The German Language Worldwide: a New Data Set on Language Learning

1. Introduction

Proficiency in a language different from one’s mother tongue can be due to different motives and related to different benefits. Language learning can be an investment good (with language learning enhancing individual human capital) or a consumption good.¹ The latter can be related to personal motives, which comprise cultural interests, like the interest in the respective culture or a general interest in language learning, and motives related to the social environment, like learning the partner’s or a friend’s mother tongue. Language learning as an investment in human capital can refer to the domestic labour market if the foreign language is important vis-a-vis trading partners or customers, eases the communication within the firm, or more generally leads to a wage premium. It can also be related to better perspectives on a foreign labour market and thus be linked to migration.

While empirical evidence on the importance of the consumption motive is mostly missing, there is a literature about the effects on labour market outcomes. Empirical evidence related to the domestic labour market, however, is mixed. While in the USA, foreign language proficiency does not contribute to higher wages (Fry and Lowell 2003), especially English proficiency in non-English-speaking countries generates wage premia (Lang and Siniver 2009; Toomet 2011; Ginsburgh and Prieto-Rodriguez 2011). On the contrary, Stöhr (2015) finds evidence that a return for other foreign languages exists in Germany, but only in occupations related to trade. Isphording and Otten (2013) show significant wage premia for English, French, and German for immigrants to Spain and relate it to the shortage of language proficiency in the Spanish work force.²

Contrary to the mixed evidence for the domestic market, there is evidence of positive effects of migrants’ proficiency in the language of the destination country. Proficiency helps migrants to overcome barriers in social and economic integration and hence reduces costs of migration. Language proficiency improves labour market outcomes of migrants by increasing earnings (see Dustmann and van Soest 2001, 2002) and employment probability (Dustmann and Fabbri 2003). Furthermore, proficient migrants are more successful in social integration into destination countries, as the probability of intermarriage increases, and they are less likely to live in an ethnic enclave which complements their economic integration (Bleakley and Chin 2010).

These studies about the economic benefits of language proficiency have mostly relied on self-reported measures and have linked them to other individual-level socio-economic or work-related data. The process of language learning has so far been largely ignored. This article aims at closing this gap for the German language by presenting data about language learning for the time period 1965–2014. Data are from the annual reports of the Goethe-Institut (GI),³ a German association which promotes the study of German language and culture all over the world.⁴ The GI has institutes which, besides cultural events, offer language courses and widely recognized standardized exams.

Exceptions of papers with a focus on language learning are Egger and Toubal (2016) and Melitz and Toubal (2014) who consider the effect of common spoken languages and acquired language skills on bilateral trade flows. They use language data for one point in time, however, which does not allow studying variation across time. Furthermore, Aparicio Fenoll and Kuehn (2016) use compulsory language learning at school as a measure for language skills and relate it to migration flows within the European Union (EU). While their data vary across time, acquired language skills likely go beyond compulsory language learning.

The small number of studies on language learning is mostly due to the lack of data. For the German language, for example, there are reports about language learning in schools, universities, and at the GI which have been published every 5 years since the early 1990s by a group of institutions consisting of the German Academic Exchange Service (DAAD), the GI, the Federal Foreign Office (FFO), and the Central Agency for German Schools Abroad (ZfA) (Ständige Arbeitsgruppe Deutsch als Fremdsprache 2003, 2006; Netzwerk Deutsch 2010; Auswärtiges Amt 2015). While they provide qualitative insights, they are not suited for a quantitative analysis, as data are available only for selective periods, often based on estimates from experts and subject to frequent revisions.

