copy the linklink copied!8. Conclusion and general policy recommendations

The concluding chapter of this study broadly summarises the main findings and takeaways from the set of case studies discussed in previous chapters, building on the analytical framework developed in the introductory chapter. It also develops a set of policy recommendations that aim at strengthening the governance of skills policy across OECD countries. More specifically, it recommends to establish co-ordinating committees with a meaningful mandate and clear internal governance structures, to promote the involvement of non-governmental stakeholders while managing the risk of undue influence of special interests, to support the establishment of information management systems that provide usable and relevant information to stakeholders, and to invest and commit diversified fiscal resources needed to strengthen skills policies.


copy the linklink copied!Overview and introduction

This report has provided a comprehensive analysis of the governance of skills policies in selected OECD countries. The introductory chapter laid out a general theoretical framework, including a set of preliminary policy recommendations developed in the context of the OECD Skills Strategy 2019 (OECD, 2019[1]). The purpose of this final concluding chapter is to reflect on the validity of these policy recommendations given the insights from the detailed case studies of Estonia, Germany, Korea, Norway, Portugal and the United States that make up the main part of this report. It brings together the most important insights from these case studies, and considers how they help fine tune and further develop the preliminary policy recommendations from the introductory chapter of this report. The concrete policy recommendations from each case study are used to develop broader and more general policy recommendations that hold across country cases, building on the positive experiences of these countries in pursuing new approaches to skills policy.

In order to facilitate the practical application of the learnings contained in this report, a self-assessment tool for policy makers and stakeholders is provided in Annex A. Based on a number of “yes/no” questions informed by the recommendations presented in this chapter, the tool enables all interested parties to explore the capacity of a particular country to strengthen the governance of its skills policy regime.

It is important to highlight that this report does not aim to highlight or promote a single “best practice” model of how the governance of skills policies should be organised, as each country case has idiosyncratic characteristics such as particular economic and welfare state institutions, particular traditions of engaging and interacting with non-governmental stakeholders, different approaches to the co-ordination of activities across levels of government and, more generally, different policy-making practices and cultures.

Given these idiosyncrasies, simply transferring “best practice” models from one country context to another is not always advisable or feasible. Still, despite these differences, in many cases mutual learning is possible, and good governance practices in one country can be transferred to others (sometimes in a modified form), especially when countries share similar traditions and institutions.

The introductory chapter identified four challenges that need to be met to strengthen the governance of skills systems, which have served as the conceptual backbone for the case studies:

  • Promoting co-ordination, co-operation and collaboration across the whole of government.

  • Engaging with stakeholders throughout the policy cycle.

  • Building integrated information systems.

  • Aligning and co-ordinating financing arrangements.

Instead of simply revisiting these challenges, which were already discussed extensively in the introductory chapter, this concluding chapter first reflects on how they played out in the concrete case studies, and which additional insights can be gained. Subsequently, it presents a set of general policy recommendations that directly refer to the four dimensions of governance. They should be regarded as refined, empirically enriched and contextualised further developments of the initial policy recommendations presented in the introductory chapter.

copy the linklink copied!Challenges in the governance of skills systems: Insights from the case studies

Promoting co-ordination, co-operation and collaboration across the whole of government

The promotion of co-ordinated policies across levels of government, as well as across different ministries, is crucial to strengthen the governance of skills systems. Different government agencies, ministries and organisations need to send out a coherent message to stakeholders and the broader society to facilitate sustainable and reliable stakeholder engagement. Achieving such a whole-of-government approach may sound simple as, formally speaking, different levels of government and ministries within government are located in a pre-defined hierarchy of authority and accountability. In practice, however, as the case studies have shown, central governments often face numerous challenges in the implementation process.

A first example of these complexities is ultimately a consequence of the inherently limited extent to which the central and/or top levels of government can control the actions of sublevel agencies and ministries. Controlling every action of sub-agencies is simply not feasible and would lead to information and management overload at the higher levels of government. Although the establishment of data-based information management systems (see below) may mitigate this problem to some extent, there is still an inherent limitation regarding how much information can be processed at the top level. Furthermore, subnational governments, particularly in decentralised countries, have a significant degree of autonomy in setting their own priorities and agendas because their political legitimation is independent from the central government. Governmental agencies and ministries are also partly autonomous in setting their agendas as a consequence of particular kinds of expertise they have developed within the general government’s division of labour.

Hence, from the perspective of the central government, the crucial question is how to achieve a coherent strategy that respects the partial autonomy of subnational governments and sublevel agencies, and mobilises and contributes to their particular expertise, as well as ensuring that policies are coherent and co-ordinated across the whole of government. As discussed in the introductory chapter, this challenge is particularly pronounced for skills policies compared to other policy areas, as education and training policy-making competences are typically in the hands of lower levels of government, particularly in decentralised, federalist countries. Furthermore, skills policy cuts across traditional domains of government ministries, most importantly ministries of education on the one hand and ministries of labour and social affairs on the other.

