International organisations in vocational education and training: a literature review

ABSTRACT

Although international organisations (IOs) are recognised as important actors in education, their involvement in vocational education and training (VET) remains underexamined. This article reviews the current state of research on IOs in VET, focusing on the ILO, OECD, UNESCO, and the World Bank, through an integrative literature review of 174 publications. The analysis identifies three roles attributed to IOs, highlighting crucial research gaps for future investigation. First, IOs are seen as important knowledge producers on VET, demonstrated by the implicit use of IO data in academic research. However, literature that explicitly analyses how IOs produce VET knowledge is largely missing. Second, IOs are described as organisations capable of formulating specific VET agendas. While there has been a particular interest in the World Bank’s views on VET, research that analyses and compares how other IOs develop their ideas on VET remains rare. Third, IOs are recognised as actors of VET policy transfer. Although existing research highlights the complexity of international and local factors in IOs’ attempts to influence national VET policies, more comparative case studies are needed to analyse concrete IO engagements. Addressing these gaps would contribute to a research agenda that takes into account the multifaceted agency of IOs in VET.

KEYWORDS:

• International organisations
• vocational education and training
• literature review
• OECD
• world bank
• UNESCO

Introduction

International organisations (IOs) are recognised as important actors in education, especially when it comes to higher education (e.g. Bassett and Maldonado-Maldonado 2010), lifelong learning (e.g. Jakobi 2009), or the construction of international large-scale assessments (e.g. Addey et al. 2017). Some aspects of the work of IOs on education have become prominent research areas of their own, such as the Organisation for Economic Co-operation and Development’s (OECD) Programme for International Student Assessment (PISA) and its ensuing effects on national education policies (e.g. Meyer and Benavot 2013). However, IOs have not received similar widespread attention when it comes to their activities in vocational education and training (VET). This neglect is puzzling given that IOs have been active actors in VET for more than 50 years, focusing, for example, on providing policy recommendations for VET (UNESCO, and ILO 2003), developing VET capacities in low- and middle-income countries (World Bank 2021), or reviewing national VET systems (OECD 2015). Although recent years indicate a rising academic interest in IOs in VET (e.g. Valiente 2014, Preckler Galguera 2018; Vanderhoven 2023), further empirical and theoretical analysis is required to better understand how IOs engage with VET and consider their influence on national VET policies and systems.

To contribute to such a future research agenda, this article reviews and synthesises the current state of research on IOs as actors in VET, considering literature from multiple disciplines, including VET research, education, political science, history, and others. Following Torraco (2005, 2016), I conduct an integrative literature review, which ‘reviews, critiques, and synthesizes representative literature on a topic in an integrated way such that new frameworks and perspectives on the topic are generated’ (Torraco 2016, 1). By analysing how IOs have so far been researched as actors in VET, this literature review identifies key research gaps and new perspectives for future research on IOs in VET.

IOs can be defined as formal intergovernmental organisations that are established by an international treaty between three or more states and that have a permanent organisational bureaucracy (see also Niemann, Martens, and Kaasch 2021, 6). This definition clearly distinguishes IOs from international non-governmental organisations, informal international institutions or transgovernmental networks that include both public and private actors. While these actors may also be relevant to VET, IOs understood as formal intergovernmental organisations are a particularly important type of actor in shaping and influencing global and national education policy (Martens and Windzio 2022).

In conceptualising IOs as actors, this article follows a constructivist approach to IOs. A constructivist perspective emphasises that IOs are not only instruments of their member states (as realist theories claim) or arenas for state cooperation (as institutionalist theories argue), but also exhibit agency of their own (Martin and Simmons 2013). Although IOs are constituted and funded by states, constructivist research in international relations has shown that IOs have a degree of autonomy from their member states by way of their organisational bureaucracies that allow them to act as autonomous actors in their own right (Barnett and Finnemore 1999, 2004). This autonomy has been argued to reside in the legitimacy of a legal-rational type of authority that IOs embody as bureaucracies, as well as in their control over technical expertise and information (Barnett and Finnemore 1999, 707). Scholarship in international relations has researched the multifaceted ways in which IOs engage as actors in world politics, emphasising, for example, their role as norm teachers (Finnemore 1993), as standard-setters (Broome and Quirk 2015) or as constructors of policy problems (Hülsse 2007). An interest in analysing IOs as actors is also visible in the field of education, where scholars have described IOs, for instance, as nodes for policy diffusion (Jakobi 2012) or as discursive forces, global networkers and coordinators (Shahjahan 2012).

This literature review extends the analysis of IOs as actors in world politics and education to the field of VET. By examining the ‘roles’ attributed to IOs in VET, it analyses what the academic literature perceives as relevant expressions of IOs’ agency. Thus, it provides a more comprehensive picture of IOs as VET actors than existing research. So far, the academic literature has rarely analysed IOs in VET as central research objects in their own right. Important exceptions are provided by Tikly (2013), Niemann (2022) and Vanderhoven (2023), who have taken up a comparative perspective on IOs’ views on VET. However, even these works can only capture a particular aspect of IOs as VET actors. The distinct value of this literature review lies in providing a comprehensive synthesis of the different actor roles that have been explicitly and implicitly attributed to IOs in VET across a large number of publications. By analysing both explicit and implicit role attributions, it becomes possible to consider how IOs are discussed in the literature even when they are not the central focus of research. The resulting synthesis of roles not only condensates existing research, but also highlights major research gaps and can thus guide further research on IOs as VET actors.