Instead of focusing on the process of language learning, the relation between language skills and outcomes has often been captured by measures of linguistic distance between individuals’ mother tongues and the languages spoken in the countries with which they want to interact. Related to migration as a potential motive, some studies control for common language as determinant of migration flows (Pedersen et al. 2008; Mayda 2010; Grogger and Hanson 2011; Ortega and Peri 2013), while others include more accurate measures of linguistic distance. Belot and Hatton (2012) for example use measures based on the linguistic family tree, while Belot and Ederveen (2012) and Adserà and Pytliková (2015) estimate the effect of linguistic distance on international migration flows with two, more fine-grained indices developed by linguists: the Levenshtein Index based on phonetic similarity and the Dyen Index based on similarity of sample words in Indo-European languages. Chiswick and Miller (1998) measure linguistic distance by an approach based on language scores of native English speakers studying foreign languages which indicate difficulties of studying a particular language.

Similar considerations as for migration flows are applicable to international trade flows. The same measures for linguistic distance or a common language as in the context of migration have been used in empirical applications (Hutchinson 2005; Melitz 2008; Lohmann 2011; Isphording 2013).

All these distance-based indicators reflect two aspects of languages. First, they serve as proxies for the costs of acquiring language skills: languages linguistically more distant from the mother tongue are more difficult to learn. Secondly, linguistic distance is in part an explanation of cultural differences (Belot and Hatton 2012) and is related to long-term cultural and historical connections between countries. Not too surprisingly, a negative relationship between linguistic distance and migration flows as well as trade has been established.

Linguistic distance is however not something invariant as often assumed, at least not on an individual level. By acquiring a foreign language, individuals can reduce the barriers resulting from that distance. The learning costs are likely related to linguistic distance, but so might also be the benefits.

The data presented here provide a detailed overview of language learning worldwide over the past five decades for the German language. Beside the number of institutes and their geographic distribution across countries, that is the supply of language learning opportunities, the demand as documented by the numbers of course and exam participation in the institutes abroad is presented. In addition, data related to the institutes located in Germany complement the picture. These data allow understanding the development of language learning across time with a focus on similarities and differences across regions.

The rest of the article is structured as follows: Section 2 describes the history and the goals of the GI in more detail. Section 3 provides information about the three constructed data sets. Section 4 presents a detailed picture about language learning worldwide over the past five decades with a special focus on regional differences. Section 5 adds the development of learning the German language in institutes in Germany. Section 6 concludes.

2. The GI

The GI is a German association that promotes German language and culture. It is one of the main actors of the foreign cultural policy and therefore closely related to the German government, in particular to the FFO. The GI was founded in 1951 as a successor of the ‘Deutsche Akademie’ (German academy) to rebuild and renew the infrastructure for foreign cultural policy after the Second World War.

In 1965, the GI published an annual report for the first time. At that time, the number of institutes grew fast, which was also attributable to the integration of other cultural institutes of the FFO into the GI. While in the beginning the GI concentrated on language teaching, in 1961 cultural promotion became a second goal. At the end of the 1960s, the GI got more closely connected to the FFO to carry out its foreign cultural policy. During that period, foreign cultural policy evolved as a third component of foreign policy beside diplomacy and foreign-trade policy (Singer 2003).

In 1970, a new concept of foreign cultural policy was introduced which emphasized cultural cooperation and exchange, instead of cultural export. Initiating and supporting interactions between cultures became the main objective of foreign cultural policy. Within this concept the term ‘culture’ was defined more broadly, including science, education, and media. The FFO acted as a coordinator of different organizations which carried out different parts of foreign cultural policy, among them the GI (Auswärtiges Amt 1970; Schneider and Schiller 2000; Singer 2003).

A framework treaty between the GI and the FFO, signed in 1976, states the relationship between the two institutions. The treaty ensures the content-related autonomy of the GI. The current version of the treaty from 2004 clarifies explicitly the main duties of the GI: first, to promote the German language by providing courses, training for teachers, and scholarships for language learning. Secondly, to support international cultural cooperation by involving cultural actors of the destination countries in cultural events and exchanges. Thirdly, to inform about social, political, and cultural life in Germany by maintaining libraries with German books and media and by organizing cultural exchange with Germany (Goethe-Institute and Auswärtiges Amt 2004). While the GI is mainly funded by the German government, language courses are financed by course fees (Goethe-Institut 2014).

