The internationalization of European higher education institutions

Grahams HD:Users:Graham:Public:GRAHAM'S IMAC JOBS:14584 - EE - BONACCORSI:BONACCORSI 9781782540717 PRINT (M3313) (G)

138

5. The internationalization of European higher education institutionsMarco Seeber and Benedetto Lepori

INTRODUCTION

The notion that higher education has become more international, and should become even more, has been repeatedly affirmed in the recent debate. However, studies on this issue have been limited to analyses at aggregate country level or to case studies of small samples of higher educa-tion institutions. The EUMIDA project provides an opportunity to fill a gap in our knowledge of the internationalization phenomenon. By focus-ing on microdata it is possible to investigate which factors are associated with high and low levels of internationalization, at the country level but also at the level of individual institutions.

Internationalization is a complex construct, which requires careful defi-nition and can be measured through several indicators (Horn et al., 2007). Several studies, as well as public policies, have focused on temporary mobil-ity. Instead, we focus on indicators of long- term mobility such as the share of foreign undergraduate students, PhDs and academic staff. These indica-tors reflect pivotal effects in academic activity; thus, any relevant process affecting their composition is also likely to affect institutional functioning.

The chapter is organized as follows. In the first section we introduce the conceptual framework by describing the main rationales for interna-tionalization and discussing several explanatory variables at country and institutional level. The empirical section describes the level of internation-alization of European higher education institutions, while the conclusions discuss the main findings and the open issues for further research.

INTERNATIONALIZATION: FROM INDIVIDUAL EXPERIENCE TO SYSTEM POLICIES

The competition for the most talented students and researchers that char-acterized the American system and was a key element of its success is now

BONACCORSI 9781782540717 PRINT (M3313) (G).indd 138 22/11/2013 14:00


The internationalization of European higher education institutions 139

extending to a global scale (Wildavsky, 2010). This complex process takes place within the larger processes of migration, yet with its own dynamics and intrinsic rules, determined by the interaction of individual, institu-tional and national factors.

For a long time internationalization was mostly the outcome of indi-vidual choices of short- or long- term mobility, while the institutional and political support was not an issue. The expectation of mobility for students and researchers is driven by the perception that the international experi-ence increases exposure to new skills, ideas and ways of working, facilitat-ing the transfer of knowledge and creativity. Students and scientists also migrate, looking for better career opportunities, work conditions, facilities and higher status.

In the last decades, however, the internationalization of research and teaching activities has become a policy and institutional priority in its own right, on the assumption that it can generate positive spillovers in terms of quality and give access to financial, instrumental and human resources (Geuna, 1998a, 1998b; De Wit, 1999; Primeri, 2008). At the institutional level, internationalization can be conceived as the process of integrating an international and intercultural dimension into the teaching, research and service functions (Knight and De Wit, 1995). Several economic, political and cultural rationales may justify the support to the international dimen-sion, which after all is directly linked to the early history and development of universities, as the concept of ‘universe’ is inherent in the same name ‘university’ (Knight, 1997).

The circulation of highly skilled people is often depicted as a globaliza-tion phenomenon, unavoidable and beyond control of single entities. In reality, nation states and supranational institutions are deeply involved in this process, explicitly promoting selective immigration policies to attract excellent students and researcher.

THE RELEVANCE OF INTERNATIONALIZATION TO EUROPEAN HIGHER EDUCATION

The European Union (EU) has promoted several initiatives to facilitate the mobility of researchers within the framework of thee European Research Area (ERA),1 even if there are still several obstacles for mobility of researchers in Europe (MORE, 2010).

Despite the ambitious goal to become the most innovative area in the world, Europe has still to face an asymmetry with the USA in terms of flows of PhD students and researchers (Moguérou, 2005). Moreover, the production of PhDs in Asian countries has dramatically increased over


140 Knowledge, diversity and performance in European higher education


the last decades, while in Europe it has levelled off, especially in the scien-tific and technical disciplines (ibid.).

Researcher mobility was initially intended to achieve a better alignment between supply and demand for researchers across Europe. However, the promotion of policies to enhance mobility across countries is not without risks. The received wisdom, implicit in the brain drain debate, suggests that host countries are the net beneficiaries of highly skilled migration, capable of ‘skimming’ the cream of employees in the competitive global markets (Ackers and Gill, 2007). The concepts of brain drain or brain gain are too simplistic to capture this dynamic, as countries will be simultaneously sending and receiving talents. But, as in a strategic game, each country attempts to pass on to others the cost of migration by obtaining from lower levels the input to fill up the gaps left by emigrants to its own skills base. Flows are not occurring randomly, rather they have a clear direction from lower to higher places, in terms of income and reputation. In Europe, the sending countries have lost significant numbers of scientists since the 1990s, and there is a sense that flows are largely unidirectional and go from ‘less attractive’ regions to more competitive places (Meyer et al., 2001).

In sum, the sustainability and management of human capital is problem-atic both in sending and receiving countries. If the former countries face a potential brain- drain issue, in the latter the continued ability to attract stu-dents and recruit researchers from abroad reduces the need to ensure the scientific career to be appealing to home- grown researchers, and ultimately threats the sustainability of science. There may be an underlying tension in the European Union research policy: on one hand, the commitment to the freedom of movement of European citizens and individual equity in employment opportunity; on the other, concerns about the consequences of free and open labour markets.

