2022/2023, compared with 87.6% for Canadian students. For the 2018/2019 entry cohort, the four-year graduation rate was 47.8% for international students and 44.1% for Canadian students. For the 2016/2017 cohort, the six-year graduation rate was 72.5% and 73.1% respectively.
  • France. The SIES note of January 2025 reports that in the bachelor's (licence) cycle, degree-mobile foreign students had a higher success rate than both foreign residents and French students. Most of the 10.4 percentage-point gap between French students and foreign residents was explained by age, type of secondary diploma and social origin; at equal characteristics it fell to about one point. At master's level, French students had higher success rates than both foreign groups, and about half of the eight-point gap with foreign residents remained after accounting for those characteristics.
  • Six-country synthesis. The OECD's 2026 report concludes that international students often have higher completion rates than domestic students, but lower first-year progression or higher initial dropout in several countries. It names financial pressure, stress and wellbeing, unmet expectations, language and study-permit issues as key drivers of dropout.

Two lessons follow. First, a single "international students" outcome figure can average away large differences between degree-mobile students and long-term foreign residents, as the French data show. Second, early progression and final completion can point in different directions, so both are needed.

Who can become an international student? Access and affordability

An international intake reflects who could afford to apply, not only who was admitted. Netz (2021), introducing a special issue of Higher Education, summarizes research showing that the likelihood of studying abroad depends on socio-demographic characteristics, including social origin, as well as on factors such as the availability of institutional or state funding. If that pattern holds for degree mobility as well, a campus with many nationalities could still draw mainly on relatively advantaged families in each country; this is an inference, not a finding of the studies reviewed.

Fees and funding also shape the composition of the international population. The OECD's 2026 report notes that in France and the Netherlands, students from outside the European Economic Area pay much higher fees on average than domestic and other EEA students, and that reliance on international tuition revenue has created financial vulnerabilities in several systems, particularly Australia, Canada and the United Kingdom. When fee income is a major motive for recruitment, the incentive is to maximize numbers from the largest paying markets, which can work against both spread and access. Indicators worth publishing alongside the enrollment share include the proportion of international students receiving need-based or merit-based institutional aid and the number of scholarship places by region of origin. Students looking at the funding side can start with our guide to scholarships and financial aid for international students.

Participation and belonging

Being enrolled is not the same as taking part. The OECD's 2026 report finds that international students' sense of belonging is often equal to or higher than that of domestic students, but that isolation and weak contact with domestic peers remain common. It cites Eurostudent survey data in which 14% of international students in France and 15% in the Netherlands felt isolated from fellow students, compared with 8% and 5% of domestic students.

Peer-reviewed studies show that averages hide variation within the international population. In a German nationwide panel of international students (N = 2,998), Thies and Falk (2023) report that more frequent participation in extracurricular activities was associated with a stronger sense of belonging, but that this association became non-significant once contact with other students, academic integration and psychological adaptation were controlled; contact with German and multinational students was important, and patterns varied by region of origin and language of instruction. In survey data from a large university in southern Ontario, Mullen and Li (2025) found that Chinese international students reported a greater sense of belonging than domestic students, South Asian students did not differ, and students from other regions reported a weaker sense of belonging; experiences of on-campus racism and discrimination were associated with lower belonging. A single "international students" satisfaction score would hide these differences. How to measure multicultural inclusion, and what support services universities should provide international students, are covered in two companion articles.

Reporting limits: small numbers and nationality as personal data

Detailed diversity reporting runs into privacy limits. Under the EU General Data Protection Regulation, a person is identifiable if they can be identified directly or indirectly, including through factors specific to their cultural or social identity (Article 4), and Article 9 restricts processing of data revealing racial or ethnic origin. Nationality is not itself listed in Article 9, but in a small program, a table showing one student from a given country together with level and field may identify that person. Official statisticians handle this with suppression and rounding: Statistics Canada's tables, for example, mark some cells as suppressed to meet the confidentiality requirements of the Statistics Act and others as too unreliable to be published.

For universities, this means that the most diversity-revealing figures (small nationality groups, completion by country of origin, belonging by region) often cannot be published at full detail. Reasonable practice is to report by region of origin or concentration index rather than full country lists, to suppress or combine small cells, and to say openly that this has been done. Readers, in turn, should treat a perfectly granular nationality breakdown for a small institution with some caution. This section describes general practice and is not legal advice.

An editorial framework: what to ask behind the headline share

The checklist below is an editorial framework proposed by this article for readers and institutions. It is not part of the UNIRANKS methodology and does not describe how UNIRANKS evaluates universities.

