EMPLOYABILITY AS A GLOBAL NORM: COMPARING TRANSNATIONAL EMPLOYABILITY POLICIES OF OECD, ILO, WORLD BANK GROUP, AND UNESCO

This article analyses transnational policies on employability to understand the development of employability as a global norm and reflects about its consequences for stakeholders in the field of lifelong learning.


Introduction
The global skills mismatch, low-paid and low-productivity jobs, problems of the informal sector, large-scale unemployment, and resulting socio-economic and political problems have made employability a top priority on the policy agenda of most policy actors. The structural shifts in the labour markets led by the Industrial Revolution 4.0 (focussing on artificial intelligence) and Globalisation 4.0 (Schwab, 2019) have forced policy actors to reconsider education policies. Governments bank upon education, especially lifelong learning policies, to ensure that individuals who get educated also get employed and remain on the labour market for as long as possible. Further, they must continue to engage constructively with the society so that the cost of ensuring their welfare remains low. In most countries, the implementation of such education policies remains a challenge due to a lack of resources, expertise, and even political will. Consequently, key global players in the field, including the OECD, the World Bank (hereafter WB), the ILO, and the UNESCO, play a major role because unlike states, these organisations have data, resources, expertise, willingness, and stability for devising long-term policy solutions and managing stakeholder interests in an efficient manner.
Despite the fact that lifelong learning policies were first devised to promote employability (OECD, 1996), orienting lifelong learning policies towards employability has met strong resistance from many stakeholders in the field. The paper reflects about the possibilities for such stakeholders by analysing the following questions: 1. What constitutes the global norm of employability and why? 2. How did it develop? 3. What policy choices do states and sub-state stakeholders have in relation to adopting or refusing this global norm?
Selected policy documents with a global scope and an instrumental role in the development of the norm from key policy actors including the OECD, the ILO, the WB, the UNESCO, and other international organisations and platforms are compared because of their pivotal role in the process. Further comparison along time between 1992 (when employability policies started to develop) and 2018 (when the latest available empirical data were used as sources for this paper) is made. Finnemore and Sikkink's model of norm dynamics (development and change in norms) provides the conceptual framework.
The paper is divided into six sections. The first section includes the problem description, research questions, methodology and the design for the paper. The second section includes a review of existing literature about employability as a concept and is used afterwards to highlight the difference between understanding employability as a concept (embedded in research, professional understanding, and practice at sub-state level) and as a global norm (embedded in transnational policies). The third section explains the conceptual framework. Section four includes a description of employability as a global norm, its links to lifelong learning, a chronological mapping of documents, and the comparative analysis. Discussion and conclusion follow in Sections five and six, respectively.

Employability
The concept of employability is not new, but its connotations have changed over time. In the 1940s, it was defined in terms of the individual's ability to work based on age, capability, and family commitments (Gazier, 1999). In the 1960s, health, disability, and social background became relevant considerations. In the 1980s and 1990s, individuals' productivity; their cumulative marketability (income that they can earn) based on human, social, and cultural capital; and meaningfulness of employment for them became predominant notions (Gazier, 1999). Thus, employability can be defined as an individual's ability (depending upon various considerations) to be on the labour market (Hillage & Pollard, 1998;Brown, Hesketh & Williams, 2003;Brown & Hesketh, 2004;Berntson, Sverke & Marklund, 2006;McQuaid & Lindsay, 2005;Yorke, 2006;McGrath, 2009;Wilton, 2011). The individual bears the responsibility as well as enjoys the freedom to shape this ability (Wilton, 2011). Failure to become employable might not only depend on the choices of the individual but also on their context. Employability can hence be relational (Gazier, 1999;Clarke, 2008;Wilton, 2011).