As the case studies have shown, there are no easy solutions. One strategy might be to strengthen hierarchical governance modes, i.e. by simply committing lower levels of government and ministries to comply with a national strategy set at the top level. The obvious downside of this strategy is that lower levels of government are likely to voice open opposition to such a hierarchical strategy if they are not involved in its design. They might even make use of their partial autonomy and undermine its implementation. To a significant extent, these actions of local governments are related to the fact that they usually have better access to local knowledge and regional needs than central governments. Eventually, a hierarchical strategy of policy co-ordination is likely to fail as government agencies and subnational governments “dig themselves in”, and are more likely to engage in “turf wars” with their counterparts rather than implementing the overall strategy.

Against this background, the strategy that most governments in the cases studies choose is one that combines hierarchical modes of steering with engagement of a range of governmental stakeholders (see further below for a discussion of the involvement of non-governmental stakeholders). For example, the Norwegian Skills Policy Council includes representatives from different ministries and local governments (counties). The Alliance for Initial and Further Training in Germany (Chapter 3) brings together different federal ministries (education, labour) with representatives of the equivalent ministries from subnational governments (Länder). ANQEP, the Portuguese National Agency for Qualification and Vocational Education and Training (Chapter 6) pursues a similar approach.

Setting up new committees to co-ordinate and implement a whole-of-government-approach to skills policy is an obvious response to the abovementioned challenges. The risk of such a strategy, however, is that committee structures proliferate and become overly complex. In the worst case, this may lead either to gridlock or ineffective co-ordination, when newly established structures are regarded as “talking shops” rather than effective decision-making bodies. Hence, the challenge for pursuing a whole-of-government-approach through the establishment of new governance structures is to find the right balance between hierarchical steering, which is needed to some extent to further the development of a joint strategy and – if needed – to break potential gridlock in committees, and collaborative decision making.

A second challenge related to the implementation of a whole-of-government-approach in skills policy is to keep skills policy at the top of the government’s agenda. Arguably, this is more of a challenge for the experts and stakeholders involved in skills policy rather than the top level of central government itself, since the latter is ultimately in charge of setting the national agenda. Broadly speaking, the case studies revealed a particular dynamic regarding the timing of skills policy becoming top of the government’s agenda: when governments face a real or perceived crisis (e.g. low educational achievements among adults in Portugal or among the elderly in Korea, high levels of youth unemployment in Germany) they are more likely to prioritise skills policies, compared to when there are few problems.

The issue with this problem-driven approach to skills policy is that the associated policy responses are likely to be less comprehensive and more selective, responding to particular issues at a certain point in time rather than promoting a more encompassing skills policy agenda. The crisis-driven approach to skills policy is also likely to affect the degree of involvement and commitment of governmental and non-governmental stakeholders. During crisis periods, commitment is likely to be higher, related to and driven by the central government’s efforts to put this issue at the top of the agenda. However, once the immediate crisis is over, joint commitment is likely to reduce.

The crisis-driven approach is likely to be counterproductive and ineffective in the long term. If governments prioritise skills policy at times of crises, when resources are scarce, policy responses are likely to focus only on the most pressing short-term problems, thus neglecting structural and long-term challenges to the skills system. The question for the top level of governments is how to avoid falling into the crisis-driven policy response mode, and instead to prioritise skills policy when times are good. When resources are less scarce, governments have more leeway to adopt and pursue long-term strategies that tackle structural issues in the skills system, such as lifelong learning or the integration of disadvantaged youth who face problems accessing the skills system even in good economic times.

Engaging with stakeholders throughout the policy cycle

Governments strive to reach out and involve non-governmental stakeholders throughout the policy cycle, i.e. from policy development to implementation. As discussed in greater detail in the introductory chapter of this report, the involvement of non-governmental stakeholders has many advantages: most importantly, it promotes a sense of ownership among stakeholders, which eases potential problems during the implementation stage of policies. Non-governmental stakeholders also bring particular kinds of expertise to the table, which would be otherwise difficult for governmental actors to access and use. This broadening of the information base of policy decisions is useful to design more effective and legitimate policies.

However, there are also several challenges to engaging with stakeholders. One of them, as mentioned in the introduction of this report, is to identify and map out the relevant stakeholders in a given skills system. The case studies have shown that policy makers generally have a comprehensive overview of the different actors involved in the skills system, in particular regarding organisational representatives of labour market interests, such as employers’ associations, trade unions or other professional associations. What can be more of an issue is how to distinguish between relevant and less relevant actors, which is – to some extent – also a decision that depends on the prevailing political context.