INTERNATIONALIZATION AND THE QUALITY OF RESEARCH

Attracting foreign researchers is commonly perceived as a recipe for increasing research quality, as international researchers generate a creative diversity (Manrique and Manrique, 1999; De Wit, 2002; Sheppard, 2004; Hser, 2005; Altbach, 2006; NAFSA, 2006; Stromquist, 2007). In general, opening the research system to foreign researchers facilitates the accumu-lation of talent in places with a high intensity of scientific research and encourages the creation of centres of excellence (Viljamaa et al., 2009). Several arguments support the existence of a positive association between internationalization and research quality.




A very low level of international staff may result from closeness and isolation, and many studies have proven that closeness is detrimental to research quality. Favouring internal, as opposed to external, knowledge exchanges preserves the existing institutional culture and status quo, leading to intellectual and organizational inertia (Leslie and Fretwell, 1996), and ultimately affects the output and quality of the research work (Rosenkopf and Nerkar, 2001; Rosenkopf and Almeida, 2003). Academic inbreeding has long been assumed to have a damaging effect on scholarly practices and achievements because it gives rise to aca-demic parochialism. When a university hires its own PhDs, there will be an overemphasis on the reproduction of locally learned knowledge, practices, as well as a consolidation of social structures in the organi-zation. This may slow or block new or alternative approaches to the creation of institutional knowledge, limiting institutional change and ultimately contributing to the ossification of the organization (European Commission, 1995). In many countries inbreeding is very common; esti-mates suggest that the level of academic inbreeding is high in Portugal (91 per cent), Spain (88 per cent), Italy (78 per cent), Austria (73 per cent) and France (65 per cent); it is medium in Norway (56 per cent), Belgium (52 per cent), Finland (48 per cent), while it is low in the Netherlands (40 per cent), Denmark (39 per cent), Sweden (32 per cent), Switzerland (23 per cent), UK (5 per cent) and Germany (1 per cent). Overall, scien-tific productivity correlates negatively with the percentage of inbreeding (Soler, 2001). In Mexico, inbred faculties generate on average 15 per cent less peer- reviewed publications, they are about 40 per cent less likely to exchange information of critical relevance to their scholarly work with external colleagues and academic inbreeding appears to be detrimental to scientific output even in leading research universities (Horta et al., 2010). In Italian universities the higher the share of researchers born in the province of the university and the worse the scientific performance of the institution (Reale and Seeber, 2011), while universities where cro-nyism is widespread are characterized by even poorer academic perform-ances (Durante et al., 2011).

Foreign staff may be valuable because of direct positive spillover on quality. Studies on the research performance of foreign researchers have shown the importance of the contribution of foreign born researchers to US science (Levin and Stephan, 1999; Kerr, 2008; Black and Stephan, 2010). International faculty members are significantly more productive in research than US citizen faculty members, but less productive in teach-ing and service (Hunt, 2009; Mamiseishvili and Rosser, 2010). Recent literature suggests that hiring external researchers into existing environ-ments is important for the ability of organizations to generate and access




new knowledge. Researcher mobility is more likely to result in interfirm knowledge transfer (Song et al., 2003), and hiring star scientists can reshape the direction of research organizations (Song et al., 2003; Lacetera et al., 2004). Internationalization stimulates linkages to external actors, which allow the university to understand their context and identify where resources are available.

Foreign researchers are expected to apply in large number to excellent universities, just because these institutions are attractive and grant a high status (Geuna, 1998b). Horn et al. (2007) analysed the internationaliza-tion of 77 research universities in the United States, and found that the top 11 institutions with respect to internationalization are also ranked among the top 11 by quality. Even so, one should be careful in identify-ing a causal relationship between hiring more foreign staff and increased research performance. A ‘Matthew Effect’ could be in place. In such perspective, internationalization would be an instance of the virtuous circle of resources, prestige and performance, as students and staff apply in large numbers to wealthy and performing institutions (Horn et al., 2007).

DETERMINANTS AND DYNAMICS OF INTERNATIONALIZATION

Some elements can be identified that are expected to affect the level of internationalization at country and institutional level.

Country Factors

Financial resourcesThe level of available financial resources, and in particular the resources invested by the state, is a major issue in internationalization (Egron- Polak and Hudson, 2010). For instance, the Federal Institute of Technology in Zurich (ETH) receives almost tenfold public funds than its Portuguese equivalent (Horta, 2010) and it is able to recruit many researchers (42 per cent) even from research- intensive countries such as the USA, Germany, UK and France (Horta et al., 2010).

Academic labour markets and recruitment proceduresSystems are characterized by different approaches and practices in the recruitment of researchers, which may favour or hinder access of foreign researchers. In some countries most recruitment occurs within the insti-tutional boundaries, whereas in other systems external markets prevail




and career advancement involves the praxis or law obligation to apply for posts in universities other than the current place of work (Musselin, 2003). New posts may be publicly advertised with different intensity and content: in the group website, by the university portal or even at the min-istry level; furthermore, the national language can be used, the English language, or both. Recruitment procedures can be highly formalized or not (EURYDICE, 2008). Standardized procedures and rules aim at an ex ante, centralized control of entry quality; in other systems, chairs and faculty level have more discretionary power in the choice of criteria, and likely more responsibility ex post for the scientific production of the selected candidate.

LanguageLanguage hinders the inward mobility in countries where less widely spoken languages are dominant and where the use of major languages is not typical in the academia. There are considerable differences between the Member States in how they deal with languages. In large countries, such as Germany, Italy, Spain and France, foreign students and scientists are still, more or less, expected to speak the national language. The situation is quite different in smaller countries, such as the Netherlands or Finland, where foreign academics are not expected to be fluent in Dutch or Finnish and where English is widely spoken among researchers. Some systems with less endowments necessarily exploit the advantage of a widely spoken language, as in the case of Spain and Portugal toward Latin American countries. There are also important disciplinary variation in the use of national language as opposed to English, as it happens, for instance, in law and humanities where it is more common to publish in national journals than in the hard sciences.