This literature review considers IOs as a general type of actor, but also focuses on the four dominant IOs in VET: The OECD, the International Labour Organization (ILO), the United Nations Educational, Scientific and Cultural Organisation (UNESCO), and the World Bank. These organisations have been selected because of their significant historical and geographical involvement in VET. First, they have a long history of engagement in VET, dating back to the 1960s (World Bank), the late-1940s (OECD,¹ UNESCO), and even to 1919 (ILO). Although VET did not always play a major role within these IOs during this time (the World Bank, for example, significantly decreased its funding for VET in the 1980s), the historical preoccupation with VET sets these organisations apart from other IOs. Second, these four organisations are the major education IOs that operate on the global level (Niemann, Martens, and Kaasch 2021).² While regional IOs such as the Asian Development Bank or the African Development Bank, and supranational organisations like the European Union (EU) also engage in VET, globally operating IOs hold particular importance due to their comprehensive geographical scope and potential influence. Moreover, while the impact of the EU on the VET policies of its member states has turned into research agenda of its own (see, for example, Scheuch et al. 2021), the same cannot be said for IOs at the global level.

This literature review focuses on VET, defined as ‘[p]ost-compulsory education and training, excluding degree and higher level programmes delivered by higher education institutions, which provides people with occupational or work-related knowledge and skills’ (MacKenzie and Polvere 2009, 73). Some definitions of VET make a closer distinction between initial and continuing VET, with the former focusing on VET that takes place before people enter the labour market, and the latter encompassing various types of skills development after initial VET (Cedefop 2014, 117, 51; Field et al. 2009, 18). For the purposes of this literature review, the main focus is on initial VET, as this aspect characterises VET as a distinct educational sector, distinguishing it from compulsory and higher education. Literature that focuses mainly on the role of IOs in related but different fields, such as lifelong learning or adult education will only be considered if it directly relates to VET.

The first section of this article describes how I collected and analysed the literature for review. The second section presents the synthesised results of the analysis. I argue that the existing literature attributes three main roles to IOs in VET and highlight crucial research gaps for further investigation in relation to each of these roles. First, IOs are seen as important knowledge producers for VET, demonstrated by the implicit use of IO data and knowledge in academic research. In contrast, research that explicitly analyses how IOs produce VET knowledge is largely missing. Second, IOs are described as organisations with specific VET agendas. While there is a particular interest in the World Bank’s VET agenda, research that analyses and compares how IOs develop their VET ideas remains rare. Third, IOs are recognised as actors of VET policy transfer that influence national VET systems and policies. While existing research highlights the complexity of international and local factors of VET policy change, it should be complemented by more case studies focusing on the mechanisms through which IOs engage in specific countries. To address these research gaps, the conclusion of this article provides further direction for future research on IOs as actors in VET.

Data selection and analysis

To select the literature for review, I researched three databases (Web of Science, Scopus, and Google Scholar) with a set of keywords in different combinations (‘international organ*’, ‘vocational’, ‘VET’, ‘TVET’, ‘UNESCO’, ‘ILO’, ‘OECD’, and ‘World Bank’).³ These three databases were selected due to their complementary coverage of literature. While the Web of Science Core Collection covers the period from 1990 onwards, content on Scopus extends in principle back to 1788. Both databases have policies guiding their content selection processes to ensure the inclusion of reliable sources. Searches in both databases focus on keywords appearing in the title or abstract of a publication. While Google Scholar also considers a large time frame, it conducts a full text search, thus having a particular extensive coverage.⁴ However, compared to Web of Science and Scopus, its criteria for inclusion are not transparent. In addition, I used a snowball sampling strategy to identify relevant cited literature in previously selected texts and, in the case of handbooks or edited volumes, surveyed adjacent chapters for their possible relevance. This snowball sampling method allowed for a more targeted approach to literature selection, enabling the inclusion of relevant publications that did not appear in any of the database searches.

To be included in the final text corpus, the literature had to meet the following selection criteria:

• Publications must be written in English.

• Publications must be books, book chapters, or journal articles.

• Publications must make a connection between IOs and VET. Publications are not included in the corpus if they contain separate keywords on IOs and VET without relating the two to each other.

Using these search strategies and criteria, 174 publications were included in the final text corpus with 80 from Web of Science and Scopus, 47 from Google Scholar, 46 from snowball sampling, and one article selected manually.⁵ The corpus includes literature from eight disciplines, including VET research (108), education research (46), political science (8), history (6), sociology (2), economics (2), development studies (1), and life course research (1) (see Appendix).⁶ Three texts were published between the years 1971 and 1988, nine in the 1990s, 42 in the 2000s, 72 in the 2010s, and 48 since 2020, indicating a growing academic interest in IOs in VET. The text corpus included 11 books, 80 book chapters, and 83 journal articles. 23 publications were written by or in collaboration with then IO staff members, demonstrating the interest of IOs to publish in academic outlets (see also section 3.1 of this article). Publications also varied with regard to their type, ranging from case studies and presentations of specific research projects to comparative meta-analyses.

To analyse the text corpus, each document was entered into the qualitative text analysis programme MAXQDA and coded according to two guiding questions, derived from the constructivist perspective on IOs outlined in the introduction. The first question was ‘What roles are attributed to IOs in VET?’ This question focuses directly on the various possible dimensions of IOs as VET actors. The second question was ‘How are IOs in VET researched?’ This question considers the different theoretical and methodological choices in studying IOs as VET actors. Answers to these questions were coded following a two-coding-cycle-model (Linneberg and Korsgaard 2019). In the first coding cycle, each text was approached with the two guiding questions, which allowed for the generation of preliminary inductive codes as answers to the questions. The second coding cycle explored patterns among the codes from the first cycle and combined them into larger categories. This made it possible to refine the first cycle codes and use the newly established categories in the further coding process. This second coding cycle was conducted after a significant number of documents (about half of the text corpus) had been coded in the first coding cycle. Codes focused on sentences and paragraphs, thus analysing the variety of document types at the textual level. Relevant information on the context of the documents was taken into account during the interpretation of the results. A coding example is presented in Table 1. In addition, the text corpus was also manually filtered for specific characteristics, such as studies on the influence of different IOs in specific contexts.

Table 1. Coding example.