There are certain trade-offs to be considered in this regard. A voluntaristic “laissez-faire” strategy would consist of government actors simply extending an invitation to all stakeholders that wish to contribute and get involved. This strategy has the advantage of being inclusive in the sense that government actors do not make preliminary judgements on which stakeholders should or should not get a voice at the table. The case of lifelong learning in the Korean city of Suwon (Chapter 4) is a good example of such an approach. In this case, numerous local stakeholders and interested individuals are involved in the further development and design of lifelong learning programmes.

The other cases studied in this report represented a different approach more rooted in the “corporatist” tradition of interest mediation (Schmitter, 1979[2]; Streeck and Schmitter, 1985[3]). According to this model, stakeholder involvement is more selective in the sense that its goal is not to maximise the number of stakeholders involved in decision making, but rather to reach out and include those stakeholders that can legitimately speak and represent different functional interests in society. Thus, the various councils and committees analysed in the case studies of this report usually include representatives from peak organisations of business and labour, as well as some selected other societal stakeholders (e.g. independent experts or school representatives). However, in some cases such as the Norwegian Skills Policy Council (Chapter 5), policy makers involve strategically selected stakeholders that are traditionally not part of the corporatist system in order to tackle new questions.

The notion of corporatism entails the idea that all relevant sectors of society are represented and that decisions are made by consensus. However, the political context in which new committees and councils are established is also important. In the case of Germany, for instance, the first National Pact for Vocational Training and Qualification of Skilled Workers – a predecessor of the Alliance for Initial and Further Training, which is the focus of the German case study – only included representatives from government and business. Union leaders refrained from participating due to a previous government decision to not introduce a training levy. This example shows the challenge of how to convince non-governmental stakeholders to contribute to and participate in joint decision making independent of the prevailing political context.

A further issue relates to different traditions of interest mediation in different countries. The country sample for this study included countries with a long tradition of social partnership (i.e. co-operation and joint decision making between employers and unions in wage bargaining, training and other policies) such as Germany and Norway, and those with a less entrenched tradition of corporatist decision making such as Portugal and the United States. There is a broad challenge of how to ensure that decisions made at the peak level of national skills councils are transferred and translated to the membership base of intermediary associations to ensure that they are effectively implemented. However, this is more relevant in the case of countries without a strong tradition of social partnership.

The crucial factor is how the top level of associations communicate and interact with their members. If the top level of associations, which usually participates in the skills councils analysed in the country cases, is disconnected from its membership base, the political legitimacy of skills councils is put into question as it ultimately depends on their responsiveness to the needs of individuals and employers in the labour market. If stakeholders are not effective representatives of societal needs and demands, corporatist decision-making bodies can easily deteriorate into privileged points of access for special interest groups to capture public institutions. Vice versa, in order to be legitimate forums for decision making, actors that participate in corporatist skills councils need to be able and willing to effectively implement joint decisions (Streeck and Schmitter, 1985[3]).

The final challenge related to engaging stakeholders that emerged in a number of case studies is how to ensure that governance structures involving stakeholders are built on a permanent, sustainable basis. As laid out in the introduction of this report, involving stakeholders in decision making requires the building of mutual trust between governmental and non-governmental actors, as well as between stakeholders representing usually opposed societal interests (e.g. business and labour). Building this trust takes time.

The long-term sustainability of governance structures may be threatened for different reasons, for example because of the scarcity of fiscal resources (as in post-austerity Portugal) or because policy makers have decided to establish these structures for limited periods of time (as in Germany). Furthermore, the composition of governance bodies, as well as their mandate and mission, may change over time, which negatively affects the long-term foundation for meaningful stakeholder engagement.

Building integrated information systems

The goal of building integrated information systems in the area of skills policy holds multiple challenges, but there are also a number of opportunities for improving and strengthening governance structures. Once established, integrated information systems can provide policy makers with more detailed data on educational outcomes and trajectories, in particular the effects of policies and programmes. This kind of information is crucial to assess whether existing policies effectively address a particular problem, or whether they need to be amended. Information systems that are actively used and accessible to practitioners such as the Early Warning Indicator System (EWIS) in the US state of Massachusetts (Chapter 7) can further strengthen the governance of skills systems by providing actors with immediate feedback. However, in order to be a reliable source for decision making, data need to be of high quality and very accurate.

The establishment of information systems can also facilitate collaboration between stakeholders. In a world where information about the education system and the labour market involves a high degree of uncertainty, effective collaboration between policy makers and stakeholders may be blocked by competing problem definitions. Rather than devising solutions to these problems, stakeholders engage in conflicts about the nature and extent of the underlying problem. In the German case, for instance, employers and unions have long disagreed about the actual number of youth in need of a training place in the apprenticeship system (Busemeyer, 2009[4]). As this case study has shown, the German Alliance for Initial and Further Training, established in 2014, represented an important space and opportunity for actors to develop an agreement about the data sources to be used and how they should be interpreted. Similarly, in the Norwegian case, the Future Skills Needs Committee that advises the national Skills Policy Council is instrumental in identifying future labour market needs, and relies on scientific analyses of current trends and multiple sources of data in combination with expert assessments from stakeholders from the corporatist system. The committee thereby tries to reconcile scientific analysis with stakeholders’ (and therefore practitioners’) own assessments and know-how to facilitate compromise in the policy reform process and prevent reform deadlocks. Another example for such an institution is the System for the Anticipation of Labour Market Needs (SANQ) in Portugal.