Editorial framework: evidence to read alongside an international student share
DimensionQuestion to askExample evidence
DefinitionIs "international" based on prior education, residence or citizenship?Stated counting rule and census date
CompositionAre degree students, exchange students and short visitors separated?Separate counts by mobility type
SpreadHow concentrated are source countries?Top-source share; Herfindahl-type index or effective number of countries
Level and fieldWhere in the institution are international students enrolled?Shares by bachelor's, master's, doctoral level and broad field
OutcomesDo international students progress and graduate?First-year continuation and cohort completion, split by mobility type where possible
AccessWho can afford to come?Aid and scholarship coverage; fee structure
ParticipationDo international students take part in academic and campus life?Disaggregated belonging and isolation measures; participation in activities
PrivacyHow are small groups protected?Stated suppression or rounding rules

Our interpretation

The evidence reviewed here supports the thesis that a high international share does not by itself establish a diverse or inclusive campus. The share depends on a counting rule; it says nothing about spread across countries; it may combine full-degree and short-stay students; and it is silent on outcomes, affordability and participation. The research on concentration is still thin and context-specific (one U.S. state's doctoral programs, one Dutch university), so it does not justify a universal threshold, but it is consistent with the view that cohort composition is worth monitoring. The official completion data caution against assuming either that international students struggle or that they thrive: the answer varies by country, level, stage of study and type of foreign student.

UNIRANKS describes its Internationalization & Diversity dimension as evaluating "both the international composition of the university and the practical opportunities available for students and academics to participate in global education." Composition and participation are different things, which is the distinction this article draws. The UNIRANKS methodology page does not publish indicator definitions for this dimension, and nothing in this article describes how UNIRANKS measures it.

Practical steps for universities and evaluators

  1. State the definition behind every international student figure (prior education, residence or citizenship), the census date and whether exchange and short-term students are included.
  2. Publish the distribution by source region and a concentration measure, such as the top-source share or an effective number of source countries, alongside the share of enrollment.
  3. Break the international share down by level of study and broad field, so readers can see where international students actually study.
  4. Report first-year continuation and cohort completion for international students, separating degree-mobile students from long-term foreign residents where the data allow.
  5. Disclose fee and aid information that shows who can access the institution from abroad, such as scholarship coverage by region of origin.
  6. Disaggregate belonging or isolation survey results by region of origin and level where sample sizes permit, and explain when they do not.
  7. Apply and publish small-cell suppression or rounding rules before releasing nationality-level data.
  8. For evaluators and prospective students: treat a single international share, or a count of nationalities, as a starting point and ask for the evidence in the framework above.

Limitations

This review draws on six peer-reviewed articles, of which three were read in full (Khoo, 2026; Haelermans et al., 2026; Netz, 2021) and three from their abstracts only (Bound et al., 2021; Thies and Falk, 2023; Mullen and Li, 2025). Claims from abstract-only sources are limited to what the abstracts state. Official statistics come from Canada and France, plus OECD syntheses; the OECD report covers six high-income host countries, and its belonging figures summarize national surveys with different questions and samples. Definitions differ across all of these sources, so the figures are not directly comparable with each other. The two concentration studies cover specific settings (doctoral programs in one U.S. state and bachelor's programs at one Dutch university) and should not be generalized to all institutions. The concentration measures and the hypothetical example are editorial illustrations, not findings. An Australian discussion paper on diversification is cited only for its existence and purpose, as its full text could not be retrieved for this review.

Related reading on UNIRANKS

This article is part of the Internationalization & Diversity research series. The companion articles referenced above explain how incoming and outgoing exchanges should be counted, how inclusion can be measured and what support international students need.

Editorial disclosure

UNIRANKS publishes university rankings, and its methodology includes an Internationalization & Diversity dimension. This article reviews independent research on international student diversity beyond enrollment numbers. It does not describe how UNIRANKS scores institutions.

Research References

  1. Bound, J., Braga, B., Khanna, G., et al. (2021). The Globalization of Postsecondary Education: The Role of International Students in the US Higher Education System. Journal of Economic Perspectives, 35(1), 163-184. https://doi.org/10.1257/jep.35.1.163 (verified 21 November 2026)
  2. Haelermans, C., Hendrikse, A., Smeets, C., et al. (2026). The impact of international classroom diversity on student performance: evidence from a Dutch University. Higher Education, online first. https://doi.org/10.1007/s10734-026-01644-y (verified 21 November 2026)
  3. Khoo, P. (2026). A Peer Like Me? Peer Effects Among International Students in Doctoral Programs. The B.E. Journal of Economic Analysis & Policy, 26(1), 161-198. https://doi.org/10.1515/bejeap-2025-0135 (verified 21 November 2026)
  4. Mullen, A. L., & Li, Y. (2025). Belonging beyond borders: a comparative analysis of international and domestic students at a Canadian university. Higher Education, 90(3), 801-818. https://doi.org/10.1007/s10734-024-01351-6 (verified 21 November 2026)
  5. Netz, N. (2021). Who benefits most from studying abroad? A conceptual and empirical overview. Higher Education, 82(6), 1049-1069. https://doi.org/10.1007/s10734-021-00760-1 (verified 21 November 2026)
  6. Thies, T., & Falk, S. (2023). International Students in Higher Education: Extracurricular Activities and Social Interactions as Predictors of University Belonging. Research in Higher Education, 64(8), 1143-1164. https://doi.org/10.1007/s11162-023-09734-x (verified 21 November 2026)

Sources

Share this article