Questions	(1) What roles are attributed to IOs in VET? (2) How are IOs in VET researched?
Examples	Example 1: ‘This paper compares within-country programmes of initial vocational education and training (IVET) in Austria, the Czech Republic and Germany and their outcomes. Specifically, it aims to analyse and compare the labour market success of graduates of different tracks at the ISCED 3 level in both early and later stages of their careers. The comparison is based on the analysis of PIAAC 2013 OECD study data’ (Hoidn and Šťastný 2021, 1). Example 2: ‘Within the OECD, a “Group of National Experts” guides VET research on vocational education and training. This group reports to the OECD Education Policy Committee and the Centre for Educational Research and Innovation Governing Board and undertook the OECD project Learning for Jobs: OECD Reviews of Vocational Education and Training (hereafter Learning for Jobs) between 2007 and 2010’ (Legrand and Vas 2014, 240, original emphasis).
Explanation of coding	Both examples have been coded under the category of ‘IOs as knowledge producers’ as a response to question 1. Since the first example attributes the role of IOs as knowledge producers in an implicit way by using IO data as research evidence, it was sub-coded as ‘implicit’ in response to question 2. The second example explicitly describes knowledge production by IOs and was thus coded as ‘explicit.’

This coding process enabled a systematic analysis, leading to a synthesis of the existing literature on the roles of IOs in VET. A major component of an integrative literature review, ‘synthesis’ is defined by Torraco as the process that ‘weaves […] streams of research together to focus on core issues rather than merely reporting previous literature’ (Torraco 2005, 362). By offering a new conceptual classification of existing research and drawing out crucial research gaps, this study gives direction for developing a future research agenda on IOs as actors in VET.

Results

The analysis identified three main roles that are implicitly or explicitly attributed to IOs in VET in the academic literature. First, IOs are seen as knowledge producers on VET, a role that is most often implicitly acknowledged by using VET data and knowledge produced by IOs as evidence in academic research. Second, IOs are described as organisations that develop their own specific VET agendas. Finally, IOs are regarded as actors of VET policy transfer that can potentially influence national VET systems and policies. These three roles are presented in the following sections, which summarise and discuss the main findings from the literature and highlight key research gaps.

IOs as knowledge producers on VET

The first role attributed to IOs in the reviewed literature concerns their function as knowledge producers on VET. This role is related to more general research findings in the field of international relations, which have argued that expertise is a key source of authority for IOs (Barnett and Finnemore 1999, 2004). Because IOs are created to address complex tasks of international scope that cannot be solved by the actions of individual states alone, IOs acquire and produce specialised knowledge in particular issue areas. In general, knowledge production in IOs takes multiple forms, ranging from data collection and the generation of datasets that are used by researchers and governments alike to concrete research outputs in the form of self-published reports or studies in academic journals. The success of IOs as knowledge producers is illustrated in a bibliometric analysis by Zapp (2018), who finds that the number of scientific publications published by IOs has steadily increased since the 1980s and especially since the early 2000s. In terms of the sheer volume of scientific output, IOs often rival or even surpass universities and research institutions.⁷ Thus, the generation of specialised expert knowledge is a key component of the work of IOs in general.

In the context of VET, the role of IOs as knowledge producers is recognised in both implicit and explicit ways in the reviewed literature. In most cases, this role is implicitly acknowledged as researchers use data and knowledge produced by IOs as evidence in their research. The OECD is a major reference point in this respect. For example, data from the OECD’s Programme for the International Assessment of Adult Competencies is used by both Choi et al. (2019) and Hoidn and Šťastný (2021) to analyse the effect of VET on individuals’ labour market success in different countries. Other OECD data sources used in the analysed literature include its Teaching and Learning International Survey (Russo, Serafini, and Ranieri 2019), the annual Education at a Glance reports (Maclean and Pavlova 2013; Sweet 2013; Zancajo and Valiente 2018), or its policy reviews of VET and adult learning (Winch 2013). In addition to the data collected by the OECD, it is also referenced when it comes to defining key terms in VET, such as ‘career guidance’ (Watts 2013) or ‘VET’ itself (Höckel 2012).

UNESCO is another important provider of knowledge on VET with one of its most influential frameworks, the International Standard Classification of Education, being widely used in the literature to frame research questions and data selection. In addition, statistical data on VET enrolment produced by the UNESCO Institute for Statistics is drawn upon by multiple publications (Cardoso 2009; Maclean and Pavlova 2013; Rodgers and Boyer 2006).

Referencing, citing, and using data and knowledge by IOs not only provides important and often hard-to-find evidence for academic researchers, especially in contexts where national data is of poor quality or non-existent. It also reinforces the role of IOs as credible producers of specialised knowledge on VET. Indeed, IOs themselves are adept at disseminating their research in academic outlets. 23 publications in the reviewed literature were written by or in collaboration with then IO staff members, often using IO datasets. For instance, Gill et al. (1999), then staff members at the World Bank and at the ILO, draw on a joint World Bank-ILO study to analyse obstacles on VET policy reforms in 19 countries. Singh (2017), from the UNESCO Institute for Lifelong Learning (UIL), analyses ‘alternative transitions’ from VET to further learning, drawing on 33 case studies compiled by the UIL to analyse whether national qualifications frameworks support alternative routes for youth in crisis. Barcucci, Zanola, and Axmann (2017) from the Employment Policy Department of the ILO draw on the ILO’s School-to-Work Transition Surveys to analyse labour market outcomes of VET graduates. These publication activities by IO staff members in academic outlets further contribute to the portrayal of IOs as credible VET research organisations, capable of engaging as peers in VET research.

