Systematizing internationalization policy in higher education : Towards a typology

The internationalization of higher education has become a strategic priority for governments because of the benefits expected to derive from it in the economic, political, socio-cultural and academic spheres. Nevertheless, a review of the academic literature on internationalization in higher education reveals that the concept of “internationalization” has been understood and applied in a variable manner. Considering that internationalization is a central theme in higher education research and practice, this state of affairs is problematic. In response to this situation, this article proposes the establishment of a conceptual typology for the analysis of national internationalization policies as a solution. In turn, such a heuristic device will ease knowledge accumulation, cross-case comparability and disciplined use of terms and procedures. The paper ends by dispelling common concerns related to classifications.


Introduction
The internationalization of higher education emerged as a key topic in research and policy debate in the 1990s (Matthews & Sidhu, 2005, Guruz, 2008) when the link between this process and its possible economic, political, socio-cultural and academic benefits became apparent (van der Wende, 2001).The latest Global Survey on the Internationalization of Higher Education revealed that internationalization "remains, or indeed grows in importance for higher education institutions" (Egron-Polack & Hudson, 2014, p.6). Internationalization has, thus, come to be seen as the "central motor of change" in higher education (ibid, p.5).Consequently, policies that promote this process have developed from ad-hoc to more structural measures which aim to have a deep impact on national higher education systems (Teichler, 2009;Brandenburg & de Wit, 2011).This development has been mirrored in the scholarly arena where the internationalization of higher education has assumed greater importance as a research topic (Kreber, 2009).For instance, a study on the specific research topics addressed in the articles of Higher Education Policy -one of the leading journals in higher education research -revealed that 'internationalization' has been the most studied research area in tertiary education policy (Huisman, 2008).Nevertheless, surveying the literature on internationalization reveals that "since the Second World War the concept has been understood and applied in a variable fashion" (Callan, 2000, p.16).This failure to sort out and clarify the concept of internationalization has negative implications for knowledge accumulation and cross-case comparability, hence, jeopardizing our ability to make systematic findings.With this state of affairs in mind, the present article pleads for a more nuanced understanding of 'internationalization', and the policies that promote the process, so that it becomes a fact-finding category with adequate discriminating power.This paper suggests a way forward by proposing the construction of a typology of internationalization that helps to clarify the concept and against which national policy efforts can be assessed.Indeed, such heuristic tools have already proven to be useful in comparative research because they allow scholars to map the empirical distribution of cases and, thus, encourage rigor and enhance transparency in case selection (Elman, 2005).In higher education research, classificatory frameworks have been developed for systematizing policies related to higher education regimes (Triventi, 2013), institutional internationalization (Childress, 2009;Bartell, 2003;Knight, 2003a), educational monitoring (Richards, 1988), international education hubs (Knight & Lee, 2014), international scholarship programs (Perna et al., 2014) and institutional differentiation (Jones, 2008;van Vught et al., 2005;Ziegele, 2013).The paper argues that constructing a similar typology of internationalization strategies of higher education on the national level would enhance the transparency of both academic endeavors and policy practice, and thus aid systematic knowledge accumulation and the design of policies that are wellequipped to deliver the intended outcomes.The paper proceeds as follows.The first section will review the literature on internationalization in higher education trying to point out the problematic issues in existing research.The second section argues for the benefits of constructing a typology that combines the conceptual and empirical levels of analysis.In particular, such a tool would help to systematize current understandings of internationalization in higher education, and would provide a frame in which past and future research can be located and interrelated.Finally, the paper reviews some of the most common points of criticism with regards to classifications and tries to provide possible solutions or counter arguments.