Share of foreign populationThe number of international undergraduate students may be affected by the share of foreign population, as they are often the children of foreign citizens. For these reason, it is important to disentangle the number of international students that comes from another country precisely for stud-ying, and disentangle the two types of international students. This issue is discussed in the empirical section.

Institutional Factors

Universities with better research quality and reputation are expected to be more internationalized, because of the mechanisms described in the previ-ous paragraph.




Larger universities are also expected to have a larger share of interna-tional staff and students. In fact, the degree of institutionalization of the international dimension changes from one university to another (Knight and De Wit, 1995; Qiang, 2003). The small institutions’ priority is outgo-ing mobility of students, while for medium- large ones it is research col-laboration. In general, the larger the institution, the higher the importance of internationalization for its leaders (IAU, 2010).

It is uncertain whether older universities should display higher or lower share of international people. On the one hand, age can be positively cor-related to prestige and status, thus older universities would result more attractive and internationalized. On the other hand, emphasis on interna-tionalization has increased in the last decades and young universities of recent expansion might show higher share of foreigners.

Research- oriented HEIs are expected to look for the best researchers worldwide and to be more attractive, as researchers usually prefer doing research than teaching, and because success in research is the main source of status. Then, by a similar argument, teaching- intensive HEIs should be less attractive and internationalized.

EMPIRICAL ANALYSIS

The empirical analysis is based on EUMIDA, OECD and Eurostat data-sets. We consider the smaller dataset of research- active institutions in EUMIDA (RAIs), which includes more detailed information. However, since the availability of the data varies across countries from staff to stu-dents, the analysis will be carried out in several steps, depending on the coverage of data.

An important methodological issue relates to the distinction between foreign and mobile students (UOE, 2006) (Figure 5.1). Foreign students are defined as non- citizens of the country in which they study. Most coun-tries have data on country of citizenship, which in most cases is a clear and well- defined demographic variable. However, the data on foreign students are inappropriate to measure student mobility as they capture immigra-tion flows in general – not all foreign students have come explicitly with the intention of studying. Instead, mobile students are precisely defined as foreign students who have moved to another country with the objective of studying.2

Identifying mobile students is not always straightforward. Actually, within the European Union there are no visa requests and this makes it impossible to derive numbers of mobile students from visa statistics. Thus, each country employs the operational definition of mobile students that




capture the definition above in the best way. The two most widely used are: (1) students who are not usual residents of their country of study, that is, those who have recently moved to the destination (host) country from somewhere else; and (2) students who received their prior qualifying education in another country. Still, several countries did not provide data on mobile students, but on foreign students instead. Table 5.1 summarizes the information on data availability, and whether the data collected refer to mobile or foreign students.

Data were collected on the main features at country and institutional level that can be associated to the level of internationalization.

In order to measure the attractiveness of the system in terms of finan-cial resources available we consider the total intramural R&D expenditure (GERD) in the HE sector in terms of purchasing power standard (PPS) per inhabitant at 2000 constant price (source: Eurostat).

It was possible to identify country practices as to two aspects of the recruitment process that are particularly relevant for internationalization: (1) whether new positions are advertised; (2) degree of formalization of the recruitment process (EURYDICE, 2008). As for the public advertise-ment there are no substantial differences across the considered countries, since almost everywhere the central (ministry) or institutional (university, department) actor publicly advertises positions. On the other hand, there

Criteria

Citizenship

Usual residence

Population of students

Foreignstudents

Nationalstudents

Mobilestudents

Resident foreignstudents

Resident nationalstudents

Homecomingnationalstudents

Source: UOE (2006).

Figure 5.1 Student categories




Table 5.1 Type of data and availability in EUMIDA database

Country Academic Staff

Students ISCED 6

Students ISCED 5

Type of Data

Austria No Yes Yes Mobile studentsBelgium No Yes Yes Only data on Flemish

universities; mobile students

Bulgaria No Yes Yes Mobile studentsCyprus No Yes Yes Foreign studentsCzech Republic No No No No individual HEI

data – aggregate data on foreign students

Estonia No Yes Yes Mobile studentsFinland No Yes Yes Foreign studentsGermany Yes Yes Yes Mobile studentsGreece No No Yes Foreign studentsHungary No Yes Yes Foreign studentsIreland No Yes Yes Mobile studentsItaly Yes Yes Yes Foreign studentsLatvia Yes Yes Yes Mobile studentsLithuania Yes Yes Yes Mobile studentsLuxembourg No Yes Yes Foreign studentsMalta No Yes Yes Foreign studentsNetherlands No No Yes Foreign studentsNorway No No No No individual HEI

data – aggregate data on foreign students

Poland No Yes Yes Foreign studentsPortugal No Yes Yes Foreign studentsRomania No Yes Yes Foreign studentsSlovakia No Yes Yes Foreign studentsSlovenia Yes Yes Yes Foreign studentsSpain Yes Yes Yes EUMIDA data not

fully reliable. Foreign PhD students; mobile ISCED 5 students

Sweden No Yes Yes Foreign studentsSwitzerland Yes Yes Yes Foreign PhD students;

mobile ISCED 5 students

United Kingdom Yes Yes Yes Mobile students

Sources: EUMIDA deviant cases explanation; UOE manual (2006) and OECD education database(2003).




are relevant differences as to the mode of recruitment. In some countries the selection occurs via a competitive examination whose rules and steps are formally defined. In Italy, for instance, both Ministry and institutions define criteria and procedures of a competitive examination. In Spain and France, institutions must follow procedures established at central level and monitor compliance with these procedures.