Contextual factors influencing an individual's ability can be categorised as a) demand factors, such as jobs available and labour market conditions such as competition, rules and regulations, macroeconomic factors, and the like; and b) supply factors, such as generation of individual assets (knowledge, skills, competencies) and factors (like inequality) affecting it, an individual's personal circumstances detrimental for gaining or losing employment, and the like. (Hillage & Pollard, 1998;Peck & Theodore, 2000;Grip, Loo & Sanders, 2004;McQuaid, Green & Danson 2005;Gore, 2005;Wilton, 2011). Education and labour market-related choices of individuals shape their employability and might be guided by the return on investment they make (Peck & Theodore, 2000;McGrath, 2009). This return on investment approach is narrow and has been specifically problematic for the following reasons. 1. It has led to a vast body of research, literature, and practices regarding supply-side corrective measures (educational offers) to deliver 'employability skills' by changing curricula, regulations, staff reorganisations, linking industry and education, and the like, which contradicts the essence of employability itself. The assumption about the possibility to deliver fixed 'employability skills' (as products) makes the whole concept of employability (a constantly changing flexible process) irrelevant. The whole narrative around it thrives on the biased assumption that employability is the individual's responsibility, meaning the individual must make choices for becoming employable. This paper shows that employability policies include individual responsibility as only one of many components. Therefore, limiting the efforts towards employability to supply-side solutions through lifelong learning and education is a short-term, unrealistic strategy and wastage of resources (see Section four). 2. It has hindered the shift in the approach from teaching to learning. Providers assume and argue that they should teach individuals to become employable and thus, limit their free choices for learning. This notion about the flexible nature of employability and the non-existence of specific employability skills is well described in transnational policies on employability (see Section four). 3. At the macro level, it could be difficult to calculate accurate individual investments and returns on investment because of the intangible aspects of employability like social and emotional costs and benefits (Wolf, 2002).
As a policy norm, employability is a solution for multiple socio-economic problems and a driver for unlimited but sustainable growth. The same is discussed in detail in Section Four.

Conceptual framework
The conceptual framework of this article draws on the literature about norm dynamics (emergence, establishment, and change in norms).
Norms can be defined as shared understandings and behavioural standards manifested as rights and responsibilities of stakeholders involved (Krasner, 1982;Florini, 1996;Finnemore, 1996;Finnemore & Sikkink, 1998;Krook & True, 2010). They produce order and stability, regularise stakeholder behaviour, and limit alternatives for policy choices (Finnemore & Sikkink, 1998). The development, survival, and power of norms therefore requires that they are based on a stable agreement among stakeholders (Krasner, 1982;Gilardi, 2012). Stakeholders might resist or refuse to accept norms if they do not appear to benefit them. Norms are thus portrayed as a positive sum agreement among all stakeholders (especially the most powerful ones) and usually remain vague to accommodate conflicting stakeholder interests (Krook & True, 2010). Norms might emerge as an institutionalised complex, collection, or cluster to portray a complex set of interrelated problems and solutions (Moore, 2012;Winston, 2018). Sustainable development is such an example, with many correlated norms about economic growth, social inclusion, environment, and the like woven together to represent a certain notion of development.
Norms could be constitutive (standardising behaviour in new situations without much precedent, for instance cyber norms when computers and internet were introduced), evaluative/prescriptive (standardising behaviour based on what should ideally be done in a given situation), or regulative (ordering and constraining the behaviour of stakeholders) (Finnemore & Sikkink, 1998). They are relevant for actors that take policy decisions based on what is the socially determined appropriate behaviour (logic of appropriateness) in a certain situation rather than thinking about the consequences (logic of consequences) of their choices (March & Olsen, 1998;Checkel, 2005;Moore, 2012;Gilardi, 2012). However, in the long run, conforming to what is appropriate rather than thinking about short-term consequences might also be conforming to the logic of consequences because legitimacy generated by confirming to the so-called appropriate has its own benefits through image creation, legitimacy, and the like. Finnemore and Sikkink (1998) describe norm dynamics (creation and development of norms) in three different stages: norm emergence, cascade, and internationalisation (Finnemore & Sikkink, 1998) (see Figure 1). : Finnemore & Sikkink, 1998:896] norm emergence may be facilitated by human agency, indeterminacy, favourable occurrences, and positive linkages among promoted or existing norms and/or values (Checkel, 1998;Finnemore & Sikkink, 1998;Carpenter, 2007;Fukuda-Parr & Hulme, 2009;Krook & True, 2010). They are promoted by norm entrepreneurs (actors that promote them) on different platforms till a tipping point (acceptance by a critical mass or about one-third of potential acceptors) is reached (Finnemore & Sikkink, 1998).

Figure 1 -Emergence and Development of a Norm (Norm Dynamics). [Source
The second stage in norm dynamics, or cascading, begins after the tipping point is reached (Finnemore & Sikkink, 1998). Stakeholders who accept the norm are rewarded, whereas those who resist are punished (Finnemore & Sikkink, 1998). Reward and punishment depend upon the sensitivity of stakeholders, and even a symbolism, such as negative international reaction against states, might be consequential (Gilardi, 2012).
When the norm is adopted by most stakeholders, the debate about why it should be adopted comes to a halt, and its adoption is taken for granted (Finnemore & Sikkink, 1998). In this third stage of norm dynamics, called norm internalisation, the burden of proof (argument for not accepting the norm) shifts from norm entrepreneurs and followers to those who resist it (Finnemore & Sikkink, 1998;Gilardi, 2012).