In contrast to the implicit acknowledgement and reinforcement of IOs as knowledge producers on VET, research that explicitly addresses how and with what effects IOs produce knowledge is less common. The few publications that focus on the social and political role of IO knowledge production are mainly descriptive. For example, Holmes and Maclean (2008a, 2008b) provide a brief historic overview of different IOs’ involvement in VET research, focusing on UNESCO, the World Bank, ILO, OECD, and the EU. Middleton and Ziderman (1997) give an overview of how the World Bank has conducted policy research on VET, linking it also to shifts in the World Bank’s broader views on VET. A more critical position is taken by Kirpal (2008), who argues that VET research produced by the OECD, the ILO and the World Bank is driven by policy interests rather than academic standards and that ‘reports and studies need to be interpreted and valued against the purpose they follow in the light of each organisation and its particular role, aims and objectives’ (Kirpal 2008, 309). Similarly, in their analysis of the OECD’s influence on Australian VET, Legrand and Vas (2014) highlight the dominant position that the OECD has in the production of labour market data, noting that ‘what is known and what is knowable is largely shaped by what the organisation determines as worthwhile’ (Legrand and Vas 2014, 326, see also Crouch, Finegold, and Sako 1999, 31).

Despite these examples, an explicit engagement with IOs as knowledge producers has so far largely failed to address crucial questions about IOs’ knowledge production in the context of VET. Future research should address how IOs decide what to research and what data to collect. Investigating the factors that shape decision-making processes within IOs would shed light on the social and political character of IO knowledge production on VET, while also allowing VET researchers to work with data produced by IOs in a more transparent way. Moreover, research should analyse how data and knowledge produced by IOs are used in national VET policy discourses. Such research could analyse whether IO knowledge activities have an impact on domestic policymaking, while also considering how and for what purpose knowledge produced by IOs is instrumentalised in political debates.

IOs’ agendas on VET

The second role attributed to IOs in the reviewed literature describes them as organisations capable of formulating specific VET agendas, thus focusing on how IOs think about VET.⁸ The following sections present the findings of the literature review for each of the four IOs, starting with the World Bank, which has received the most scholarly attention among the four IOs in terms of its views on VET. The final section considers publications that compare the VET agendas of different IOs.

World Bank

Founded in 1944, the World Bank has provided financial and technical support for VET in low- and middle-income countries since the early 1960s, with its first education loan going to a VET programme in Tunisia in 1963 (Niemann 2022, 135). Several publications focus on the World Bank’s changing VET agenda over the years. Heyneman (2003) traces the World Bank’s historical education policies from 1960 to 2000, describing how the Bank’s approach to education evolved from prioritising VET in the 1960s to neglecting it in favour of basic and academic education in the 1980s. While the Bank’s initial focus on VET aimed to ensure that projects were adequately staffed (following a so-called ‘manpower forecasting model’), the introduction of the rates of return methodology in the 1980s would provide a stronger economic justification for investing in basic and academic education. Heyneman emphasises the internal organisational struggles surrounding this shift in lending and the ‘Bank’s tendency to become “captured” by single methodologies’ (Heyneman 2003, 315). In a similar vein, Mundy and Verger (2015) explain the Bank’s shift in prioritisation from VET to basic education by a change in organisational culture brought about by Bank economists turning to rates of return analyses in the 1980s. In addition, they also highlight that the Bank’s education sector staff used rates of return analysis to defend its education agenda against widespread criticism of the Bank for its structural adjustment policies in the early 1990s (Mundy and Verger 2015, 12).

Focusing on the World Bank’s VET agenda of the 1990s and its historical influences, Bennell, Paul and Segerstrom (1998) argue that the Bank’s reduction of funding for VET was ‘based on the proposition that the governance, funding and provision of VET is best left to individuals, enterprises, and private sector training institutions with government intervention kept to a minimum.’ The authors criticise this position for being ‘seriously flawed’ when considering it in relation to the VET policies of industrialised economies (Bennell, Paul and Segerstrom 1998, 271). They also question the Bank’s interpretation of rates of return studies, arguing that ‘in the large majority of developing countries for which data are available, social rates of return to school-based vocational education are not significantly lower than for general secondary schooling’ (Bennell, Paul and Segerstrom 1998, 273).

The historical development of the World Bank’s VET agenda is also covered to a lesser extent in other publications. Holmes and Maclean (2008b, 76–77) provide a brief overview, focusing on key research reports. Köhne and Stockmann (2008, 150–52) outline the World Bank’s position on VET since the 1990s, summarising criticisms of the Bank’s demand-driven training model, while King (2014, 40–42) briefly recapitulates the major shifts in the World Bank’s VET agendas, highlighting criticisms of the Bank’s dismissal of school-based VET.

Several publications critically examine the extent to which VET has returned as a priority for the World Bank since the 2000s, with a particular focus on the World Bank Education Strategy 2020 Learning for All: Investing in People’s Knowledge and Skills to Promote Development (published in 2011). Analysing the frequency of terms used in the document, Nordtveit (2012) finds that the strategy’s supposed focus on skills is not reflected in its vocabulary. Ngcwangu (2015) criticises the strategy’s market-driven approach, noting that it ‘advances an agenda of individual responsibility for skills training in which one’s productivity is based on individual choices’ (Ngcwangu 2015, 29). In doing so, the World Bank excludes ‘the possibility of alternative conceptualisations of TVET, which prioritise social justice and sustainable development’ (Ngcwangu 2015, 25). Providing a historical overview of skills development in Africa, McGrath (2023) argues that despite the World Bank’s prioritisation of primary education over VET in the 1990s, it ‘continued lending to VET projects to some extent, largely because this is what their clients wanted as African governments still sought to respond to youth unemployment through VET. Thus, the ideology of the Bank’s research needed to bend to the realities of its lending’ (McGrath 2023, 30).

Overall, existing research on the World Bank’s VET agendas is shaped by an interest in the historical changes of its VET funding, which are most often explained by a changing organisational culture within the Bank. In addition, several publications have criticised the Bank’s past and recent VET strategies for their market-oriented approach. Further research could yet provide a more systematic analysis of the Bank’s recent agendas, focusing on how VET has been perceived within the Bank in the last years.

Organisation for Economic Co-operation and Development (OECD)

The OECD was established in 1961 as the successor organisation to the Organisation for European Economic Co-operation (OEEC), which was created in 1948 to support the implementation of the Marshall Plan in Europe. Although primarily conceived as an economic organisation, the OECD has focused on education since its inception and has become a significant actor in the global governance of education (Rizvi and Lingard 2009; Sellar and Lingard 2013).