Building up integrated information systems in the area of skills policy can be challenging from a purely logistical and administrative perspective compared to other policy areas. This is because skills policies span multiple sectors of governmental activity, most obviously education and labour market policy, but also issues related to financing, taxation and economic policy making. Furthermore, governance challenges emerge mostly at the intersections of these subsystems rather than within them, for example when individuals move from general to vocational or higher education, when individuals transition from the education system to the labour market, or when skills investment policies need to be matched with financing and taxation issues.

Assembling data on educational transitions requires the tracking of educational and labour market careers over long periods of time as individuals proceed through different sectors of the skills system. In terms of data collection, this requires co-ordination and collaboration between different sectors and statistical agencies, i.e. education statistics, labour market and unemployment statistics as well as social security administration and, potentially, tax authorities. Many countries already collect statistics on these subsectors, but they are not harmonised and synchronised to be able to provide a comprehensive assessment of the skills system as a whole. Furthermore, even when the system infrastructure is up and running, participating educational institutions, such as schools, need the required resources (i.e. in terms of expertise) and time to reliably enter the data into the system.

The Estonian Education Information System (EHIS) (Chapter 2) has to a large extent achieved and implemented this goal. In Estonia, each citizen has an individual identification number that is always included in statistical information. This allows the tracking of the development of individuals across time, which most other OECD countries do not have. This example shows that the establishment of comprehensive information management systems requires a certain permissive and supportive culture among both the stakeholders involved in the governance of the system as well as the population at large. In other countries such as Germany, this kind of culture is less prevalent.

Across countries, the establishment of integrated information systems has the potential to strengthen the governance structures of skills system, but it is not the only solution to persisting challenges. As became clear in the US case study, information systems must be designed in a way that maximises their usefulness and therefore the likelihood that they will be actually used by practitioners, stakeholders and policy makers. In this regard, involving stakeholders is, again, crucial. The challenge for governments is to strike the right balance between what is feasible from a technological point of view and what is sensible and useful from the point of view of the actors directly involved in the system.

Aligning and co-ordinating financing arrangements

Across the country cases in this report, there was a central question of how to ensure that the establishment of new governance structures is accompanied with the necessary fiscal resources. The socio-economic context in which skills policies unfold clearly plays an important role in this regard. For instance, financial constraints may more be binding in the case of Portugal, which is still dealing with the aftermath of the debt crisis, compared to Germany or Norway, which benefit from easier economic circumstances. Thus, within public budgets the task for central governments is to mobilise and safeguard fiscal resources to be invested in skills policies against competing interests that may dominate in the short term. As mentioned above, skills policy is more likely to rise to the top of the government’s agenda in times of crisis, even though resources are scarce and need to be concentrated on fighting immediate short-term problems. Hence, the challenge for governments is to effectively commit to long-term strategies that ensure sufficient fiscal resources to address long-term and more structural problems in skills policies.

A second question is how to mobilise additional sources of funding besides public money. This is less of an issue in countries with a long tradition of employer involvement in vocational education and training such as Germany and, to some extent, Norway. It is more of a challenge in countries that lack such a tradition, for example Portugal. In these cases, the state has to step in to provide subsidies that encourage employers to participate in training. However, there are also complexities associated with this strategy, as employers might become accustomed to these subsidies and therefore even less willing to contribute in the future.

At the level of individual citizens and households, skills policies should set incentives for employees and citizens to participate in training, particularly lifelong learning. The Korean case study is a good example of how the provision of an encompassing lifelong learning system in combination with financial support for lowering the costs of education can lead to a high level of participation in lifelong learning activities by citizens. Aligning financial incentives across the potential sources of funding – public budgets, employers, individual citizens and other private sector stakeholders – should aim to lower barriers of access to training and channel resources to the sectors of the skills system where labour market needs are likely to be greatest. Establishing integrated information management systems will support this process.

copy the linklink copied!Policy recommendations

The following section develops and presents a number of recommendations for policy makers related to the four main challenges in strengthening the governance of skills systems identified at the beginning of this chapter. Each subsection presents one main recommendation aligned with the four challenges, as well as a number of more detailed recommendations related to the main recommendation.

