Destination cities of European exchange students

Abstract

The Erasmus programme is generally considered the flagship of intra-European exchange programmes in higher education, with more than three million participants since 1987. Whereas a number of studies investigated the determinants of student mobility decisions, no knowledge exists on the main destination cities of European exchange students. Our research note aims at filling this gap in the academic literature. Making use of a unique data-set from the European Commission containing micro-level data on the full population of Erasmus students for study purposes in 2012–2013 (n = 211,267), we provide a descriptive overview of the spatial distribution of Erasmus students at the city level. The results reveal that European exchange students are mainly attracted by capitals and second-tier metropolitan cities. Furthermore, the analysis reveals significant variation regarding the main region of origin of mobile students within most destination countries.

Keywords:

• Erasmus student mobility
• destination choices
• cities
• spatial distribution
• European Union

Introduction

Over the past few decades, international student mobility and migration significantly increased. Whereas in 1975, 0.8 million students were enrolled outside their country of citizenship, this number increased to 4.5 million in 2012 (OECD, 2014). Globally, Europe is the main destination of international students, hosting 48% of all international students (OECD, 2014). In contrast to other world regions, the most common form of student mobility in Europe is credit mobility, whereby students go abroad for a limited period of time in the framework of an exchange programme (Brooks & Waters, 2011). This is principally the result of the Erasmus programme, the largest European student exchange scheme for higher education students. Since its initiation in 1987, more than three million students have studied in another European country within this framework (European Commission, 2014a). Today, more than 4000 institutions from over 30 countries participate, and its annual budget exceeds 450 million euro (Souto Otero, Huisman, Beerkens, De Wit, & Vujić, 2013). In sum, these impressive numbers indicate that international students now form an intrinsic part of the ‘new European map of migration’ (King, 2002).

Despite the substantial number of students moving internationally as well as the importance attached to student mobility at a political level (see e.g. Brooks & Waters, 2011; Findlay, 2011; Van Mol, 2014), international student mobility has long been neglected by migration scholars (Findlay, King, Stam, & Ruiz-Gelices, 2006; King & Raghuram, 2013). Consequently, much remains to be done. One of the main lacuna in the emerging literature concerns information on destination cities of exchange students (Insch & Sun, 2013; Llewellyn-Smith & McCabe, 2008). This is partly due to data limitations. When providing contextual overviews of the Erasmus programme, educational practitioners, politicians and scholars generally rely on the annual statistics published on the website of the European Commission. These statistics cover the main home and host universities as well as Erasmus student flows between countries. In scholarly terms, they allow us to investigate how specific characteristics of higher education institutions and countries explain student mobility flows within Europe (e.g. Rodríguez González, Bustillo Mesanza, & Mariel, 2011). Data on the main destination cities of Erasmus students, however, are non-existent today.

Gaining insight into the spatial distribution of Erasmus student mobility at the city level, is relevant for advancing our understanding of the mobility decision process of exchange students. In the literature on international migration it is suggested that people are attracted by countries rather than by particular localities, as individuals generally move to localities where there are job opportunities (e.g. Geis, Uebelmesser, & Werding, 2013; Hofmann, 2015; Moral-Pajares & Jiménez-Jiménez, 2014; Palmer & Pytliková, 2015). Nevertheless, we argue this might not hold true for exchange students. After all, it has been amply demonstrated that Erasmus students are mainly motivated by experiential instead of academic goals (e.g. Findlay et al., 2006; Teichler, 2004; Van Hoof & Verbeeten, 2005; Van Mol & Timmerman, 2014). Analyses at the macro-level seem to confirm this trend, revealing that Erasmus student mobility is biased towards Mediterranean countries, which would be attractive because of their climate (Rodríguez González et al., 2011). We expect that besides the characteristics of host institutions and countries, students also consider characteristics of host cities when making mobility decisions. After all, the host city is the physical environment where the educational experience will take place (Cubillo, Sánchez, & Cerviño, 2006). Descriptive information on the main destination cities of Erasmus students might thus stimulate future research, opening possibilities to broaden existing frameworks explaining student mobility flows by institutional and country-level factors (e.g. Mazzarol & Soutar, 2002), by adding a crucial intermediate context in the decision-making process, namely the destination city. In addition, such descriptive information is helpful for researchers empirically investigating the dynamics of intra-European student exchanges, as it allows us to situate particular fieldwork settings within the broader European context. With this research note, we provide such a descriptive overview, focusing on the general attractiveness of destination cities of Erasmus students as well as the relative popularity of each city according to the students’ region of origin.