The choice of platform, or venue, plays a crucial role in the process. It includes membership, mandate, output status, procedures and legitimacy (Coleman, 2011).
Membership refers to considerations about who is active on the platform and in what capacity (Coleman, 2011;Krook & True, 2010). Specific and clear norms can be formulated with a small number of homogenous stakeholders, large power differences, top-down power relations, and pro-norm human actors (experts, bureaucrats, etc.) (Beach, 2004;Coleman, 2011). On the contrary, a large number of heterogeneous stakeholders with limited power difference, and anti-norm human actors lead to ambiguous norms masking disagreements (Beach, 2004;Coleman, 2011).
The mandate (focus and scope of discussions) and output status (product) are closely linked factors. Limited mandates with single or few issues and a small scope often facilitate intense discussions. Larger mandates with multiple issues and a broad scope may lead to ambiguous and complicated agreements, bargaining, trade-offs (compromises), logrolling (exchanging favours), and the like (Koh, 1997;Coleman, 2011). Output status (or outcomes) of negotiations may be binding or non-binding on the stakeholders, may be visible or unnoticed, and may come in various forms like reports, treaties, and laws (Coleman, 2011).
Procedures regarding what happens when, how, and who regulates it, include the dominance of pro-and anti-norm stakeholders in procedures like drafting, bundling of items to be discussed, agenda for discussion, chronology, and sequencing of items and procedures for arriving at decisions (like voting of various forms, veto and consensus) (Kauffmann, 1996;Coleman, 2011).
legitimacy (acceptance) of the venue refers to whether the stakeholders consider it an appropriate (or inappropriate) forum for the discussion about the norm (Coleman, 2011). It is usually based on precedents (Coleman, 2011). Fruitful negotiations in the past related to similar or related norms can increase the credibility of a venue and favour norm acceptance (Coleman, 2011). Contrarily, negative connotations like failed negotiations may lead to less credibility, mistrust among stakeholders, and influence mobilisation against norm acceptance (Coleman, 2011).
Norms change constantly as a result of their competition with other rising or declining norms; changes in meanings associated with them or their components; and changes in context, positions, and internal dynamics of norm entrepreneurs and stakeholders (Krasner, 1982;Florini, 1996, Finnemore & Sikkink, 1998Krook & True, 2010). Boomerang effects or the involvement of transnational advocacy networks (Tans) to create pressure on potential norm accepters for accepting a norm might induce changes in norms due to the internal dynamics of these Tans (Keck & Sikkink, 1998;Krook & True, 2010). Thus, norms change constantly while they are promoted and adopted (Sandholtz, 2008;Wiener, 2009;Krook & True, 2010).

Development of employability as a global norm: mapping and comparison of policy documents
The OECD has been the most influential actor in relation to employability policies. In 1992, OECD member countries approached it for policy solutions to deal with the challenges of an increasing social welfare burden. These countries were marred by high unemployment and precarious, low-paid, low-productivity jobs with poor working conditions. Consequently, the OECD proposed the 1994 Jobs Strategy, featuring macro and microeconomic policies to fuel limitless but sustainable economic growth (OECD, 1994).
The OECD proposed structural and macroeconomic reforms (e.g. tax reforms); engagement of the private sector as employer and inves-tor; active labour market policies; cuts on social security (especially in Europe); measures to encourage people to work (even in low-paid jobs); and reduced reliance on social benefits. It also proposed to link all nonwork (not depending on labour market participation in any way) social benefits to work. The disadvantaged and marginalised were an exception, but attempts to engage them with the labour market were also proposed (OECD, 1994). Further, wages (in favour of employers) and working conditions (in favour of employees) were to be made flexible (OECD, 1994). Self-directed lifelong learning was proposed to equip the individual for coping with technological dynamism and structural shifts in labour markets caused by contextual changes (OECD, 1994). Education quality, early childhood education, school-to-work transitions, investments in equipping individuals (especially the least qualified and disadvantaged ones) with relevant skills, and the integration of education in national qualification frameworks were emphasised (OECD, 1994). Industry-education linkage, standard assessments for training, recognition of prior, non-formal, and informal training, and mobility were suggested, too (OECD, 1994).
Individuals were supposed to take advantage of these reforms, actively participate in the labour market, constantly engage in self-directed lifelong learning, and transform themselves into a highly productive human resource (OECD, 1994). In case individuals failed to make adequate choices or lagged behind due to disabilities or marginalisation, all stakeholders, particularly the state, was to support them (OECD, 1994). Thus, the state had to bear the responsibility of engaging other stakeholders and creating a conducive environment around the individual in which they could make free choices to shape their career (OECD, 1994).