From a historical perspective, Granata (2022) analyses how the OEEC/OECD promoted VET in post-war Mediterranean Europe, noting that technical and vocational education was seen as ‘one of the key means for preparing citizens for modern society in a Cold War context’ (Granata 2022, 101). Also considering the OEEC, Bürgi (2019) focuses on the European Productivity Agency (EPA), which was established by the OEEC in 1953 to promote productivity in Europe. Bürgi notes that educational issues, including VET, were one of the pillars of the EPA’s programme with the EPA teaching union leaders about VET, producing VET publications with a particular focus on transforming rural workers into industrial workers, and organising internships for people from African, Asian and Latin American countries (Bürgi 2019, 28, 32–33).

However, research that considers more recent OECD agendas on VET remains rare. One important exception by Valiente (2014) focuses on the OECD’s skills strategy Better Skills, Better Jobs, Better Lives: A Strategic Approach to Skills Policies (2012). Valiente notes that the strategy is based on a human capital approach in which investment in skills is seen as contributing to economic growth, reflecting ‘the OECD’s traditional economistic focus on education’ (Valiente 2014, 3). At the same time, Valiente observes that the strategy also incorporates elements from the new political economy of skills approach ‘and its emphasis on the institutional factors that shape the demand for skills in capitalist societies’ (Valiente 2014, 3). This is reflected, for example, in the strategy’s call not only to invest in skills, but to also to make effective use of available skills. Valiente sees this change in perspective as an important policy innovation, as it shifts the responsibility for unemployment from individuals and the education system to the state and firms.

In light of the COVID-19 pandemic, Avis et al. (2021) note that the OECD calls for a ‘resilient VET’ model which ‘proposes building on emergency responses to strengthen such elements as digitisation, short-term training, and recognition of prior learning and other tools to support the mobility of labour’ (Avis et al. 2021, 11). However, the authors criticise that these measures ‘provide hollowed-out forms of VET for an atomised workforce’, with the risk of ‘further narrowing of the scope and possibilities of access to VET, enhancing its already profoundly unequal outcomes’ (Avis et al. 2021, 11, 2).

Overall, compared to literature that focuses on the OECD’s role in education more broadly (see, for example, Sorensen, Ydesen, and Robertson 2021; Ydesen 2019), it is striking that the OECD’s views on VET have not received similar attention. Further research is needed not only on the historical role of VET in the OECD, but also on its current VET agenda – especially in light of the findings that the OECD has begun to widen its human capital approach while also promoting a ‘resilient VET’ model during the COVID-19 pandemic.

United Nations Educational, Scientific and Cultural Organisation (UNESCO)

Founded in 1945, UNESCO is one of the most important IOs specialised in education. It is particularly active in VET through its UNESCO-UNEVOC International Centre for Technical and Vocational Education and Training, established in 1999 and based in Bonn, Germany, since 2002.

Despite the importance of VET to UNESCO, research that directly addresses its VET agendas is rare. An exception is provided by Hollander and Mar (2009), who describe how UNESCO became active in VET after its first International Congress on Technical and Vocational Education in 1987, noting that UNESCO’s role in VET grew after subsequent congresses and with the rise of the Education for All agenda. They argue that UNESCO has adopted a holistic view of VET, seeing it not only as a means to economic growth and employability, but also ‘as a preparation for responsible citizenship and an instrument for promoting sustainable development’ (Hollander and Mar 2009, 43–44). In short, ‘UNESCO sees the role of TVET as educating the whole person, not just the worker’ (Hollander and Mar 2009, 44). The authors argue that this holistic view stems from the 1996 report Learning: The Treasure Within, written by the International Commission on Education for the Twenty-First Century and distinguishes UNESCO’s work on VET from that of other organisations.

Some publications focus on specific UNESCO policies on VET. For Hawley (2009), the UNESCO/ILO report Revised Recommendation Concerning Technical and Vocational Education (2001) understands VET as occupational preparation for young people over the age of 15, while also linking VET to poverty alleviation, sustainable development, and the facilitation of lifelong learning. Hawley notes that the report ‘is probably most similar to the view of “career and technical education” in the United States, stating that technical skills and vocational competencies should be taught within the context of general education’ (Hawley 2009, 518). In a study of UNESCO’s influence in VET in Botswana and Namibia, Preckler Galguera (2018) presents an overview of UNESCO’s general scope and objectives in education and describes key elements of UNESCO’s 2016–2021 VET strategy.

Overall, however, research on UNESCO’s views on VET is not only scarce, but also remains selective, focusing on specific VET documents, rather than explaining the larger historical development of VET within the organisation. Further research in this direction could particularly address how organisational VET discourses within UNESCO have influenced its official positions.

International Labour Organization (ILO)

Founded in 1919 as an agency of the League of Nations, the ILO is one of the oldest existing IOs. Already in 1919, the preamble of the ILO constitution stressed the importance of ‘promoting the organisation of vocational and technical education.’ Unique for an IO, the ILO has a tripartite structure that involves representatives of governments, employers, and workers in its decision-making process.

The early importance of VET for the ILO is addressed in some historical studies, although usually in passing. Mechi (2013) mentions the ILO’s 1948 ‘Manpower Programme’, its first major technical assistance programme, which focused on ‘vocational training and retraining’, among other things. Maul (2012) highlights that ‘[a]pproximately half of all the money spent by the ILO in the 1950s was used on activities in this field’ (Maul 2012, 132). In a volume on the history of the ILO, Alcock (1971) describes how the ILO and UNESCO competed in the 1950s over which organisation should be responsible for VET. While the ILO considered VET to be part of its mandate, UNESCO thought of VET as belonging to the general education system and therefore to its sphere of competence. As governments increasingly turned to UNESCO rather than the ILO for assistance, the ILO was keen to keep VET under its wing in order to continue its manpower programmes. Eventually, in 1954, an agreement was signed ‘under which matters invoking the word “training” were likely to fall to the ILO, and “education” to UNESCO’ (Alcock 1971, 248).