Data and methods

Our analysis is based on micro-level data from the European Commission, covering the full population of Erasmus students in the 2012–2013 academic year by both institution of destination and of origin (European Commission, 2014b). Our overview is restricted to student exchanges for study purposes (n = 211,267), as student mobility for work placements might follow different patterns. For each destination institution, we mapped the spatial location. Thereafter, we aggregated incoming student numbers for institutions located in the same city. Some of the localities were very small towns located in the immediate environment of (very) large cities. As we expected students going to these small locations are mainly attracted by these nearby larger cities instead of the small locality, we aggregated them with the larger cities if the distance between both localities was less than 10 km.

The relative popularity of each city according to students’ region of origin in each city is calculated as the highest ratio between the actual percentage of Erasmus students from a region of origin in that city and the expected percentage of Erasmus students from that region. This expected percentage is calculated as if all students by region of origin would be distributed equally over all cities. The expected distribution differs per country as Erasmus students are not eligible for an exchange in their own country of origin.

Findings

Figure 1 shows that students go to a great variety of destinations, 884 locations in total. The top-20 destinations of Erasmus students are (in descending order): Madrid (6697 students), Paris (6423), Barcelona (3801), Lisbon (3693), Valencia (3434), Istanbul (3395), Berlin (3230), Prague (2949), Vienna (2689), London (2616), Budapest (2522), Rome (2508), Milan (2388), Warsaw (2108), Seville (2079), Stockholm (1979), Granada (1960), Lyon (1928), Dublin (1901) and Copenhagen (1901). An overview of the 50 most popular cities can be consulted in Table 1.

Figure 1. Number of Erasmus students* by destination city, 2012–2013. Source: European Union Directorate-General for Education and Culture – Erasmus Mobility Statistics 2012–13.

Note: *Erasmus student exchanges for study purposes only (n = 211,267).

Table 1. Top-50 destination cities of European exchange students, 2012–2013.

Rank	City	Number	Rank	City	Number
1	Madrid, ES	6697	26	München, DE	1596
2	Paris, FR	6423	27	Krakow, PL	1553
3	Barcelona, ES	3801	28	Glasgow, UK	1336
4	Lisbon, PT	3693	29	Lille, FR	1235
5	Valencia, ES	3434	30	Amsterdam, NL	1213
6	Istanbul, TR	3395	31	Toulouse, FR	1197
7	Berlin, DE	3230	32	Göteborg, SE	1167
8	Prague, CZ	2949	33	Vilnius, LT	1145
9	Vienna, AT	2689	34	Wroclaw, PL	1143
10	London, UK	2616	35	Salamanca, ES	1130
11	Budapest, HU	2522	36	Brno, CZ	1128
12	Roma, IT	2508	37	Oslo, NO	1126
13	Milan, IT	2388	38	Ljubljana, SI	1094
14	Warsaw, PL	2108	39	Groningen, NL	1086
15	Sevilla, ES	2079	40	Bordeaux, FR	1056
16	Stockholm, SE	1979	41	Gent, BE	1031
17	Granada, ES	1960	42	Montpellier, FR	1007
18	Lyon, FR	1928	43	Coimbra, PT	995
19	Copenhagen, DK	1901	44	Turin, IT	948
20	Dublin, IE	1901	45	Leuven, BE	938
21	Helsinki, FI	1857	46	Firenze, IT	934
22	Brussels, BE	1791	47	Utrecht, NL	928
23	Aarhus, DK	1767	48	Manchester, UK	902
24	Bologna, IT	1666	49	Lund, SE	893
25	Porto, PT	1638	50	Grenoble, FR	883

Source: European Commission, authors’ own calculation.