Establish co-ordinating committees with a meaningful mandate and clear internal governance structures

In order to promote co-ordination and collaboration across levels of government and to implement a whole-of-government-approach to skills policy, it is advisable to establish committees (skills councils or similar co-ordinating committees or cross-departmental institutions) that span different levels and departments of government, as well as involve non-governmental stakeholders. The potential downside of establishing this kind of inter-governmental committee is a proliferation of decision-making bodies that could, in the worst case, weaken rather than strengthen the governance of skills policy. To prevent this from happening, some key factors should be considered.

Define a clear mandate

Skills councils should have a clearly defined mandate. This may be either broad (e.g. devising a national skills strategy) or more limited (e.g. coping with a particular problem such as the integration of refugees into the training system), but it should be clearly defined to provide stakeholders with a clear orientation. Furthermore, the mandate should be defined so that potential overlap with existing institutions and decision-making structures is minimised. In cases where overlap is unavoidable, the mandate should clearly define the relationship and mutual accountability between newly established and existing institutions. These issues may appear self-evident; however, the case studies in this report include a number of examples where there are conflicts between old and new institutions about the proper division of labour.

Defining a clear mandate for newly established councils and committees may be more challenging in countries with a strong tradition of corporatism and social partnership as there are already a number of existing bodies with respective mandates. In these country contexts it is particularly important to highlight the added value of additional committees to the overall governance of the system so that stakeholders are effectively motivated to commit resources. In countries with a more pluralist and liberal tradition of interest mediation, competing mandates of existing committees are less of an issue; however, in comparison to corporatist countries, newly established committees might face more obstacles in filling their mandates with substance (see next point).

Define mandate with real substance

The mandate for skills councils and inter-departmental committees should have real substance, which means that they need a certain degree of autonomy and leeway in setting their agenda, as well as policy-making authority. On the part of the central government, delegating some decision-making responsibility to skills councils involves a certain willingness to trust the involved stakeholders, both governmental and non-governmental. If stakeholders realise that their participation in skills councils has an impact, they are more likely to commit and contribute constructively. In contrast, if setting up new skills councils does not provide added value as the established institutions are reluctant to concede any of their control, they are more likely to deteriorate into mere “talking shops” with little impact on strengthening the governance of skills systems.

Ensure strong internal governance structures

In order to work effectively, skills councils and similar committees need clear and strong internal governance structures. Decision-making processes within the skills council need to be agreed on, ideally with the goal of maximising the commitment and involvement of stakeholders. Important issues include who sets the agenda (the central government or the committee as joint decision-making body) and how decisions are made (by consensus or majority).

The case studies indicate that a division of labour between the top political level and a working level is effective. Less frequent meetings of the top level (e.g. heads of ministries and peak level associations) set the strategic priorities for the work of skills councils and provide an effective means to enhance the visibility of skills policy in the media. At the working level, experts and advisors should meet regularly to implement the top level’s strategic priorities. These regular meetings also facilitate the development of a joint problem-solving perspective. Having two levels of decision making may also help to mitigate conflicts, with political level conflicts delegated to the working level for further discussion. Vice versa, if there is disagreement at the working level, political leaders ultimately decide on how these should be solved.

Promote the involvement and commitment of non-governmental stakeholders, while managing the risk of undue influence from special interests

The case studies have shown that skills councils and similar committees typically involve a number of non-governmental stakeholders, but there are significant differences regarding the relationship between governmental and non-governmental actors, with some councils more dominated by state actors, and others more equally balanced in the spirit of corporatist decision-making. A crucial insight from the case studies is that stakeholder involvement is essential to strengthen the governance of skills policies, for the reasons mentioned above. However, stakeholder involvement needs to be meaningful for stakeholders to commit and contribute, and governance structures need to be designed to prevent special interest groups from capturing public institutions and gaining privileged access. To achieve this delicate balance, the following recommendations are made.

Promote meaningful stakeholder involvement

Governance structures should be designed to allow stakeholders to participate and contribute in a meaningful way. The primary purpose of stakeholder engagement should not be to legitimise decisions and policies that were de facto decided beforehand. This kind of perfunctory involvement does not set strong incentives for stakeholders to commit themselves and provide resources. It also makes it difficult for intermediary associations such as employers’ associations and unions to mobilise individual members (employers, workers) to participate in the production of collective goods, such as skills.

To achieve meaningful stakeholder involvement, stakeholders should be given the opportunity to have an input in the process of agenda-setting in joint skills councils. Although policy makers define the mandate (see previous point) and the overall strategic framework of skills councils and inter-departmental committees and agencies, these governance structures should be sufficiently open to allow for and even promote bottom-up initiatives, giving stakeholders the autonomy and ultimately the resources to try out innovative policy ideas. The German Alliance for Initial and Further Training, for instance, allows involved stakeholders to develop and experiment with practical solutions to concrete problems in labour market placement services.