To fulfil these duties, the GI has institutes all over the world. In 2014, there were 149 institutes in 94 countries of which 126 institutes offered language services (see Figure 1). In addition, there were 12 institutes in Germany. There is no official guideline for location choices of new institutes. According to GI officials, openings and closings of institutes take place in consultation between the FFO and the GI. The process of opening new institutes starts with suggestions for locations mostly by the GI, the FFO, or members of the legislature. In a next step, the GI and the FFO discuss the locations on the basis of their general objectives taking into account legal, political, and social aspects of the potential host country. Furthermore, aspects related to the overall security in the host country and global developments play a role, as well as considerations about the larger region. Very often location decisions can be seen as reactions to changes in the political, social, or economic situation.

Over the past 50 years, the number of countries with at least one institute has continuously increased, while the number of institutes has fluctuated between 130 and 160 for most of the time. Figure 2 illustrates these developments for the period 1965–2014 and shows that language services are widely offered over the whole period.⁵

There might be self-selection of participants based on the following three characteristics: willingness or ability to pay, location, and age. Selection on willingness to pay could occur if the prices of courses at the GI differed significantly from the costs of other equally suitable learning options. Courses could be more expensive if one considers the GI as a premium provider of language courses because it is a semi-official German organization with a long tradition and a good reputation. Courses could also be less expensive because the majority of the GI’s funds come from the German government. Counterarguments for both lines of reasoning can, however, be found. Historical price data on language courses are not available. If one looks at current prices instead, no clear pattern emerges. In particular, the prices do not indicate that the GI is usually the most expensive provider in the market. At the same time, language courses are priced to be self-financing, that is not financed by government funding. It can therefore not be expected that the courses of the GI are cheaper than comparable courses offered by other providers.

Institutes are usually located in capitals and other major cities. The lack of institutes in rural areas is likely to lead to an under-representation of language learners from these areas among participants at the institutes. However, the bias need not be as large as one would initially expect: Institutes offer both extensive and intensive language courses. Extensive courses are based on weekly lessons and last for several months, but intensive courses are taught en-block. Participants of intensive courses do not necessarily have to live in the vicinity of the respective institute. They may also stay there for the duration of the course only. This holds in an analogous way for participation in exams.

Admittedly, language services offered by the GI are only one way for adults to acquire skills in the German language. Naturally, there are a large number of other language learning opportunities, including universities, private language schools, and Internet platforms. This multitude of alternative learning opportunities might give rise to additional concerns regarding the self-selection of language learners into courses offered by GI and possible changes across time. The language courses taught by the GI are mainly offered as traditional ‘offline’ forms of language learning. Since 2010, this has been complemented however by online and blended-learning courses, which combine traditional and online learning (Goethe-Institut 2011). At the other end of the spectrum are institutions which only offer online courses. The latter kind of courses may be more attractive to a younger generation of language students, which is more familiar with using the Internet in general. While this difference may lead to an over-representation of older participants among the participants in language courses at the GI, the advent of online language learning platforms in the late 2000s falls in the very last years of our data. Given the nature of our macro-level data, we can only speculate about potential changes in the composition of participants over time and regions and, more generally, in the role of the GI.⁶ The data allow, however, for a detailed picture of the overall development over a long time period and for a large number of countries.

3. Data sets

Since 1965, the GI has continuously published annual reports in which activities of each institute including data about language course and exam participation are reported. These annual reports are publicly available. We digitized this information to construct three data sets.⁷

While the first one focuses on the presence and type of activities of the GI in each country, the two other data sets contain data about the extent of language services abroad and in Germany. The following provides some more details in terms of years and countries as well as content of the three data sets.