In other countries the recruiter enjoys more discretionary power and flexibility (Table 5.2). Accordingly, a dummy variable was set up to dis-tinguish discretionary flexible systems of recruitment (1) from formalized ones (0).

As to the language variable, the value 1 is assigned to countries where English, Spanish or Portuguese are the main spoken languages, and to small countries in which the language of one or more larger neighbour-ing country is commonly used, such as in the Baltic countries (Russian), Switzerland (French, German, Italian), Belgium (French, Dutch, German), Austria (German), Luxembourg (French and German). Table 5.3 summa-rizes the values for the indicators considered at country level.

With respect to factors related to individual higher education institu-tions, five indicators were constructed (Table 5.4).

Research quality is measured using the Leiden university ranking indi-cator. This indicator is robust to annual variability, as it considers the

Table 5.2 Recruitment process: public advertisement and selection practices

Public Advertisement

No Yes (central)

Yes (institutional)

Yes (central 1 institutional)

Formal competitive examination

No IE, NL, SI, FI, SE, UK, NO

BEL, DK, DE, EL, CY, LU, MT, AT, PL

Yes (central)

ES, FR

Yes (institutional)

BG PT

Yes (central 1 institutional)

CZ, EE, IT, LV, LT, HU, RO, SK

Source: Our elaboration on EURYDICE (2008).




Table 5.3 Country features associated withthe level of internationalization

Financial Resourcesa

Recruitment Modeb

Language Foreign Population (%)c

Austria 191 1 1 10Belgium 115 1 1 9Bulgaria 4 0 0 0Cyprus 37 1 0 16Czech Republic 44 0 0 3Denmark 211 1 0 5Estonia 80 0 1 17Finland 175 1 0 3France 101 0 1 6Germany 112 1 0 9Greece 61 1 0 8Hungary 30 0 0 2Ireland 124 1 1 13Italy 84 0 0 6Latvia 37 0 1 18Lithuania 59 0 1 1Luxembourg 54 1 1 43Malta 30 1 0 4Netherlands 194 1 0 4Norway 182 1 0 7Poland 26 1 0 0Portugal 83 0 1 4Romania 14 0 0 0Slovakia 18 0 0 1Slovenia 47 1 0 3Spain 76 0 1 12Sweden 226 1 0 6Switzerland 221 1 1 22United Kingdom 122 1 1 7

Notes:a. Total intramural R&D expenditure (GERD) in HE sector, by PPS per inhabitant at

2000 constant prices. Year 2008 (Greece: year 2007).b. EURYDICE (2008).c. Share of foreign population. Year 2008.

Sources: Eurostat Science, technology and innovation online database; Eurostat population statistics; Norwegian Institute of Statistics (Norway); Swiss Institute of Statistics (Switzerland).




productivity of the last eight years (2000–07), and comprehends the largest sample of European HEIs (n 5 250).

Size is measured with the number of total staff. This measure was preferred to the number of students, which can be biased by the large disciplinary variability in students to professor ratios, and also preferred to the mere number of academics, which does not fully consider the impact of the student number on the mass and volume of activity of the institution.

The number of PhD students over undergraduate students is an index commonly used to estimate research intensity (Bonaccorsi et al., 2007). Teaching intensity is measured by the number of undergraduate students per unit of academic staff.

THE INTERNATIONALIZATION OF EUROPEAN HIGHER EDUCATION SYSTEMS AND INDIVIDUAL INSTITUTIONS

Internationalization is described by employing Eurostat and OECD data for national aggregates and EUMIDA data for individual HEIs. Table 5.5 describes the degree of internationalization of the EUMIDA countries by the share of foreign academic staff, ISCED 6 (PhD) stu-dents and ISCED 5 (undergraduate) students. The level of internation-alization is often the highest as to the PhD level, followed by researchers and undergraduates.

Table 5.4 Institutional indicators

Indicator Type of Data and Source

Research quality Leiden ranking size- independent indicator CCP/FCS based on field normalized average impact. Year 2008

Size Number of total staffa

Age of the institution Year of foundationa

Research intensity Ratio number of PhD students/number of undergraduate studentsa

Teaching intensity Ratio number of undergraduate students/number of academic staffa

Note: a. Data from EUMIDA.




Table 5.5 Internationalization of European higher education systems

Academic Staff Students ISCED 6 Students ISCED 5

No. HEIs

Totald % Intnl

No.HEIs

Total % Intnl

No. HEIs

Total % Intnl

Austria 20 706 25 17 310 23 67 252 829 15Belgium 21 708 5b 17 19 068 34 41 252 369 8Bulgaria 12 235 1a 29 3666 1 29 207 585 2Cyprus 397 3 364 7 7 13 616 11Czech Republic

16 879 4a 25 096 10*** 210 146 8***

Denmark 11a 11 6251 4 14 116 437 7Estonia 3577 2a 5 2357 3 7 44 155 2Finland 16 595 18 20 792 9 47 275 674 4France 7b 68 696 25b 1 548 740 13c

Germany 109 128 026 11 131 88 095 21 300 1 848 740 10Greece 21 547 38 557 549 4Hungary 15 233 2a 24 6897 7 61 366 815 4Ireland 9319 19 5756 25 21 149 280 5Italy 80 97 606 3 79 39 238 7 78 1 914 054 3Latvia 20 4381 0.6 20 2152 1 20 88 758 1Lithuania 7743 0.6a 14 2572 1 15 138 388 1Luxembourg 63a 1 250 80 1 3276 42Malta 690 0a 1 72 6 1 8700 7Netherlands 22 964 7849 54 603 167 6Norway 13 085 11a 7883 29c 126 508 7c