Limit the overall number of stakeholders involved

Meaningful stakeholder involvement implies that not all potential stakeholders can and need to be involved. As discussed above, the identification of relevant (and less relevant) stakeholders is an important part of the process of designing effective governance structures. The exact number of stakeholders to be involved depends on the country context, i.e. the size and diversity of the labour market and the economy as a whole, political institutions (centralised vs. decentralised), and existing relationships between business and government. However, it is important to understand that maximising effective stakeholder involvement is not the same as maximising the number of stakeholders.

Simply maximising the number of stakeholders could, in the extreme, lead to a weakening of stakeholder engagement as there would be very little leeway for individual stakeholders to have a measurable impact on the policy process. In such an extremely pluralist setting, the influence of stakeholders would be commensurate to their economic power, leading to a de facto concentration of political influence in the hands of the economically powerful. To a certain extent it is legitimate that economic power correlates with political influence. However, as the economic theory of collective action argues (Olson, 1965[5]), special interest groups (i.e. small groups with clearly defined economic interests) have more economic and political clout in a pluralist setting than groups representing diffuse, collective interests. Hence, governance structures should be established to ensure that stakeholders representing economically less influential individuals and groups are also given a seat at the table. Typically, these stakeholders would be labour unions, but representatives from small and medium-sized enterprises, parental groups or other non-governmental civil society organisations could also be included.

Putting an upper limit on the number of stakeholders to be involved also ensures that joint skills councils can develop into forums where real deliberations take place, rather than a simply superficial exchange of positions. A smaller number of stakeholders also promotes the development of mutual trust between the involved actors, which is an important precondition for effective and legitimate decision making in these councils. Finally, it is important to note that limiting the number of stakeholders at the top also promotes organisational development at lower levels. For instance, if unions, employers or local governments are only allowed to send one or two delegates each to the top level of decision making in skills councils, they are pressed to organise, mobilise and interrogate their individual members and lower-level member associations. This is likely to be more effective in co-ordinating the different viewpoints and positions of individual members and lower-level associations in a certain part of the economy, rather than giving each actor a direct seat the table at the top level.

Prevent gridlock and manage the risk of undue influence by special interests

When delegating substantial decision-making authority to stakeholders, governments need to ensure that this does not lead either to gridlock – i.e. when stakeholders mutually block each other and therefore the whole process – or capture by special interests – i.e. when stakeholders “hijack” joint decision-making bodies to promote their own particularistic interests and positions rather than collective concerns. Both are prevented when and if the government effectively retains the ultimate political supremacy over the whole process. In a sense, governmental actors therefore need to remain “first among equals” among the stakeholders involved in joint decision-making bodies, walking a thin line between granting autonomy to non-governmental stakeholders on the one hand, while retaining the political capacity and authority to take over the whole process should problems develop on the other.

Retaining ultimate control over the process allows governments to cast a “shadow of hierarchy” (Scharpf, 1997[6]) on negotiation partners, which may be more or less constraining. For instance, the government could delegate certain tasks (defining a clear mandate, see above) such as reforming occupational standards in a particular economic sector to stakeholders, providing them with effective decision-making powers on the condition that they achieve a consensual solution. If, however, individual actors block joint decision making to maximise their own particularistic goals, the government should be able to step in to prevent the process from being unduly influenced by special interests.

Pursuing such a “neo-corporatist” approach requires a certain degree of self-restraint on the part of government, as well as a certain understanding on the part of stakeholders that participating in joint decision-making bodies is not akin to transactional bargaining, but ultimately aimed at developing instruments and strategies for collective problem solving.

Allow stakeholder engagement to develop over time

Building trust among non-governmental stakeholders, as well as between public and private actors, takes time. Countries with a long tradition of social partnership such as Germany and Norway find it easier to mobilise the commitment of private actors and stakeholders to strengthen the governance of skills policy. However, even in these countries, new committees such as the Skills Policy Council in Norway and the Alliance for Initial and Further Training in Germany need time to develop their organisational identity. In countries without a long tradition of social partnership, this process is likely to take even longer. However, as the case study of Portugal has shown, it is not impossible to establish working corporatist institutions if the government stays committed to providing the necessary resources over several years (see also below).

The success of new governance bodies such as skills councils are often the result of engaged stakeholders that are used to systems of social partnership and stakeholder involvement. Building on previous experience, they actively try to shape the council’s and committee’s mandate. This kind of stakeholder activism might be lacking (or less pronounced) in countries that do not have the same strong tradition of corporatism. Such problems are visible regarding expanding the Norwegian system of corporatism to regional skills policy (i.e. vertical co-operation), where most local authorities have to be actively convinced to put skills policy on their agenda. Ultimately, countries with a shorter history of social partner bargaining and corporatism have to be prepared to provide substantially more time and resources to allow for the success of similar policy measures.