Holmes and Maclean (2008b) stress that despite VET’s early importance to the ILO, research on VET has remained a low priority at the ILO until the 1980s. Since then, ILO publications have stressed ‘the need for a reorientation of training from preparation for wage employment towards skills development for self-employment’ (Holmes and Maclean 2008b, 77). Kirpal (2008) makes a similar observation, noting that the ILO’s VET strategies and programmes are characterised by a close link to human resource development, requiring collaboration with businesses and employer associations, and an orientation towards cultivating skills for self-employment (Kirpal 2008, 307).

In an article on the global diffusion of the recognition of prior learning, Maurer (2021) argues that ‘[t]he ILO is the international organisation that has been most systematically concerned with questions of VET globally for the past several decades, yet it has received surprisingly little scholarly attention’ (Maurer 2021, 476). Maurer emphasises that the ILO’s general focus on labour rights and social justice provides the context for its approach to VET, visible, for example, in ‘its efforts to improve working conditions in the informal labour sector, which explains why it has long been more interested in promoting work-based forms of learning than, for example, the World Bank’ (Maurer 2021, 476).

Considering the ILO’s role in education more broadly, Niemann (2022) emphasises that the ILO – like UNESCO – follows a ‘holistic and humanistic leitmotif in education’ (Niemann 2022, 149). Niemann traces this leitmotif over time, noting that ‘the assurance of equality of educational and vocational opportunity’ was defined as one of ILO’s tasks in its Declaration of Philadelphia in 1944 and that the ILO already collaborated with UNESCO on technical and vocational training in 1948 (Niemann 2022, 149). Niemann notes that the ILO focused on VET in its Decent Work Initiative in 1999, with which ‘the ILO stipulated that education systems should be designed to enable knowledge transfer and skill formation, especially for disadvantaged socioeconomic groups’ (Niemann 2022, 150). Overall, for Niemann, the ILO follows a human rights approach to education, seeing it not only as a qualification for the labour market, but also as a contribution to personal development, active citizenship, and social cohesion. In this context, improving the work of teachers has been a key concern of the ILO (Niemann 2022, 152). This aspect is also considered by Comyn (2022), who (himself an ILO staff member) focuses on the ILO’s role in addressing the employment and professional development of TVET teachers. Comyn finds that through various activities such as capacity building, providing normative guidelines, and training, ‘[t]he ILO actively promotes the introduction of competency- or outcome-based education and training’ (Comyn 2022, 93).

In summary, research on the ILO has so far highlighted the ILO’s general positions on education and its tendencies to focus on the informal side of VET. However, despite the long-standing importance of VET to the ILO, a comprehensive historical reconstruction of the ILO’s views on VET is still lacking.

IOs’ VET agendas in comparison

Beyond research on the VET agendas of individual IOs, publications by Tikly (2013), Niemann (2022), and Vanderhoven (2023) take up a comparative perspective on IOs’ views on VET. Tikly (2013) and Niemann (2022) focus on the background ideas of IOs’ agendas on VET and education, identifying two major approaches. Tikly (2013) contrasts a human capital approach with a sustainable development approach, attributing the former to the World Bank’s rationale for investing in VET as a means of supporting economic growth, and the latter to UNESCO’s ‘more human-centred view of TVET as a means for supporting sustainable development’ (Tikly 2013, 3). Similarly, but considering IOs’ educational leitmotifs more broadly, Niemann (2022) argues that ‘a dualism between economic utilitarian views (World Bank and OECD) and idealistic humanistic (UNESCO and ILO) views can be identified’ (Niemann 2022, 153). However, while Niemann posits the ILO as an idealistic humanistic IO, Tikly argues that it follows a human capital theory approach (Tikly 2013, 5). Both authors see their distinctions as ideal types and as ‘complementary rather than mutually exclusive’ (Niemann 2022, 132), with IOs being able to be informed by multiple approaches ‘albeit with differing degrees of emphasis’ (Tikly 2013, 4). Whereas Tikly concludes by arguing for a human capabilities approach as a third alternative, Niemann is interested in tracing and comparing the different educational leitmotifs of IOs over time, noting that ‘IOs involved in the field of education have started to take a more integrative, ideational approach’ (Niemann 2022, 129).

Considering the question whether there exists a ‘a global apprenticeship agenda’, Vanderhoven (2023) analyses and compares recent apprenticeship discourses of the EU, the World Bank, the ILO, and UNESCO. Vanderhoven finds that apprenticeships ‘act as a somewhat polysemic policy object that can be discursively bent to serve the interests and identities of different IOs’ (Vanderhoven 2023, 12). While there is ‘evidence of significant ideational, theoretical, and technical convergence’ across IOs that draw on international political economy perspectives (the EU, ILO, and OECD), Vanderhoven also observes important organisational differences, arguing that ‘distinct institutional foci remain which challenge a totalising convergence narrative’ (Vanderhoven 2023, 11).

Overall, research on the different VET agendas of IOs shows that several IOs have not yet been adequately analysed. While the World Bank has received the most attention, more research is needed to study how the OECD, the ILO and UNESCO have developed their views on VET. Comparative research on the different VET agendas of IOs is also rare. Vanderhoven (2023, 12) provides an important first step in this direction, while also emphasising the need for ‘[a] more fine-grained understanding of IOs’ ideational work’. This call should be extended to also include the analysis and comparison of the historical development of IOs’ views on VET. Such a perspective would be particularly valuable since most of the reviewed literature merely describes IOs’ positions on VET, rather than explaining how they were formed. Thus, it is still unclear why certain educational leitmotifs have become dominant within IOs. Indeed, IOs are often treated as monolithic actors, rather than as complex organisations in which official positions are created, debated, and possibly contested. Such a perspective largely ignores the processes, actors, and discourses within IOs that shape VET. Future research should analyse these inner workings more closely to provide a comprehensive perspective on how IOs engage with VET as organisations.