Poland 77 720 1a 75 28 880 2 91 1 189 103 1Portugal 7a 17 14 968 12 18 189 289 5Romania 26 483 0.1a 32 22 197 3 53 877 068 1Slovakia 7357 1a 24 10 417 7 33 220 102 3Slovenia 4 5872 3 4 1752 6 4 89 110 1Spain 47 80 514 1 47 66 404 20 47 1 120 831 2Sweden 20 779 21 16 951 25 41 346 740 8Switzerland 12 21 575 51 12 19 076 47 36 153 519 14United Kingdom

129 137 278 23 144 95 256 42 146 2 151 145 14

Total 272 768 722 773 543 014 1270 15 063 693

Notes:a. Sources: Eurostat, Science, Technology and Innovation in Europe. Year 2006, with the

exception of Denmark, Latvia and Portugal (2005) and Norway (2003).b. Sources: Erawatch, Country Profiles; information on the labour market for researchers.

Year 2008, with the exception of France (2007).c. Source: OECD database.d. When no HEI is selected, the totals come from OECD database; in the other cases totals

refer to the sample of institutions in EUMIDA, which always includes at least 90% of the number of ISCED 5 students.




PATTERNS OF INTERNATIONALIZATION IN EUROPEAN COUNTRIES

Academic Staff

Among the medium- sized and large countries, Switzerland emerges, by far, as the more internationalized system, with 51 per cent of foreign academic staff; UK follows (23 per cent), then Germany, Denmark and Norway (11 per cent), Portugal and France (7 per cent) and Belgium (5 per cent), while the rest of the sample have marginal share of foreign academics (Figure 5.2).

The most internationalized countries clearly sum up many factors of attractiveness: as an example, Switzerland is characterized by abundant resources and high reputation, three main European languages are spoken and English is commonly used in academia. The United Kingdom is a prestigious system, where English, the main language of modern science is

0.00ES LT LV IT SI DE UK CH

0.20

0.40

0.80

0.60

1.00

For

eign

aca

dem

ics

Country capital

Note: Box plots identify the mean value (horizontal line), the interval between first and third quartile (shaded box), and outliers (circles, more than one standard deviation, and stars, more than two standard deviations).

Source: EUMIDA.

Figure 5.2 Share of foreign academic staff by country




spoken. German resources and the system overall reputation are similar to the UK, but its language is not as common as English. Norway and Denmark invest large amounts of resources, English is commonly used but apparently it is still to some extent peripheral. As to the bottom end, Mediterranean and Eastern countries have less resources, a somewhat less reputed scientific production, and English is not commonly used in academia.

Formalized competitive systems of recruitment are more common in countries with low levels of internationalization. This suggests that despite the underlying goal of objective assessment and meritocracy, this mecha-nism, characterized by rules and bureaucratic procedures, may create obstacles to the access of foreign applicants.

It is difficult to disentangle the contribution of each factor to attractive-ness as, quite interestingly, they often occur together: systems with many resources tend to speak English in academia, tend to be highly reputed and to adopt the discretionary recruitment approach.

EUMIDA microdata show that in the sample of higher education insti-tutions the variation in the level of internationalization is mostly linked to differences between countries (57 per cent) rather than within countries (43 per cent, ANOVA test, significance ,2.2e–16***). This suggests that country factors are more important in driving internationalization than factors related to individual higher education institutions. Countries with low levels of internationalization are also rather homogeneous across uni-versities, with the exception of Slovenia (including only four cases) and Italy (including, however, three outliers).3

Swiss universities all show very high shares of foreign academics. The universities in Germany and the UK display more variation. In Germany there are five small, specialized HEIs with much larger shares of internationals. In the UK, the London Business School (80 per cent) is a clear outlier, whereas the upper end of the distribution is occupied by high- standing and/or specialized institutions, such as the London School of Economics and Political Science, the School of Oriental and African Studies, Imperial College of Science, Technology and Medicine, the School of Pharmacy, the University of Cambridge, the University of Essex, the University of Oxford. The HEIs on the bottom end are mostly Fachhochschulen in Germany and university colleges in the UK.

Among the largest countries, there is no strong and significant asso-ciation (correlation tests) between internationalization and the size, age, research orientation and teaching intensity of the university. The only exception is the UK. Once a few outliers are excluded from the sample,4 it clearly emerges that larger and more research- intensive universities tend to be more internationalized (Figures 5.3 and Figure 5.4) whereas the teaching- intensive HEIs are less internationalized.




0.000 2.00 4.00 6.00 8.00 10.00

0.1000

0.3000

0.2000

0.5000

0.4000

0.6000

For

eign

aca

dem

ics

Size (1000 units)

R2 Quadratic = 0.171

Figure 5.3 UK: relationship between size (1000 units of staff scale) and internationalization

0.000.00 0.0500 0.1000 0.1500 0.2000 0.2500 0.3000

0.1000

0.3000

0.2000

0.5000

0.4000

0.6000

For

eign

aca

dem

ics

Research

R2 Quadratic = 0.653

Figure 5.4 UK: relationship between research intensity and internationalization




PhD students

The share of international PhDs tends to be larger than the share of foreign academic staff and undergraduate students. In fact, early career stages are the more mobile as many researchers will benefit from the inter-national experience to spur their career in the home country (Musselin, 2004).