In 1996, the OECD published the first consolidated policy on lifelong learning and received a five-year mandate to further develop the same (Ehlers, 2019). Mapping and comparison (Table 1, 2, and 3) show that employability evolved from a policy solution in 1994 to an essential strategy for sustainable development in 2018. Table 1 shows the emergence of employability as a global norm between 1994 and 2010. The OECD proposed it in 1994, reviewed it in 1998, and assessed it in 2006(OECD, 1994. Despite acknowledging a drop in unemployment, the OECD observed that implementation lags still led to limited outcomes (OECD, 2006). The OECD thus reorganised and restructured the policy to facilitate its implementation in 2006 (OECD, 2006). Comparison reflects the shift in focus of policy recommendations from formulation to implementation and evaluation.
The influence of the OECD was limited to high-income countries, but the UNESCO and the ILO included elements of the OECD strategy in their policy recommendations to low-and middle-income countries struggling with poverty, unemployment, and many other challenges similar to the ones faced by high-income countries in the 1980s (UNESCO, 2001;ILO, 2004). The proposals of the UNESCO and the ILO changed the narrative around employability and included a rights-based and social-justice-oriented approach to fit the contexts in low-and middle-income countries (Table 1). However, many lowand middle-income countries adopted the idea (not reform) of lifelong learning ( Jakobi, 2012) but continued to focus on primary education and gender parity in education policies, thereby failing to integrate their policies in favour of employability since their resources were diverted to achieve the Millennium Development Goals (Singh, 2018).
During the financial crisis, the ILO brought representatives of different social partners together, highlighted the failure of past policy packages for socio-economic development in low-and middle-income countries, raised the issue of ensuring decent work and social justice, and called upon the high-income G20 countries to rescue low-and middle-income countries (ILO, 2008a;2008b). The ILO proposed the implementation of the OECD strategy in low-and middle-income countries to ensure that their recovery from the financial crisis was accompanied by their recovery from a long-term jobs crisis (resulting from past policy failures) (ILO, 2009).
In 2009, the G20 countries invited the ILO to formulate a comprehensive strategy with an approach focussing the move from employment to employability (G20 Research Group, 2009). The ILO engaged representatives of all possible stakeholders and drew on its own recommendations to formulate a comprehensive strategy, which was in no way different from the core OECD strategy (ILO, 2010;OECD, 1994). The only concrete differences included references to stakeholders from low-and middle-income countries, other international organisations, and regional and multilateral institutions, as well as some keywords closely connected to the concerns of the low-and middleincome countries. In other words, the OECD strategy was extended to low-and middle-income countries between 2008 and 2010 by the ILO, using the financial crisis as a window of opportunity. The documents in Table 1 deal with guidelines for policy formulation and abstract ideas for their implementation.
The policies between 2010 and 2016 (Table 2) were evidence-based, featuring concrete recommendations about implementation, performance evaluation, and benchmarking in favour of employability. The WB issued the STEP (Skills Towards Employment and Productivity) framework for performance measurement and benchmarking in 2010 (The World Bank, 2010). An overall education policy from the WB followed in 2011 aiming at 1) introducing concrete reforms in the education systems of low-and middle-income countries and 2) building a global knowledge base based on innovation and empirical data (The World Bank Group, 2011). In 2013, the OECD and the WB introduced more benchmarking and performance indicators in collaboration with the European Training Foundation, the ILO, and the UNESCO for comparing low-income countries; against the OECD standards (  In 2016, the ILO, the OECD, the WB, and the IMF came up with a consolidated policy on employability, signifying the standardisation and convergence of their employability policies and their alignment with the international development agenda (OECD, ILO & The World Bank, 2016). Employability was thus directly or indirectly pushed on the agenda of most stakeholders who accepted the advice of any of these organisations or agreed to the international development agenda manifested as Sustainable Development Goals (SDGs, 2015). Policies after convergence (post-2016) include concrete guidelines for implementing policy reforms to promote employability as a component of the sustainable development agenda (see Table 3  Comparison of selected activity by policy actors ( Figure 2) shows that the cumulative activity to promote employability as a global norm at the international platforms was quite low till 2006, became considerably high in favour of convincing and pressurising states to implement employability policies during 2010-2016, and started declining afterwards (post-2016) even though still remaining relatively high. Selected Activities by International Organisations Year of Activity Year of Activity

Discussion
Comparison and mapping shows that despite diff erent challenges and policy objectives, the core policy on employability (as proposed by the OECD) did not change. However, modifi cations were introduced for implementation based on contextual considerations, empirical evidence, inclusion of heterogeneous stakeholders, need for legitimacy, and oth-er factors characterising the change in the venue of policy formulation (OECD, 2006(OECD, , 2018ILO, 2010).