IOs as actors of VET policy transfer

A third role in the reviewed literature describes IOs as potential influencers of national VET policies and systems. The literature often frames IOs as key actors of VET policy transfer, for example, by emphasising that ‘[t]oday, policy transfer in TVET is particularly being driven by international organisations’ (Maurer and Gonon 2014, 17) or by highlighting the need to examine ‘the underlying assumptions among representatives of international organisations who are steering and managing policy transfer processes’ (Barabasch, Bohlinger, and Wolf 2021, 352).

Despite this framing, only 19 publications in the reviewed literature address how IOs influence specific national or regional VET policies (see Table 2 for an overview). Ten publications focus on the World Bank compared to four for UNESCO, three for the OECD, and two for the ILO. Six of these publications focus on Africa, five on Asia, four on the Americas, three on Europe and one on Australia. Notably, most of these studies have been published since 2010, with seven since 2020 alone, demonstrating an increasing interest in IOs as actors in VET policy transfer.

Table 2. Studies focusing on IO activities in specific regions and countries.

	Africa	Americas	Asia	Europe	Oceania
ILO		Brazil (da Silva 2022), Mexico (da Silva 2022)	Bangladesh (Maurer and Morshed 2022)
OECD				Ireland (Walsh 2011), Mediterranean Europe (Granata 2022)	Australia (Legrand and Vas 2014)
UNESCO	Botswana (Preckler Galguera 2018), Namibia (Preckler Galguera 2018)	Brazil (da Silva 2022), Chile (Zancajo and Valiente 2018), Mexico (da Silva 2022)	Asia-Pacific Region (Chinien et al. 2009)
World Bank	South Africa (Ngcwangu 2015), Sub-Sahara Africa (Bennell 1996, 2021, 2022)	Chile Valiente, (Sepúlveda, and Zancajo 2021), Colombia (Díaz Ríos and Urbano-Canal 2021)	Bangladesh (Maurer 2012), China (Müller 2021), Indonesia (Sensenig 2012), Sri Lanka (Maurer 2012)	Romania (Abhishek 2022)

The majority of these studies affirm that IOs have an influence on national VET. For example, Maurer and Morshed (2022) describe how a VET reform project in Bangladesh funded by the EU and implemented by the ILO between 2007 and 2015 was a ‘critical intervention’ that led to Bangladesh’s 2011 Skills Development Policy, which would ‘re-define the key aspects of VET in Bangladesh’ (Maurer and Morshed 2022, 5). Preckler Galguera (2018) finds that ‘UNESCO has had a direct and positive influence on the development of national TVET systems in Botswana and Namibia’, noting, for instance, the introduction of a new approach to curriculum development in both countries (Preckler Galguera 2018, 201). Da Silva (2022) argues that in the cases of Brazil and Mexico, IOs ‘have actively acted in establishing partnerships with public and private institutions, which are transformed into great experimentation and tropicalization laboratories’ and thus contribute to establishing ‘educational standards, experiences and instruments that directly impact on school realities and training models for the world of work’ (da Silva 2022, 75).

However, several studies emphasise that IO policy transfer in VET should not be understood as a one-way street. Legrand and Vas (2014) highlight that ‘the ideas and recommendations that proceed from OECD reporting are mediated through a matrix of institutions in Australia at a state and federal level’, demonstrating that policy transfer works as ‘process of ideational influence rather than outright direction’ (Legrand and Vas 2014, 242). In the case of Chile, Valiente, Sepúlveda, and Zancajo (2021) show that governments with different ideological orientations followed different IO recommendations:

While progressive democratic governments adopted the more egalitarian and humanistic development paradigms promoted by ECLA [United Nations Economic Commission for Latin America and the Caribbean], the right-wing governments of the dictatorship opted for the neoliberal recipes promoted by multilateral financial organisations like the World Bank.

(Valiente, Sepúlveda, and Zancajo 2021, 289)

Thus, the authors argue that ‘global and local drivers of VET policy changes do not operate separately but are reciprocally interrelated’ (Valiente, Sepúlveda, and Zancajo 2021, 289). In a similar vein, Díaz Ríos and Urbano-Canal (2021) argue that Colombia, despite being a continuous recipient of financial and technical assistance by the World Bank, has also rejected some of the Bank’s recommendations. Acceptance of IO recommendations is here explained by local power constellations between domestic supporters and opponents of policy transfer. The importance of a willing local policy elite is also stressed by Walsh (2011), who analyses how the OECD influenced policy changes in technical education in Ireland in the 1960s. While the OECD acted as ‘an important catalyst’ by encouraging far-reaching reforms, its recommendations were successful because they corresponded with changing attitudes towards technical education within Irish policy elites (Walsh 2011, 380). Finally, in the case of China’s introduction of vocational qualification certificates, Müller (2021) finds that the World Bank ‘attempted to play the role of a neutral international principal, but could only do so in limited ways and insofar as its preferences were compatible with those of the leadership core of the Party-state’ (Müller 2021, 1–2). These studies suggest that a comprehensive analysis of IO influence on national VET policies and systems should closely consider the interplay between the international and local level.

In addition to assessing the impact of IOs on the national level, the reviewed literature also mentions different instruments that IOs use to exercise influence on national VET. For example, Hollander and Mar (2009) describe that UNESCO-UNEVOC assists its member states through capacity building and seminars, advisory services, knowledge sharing, research and publications, global networking and the promotion of partnerships. Maurer (2021) argues that the ILO and UNESCO have used standard setting, monitoring and technical assistance to promote the dissemination of the recognition of prior learning. Legrand and Vas (2014) study how the OECD’s peer review mechanism and policy recommendations have influenced Australian VET policy.