The first data set contains information about the presence of the GI on the city level for the period from 1965 to 2014. For each city-year combination the data set provides data about the types of institutes, their offer of language services, as well as their opening and closing years. Over the analysed 50 years, the GI has been present in 272 cities in 109 countries.

The second data set contains information about language course and exam participation at institutes abroad. From the annual reports of the GI, we construct three variables for different time periods. First, from 1986 to 2014 the GI reported for each institute and year the number of participants in standardized exams (‘zentrale Prüfungen’) which are widely recognized, for example for language requirements in universities. These exams can be categorized in exams for children, adolescents, and adults, and for professional use, and they are further differentiated by the level of language skills. The exams can be taken by course participants, but this is not obligatory; they can also be taken by individuals who have not attended a language course at the GI. Only in very few years numbers were reported separately for each type of exam, while for most years only aggregate information was available. Hence, we only use aggregate numbers for exam participation per year and institute.

Furthermore, there is information on course participation by two variables per institute and year: the number of registrations in language courses for the period 1990–2014 and an indicator, which we construct, for sold course units for the periods 1972–1989 and 1997–2014.

As to registrations, courses are organized in course periods, mostly two periods (semesters) or three periods (trimesters) per year. While from 2000 onwards the annual number of registrations is reported in the annual reports, in the years 1990–1999 only average numbers of students per course period were published. For this period, we construct the number of registrations per year by multiplying the average number of students per course period with the number of course periods.

Concerning sold course units, information was reported in the year 2006 and from 2009 onwards. For the years 1972–1989 and the remainder of the years between 1997 and 2014, we construct the measure for sold course units as follows: sold course units = total number of lecture units * average course size, where the number of lecture units⁸ is the sum of the units all teachers taught at an institute within a year, and the average course size is the number of students divided by the number of courses.

Total course and exam participation at institutes worldwide reached their maximum for all three variables in 2014 with 287,630 exams, 229,702 registrations, and 17,113,040 sold course units.

The third data set provides detailed information about the annual number of registrations in language courses in Germany by nationality.⁹ Data are available for the period 1966–2014. In total, course participants with around 200 different nationalities studied the German language at the GI in Germany over the whole time span.

4. Language Learning across Time—Regions Abroad

The almost continuous expansion of the presence of the GI over the past five decades as shown in Figure 2 hides important heterogeneities across time and across regions. Furthermore, not only the presence of institutes, that is the supply of language learning opportunities, captures the spread of the German language worldwide but also the demand as documented in the numbers of registration and exam participation is an important indicator. In the following, the regional distribution of institutes and data about registrations in these institutes are presented including their evolution over time.

4.1 The worldwide presence of the GI

The expansion of the GI can be described more precisely if we look at the evolution across regions in Figure 3, which shows the number of countries with institutes and the number of institutes separately for different regions for the period 1965–2014.¹⁰ The assignment of the countries to regions according to the regional organization by the GI since 2008 can be found in Table A1.

Europe (inclusive Central Asia) is divided into five regions, which follow three different patterns (see Figure 3a and b). First, Southwest and Northwest Europe as the ‘old Europe’ had a relatively high number of institutes before 1990. This number decreased in the 1990s, especially in Northwest Europe. The number in Southwest Europe decreased only slightly before increasing to the old level again. The number of countries in these two regions was fairly stable throughout the whole period. Secondly, in Southeast Europe the number of institutes was quite stable during the whole period, while the number of countries began to increase in 1990. The latter reflects the expansion of the GI to (former) socialist countries, like Bulgaria, Romania, and the successor states of Yugoslavia. At the same time, some institutes in this region were closed, in particular in Greece. Thirdly, Central Eastern Europe and Eastern Europe and Central Asia reflect even more the expansion of the GI around and after the fall of the iron curtain. While before the end of the 1980s (almost) no institutes were located in these regions, the number of institutes and countries has increased sharply since then.