Support the establishment of information management systems, but ensure that they provide usable and relevant information to stakeholders and policy makers

The case studies on Estonia and the United States focused explicitly on the role of comprehensive education information management systems in strengthening the governance of skills policy. However, information and data management systems also play a crucial role in establishing a common knowledge base among stakeholders in the other case studies. Overall, governments should invest more in building capacities in information and knowledge management, but they must pay particular attention to the usefulness of the collected data for all concerned. This can be ensured by involving stakeholders (i.e. the users of the data to be collected) in the design of the systems.

Involve stakeholders in the design of integrated information systems

The engagement of stakeholders throughout the whole policy cycle helps to mobilise and commit societal resources, which strengthens the overall governance of skills policy. In this spirit, stakeholders should also be involved in designing and updating integrated information systems that aim to inform governance decisions. Stakeholders, both governmental and non-governmental, can help to identify the different kinds and types of data needed from their perspective to better inform governance decisions.

Involving stakeholders in the design and upgrading of information systems increases the likelihood that these systems will be actively used by the stakeholders in the system. However, it is important to avoid a situation where actors (from students, parents and teachers to workers and firms) perceive information management systems as systems of control that outside actors might use to monitor and constrain their actions.

Use information management systems to inform rather than automate governance decisions

Integrated information systems have great potential to improve governance decisions by reducing uncertainties, for example regarding the effects of policy choices on labour market outcomes. However, opposition to the introduction of integrated information systems is more likely to occur when the public as well as the actors directly involved in the skills system perceive them (rightly or wrongly) as premeditating or even “automating” decisions that should be taken by stakeholders themselves.

Therefore, making use of information management systems to inform joint decisions by stakeholders is a more effective and legitimate strategy to strengthen the governance of skills systems than using these systems to circumvent collective decision making. This is largely because collective decision making usually involves a significant degree of redistributive bargaining that cannot be “outsourced” to automated systems. Furthermore, interactions between stakeholders can and do contribute to the development of a joint problem-solving perspective, at least in the long term. This requires continuous interaction between the actors involved, which again cannot be automated. However, using information management systems to establish and further develop a joint and shared knowledge base is useful in supporting interactions between stakeholders as it contributes to reducing contentious distributive conflicts.

Even when data are available, policy makers and end users on the ground need to be able to effectively use these data. The case studies of Estonia and of Massachusetts in the United States have shown that data are often available, but that their usage is much lower than hoped for by the government. Easy access to data, which might include visualisations, is important to get decision makers and practitioners to actually make use of the data. User-friendliness on both sides – for those who enter the data and for those who analyse and use the data – is critical for making the establishment of information management systems a sustainable success.

Make use of different kinds of data

Opposition to the broad introduction of information management systems may also occur because most of the data collected in these systems are quantitative in nature, i.e. based on the large-n measurements of, for example, competencies, educational attainment and labour market outcomes. Practitioners with considerable work experience in the field may regard quantitative data on student achievement as poor proxies for their own professional (qualitative) assessment of a certain student. Hence, information management systems that rely solely on quantitative sources of data may easily be perceived as undermining professional standards, as well as the status of professionals themselves.

Information management systems should be built so that they are able to collect and process data from different sources, including qualitative assessments from professionals. This is ultimately not a matter of technology, but a question of governance, i.e. which data sources will be used to achieve particular performance goals.

Invest and commit the diversified resources needed to strengthen skills policies

Strengthening the governance of skills policy takes time and commitment on the part of governments. The case studies have revealed that this can be challenging at times, depending on the country context. These challenges can be tackled when following a number of recommendations.

Provide adequate resources by setting long-term budgetary goals

As mentioned above, a significant challenge in the domain of skills policy and budgetary priorities is that governments are more likely to prioritise skills policies in times of economic crisis and fiscal austerity, i.e. when there is an acute need to counter short-term problems such as (youth) unemployment. Even though skills policy can effectively contribute to solving these issues, its effectiveness is likely to be dampened in times of fiscal austerity as there are fewer resources available for skills investments. At the same time, the individuals and groups who benefit from skills policies during times of crisis are likely those who would fare better under better economic conditions anyway. The pro-cyclical approach to skills policy also implies that skills policy is less likely to be at the top of the government’s agenda when economic conditions improve. This is unfortunate because structural long-term problems (e.g. bleak employment prospects for disadvantaged youth or a lack of further training opportunities for the long-term unemployed) are likely to persist. In principle, governments are more likely to be able to mobilise fiscal resources for skills policy in good economic times when the political incentives to do so are weaker.

Hence, the first recommendation to weaken the pro-cyclical approach to skills policy is to set budgetary priorities according to long-term strategic goals that are jointly decided in collective decision-making bodies such as the skills councils and inter-departmental committees mentioned above. These strategic goals should be focused on structural long-term challenges in the system and supported by the data collection in integrated information systems. Funding for skills policy could and should include a cyclical component to allow policy makers to respond to short-term crises, but the bulk of funding should be devoted to solving structural issues.

