Policies for Developing Advanced STEM Skills in Higher Education

3. Developing a strategy for advanced STEM skills

Abstract

This chapter examines how strategies for advanced science, technology, engineering and mathematics (STEM) skills can support policy co‑ordination, priority setting and continuing policy learning across education, research, innovation and employment. It presents a strategic approach for diagnosing constraints, selecting and sequencing responses, clarifying responsibilities, and aligning policy ambitions with implementation capacity and resources. The chapter concludes with eight areas for policy action that can be used to guide the design, implementation and evaluation of a strategy for advanced STEM skills.

Key issues

A strategy for advanced STEM skills can connect policy portfolios and levels of government, clarify responsibilities, combine resources and strengthen implementation. This requires a clear diagnosis, agreed priorities, defined responsibilities and adequate, sustained resources.

Policy priorities need to be realistic and supported by sufficient capacity. Ambitious objectives will have limited effects without adequate educators, technical staff, supervision, infrastructure, learner support and institutional leadership. Expanding participation, advanced study or skills-renewal provision without these conditions may limit access to high-quality learning and risk to move constraints to completion, progression, and employment.

Monitoring, evaluation and policy learning are essential to improve interventions. Monitoring can track participation, competence development, progression, skills use and renewal across fields, places and population groups. Evaluation can examine implementation, interactions and unintended effects, including whether constraints are removed or transferred. Higher education institutions support policy learning by interpreting evidence, identifying implementation challenges, improving practice and informing policy adaptation.

International mobility and collaboration matter for advanced STEM skills. As global competition for talent intensifies, strategies should connect attraction with recognition of qualifications and experience, integration into education, research, employment and society, conditions for retention and continued development, and opportunities for return and collaboration. They should coordinate responsibilities across policy portfolios, consider effects on origin and destination countries, and support mutually beneficial talent circulation and partnerships.

The development and effective use of advanced STEM skills depend on policy decisions made across education, science, technology, innovation and employment portfolios. Learners need access to suitable preparation, high-quality programmes and viable routes to advanced study and skills renewal. Higher education institutions need sustainable pedagogical, organisational and resource capacity. This includes qualified academic and technical staff; effective curriculum design, teaching, assessment, feedback and supervision; programme leadership and co‑ordination; connected academic, practical and psychosocial support; suitable infrastructure and research environments; and the data and analytical capacity to identify barriers, monitor learning and progression, and improve programmes. Employers need to recognise the resulting competence, recruit people with appropriate foundations and support their transition into work so that advanced STEM skills are used and continually developed.

A strategy for the development of advanced STEM skills can establish shared priorities, clarify responsibilities and enhance co‑ordinated action. It can support decision-making under uncertainty by sustaining long-term capacity while allowing priorities to be adjusted as needed. The contribution of a strategy depends on the choices made beneath its broad objectives. A strategy can support evidence‑based priority setting, connect complementary interventions, assign responsibilities and sustain attention over time. The STEM Education Strategic Plan presented by the European Commission illustrates these contributions. Its objectives and targets can focus political attention, support co‑ordination and encourage action towards shared objectives.

Broad strategic objectives can establish direction, but they do not determine which constraint should be prioritised, which mechanism should be addressed or which actor should act. Nor can they remove uncertainty about future technological developments, applications or demand for particular combinations of competence. The following sections set out how policymakers can diagnose constraints, select and sequence appropriate responses, align responsibilities and ambitions with implementation capacity, and use monitoring, evaluation and policy research to support continuing policy learning. They examine when a strategy for advanced STEM skills may be useful and present eight areas for policy action to guide design, implementation and monitoring.

International mobility and collaboration matters for advanced STEM skills. In a context of increasing global competition for advanced STEM talent, strategies may seek to attract international learners, researchers and professionals, but should connect recruitment with the recognition of qualifications and experience, integration into education, research, employment and wider society, suitable conditions for retention and continued development, and opportunities for return and ongoing collaboration. These elements involve responsibilities across policy portfolios. Strategies should also consider the effects of mobility on origin and destination countries and support mutually beneficial circulation and international collaboration.

Diagnosing the constraint before selecting a response

Indicators of enrolment, graduation and employment can show general patterns and changes across countries, regions and periods. When sufficiently detailed, they can show whether participation is increasing, how representation differs among population groups, whether graduate numbers are changing and whether employment is growing or declining in broadly defined fields and occupations. These indicators can identify developments that require closer examination and help monitor progress towards broad policy objectives. However, they cannot explain why these patterns have occurred. Explaining these patterns requires closer analysis of the mechanisms involved. These may arise in areas that involve different responsibilities across public authorities, higher education institutions and employers.