In Asia (without Central Asia and Middle East), there are three regions (see Figure 3c and d): East Asia, Southeast Asia (with New Zealand and Australia), and South Asia. The numbers of institutes and countries in East Asia have slightly increased over the whole period, especially after 2000. In South East Asia, the relatively high number of institutes was reduced slightly in the 1990s, while the number of countries was more or less stable. While at the beginning relatively many institutes were located in South Asia, this number decreased at the end of the 1980s, and so did the number of countries.

In North Africa (including Middle East) and Sub-Saharan Africa (see Figure 3e and f), the numbers of institutes and of countries almost coincide, which means, that the GI has no more than one institute in most countries. This number was quite stable until around 2005 when the GI started to expand in both regions.

Finally, America is divided into two regions (see Figure 3g and h), North/Central America, and South America. South America was a core region of the GI until the beginning of the 1990s with an average of at least two institutes per country. While there has been no change in the number of countries with institutes after 1966, the number of institutes has changed: it increased until 1980 and stayed constant until the mid of the 1990s. Afterwards, the GI began to close many institutes there. In North America, additionally to Canada, Mexico, and the USA, the GI has been only present in Costa Rica and Cuba for some years, but not in the remainder of Central America. The number of institutes in North America increased sharply after 1975. At the end of the 1990s, many institutes were closed.

Overall, the years from 1965 to 2014 can be divided roughly into three periods. Until the end of the 1980s, the focus of the GI was mainly on America and the ‘old Europe’, but also on Asia and Africa to a certain extent. With the change of the political landscape in the 1990s the GI expanded to former socialist countries, while simultaneously decreasing its large number of institutes in America and the ‘old Europe’. From the early 2000s onwards, the presence of the GI has been quite stable, only in Africa the GI has expanded to new countries.

It is possible to interpret this development in the light of changes to the public funds by the FFO (Schneider and Schiller 2000). In the beginning, public funds allowed the expansion of the GI. Subsequently, financial cuts and political realignment of foreign cultural policy led to closings (Singer 2003). This affected the GI mainly in the early 1990s when the GI opened many new institutes in Middle and Eastern Europe, and institutes in other regions had to close because of that expansion. At the end of the 1990s, further financial cuts led to more closings of institutes (cf. Figure 2).

4.2 Course and exam participation

While the overall number of institutes with language services has been largely constant, the numbers of course and exam participation show a different pattern with significant differences across regions. Figure 4 shows total numbers of exams, registrations, and sold course units as well as numbers for Europe with Central Asia and for the rest of the world (ROW) with some further disaggregation. For all three variables, we can see a similar trend (see Figure 4a to c): after a long period of quite stable numbers of course and exam participation, numbers have strongly increased since 2005.

In Figure 4d to f, the three variables for the European and Central Asian countries are disaggregated further according to the respective GI regions. Figure 4d shows that the largest part of the increase in exam participation after 2005 took place in Southeast Europe, but also in other regions (except Northwest Europe, and Central Eastern Europe) numbers for exam participation increased. For registrations, we see a sharp increase for Central Eastern Europe and Eastern Europe and Asia in Figure 4e, which results from the expansion of the GI in these regions in the 1990s as described in Section 4.1. Registration numbers for Southeast Europe and Northwest Europe remained quite constant in the whole period from 1990 to 2014. On the contrary, registrations in Southwest Europe decreased until 2008. When the recession set in, numbers began to increase again. Similar observations can be made for the years from 1997 onwards in Figure 4f which presents sold course units. Before 1990, only three of the regions offered language courses. In particular the number of sold course units was very high in Southwest Europe. The numbers in Southeast Europe increased from 1972 to 1989 continuously and remained constant in Northwest Europe.