Systematizing knowledge on internationalization
"Getting to the heart of what internationalization means is not a simple matter" (van Gyn et al., 2009, p.27).Throughout the years, definitions of internationalization have evolved in various ways: from focusing on a set of specific activities to be carried out by universities (Arum & van de Water, 1992) to viewing it as a dynamic process to be integrated in the wider set of organizational activities of higher education institutions (Knight, 1993); from focusing on internationalization as an institutional endeavor to viewing it as a result of broader developments and synergies between various levels of authority with the power to steer internationalization (van der Wende, 2001); and, finally, from viewing internationalization as a limited function of a university's context (Soderqvist, 2002) to viewing it as a broad and eclectic mix of policies and processes that evolve on various scales (Knight, 2003b).Crucially, mastering the literature on internationalization is made difficult by: 1. the proliferation of different labels associated with the historical development of internationalization: "re-internationalization" (Teichler, 2009), "postinternationalization" (Brandenurg & de Wit, 2011), "globalized internationalization" (Jones & de Wit, 2014); 2. the coexistence of various scales of analysis: "internationalization versus Europeanization" (Teichler, 2002), "internationalization versus globalization" (Scott, 2000), "Europeanization versus internationalization versus globalization" (Callan, 2000); 3. the development of similar parallel processes across different scales: "demonopolization", "de-institutionalization", "de-nationalization" (Kehm, 2003); and 4. the plethora of policy measures that come under the 'banner concept of internationalization' (Callan, 2000), i.e. international student mobility, collaborative research, development of curricula and strategies for teaching and learning, the establishment of institutional networks (Altbach, Reisbeg & Rumbley, 2009).Another reason, and maybe the most important one for this lack of conceptual clarity, is the perpetual quest for generalization in the social sciences.The ubiquitous use of the concept of internationalization (Teichler, 2009) has resulted in what is known "as the Hegelian night in which all the cows look black and eventually the milkman is taken for a cow" (Satori, 1970, p.64).In other words, trying to obtain worldwide applicability has led to concept stretching which, in turn, has reduced the analytical purchase (Dunlop & Radaelli, 2013) of internationalization.This is problematic, not only for academic research, but also for public and institutional policy formulation, because it suggests that scholars talk past each other, making the accumulation of knowledge increasingly difficult (ibid.).If the trend continues and "internationalization is underscored as an argument for almost any higher education reform" (Teichler, 2009, p.94), the results will invariably be disappointing.Thus, this paper argues that internationalization policies should not be sold as a panacea for all the problems of higher education.Instead, it should be acknowledged that internationalization policy for higher education is no one-size-fits all approach.Rather, it is an umbrella term that comprises various measures designed to tackle specific issues (Callan, 2000).How policy makers understand the idea of internationalizing higher education systems differs from country to country (Graf, 2009;King, 2010;Matei & Iwinska, 2014) which goes to prove that internationalization is far from being a unidimensional concept (Kreber, 2009).All in all, it is safe to argue that there is no consensus on the scope of meaning of internationalization in higher education.Considering the importance attached to this process by universities (Egron-Polak & Hudson, 2014; European University Association, 2013), governments (Jones & de Wit, 2014;Kalvemark & van der Wende, 1997;Luijten-Lub, van der Wende & Huisman, 2005;Knight, 2004), supranational organizations and institutions (European Commission, 2013;Henard, Diamond & Roseveare, 2012), non-governmental actors (British Council, 2011), and scholars (Huisman, 2008), it is puzzling to understand why there is not more agreement on the conceptual borders of internationalization.This state of affair raises a salient question: how can the conceptual clarity of internationalization be improved so as to increase its explanatory purchase in the study of higher education?This paper posits that developing a typology of internationalization is a viable solution to this conundrum.The next sections are dedicated to outlining the benefits and pre-empting the criticism for building such a heuristic tool.

Reasons for building a typology of internationalization
It has been argued that classifications are central to the social sciences as "without classification, there could be no advanced conceptualization, reasoning, language, PRADEC Interdisciplinary Conference Proceedings © 2015 Prague Development Center data analysis or, for that matter, social science research" (Bailey, 1994, p. 1).Yet, skepticism about the value of classifications is commonly expressed.In what follows, this section will outline the advantages for building a typology of internationalization policy in higher education.But before doing this, the most basic characteristics of such a classification are examined.Generally, a typology is a classificatory device that reduces the complexity of empirical phenomena by arranging cases according to theoretically significant dimensions (Collier et al., 2012).In other words, cases that score similarly on the variables of interest are catalogued as belonging to the same type.Thus, good typologies must have two characteristics: exhaustivity (ability to allocate all existing cases to an appropriate type) and mutual exclusivity (each empirical case is a member of one type only) (Bailey, 1994).Directly following from this definition are a number of advantages that make typologies a viable, if not a necessary, tool "that helps to arrange the observable empirical `mess` in more ordered, transparent and therefore comprehensible manner" (van Kersberger & Vis, 2015, p. 116).These benefits include: comprehensive and exhaustive description of types of cases, parsimonious descriptions due to reduction in complexity, identification of similarities and differences between cases that can help with case selection for research, identification of an exhaustive list of dimensions or variables of theoretical interest, introduction of types as criteria for measurement, and versatility in that it provides the possibility to showcase either concepts, empirical cases, or both (Bailey, 1994).Thus, when dealing with complex multi-dimensional concepts, a viable solution is a taxonomical exercise which "unpacks concepts and plays a non-replaceable role in the process of thinking in that it decomposes mental compounds into orderly and manageable sets of component units" (Sartori, 1970(Sartori, , p.1038)).For young research fields, as is the case of internationalization in higher education, typologies have proven to be invaluable tools for dealing with epistemological problems (Arts & Gelissen, 2002).In other words, a typology can map the conceptual dimensions of internationalization policy in a systematic way (Callan, 2000) and establish the multiple dimensions of internationalization policy in higher education.Therefore, the main contribution of building such a typology, for both scholars and practitioners, would be to enhance transparency (Ziegele, 2013).On the one hand, the theoretical contribution would be to build the foundational blocks of an analytic heuristic, as an essential tool to increase conceptual clarity and set the boundaries within which the concept of internationalization can travel (Gerring & Barresi, 2003), increase analytic differentiation so as to capture the diverse forms of internationalization in higher education (Collier & Levitsky, 1997), encourage rigor in concept formation and measurement (Collier, LaPorte & Seawright, 2012), represent the baseline for comparing diversified policy approaches and conceptual understandings (van Vught et al., 2005), and, thus, improve the likelihood of generating cumulative knowledge (Elman, 2005).On the other hand, the practical contribution of such a typology could be to increase the transparency of higher education policies and processes for students, businesses, universities and policy makers (Ziegele, 2013), to ease consortia formation between universities and mutual agreements between states (van Vught et al., 2005), and to show how the typology can be put to work by mapping national policies of internationalization.