An observed pattern may indicate a possible bottleneck, but it does not establish whether a significant and avoidable constraint exists, identify its cause or determine the appropriate response. Diagnosis should test competing explanations, establish whether and for whom a bottleneck exists, distinguish contributing constraints by field, level, population and location, and identify the mechanisms through which they operate. A sufficiently specific diagnosis can establish whether public intervention is justified, where responsibility for addressing the identified constraints lies, and which actors have the relevant instruments and implementation capacity.

The questions presented in Table 3.1 can guide diagnosing a constraint before selecting a response. The context-specific relevance of these questions depends on the responsibilities and decision-making scope of the actor undertaking the diagnosis. Public authorities and education planners may apply them to system-wide patterns, resource allocation and regulatory conditions, while higher education providers may apply them to programme design, teaching, assessment, learner support and institutional capacity. The questions remain broadly the same, but their application, evidence and possible responses depend on the actor and context.

Table 3.1. Common questions for diagnosing a constraint before selecting a response

Diagnostic question	What to examine	What the diagnosis should establish
What is the observed pattern?	Analysis may concern entry, persistence, completion, progression, recruitment, skills use, skills renewal, advancement or retention.	A precise description of the pattern, distinct from its possible causes and proposed responses. Where international mobility is relevant, analysis should also examine effects on origin and destination countries.
Who is affected?	Evidence should establish whether learners, graduates or workers with different characteristics or circumstances differ in how they are affected by the pattern.	Which groups are affected and how they are affected.
Which mechanisms could explain the pattern?	Several mechanisms may interact, and not all may be easily detectable.	The mechanisms contributing to the pattern, how they interact and whether their effects become visible at a different stage or level.
What is the evidence?	Diagnosis should combine evidence from education systems, higher education institutions, learners, graduates, workers and employers.	What does the available evidence explain; what evidence is missing and how could it be obtained.
Which policy instruments can be used?	Relevant instruments may include funding, regulation, information, guidance, quality assurance, recognition arrangements, institutional capacity building, learner support, support or incentives for employer action, and co‑ordination among actors.	Which instruments address the diagnosed pattern and underlying mechanisms and why they are more appropriate than alternative responses.
How should action be combined and sequenced?	A response may depend on complementary action by several actors. Policy should consider how the proposed intervention interacts with existing arrangements, which conditions must be established for it to operate as intended, and whether action at one point could create additional pressure elsewhere.	Which actions should precede, accompany or reinforce others and how additional pressure elsewhere can be avoided.
What capacities and resources are required?	Responsibilities are distributed and several actors may need to act where constraints are connected or mutually reinforcing. Analysis should distinguish between capacity that already exists, capacity that can be redirected and capacity that must be developed before an intervention can operate effectively or at scale.	Which actors are relevant; which capacities and resources already exist, which can be redirected and which must be developed before or during implementation.
What outcomes and potential trade-offs should be monitored?	Monitoring should examine whether the identified constraint is reduced, whether effects differ across groups, whether the intended competence and educational quality are maintained, and whether the intervention produces unintended effects or transfers pressure to another part of the educational or professional route.	The baseline, intended outcomes, distributional effects and potential trade-offs.

Source: Authors based on discussions in the Mutual Learning Group on Advanced STEM skills development in higher education.

International policy inventories can complement national diagnosis by showing which approaches have been adopted in other contexts. UNESCO’s Global Observatory of Science, Technology and Innovation Policy Instruments (GO-SPIN) and the EC-OECD STIP Compass provide information on national science, technology and innovation policies and instruments, including initiatives related to STEM skills. Used alongside national evidence and stakeholder engagement, these resources can help policymakers identify potentially relevant instruments, compare the composition of policy responses and locate areas receiving limited attention.

These resources can inform the search for possible responses, but selecting among them still requires analysis of the relevant mechanisms, responsibilities, capacity and context. The questions in Table 3.1 can help to structure this analysis. Several constraints may interact, relevant evidence may be held by actors who are not responsible for addressing the constraint, and a pattern visible in one location may originate elsewhere. Chapter 1 distinguishes individual-facing constraints from those arising within higher education institutions, employers and wider system arrangements, and applies these categories across the five bottlenecks. This distinction helps locate the constraints, identify the actors with the relevant instruments and capacity, and determine whether complementary or sequenced action is needed.

A sufficiently specific diagnosis provides the basis for evaluation. It identifies the initial situation, the mechanism an intervention is expected to change, the actors responsible for implementation and the outcomes through which progress can be assessed. The following section examines how evaluation and policy research can test these assumptions and support policy learning.

Evaluation and policy research as foundations for policy learning

Evaluation tests the explanations and expected mechanisms established through diagnosis. Outcomes in the development and effective use of advanced STEM skills are rarely explained by a single factor. As discussed in Chapter 1, learners enter programmes with different preparation, resources, prior opportunities and personal circumstances. Institutions differ in staffing, infrastructure, programme design and learner support. Employers differ in recruitment, onboarding, management and work organisation. These conditions interact, and an intervention may influence several outcomes at different points in time. Evaluation therefore needs to examine not only whether outcomes change, but also how change was expected to occur and the conditions under which the intervention was implemented.