Eastern countries show the lowest shares (2–3 per cent), then Italy and Finland (7–9 per cent), a mid- positioned group includes Spain, Germany, Austria, Ireland, Sweden (20–25 per cent); finally, a highly internation-alized group includes UK, Belgium and Switzerland (42–47 per cent) (Figure 5.5).

The most attractive countries for PhDs are roughly the same as for aca-demic staff and undergraduates: Switzerland, United Kingdom, Germany, plus Austria and Sweden. Many countries faced a growth of foreign PhDs in the last decade. Foreign PhDs are now numerous even in countries where foreign academics and students are rare. For instance, from 2000 to 2008 they grew 12 per cent to 20 per cent in Spain, 1 per cent to 7 per cent in Italy, 6 to 12 per cent in Portugal (source: Eurostat).

The internationalization of PhD students has a larger variability between countries (64 per cent), than within countries (36 per cent, ANOVA test,

0.00BG PL RO IT HU SK FI ES IE DE SE AT UK CH

0.20

0.40

0.80

0.60

1.00F

orei

gn P

hDs

Country

Figure 5.5 Share of foreign PhD students




significance ,2.2e–16***). This apparently confirms that country factors are very important in driving internationalization. There is an evident rela-tionship with investment in R&D, since all the top countries are wealthy and spend large resources on higher education and research. Flows of PhDs also show the importance of language and cultural patterns. While Switzerland mostly attracts European students, UK attracts worldwide, whereas France, Belgium, Spain and Portugal prominently attract from former colonies and countries sharing the same language (Table 5.6).

Universities of different age and size display similar levels of interna-tional PhDs. In several countries research- intensive universities are more internationalized, and an even stronger association (though negative) occurs with the teaching burden (Figure 5.6).

Table 5.6 Origin of PhD students in European countries

Total Own Country

Europe North America

Asia Africa South and

Central America

Oceania

United Kingdom

79 997 53% 18% 4% 18% 5% 2% 1%

Switzerland 18 149 54% 35% 1% 6% 2% 2% 0%France 68 995 61% 8% 1% 11% 16% 4% 0%Belgium 9768 69% 13% 0% 6% 9% 3% 0%Austria 17 280 74% 18% 0% 5% 1% 1% 0%Spain 66 533 77% 5% 0% 1% 1% 16% 0%Norway 6048 77% 12% 1% 6% 3% 1% 0%Sweden 18 775 82% 10% 0% 6% 1% 1% 0%Denmark 6055 84% 10% 1% 5% 0% 1% 0%Portugal 15 964 89% 2% 0% 1% 2% 6% 0%Czech Republic

24 334 90% 8% 0% 1% 0% 0% 0%

Finland 21 541 92% 5% 0% 2% 0% 0% 0%Hungary 7150 92% 6% 0% 1% 0% 0% 0%Italy 39 208 93% 3% 0% 2% 1% 1% 0%Slovak Republic

10 674 94% 5% 0% 0% 0% 0% 0%

Poland 31 809 97% 2% 0% 0% 0% 0% 0%

Note: The country of origin refers to the country of citizenship. No data available for Germany, Greece, the Netherlands.

Source: Authors’ elaboration, based on OECD data. Advanced research programme, non- citizen students of reporting country. Year 2008.




Students

The quality of the analysis of undergraduate internationalization is partly limited by the use of different statistical criteria, where some countries con-sider foreign students and others mobile students. Based on the available data the Eastern European, Mediterranean and geographically peripheral countries such as Finland and Ireland, show the lowest levels (2–5 per cent). A second distinct group is characterized by levels above 8 per cent, which include wealthy countries in Central Europe and the UK.

The share of foreign undergraduate students across European countries is more homogeneous than staff and PhDs5 and most variation occurs between HEIs in the same country (80 per cent, ANOVA test, significance ,2.2e–16***) (Figure 5.7).

Several outliers emerge from the graph. In most countries they are limited in number, and represented by HEIs with very peculiar geographi-cal position or discipline orientation. In Germany and the UK outliers are more common and include HEIs with peculiar discipline orientations (Table 5.7).

In Germany and UK international students are enrolled mainly in uni-versities with low teaching intensity (Figure 5.8).

0.000.00 20.00 40.00 60.00

0.2000

0.6000

0.4000

0.8000

1.0000

For

eign

PhD

s

Teaching

IT: R2 Quadratic = 0.122SE: R2 Quadratic = 0.198UK: R2 Quadratic = 0.233AT: R2 Quadratic = 0.209CH: R2 Quadratic = 0.158

ITSEUKATCHITSEUKATCH

country_code

Figure 5.6 Association between internationalization of PhDs and teaching intensity in a sample of countries




In conclusion, the most important factors associated with internationaliza-tion of staff and students are research intensity and (negatively) teaching intensity, but the strength and significance of this association change from one country to another.

THE ASSOCIATION BETWEEN INTERNATIONALIZATION AND RESEARCH QUALITY

The theoretical section pointed out arguments supporting the existence of a positive association between research quality and internationalization. In Figure 5.9, a sample of 140 universities is placed according to the share of foreign academic staff and the Leiden ranking indicator of research performance.6

Similarly, in Figure 5.10, a sample of 195 universities is used to repre-sent the relationship between the share of foreign PhD students (y) and research performance based on Leiden ranking (x).