The last row of Figure 4 disaggregates the variables on the continent level for the rest of the world. Figure 4g shows that the increase in exam participation after 2005 took mainly place in Asia and Africa with Middle East, while in America numbers only increased slightly. For course participation the picture is different. Registrations in America decreased in the 1990s and then stayed constant on a lower level. In Africa the number of registrations increased slightly, while in Asia the numbers increased sharply from 2000 onwards. Numbers for sold course units after 1997 again followed the same trends as can be seen in Figure 4i while in the period before 1990 the numbers of sold course units were quite stable in all three continents.

There has been a general expansion in total course and exam participation over the years, similar to the expansion of the GI to more countries (see Figure 2). One reason for the significant increases especially since 2005 might have been the language requirement introduced in 2007 in Germany for non-EU citizens who immigrate to reunite with their family. As the GI offers exams that are recognized for that purpose, the comparatively large increase in the number of exams could be a consequence of this requirement in non-EU countries. The exams can be taken without attending courses at the GI. It is therefore not a surprise that the numbers for course participation, that is registrations and sold course units, have increased comparatively less. Nevertheless, a slight increase can be seen for the registrations and sold course units as well, in particular for the rest of the world. As to the most recent increase in the number of exams, this is mainly attributable to a change of the structures of the exams in 2013 (Goethe-Institut 2014). Before that change, all parts (reading, writing, speaking,…) had to be taken together, while now each part constitutes a separate exam.

Figure 5 illustrates the number of registrations in each country in the year 2014.¹¹ This can be related to the number of institutes in the different countries as well as to their different sizes. In 2014, the average institute had 1963 registrations, and half of all institutes had 1500 or less registrations. On the other hand, the five largest institutes were spread all over the world: Bangkok (6800), Rome (6500), Moscow (6000), Madrid (5200), and Mexico City (5200).¹²

5. Language Learning across Time—in Germany

Opportunities to learn German do not only exist abroad. There is also the option to learn the German language in one of the institutes in Germany. Participants can either come explicitly for a course and return to their home country at the end of it; or they are migrants in Germany who want to improve their proficiency of the host country’s language.

Figure 6 plots the development of the aggregate registrations by continents. Figure 6a distinguishes between European, Non-European,¹³ and German students, who are mainly ethnic German repatriates¹⁴, and reports the aggregate of all students. The trend was similar for students from European and non-European countries and so was their relative importance over time. In the mid of the 1980s, the number of registrations by European and non-European students began to increase until a peak was reached in 1992, where 31,179 students from all over the world studied at German institutes. After 1992, language course participation decreased and stayed on a quite constant level until 2006. Since 2006, there has been a steep increase in language course participation of European and non-European students where the total number of registrations reached the maximum in 2013 with 35,468 students. In 2014, registrations by European students amounted to 13,459 and by non-European students to 20,397 from a total of about 200 countries.

In Figure 6b non-European regions are disaggregated by continents and the (former) Soviet Union. Throughout the whole period, students from Asia have constituted the largest group, while the smallest group has been students from Oceania. The number of registrations by Asian students has increased sharply since 2008 which has contributed significantly to the overall increase in course participation numbers of non-European students.

During the existence of the Soviet Union, numbers of registrations from that region were very low, but after its dissolution in 1990 the number of students increased sharply up to around 2000 registrations in 2003. Afterwards, a downward trend began, until course participation started to increase again in line with the common upward trend. At the end of the period, the number of registrations from the former Soviet Union had converged to the number of registrations by students from North America, which was relatively stable over the whole period from 1966 to 2014. The number of registrations by students from Africa and South America was of the same magnitude for most of the time. They also increased, however, following the common trend from 2006 onwards. In 2014, 8587 students came from Asia, 3311 from successor countries of the Soviet Union, 3121 from North America, 2785 from Africa, 2221 from South America, and 372 from Oceania.