Employability emerged as a global norm between 1994 and 2010. Despite the fact that the core elements of this norm were already in place by 2006, it took four years and one favourable event (financial crisis) to push it to the tipping point in 2010 (Figure 3). While the relatively small number of homogenous OECD member countries accepted a clear norm on employability in the beginning, the ILO provided the legitimacy needed to promote it globally by engaging stakeholders from a large number of heterogeneous low-and middle-income countries.  Finnemore & Sikkink, 1998: 896] The OECD acted as the norm entrepreneur, as it developed and promoted the policy. Together with the WB, it led the cascading through implementation guidelines, performance standards, indicators, establishment of comparative data platforms and analysis of data, policy reviews, measurement, and benchmarking between 2010 and 2016.
The ILO, the UNESCO, and other organisations chose to jump on the OECD bandwagon. Lack of expertise and limited means, especially after the financial crisis, contributed to an increased acceptance of employability among states based on the logic of appropriateness. After the convergence of policies in 2016, norm internalisation began.
International organisations, especially the UNESCO, aligned the norm of employability with the norm cluster representing Sustainable Development. The key policy actors managed to increase their influence through norm development and ensured that the states (primary sources for funding) and other stakeholders depended on them. By embedding and linking different norms to a norm cluster, they pushed the stakeholders indirectly to internalise all norms. Linking lifelong learning to this norm as an education strategy, for instance, made it essential for stakeholders to accept employability-oriented policies and then conform to measurement and data standards woven around the whole narrative of the norm cluster of sustainable development. This trend is visible in other policy areas as well, for instance in case of Development Education (Singh, 2018).
The existence of a limited number of key players (four) facilitated the development of a clear norm, but the engagement of these actors also caused changes. The ILO pushed for inclusion based on social justice and dialogue among social partners; the UNESCO promoted a rightsbased approach; the OECD promoted ceaseless, balanced, and inclusive growth (for optimum utilisation of resources), and the WB induced policy integration and standardised measurement (through indicators and comparative data). However, the most remarkable changes to the policy were triggered by implementation problems.
Transnational employability policies have a strong influence on education and lifelong learning policies across the globe. The OECD devised lifelong learning policies as a strategy to promote employability in the early 1990s, materialising existing ideas (since the late 1960s) of the Lifelong Education and Learner Centric Approach into a concrete policy (Ehlers, 2019). This change was marked, among other changes, with changes from a humanistic orientation to an economic orientation, from process-orientation to product or process orientation, and from teachercentric to learner-centric approaches in education (Ehlers, 2019). The most prominent change however, was the shift in focus of policies from lifetime employment to lifetime employability (Gurria, 2011, online source).
Employability was proposed as a constitutive norm (new policy solution) in the beginning. During cascading, it became prescriptive due to elements like performance benchmarks, indicators, and the like. In the post-2016 period, it has become rather regulative in nature. States conforming to the sustainability agenda find it difficult to reject employability now because it is aligned with the norm cluster of sustainable development (with many other norms including those on environment, inclusion, gender, etc.), and most stakeholders have made commitments to achieve it. Path dependency of accepting sustainable development has pushed stakeholders to promote and internalise the norm of employability. Lifelong learning is also a part of the same cluster and promoted as a sustainable policy for education, with employability as one of its foremost objectives. Thus, for the stakeholders who are not convinced by the notion of orienting policies in favour of employability, options for reconsidering effective implementation are open but the possibilities for rejecting the norm look rather doubtful.

Conclusion
A comparative approach not only enables an understanding of the similarities and differences among certain entities, processes, or phenomena, and the reasons thereof, it also facilitates the identification and mapping of change linkages among variables. From the comparison of policy documents by key global policy actors over time, it is evident that ignoring the objective of employability in lifelong learning is not an option for stakeholders anymore. The key policy actors have ensured the development and consolidation of employability as a global norm and have integrated it with the norm cluster (sustainable development). The strengthening and survival of these actors depends on the strengthening of such norm clusters and thus, they tend to weed out all stakeholders with intentions to do otherwise using strategies like performance measurement. Rejecting employability may amount to rejecting the idea of sustainability. Thus, the way forward for stakeholders who stand dissatisfied with the current notion of employability and its linkage with lifelong learning is either to influence the evolution of the norm in a favourable way or to wait for some other favourable norm to replace it. In the second case, the risk of being weeded out as unsustainable and thus irrelevant cannot be denied.