Figure 2 shows only cities receiving at least 250 Erasmus students, and indicates the capitals and second-tier metropolitan regions (as defined in ESPON, 2013), as well as the cities hosting a world class higher education institution (defined as institutions included in the top-50 of the Times Higher Education Ranking and/or Shanghai Ranking for 2013). Furthermore, the figure indicates the relative popularity of each city according to students’ regions of origin. This figure clearly illustrates that students are mainly attracted by larger European cities, instead of being attracted by the best universities. Only 5 of the 20 most popular locations (Copenhagen, London, Munich, Paris and Stockholm) host a world-class university. This suggests that international and large cities might have a larger appeal to Erasmus students instead of the prestige of the academic institution they will attend. Furthermore, the figure indicates interesting patterns regarding the region of origin of European exchange students. Students from Northern Europe are overrepresented in most cities hosting a world-class institution for higher education, but are not the most overrepresented group anywhere in Eastern or Southern European cities – with the exception of Prague. Southern European students are shown to be mainly overrepresented in other Southern European cities, as well as in Polish and Lithuanian cities. By contrast, the overrepresentation of Eastern and Western European students is more spatially dispersed. Overall, the figure reveals considerable variation in the regions of origin of incoming students within most cities.

Figure 2. Number of Erasmus students* by destination city (receiving at least 250 Erasmus students), type of city** and most overrepresented region of origin***, 2012–2013****. Source: European Union Directorate-General for Education and Culture – Erasmus Mobility Statistics 2012–13.

Notes: *Erasmus student exchanges for study purposes only. **Capitals, second-tier metropolitan regions (as defined in ESPON, 2013), and cities hosting a world-class higher education institution (defined as ionstitutions included in the top-50 of the Times Higher Education Ranking and/or Shanghai Ranking for 2013). ***Regions of origin: Northern Europe (Denmark, Finland, Iceland, Norway and Sweden), Western Europe (Austria, Belgium, France, Germany, Ireland, Liechtenstein, Luxembourg, Netherlands, Switzerland and United Kingdom), Eastern Europe (Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia), and Southern Europe (Cyprus, Greece, Italy, Malta, Portugal, Spain and Turkey). ****The over-representation per destination city is calculated as the ratio between the actual percentage of Erasmus students from a region of origin in that city and the expected percentage of Erasmus students from that region of origin. The expected distribution of all Erasmus students over all destination cities is calculated as if all students by region of origin would be distributed equally over all cities. The expected distribution will differ per country since Erasmus students are not eligible for their own country of origin. The overall distribution for all destinations is 5.6% Northern, 41.5% Western, 16.8% Eastern and 36.1% Southern European students, but for instance for German destinations the expected distribution is 6.5% Northern, 32.2% Western, 19.4% Eastern and 41.8% Southern European students, due to the exclusion of students originating from Germany who are not eligible for destinations in Germany. In the case of, for example, Berlin there are 3.230 Erasmus students of which 12.8% Northern, 45.2% Western,12.7% Eastern and 29.3% Southern European students, which gives the highest overrepresentation ratio of 1.96 (12.8%/6.5%) for Northern European students.

Finally, we investigated bilateral flows between cities. The analysis reveals an impressive number of 34,747 different bilateral flows. As can be observed in Table 2, the 20 most prominent flows are all directed towards capital cities, with the exception of the student mobility flow between Barcelona and Milan, two major second-tier cities. Interestingly, however, these 20 most populated flows only cover 2.9% of all Erasmus students in the 2012–2013 academic year. When we compare this number with the relative share of the most popular destination cities, we notice that 28% of all Erasmus students head towards the 20 most popular destinations, and 43.9% to the top-50. So whereas exchange students appear to predominantly cluster in European capitals and second-tier cities, the results suggest there is substantial variety in terms of cities of origin.

Table 2. Top-20 largest flows of Erasmus students between cities, 2012–2013.