Possible problems of typologies
Despite the outlined advantages of typologies, there are a number of voices that are more critical of such classification tools.Bailey (1994) put together a comprehensive list of common concerns expressed about classifications: they provide no explanation as they are purely descriptive, they lead to reification of theoretical constructs, they are static rather than dynamic, it is difficult to select appropriate cases and dimensions of theoretical interest, and large classifications are unmanageable while small ones are too simplistic.Next, the paper will try to disperse some of these criticisms.
To begin with, possibly the most prevalent contention is that classifications are purely descriptive devices that have no explanatory function.However, rather than dismissing typologies for being non-explanatory, scholars should remember that descriptive inferences are foundational components for explanation and theorization (Bailey, 1994;Mahoney, 2004).As Sartori (1970) reminds us, proper conceptualization must precede theorization.As concepts define the range of phenomena that fall within their scope and those that do not, their precise meaning and their frame must be clear before engaging in explanation (Sartori, 1970;Gerring, 1999;Gerrring & Barresi, 2003).
Secondly, there is concern that classifications may treat theoretical constructs that have no empirical manifestation as being real empirical cases.However, this problem can be avoided if scholars make sure to specify whether the types in a typology are based on constructs or empirical cases (Bailey, 1994).Moreover, this paper argues for the construction of a typology that combines the conceptual and the empirical level of analysis, so as to make sure that this problem is avoided.Following this strategy, it is proposed that actual national strategies of internationalization should be analyzed according to the variables of interest, thus forming empirical clusters to which conceptual labels are subsequently attached.
Next, typologies are criticized for being static rather than dynamic, and that it is difficult to select cases and variables for building them.However, these problems are not specific to classifications.Statistical techniques, such as regression analysis, are also static research approaches and have similar problems to select dimensions of interest and to justify case selection (ibid).Also, the problem of being static may represent a bigger issue for those classifications that are purely empirical, rather than conceptual or a combination of conceptual and empirical (as is the one proposed in this paper).With regards to choosing appropriate variables for analysis, the development of a typology helps to rein in the various dimensions of internationalization as it "rests on the assumption that although definitions for a given term are, in principle infinite (…), most definitions juggle and re-juggle the same set of attributes" (Gerring & Barresi, 2003, p.207).
Finally, computerized cluster methods can mitigate the contention that large classifications are unmanageable, or that scholars have to be content with small typologies that are ultimately simplistic (Bailey, 1994).Cluster analysis is a tried and tested method for reducing complexity as it helps to group empirical cases into homogenous types using as a basis the similarity of those cases on the dimensions of theoretical interest.

PRADEC Interdisciplinary Conference
Proceedings © 2015 Prague Development Center

Conclusion
The internationalization of higher education remains a messy field, as only timid attempts were made to systematize the process (Kehm, 2003).Cleaning the vocabulary associated with the process of internationalization in higher education seems to be an insufficient exercise, since the concept is used in such a variable manner that it is questionable whether it can still serve as a valuable foundation for theory building.To put it another way, measuring internationalization and its impacts is difficult unless we know what we are actually measuring.The article argued that the construction of a typology that combines the empirical and conceptual levels of analysis can represent a viable way of dealing with this problem as it would enhance the transparency of both scholarly research and policy practice.In turn, this would aid systematic knowledge accumulation that could better inform the design of policies aimed at forwarding internationalization in higher education.Moreover, the paper tried to tone down some of the most common point of criticism with regards to such heuristic tools by showing how these problems can be resolved.