Yet unclear, however, is the extent to which these different instruments relate to IOs’ impact on national VET systems and policies. One important question is whether material attempts at influence, such as technical and financial assistance, impact national VET more or differently than ideational activities, such as policy recommendations or standard-setting. The World Bank, for example, has in the past made policy implementation a condition of providing loans, but it is unclear whether this has given the Bank a particularly strong influence on national VET policies compared to other IOs. At the same time, it is necessary to consider how IO instruments are perceived and implemented at the local level. It is likely that countries with different socio-economic structures, histories, and VET systems will respond differently to various IOs and their instruments. Future research should analyse and compare how context matters in IOs’ attempts to influence national VET systems and policies, particularly through comparative case studies.

Finally, a yet underexplored question is the extent to which IO VET policy transfer is influenced by concepts and ideas promoted by member states within IOs. For example, in the context of bilateral VET policy transfer, Germany has focused on promoting dual apprenticeships, while the United Kingdom and Australia have emphasised competency-based education and training (Gessler and Peters 2020; Stockmann 2014). It would be useful to explore the extent to which these forms of VET are also promoted by country representatives within IOs, and to consider the influence of national policymakers on the activities of IOs in VET policy transfer more generally.

Addressing these issues through comparative case studies would provide a more comprehensive approach in studying IO policy transfer in VET than existing research. It would combine an exploration of how IOs develop their orientations in VET policy transfer with an analysis of the effects and success of different material and ideational instruments used by IOs, while also considering how IO activities are mediated and perceived at the national level. Thus, future research in this direction would contribute to both analysing and explaining the factors that influence and shape VET policy transfer by IOs.

Conclusion

Despite their diverse activities, a substantial research agenda on IOs in VET has yet to be established. To contribute to future research in this direction, this article conducted an integrative literature review that analysed how IOs have been studied as actors in VET across various disciplines, with particular attention to the four most important IOs in VET – the ILO, OECD, UNESCO, and the World Bank. The results of the analysis reveal three roles that are implicitly and explicitly attributed to IOs in the context of VET. First, IOs are seen as important knowledge producers for VET, demonstrated by the use of IO data and knowledge in academic research. Second, IOs are described as organisations with specific and sometimes diverging VET agendas. Finally, IOs are recognised as actors that have the potential to influence national VET systems and policies. Future research should address important research gaps that arise in relation to each of these roles and analyse how IOs generate knowledge on VET, how they develop their VET agendas, and how they shape VET policy transfer.

It is important to note that this literature review, by focusing on the roles attributed to IOs in VET, could not address all relevant issues concerning the treatment of IOs in the academic literature. For example, it did not analyse the motivations of researchers to use and refer to knowledge produced by IOs. In addition, as a literature review, this study could not itself provide an empirical analysis of IOs’ engagement with VET. However, by synthesising the three different roles and outlining key research gaps, it has provided possible directions for a future research agenda that directly addresses IOs as VET actors. Implementing such a research agenda requires opening the black box of IOs and treating them not as mere providers of VET data or as monolithic actors with fixed notions of VET, but as complex organisations that engage in VET in multiple ways. Studying IOs from such a perspective would mean paying attention to the organisational processes within IOs that are relevant to VET, as well as considering the interactions between IOs and local actors.

Arguably, the success of such a future research agenda will depend on the extent to which researchers interested in the international dimension of VET connect with already existing research on IOs. Several possibilities exist. Analysing IOs as knowledge producers on VET can benefit from literature that focuses on the production of expertise within IOs (e.g. Littoz-Monnet 2017). To investigate how IOs develop their VET agendas, insights can be gained from understanding IOs as bureaucratic organisations (e.g. Barnett and Finnemore 2004) or by analysing their everyday practices (e.g. Bueger and Gadinger 2018). And to study the influence of IOs on national VET systems, conducting country case studies and paying attention to the different mechanisms and instruments of IOs (e.g. Jakobi 2009; Martens and Jakobi 2010) allows one not only to consider whether IOs influence VET, but also to better understand how they do so.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13636820.2024.2320895.

Additional information

Funding

This work was supported by Internal Research Funding (Interne Forschungsförderung) of the Helmut Schmidt University Hamburg.

Notes

1. Then the Organisation for European Economic Co-operation (OEEC).

2. Niemann et al. (2021, 172) also mention the United Nations High Commissioner for Refugees (UNHCR) and the United Nations Children’s Fund (UNICEF) as globally operating education IOs. However, these organisations have begun to only focus on VET in recent years and their degree of VET engagement is substantially less than those of the other four IOs.

3. For Web of Science and Scopus, the search query was as follows: (‘international organ*’ AND (‘vocational’ OR ‘VET’ OR ‘TVET’)) OR (‘oecd’ AND (‘vocational’ OR ‘VET’ OR ‘TVET’)) OR (‘unesco’ AND (‘vocational’ OR ‘VET’ OR ‘TVET’)) OR (‘ilo’ AND (‘vocational’ OR ‘VET’ OR ‘TVET’)) OR (‘world bank’ AND (‘vocational’ OR ‘VET’ OR ‘TVET’)).

4. Only the first 100 results of each individual search operation in Google Scholar were screened for the analysis.

5. The initial number of results of 2126 (Web of Science) and 429 (Scopus) were first refined by research areas and then screened by applying the selection criteria. Since the analysis was conducted in January 2023, the search only covers the years until 2022 in their entirety. The text by Vanderhoven (2023), which was published after the initial analysis was conducted, was later manually added to the corpus because of its central importance to the topic.

6. The determination of a publication’s disciplinary classification was based on the publication outlet’s disciplinary orientation and its self-presentation. In case of interdisciplinary journals or books, the main criterion was the thematic focus of the text.

7. Based on an analysis of scientific outputs in the bibliographic database Scopus, Zapp notes, for example, that ‘[t]he World Bank outstrips the London School of Economics and Political Science by more than 4000 articles in the social sciences and economics, and UNESCO is head-to-head with the renowned Institute of Education London (~5300 articles)’ (Zapp 2018, 18).

8. Attributing to IOs the capacity to have views on VET is a crucial reflection of their agency, as this capacity enables IOs to engage more deeply with VET, for example, by participating in its policy transfer.