#	Connection			Students		Total
				→	←	↔
1	Paris	–	Madrid	395	362	757
2	Paris	–	London	263	221	484
3	Madrid	–	Rome	241	213	454
4	Paris	–	Berlin	234	157	391
5	Milan	–	Madrid	223	147	370
6	Rome	–	Paris	233	125	358
7	Madrid	–	Lisbon	167	121	288
8	Milan	–	Paris	193	84	277
9	Madrid	–	Brussels	142	123	265
10	Barcelona	–	Paris	160	104	264
11	Milan	–	Barcelona	140	123	263
12	Istanbul	–	Paris	145	107	252
13	Milan	–	Lisbon	140	97	237
14	Paris	–	Dublin	144	70	214
15	Rome	–	Barcelona	114	100	214
16	Vienna	–	Paris	142	69	211
17	Istanbul	–	Berlin	105	100	205
18	Madrid	–	Munich	120	85	205
19	London	–	Madrid	102	100	202
20	Lisbon	–	Barcelona	124	78	202

Source: European Union Directorate-General for Education and Culture – Erasmus Mobility Statistics 2012–13.

Discussion

Although destination city characteristics probably play a crucial factor for explaining student mobility patterns within Europe, they have been largely neglected in the academic literature. With this research note, we aimed to take a first step in improving our understanding on the (uneven) distribution of Erasmus students across destination cities. Based on a unique data-set covering the full population of Erasmus students in 2012–2013, we showed that European exchange students head to a wide diversity of destinations. Nevertheless, capitals and large metropolitan regions are clearly more attractive than smaller localities, as a significant share of Erasmus exchange students head towards European capitals or second-tier metropolitan cities. However, when considering the major bilateral flows, it can be observed that only about 3% of Erasmus students move within the 20 most prominent flows, which cover 10 European capitals, as well as Barcelona and Milan. This suggests intra-European credit student mobility is characterised by a pyramid-shape, whereby students move from a wide variety of origins towards a select number of main (large) destination cities. Furthermore, we revealed intriguing patterns regarding the zones of origin of exchange students across and within destination countries. The results suggest, for example, that the status of higher education institutions might be most important for Northern European students, as these students are overrepresented in cities hosting a world-class institution for higher education. In addition, students from Southern Europe seem to mainly move within their own region, as well as towards Eastern Europe. This pattern might be related to similarities between localities in terms of costs of living, culture and/or climate.

In sum, although presented results in this research note are descriptive, they show it is imperative to take the city level into account in future empirical research into international student mobility. In this research note, for example, we are not yet able to grasp the relative share of Erasmus students on the total student population in each city, as this requires compiling additional statistical information on the total student population in the large number of destination cities covered by the database. Nevertheless, such analysis would be highly relevant for grasping which cities dispose of the most diversified student populations (for an example on degree mobile students in the Netherlands, see Pellenbarg & Van Steen 2015), which on its turn may significantly affect student’s local experiences. Future studies should also investigate which specific characteristics make some cities more attractive than others, taking the students’ country of origin into account. Possible foci are cities’ labour market characteristics, as well as the specific amenities cities provide in terms of, for example, transport infrastructure, costs of living, social cohesion, recreation areas and nightlife (for an example on internal labour migration within Germany, see Buch, Hamann, Niebuhr, & Rossen, 2014).

From a theoretical viewpoint, it should be noted that future in-depth studies on the spatialities of student mobility might yield great potential in terms of confirming, falsifying and/or advancing classical and contemporary migration theories. When comparing the described properties of Erasmus student mobility flows, for example, with classical migration theories such as Ravenstein’s migration laws (Ravenstein, 1885) particular similarities can be detected. For example, most Erasmus students are female (European Commission, 2014a), each mobility from one locality to another has a compensating counter-current, and Erasmus students mainly move to larger cities. Nevertheless, much more can be done with the data-set at hand. A more detailed analysis would allow, for example, exploring whether students are also more likely to move short distances, and whether those from large towns are proportionally less mobile compared to students from smaller towns. Furthermore, in-depth analyses of student mobility spatialities can also inform contemporary migration theories. For example, the interplay between talent workers, including students, and cities is becoming increasingly important for remaining competitive among global knowledge economies (Florida, 2002). After all, ‘a large and constant supply of talented people is required’ for enhancing a high rate of innovation (Hansen & Niedomysl, 2009, 192). Future studies can investigate, for example, whether international students indeed move to the most innovative cities by incorporating indicators on the innovativeness of destination cities.

In conclusion, more developed insights into the decision-making process of exchange students and the importance of cities will not only feed academic research and theory-building, but will also be very informative for educational practitioners, local administrations and policy-makers.

Disclosure statement

No potential conflict of interest was reported by the authors.