In both cases better research performances are clearly associated with higher internationalization of academics (0.78***) and PhD (0.63***). For staff, positive correlations also occur at the national level, particularly in countries with high levels of internationalization: Germany (0.52***), UK (0.50***) and Switzerland (0.64). On the contrary, in the case of interna-tionalization of PhD students the within- country correlations are weak

0

01 LT

02 LV

03 PL

04 R

O05

SI06

ES07

BG08

EE09

IE10

SK11

IT12

GR13

PT14

FI

15 N

L16

MT17

HU18

SE19

DK20

AT21

BE22

DE23

CH24

CY25

UK26

LU

0.200

0.400

0.800

0.600

1.000

Inte

rnat

iona

l ISC

ED

5

Country

Figure 5.7 Share of foreign undergraduate students



158

Tabl

e 5.

7 H

EIs

out

liers

per

shar

e of

inte

rnat

iona

l und

ergr

adua

te st

uden

ts b

y co

untr

y

Cou

ntry

% In

tnl

Cou

ntry

M

ean

Tota

l St

aff

Inst

itutio

n N

ame

HE

I Fea

ture

s and

Cou

ntry

Issu

es

AT

839%

11L

aude

r Bus

ines

s Sch

ool

Smal

l spe

cial

ized

uni

vers

ity w

ith

in

tern

atio

nal o

rient

atio

nB

E66

10%

n.a.

Eva

ngel

ical

The

olog

ical

Fac

ulty

Spec

ializ

ed u

nive

rsity

BG

182%

410

Med

ical

Uni

vers

ity –

Ple

ven

Med

ical

scho

ols

1616

34M

edic

al u

nive

rsity

– S

ofia

CH

5811

%44

4U

nive

rsità

del

la S

vizz

era

italia

naL

ocat

ed in

the

Ital

ian

cant

on, n

eigh

bour

ing

with

Ital

y an

d ge

ogra

phic

ally

isol

ated

from

th

e re

st o

f th

e co

untr

yC

Y36

16%

51O

pen

Uni

vers

ity o

f C

ypru

sPe

culia

r int

erna

tiona

l orie

ntat

ion

2314

Cyp

rus I

nter

natio

nal I

nstit

ute

for t

he

E

nviro

nmen

t and

Pub

lic H

ealth

Stro

ngly

spec

ializ

ed in

stitu

tions

with

pecu

liar o

rient

atio

nD

E80

11%

353

Jaco

bs U

nive

rsity

Bre

men

(Priv

ate

HE

I)T

here

are

oth

er th

eolo

gica

l, ar

ts, m

usic

and

com

mer

cial

- orie

nted

HE

Is w

ith h

igh

leve

l of

inte

rnat

iona

lizat

ion

7746

Inte

rnat

iona

les H

ochs

chul

inst

itut Z

ittau

7244

ESC

P- E

AP

Eur

opäi

sche

Wirt

scha

ftsh

ochs

chul

e B

erlin

5839

Inte

rnat

iona

le U

nive

rsity

in G

erm

any,

Bru

chsa

l (Pr

ivat

e H

EI)

5734

Eva

ngel

ische

Fac

hhoc

hsch

ule

Nür

nber

g



159

5443

Eur

opea

n Sc

hool

of

Man

agem

ent a

nd

Te

chno

logy

, Ber

lin (P

rivat

e H

EI)

4968

Her

tie S

choo

l of

Gov

erna

nce

Ber

lin

(P

rivat

e H

EI)

4897

Staa

tlich

e H

ochs

chul

efür

Mus

ik T

ross

inge

nF

I28

5%0

Ala

nd U

nive

rsity

of

App

lied

Scie

nces

Loc

ated

in a

isla

nd in

fron

t of

Swed

en,

w

ith w

hich

the

sam

e la

ngua

ge is

shar

edG

R22

5%n.

a.Te

chni

cal U

nive

rsity

of

Cre

teSt

rong

inte

rnat

iona

l orie

ntat

ion,

part

ners

hips

and

con

solid

ated

pa

rtic

ipat

ion

in S

ocra

tes p

rogr

amH

U83 52

8%13

6 74C

entr

al E

urop

ean

Uni

vers

ity, B

udap

est

And

ráss

y G

yula

Uni

vers

ity, B

udap

est

Eng

lish

spea

king

uni

vers

ity, a

lso a

ccre

dite

d

in

the

US

and

fund

ed b

y G

eorg

e So

ros

Ger

man

spea

king

uni

vers

ityIT

644%

54In

stitu

te fo

r Adv

ance

d St

udie

sSp

ecia

lized

uni

vers

ity w

ith in

tern

atio

nal

or

ient

atio

n25

186

Uni

vers

ity fo

r For

eign

ers S

iena

U

nive

rsity

for f

orei

gn st

uden

ts in

Ital

y24

333

Uni

vers

ity fo

r For

eign

ers P

erug

iaU

nive

rsity

for f

orei

gn st

uden

ts in

Ital

yN

L43

6%19

0C

odar

ts, U

nive

rsity

for t

he A

rts

Spec

ializ

ed u

nive

rsity

3829

39M

aast

richt

Uni

vers

ityU

nive

rsity

on

the

bord

erPL

132%

901

Pom

eran

ian

Med

ical

Uni

vers

ity in

Szcz

ecin

All

the

med

ical

uni

vers

ities

in P

olan

d sh

ow

hi

gher

shar

es o

f in

tern

atio

nal s

tude

nts



160

Tabl

e 5.

7 (c

ontin

ued)

Cou

ntry

% In

tnl

Cou

ntry

M

ean

Tota

l St

aff

Inst

itutio

n N

ame

HE

I Fea

ture

s and

Cou

ntry

Issu

es

RO

182%

1280

Uni

vers

ity o

f M

edic

ine

and

Phar

mac

y

Iu

liuH

ateg

anu

of C

lujN

apoc

aM

edic

al u

nive

rsiti

es

1213

32G

rigor

e. T.