The distribution of the number of registrations is clearly right-skewed. In 2014, around 83.5% of the countries had less than 250 registrations for students studying German at an institute in Germany. A closer look at the left side of the distribution shows that many countries had even much fewer registrations. While 56.2% of the countries had less than 50 students at the GI in Germany, the median of the distribution is 29 students and the first quartile is 4 students. On the other side of the distribution, a few countries contribute a large proportion of the total number of registrations at German institutes. While many of them are European,¹⁵ only one of the larger countries is South American (Brazil with 1182 registrations) and African (Libya with 1318 registrations). Also many students come from the four Asian countries Saudia Arabia (1509), China (1319), Japan (1161), and India (756). With Mexico (608) and the USA (2000), North America also contributes a large proportion of registrations. Also, another large group of students comes from Russia, as largest successor of the Soviet Union (2028). Figure 7 illustrates the number of registrations by nationalities in the year 2014 and shows the different importance of different nationalities for the study of German at institutes in Germany.¹⁶

6. Conclusion

This article presents a comprehensive overview of German language learning for more than 100 countries (including Germany) over a period of 50 years. Our new and unique data allow providing information about the supply of language learning opportunities, that is the number and the geographic distribution of institutes, and the demand in form of course and exam registrations. These data do not only show the development of language learning for the German language across time, they also underline common trends and heterogeneities across regions.

With respect to the supply of language learning opportunities, the years from 1965 to 2014 can be divided into three periods. Before 1980s, the GI was mostly present in America and the ‘old Europe’. In the aftermath of the fall of the iron curtain, the GI expanded to former socialist countries, while simultaneously decreasing its large number of institutes in America and the ‘old Europe’. From the early 2000s onwards, the presence of the GI has been quite stable, while only in Africa the GI expanded to new countries. Interestingly, while the number of countries with a least one institute continuously increased over the whole period, the overall number of institutes fluctuated for most of the time between 130 and 160 with peaks in the early 1970s, the early 1990s, and in the recent past.

As to the demand side, course and exam participation was relatively stable at institutes in Germany and abroad in the first decades. At institutes in Germany, participation numbers began to increase in the mid of the 1980s. Since 1990 or 1995, respectively, there has been a general expansion in total exam and course participation also at institutes outside Germany similar to the expansion of the GI to more countries—even though the number of institutes has not shown the same pattern. For institutes outside Europe, this has been mainly driven by Asia where course participation numbers started to rise around 1995, and numbers for exam participation have much grown since 2005. A steep increase of registrations by Asians at institutes in Germany followed around 2008. Similarly, registrations by participants from Africa have started to increase after 2005.

In Europe with Central Asia, most changes are driven by three regions: in Southwest Europe, course participation numbers decreased from 1990 onwards and only started to increase again around 2010—about at the same time when numbers for exam participation started to increase as well. The same pattern holds for registrations by Europeans at institutes in Germany. Registrations from Eastern Europe and Central Asia have shown an upward trend since 1990 without, however, a comparable increase in exam numbers. On the contrary, exam numbers for Southeast Europe have been rising from 1990 onwards, while numbers of course registrations have not changed much. Numbers of registrations by students from the former Soviet Union at institutes in Germany confirm this rising interest in the German language in particular from 2005 onwards.

These different patterns highlight the heterogeneity across time and even more so across regions. In addition, even for the same region, differences can be observed depending on which aspects of language learning (demand-side or supply-side) and on which specific forms of participation (course or exam participation) one is focusing. While we provided some possible explanations for some of the observed patterns, for example, the language requirement introduced in 2007 in Germany for non-EU citizens who immigrate to reunite with their family, more research is needed to fully understand the reasons behind the developments and to which extent they can be attributed to policy changes in Germany or abroad or to other changes of the institutional setting or the individual motives. Also global events, like the financial crisis, might have had an effect. At the same time, the data themselves can contribute to a better understanding of the economic and cultural relations between Germany and countries and regions worldwide and their development over the past five decades.

Funding

This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, UE 124/2-1)—270886786.

Footnotes

Acknowledgement

Helpful comments by Panu Poutvaara are gratefully acknowledged.

References

Appendix

Author notes