Furthermore, setting up new governance structures and committees needs to be supported by the provision of adequate fiscal resources. The establishment of comprehensive information management systems in particular is costly and requires significant and long-term commitment on the part of government. Similarly, setting up advisory boards such as skills councils should be accompanied with the provision of personnel and financial resources to allow these committees to work effectively.

Finally, the government can provide resources in terms of dedicated staff and personnel who are tasked with organising and facilitating collaboration and co-operation across levels of government, but also between the public and private sectors. For instance, in the Korean case study of lifelong learning in Suwon City, dedicated lifelong learning educators perform important bridging functions between the stakeholders involved by facilitating access, designing lifelong learning programmes and taking care of budgetary aspects.

Tap into multiple sources of funding while keeping equity concerns in mind

To some extent, the constraints of public financing can be mitigated if governments tap into multiple sources of funding. Skills investments create concrete benefits for different sectors and stakeholders, i.e. the public sector, businesses and individual households. Each of these sectors can therefore be expected to contribute to its financing. Countries differ widely regarding the exact division of labour between these different sources, as the case studies in this report have confirmed.

Broadly speaking, funding strategies that are concentrated on one funding source are more likely to lead to negative side effects than a more balanced approach. For instance, a strong reliance on public sources of funding might be suited to prevent inequalities in the distribution of funding, but this approach might lead to particularly binding funding constraints as public budgets are increasingly under pressure. Vice versa, focusing mostly on contributions from individuals and private householders (e.g. via tuition fees) could aggravate equity concerns related to accessing education and, in the worst case, overload private households with student debt. Finally, putting too much pressure on employers to commit and contribute resources to skills development could trigger counter-productive reactions as employers might pull out of collective training schemes.

Hence, a balanced approach to making use of multiple sources of funding seems advisable. This approach allows for individual contributions when and if skills investments lead to tangible and significant benefits (i.e. wage increases) at the individual level. When mobilising funding from private households, particular attention should be paid to equity concerns to prevent credit or other funding constraints. A balanced approach to funding also requires employers (businesses) to contribute to the financing of skills development, for example by paying for work-related on-the-job training, by investing in apprenticeship training, or by contributing to the financing of out-of-firm training workshops in particular regions or sectors of the economy. Furthermore, as the Korean case study has shown, costs for training facilities can be reduced if existing facilities (schools, libraries, cafés, etc.) are used as places of learning.

Finally, a balanced approach to funding skills development requires the public sector to commit adequate and commensurate resources (see previous point), but more importantly, to counteract potential inequalities in the distribution of funding across individuals, regions and sectors. This does not necessarily imply a full equalisation in the distribution of funding, which is neither achievable, nor – in the extreme – advisable. Rather, the contribution of the public sector to funding skills policy should strive to improve the most significant and obvious imbalances in the provision of fiscal resources, which are often related to and driven by differences in socio-economic background conditions.

A concrete example of how this could be achieved is to develop funding formulas for schools and other educational institutions that do not simply provide a fixed amount of funding per student, but that weight funding according to the socio-economic profile of the student body. Another example would be to provide subsidies for education, training and lifelong learning opportunities targeted at those most in need of such opportunities.

Most countries already have these kind of subsidy regimes, but they are mostly focused on initial education and training and neglect the growing sector of lifelong learning and further education. The Korean case study in this report provides a good example of how a balanced approach to funding lifelong learning that includes both individual and public contributions can help overcome barriers to accessing this type of learning. Regarding further training in the labour market, public funding could effectively compensate for inherent biases, such as employers focusing skills investments on highly-skilled workers, or resource-rich individuals being more likely to have the necessary resources to be able to invest further in their education.


[4] Busemeyer, M. (2009), Wandel trotz Reformstau: Die Politik der beruflichen Bildung seit 1970 (Change Despite a Backlog of Reforms: the Policy of Vocational Training Since 1970), Campus, Frankfurt a.M., New York.

[1] OECD (2019), OECD Skills Strategy 2019: Skills to Shape a Better Future, OECD Publishing, Paris,

[5] Olson, M. (1965), The Logic of Collective Action: Public Goods and the Theory of Groups, Harvard University Press, Cambridge, MA.

[6] Scharpf, F. (1997), Games Real Actors Play: Actor-centered Institutionalism in Policy Research, Weastview, Boulder, CO.

[2] Schmitter, P. (1979), “Still the Century of Corporatism?”, in Schmitter, P. and G. Lehmbruch (eds.), Trends Toward Corporatist Intermediation, Sage, Beverly Hills, London.

[3] Streeck, W. and P. Schmitter (1985), “Community, market, state – and associations? The prospective contribution of interest governance to social order”, in Streeck, W. and P. Schmitter (eds.), Private Interest Government: Beyond Market and State, Sage, London, Beverly Hills, New Delhi.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at