Pop

a U

nive

rsity

of

Med

icin

e

an

d Ph

arm

acy

Iasi

SE31

9%49

1B

leki

nge

Inst

itute

of

Tech

nolo

gyTe

chni

cal o

rient

ed u

nive

rsiti

es26

2836

Roy

al In

stitu

te o

f Te

chno

logy

SK20

4%N

anD

ubni

ca T

echn

olog

y In

stitu

te in

Dub

nica

nadV

áhom

Uni

vers

ity b

orde

ring

with

Cze

chR

epub

lic

UK

6716

%18

99L

ondo

n Sc

hool

of

Eco

nom

ics a

nd

Po

litic

al S

cien

ceH

ighl

y re

pute

d sp

ecia

lized

uni

vers

ities

in

bu

sines

s, m

edic

ine

and

arts

, and

loca

ted

in

Lon

don,

tend

to sh

ow v

ery

larg

e sh

are

of

inte

rnat

iona

l stu

dent

s67

714

Lon

don

Bus

ines

s Sch

ool

6222

7T

he U

nive

rsity

of

Buc

king

ham

4981

2L

ondo

n Sc

hool

of

Hyg

iene

and

Tro

pica

l

Med

icin

e47

1596

Cra

nfie

ld U

nive

rsity

4417

9R

oyal

Aca

dem

y of

Mus

ic43

165

Roy

al C

olle

ge o

f M

usic

4275

1T

he S

choo

l of

Orie

ntal

and

Afr

ican

Stu

dies

4028

9R

oyal

Col

lege

of

Art




0.000.00 10.00 20.00 30.00 40.00

0.40

0.20

0.60

0.80

Inte

rnat

iona

l ISC

ED

5

Teaching intensity

DE: R2 Linear = 0.215UK: R2 Linear = 0.170

country_codeDEUKDEUK

Note: Teaching intensity cut at 40 on the graph.

Figure 5.8 Internationalization and teaching intensity in UK and Germany

00 0.5 1 1.5 2

10%

20%

30%

40%

60%

50%

70%

Shar

e fo

reig

n re

sear

cher

s

Performance

UK

SI

IT

DE

ES

CZ

CH

Figure 5.9 Performance and share of foreign academic staff




and not significant. Apparently the value of the international experi-ence for a PhD is mostly related to the prestige and the infrastructure of the hosting country and only secondly to the prestige of the institution, whereas academics are more sensitive to differences in quality and reputa-tion of individual institutions.

CONCLUSIONS

This chapter aimed at describing the internationalization of researchers, postgraduate and graduate students of the European higher education systems against a set of country and institutional features. The results of the analysis clearly point out some major patterns, and suggest some ques-tions requiring further research.

Internationalization is heterogeneous across countries and HEIs. The most attractive countries invest large amount of resources in R&D, are high research performing, the recruitment procedures are flexible and the national language is diffused in Europe or worldwide.

The most internationalized HEIs are research intensive, with low teach-ing burden, whereas there are small differences related to size and age of the institutions. An interesting and strong association emerges between

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

Sha

re o

f int

erna

tiona

l IS

CE

D 6

stu

dent

s

Performance

AT

BE

CH

CZ

DE

EL

ES

FI

HU

IE

IT

PL

PT

SE

Figure 5.10 Performance and share of foreign PhD students




internationalization of research staff and research performance. The lack of longitudinal data do not allow us to disentangle the causal nexus between internationalization and quality; at this point, only the existence of several interrelated processes can be speculated on to determine that correlation. In the future, a permanent collection of data may improve our knowledge on the matter.

Nations and institutions are eager to attract international research-ers. The relevance of the process is witnessed, for instance, by the notable growth of international PhD programmes. However, it is uncertain whether this process will be beneficial. In fact, the growth of PhDs from less developed countries may highlight that salaries and academic careers are not more appealing to nationals. In this scenario, the flows of foreign PhDs would follow the rules and dynamics typical of the low skilled rather than the high skilled. They would fill the need for low- cost human power, with their quality level becoming an issue of secondary importance. Internationalization may be counterproductive in the absence of system-atic efforts to attract and retain qualified national academics (Salmi, 2011). Strategies of attraction should not be driven by short- term goals or merely exploit cultural and language strategic advantages, but they must also take into account scientific motivations, the quality and variety of human capital attracted.

NOTES

1. For instance, information on mobility, opportunities to study abroad, increased assist-ance, ‘Scientific Visa’ for ‘third country’ researchers and the ‘Charter and Code’ to improve the rights of researchers across Europe.

2. The mobile student status is maintained for as long as continued education at the same level of education lasts. All tertiary programmes are considered as the same level, thus a mobile ISCED 5 becoming a PhD will remain ‘mobile’.

3. The University of Bolzano- Bozen (a German- and Italian- speaking university near the Austrian border, 26 per cent; the theological free university San Pio V in Rome, 25 per cent; the Bocconi Business School in Milan, 15 per cent).

4. Namely the London Business School, with 80 per cent international staff; the Institute for Cancer Research and the London School of Hygiene and Tropical Medicine, with very high research intensity and that cannot be considered only HEIs in the strict sense.

5. Standard deviation is the lowest both considering the entire sample of HEIs (0.17 staff, 0.17 PhDs, 0.10 students) and the selected sample with HEIs possessing all the data (0.13 staff, 0.18 PhDs, 0.12 students).

6. Performance indicator – CPP/FCS, Leiden ranking 2008; size- independent, field- normalized average impact.


The internationalization of European higher education institutions

Documents