Policy research has an important role in this process. It can develop and validate measures of learning and competence; examine curriculum, teaching, assessment and programme design; identify mechanisms affecting entry, persistence, completion, progression and skills renewal; and analyse recruitment, onboarding, work organisation, skills use, continued development, advancement and retention. It can also examine how governance, funding, regulation and institutional and organisational capacity shape these outcomes across education, research and employment. Drawing on education, labour-market, organisational, workplace, and science, technology and innovation research, it can help distinguish constraints arising from individual circumstances from those arising through institutional arrangements, employer practices or wider system conditions.

Policy learning requires sustained connections among researchers, policymakers, higher education institutions, employers, learners and workers. Evaluation questions should correspond to the constraint being addressed, while findings should be available in forms that can inform funding, regulation, programme design and organisational decisions. Incorporating evaluation into policy design can preserve baseline evidence, document implementation and create opportunities for comparison that may no longer be available if evaluation begins only after implementation.

Evaluation should be integrated into policy design rather than introduced only after implementation. Public authorities and institutions need the capacity to formulate evaluable objectives, identify the expected sequence of change, establish appropriate comparisons, examine implementation and follow outcomes beyond an intervention’s immediate effects. Without a clear account of the constraint being addressed and how an intervention is expected to affect learning, progression, employment or skills use, it is difficult to determine whether weak outcomes reflect an incorrect diagnosis, inadequate implementation, insufficient time, limited institutional or organisational capacity, or an intervention that is ineffective even when delivered as intended. Box 3.1 summarises the principal methodological challenges involved.

Box 3.1. Methodological challenges in evaluating policies for the development and effective use of advanced STEM skills

Determining whether a policy has improved the development or effective use of advanced STEM skills is rarely straightforward. Observed changes may reflect the intervention, differences among participants and institutions, wider contextual developments or interactions among these factors. Evaluation therefore needs to consider attribution, implementation, the timing of impacts and the extent to which findings apply in other settings.

• Several interacting conditions may influence outcomes. Entry, completion, progression and effective skills use are rarely explained by a single condition, making mono-causal explanations generally inadequate. The effect of an intervention may also depend on complementary arrangements. Financial support may make participation feasible but have limited effects where teaching or scheduling remains inaccessible. Employer-provided learning may develop competence without improving its use where workers lack autonomy or opportunities to apply it. Evaluation therefore needs to distinguish the effects of the intervention from those of the conditions on which its effectiveness depends.

• Impacts may take time to emerge. An intervention may influence connected outcomes at different points. Changes in preparation or guidance may affect applications and programme choice before influencing completion. Better onboarding may initially affect role clarity and workplace relationships before influencing performance, skills use or retention. Evaluation should reflect this expected sequence and allow sufficient time for later effects to become visible.

• Selection bias can affect observed outcomes. Selection bias arises when participants differ systematically from non-participants in ways that also influence the outcomes being measured. Participants in optional interventions may differ in prior attainment, motivation, available time, financial resources or access to networks. Institutions and employers adopting a policy may likewise differ from non-adopters in capacity, leadership or commitment. Observed differences may therefore reflect these pre‑existing characteristics as well as the effects of the intervention.

• Implementation may vary across settings. Interventions with the same formal design may differ in staffing, intensity, participant engagement, local leadership and access to complementary support. A weak average effect may conceal effective implementation in some settings and incomplete delivery in others. Differences in outcomes may therefore reflect variation in reach, participation, implementation quality and local adaptation, as well as the effectiveness of the intervention itself.

• Evidence on effects, mechanisms and implementation can be difficult to combine. Quantitative evidence can estimate changes in outcomes but may provide limited explanation of why they occurred. Qualitative and process evidence can reveal participant experiences, implementation differences, causal mechanisms and contextual conditions, but may not establish the scale or distribution of effects. Evaluation may need to combine these forms of evidence to determine whether change occurred, how it occurred and under which conditions.

• Longitudinal evidence may be limited. Effects on progression, employment, skills use, advancement and retention may emerge several years after participation. Short observation periods can capture immediate changes while missing whether they persist or influence later educational and occupational routes. Following outcomes over time may be constrained by fragmented data, participant attrition and difficulties linking education and employment records.

• Pilots provide limited evidence of effectiveness at scale. Pilots can show whether recruitment, delivery and evaluation arrangements are feasible and help identify possible mechanisms and necessary adaptations. However, their participants, staff and institutions may not represent those involved in wider implementation. Expansion may alter staffing requirements, participant composition, delivery quality and institutional ownership. Pilot findings therefore do not establish effectiveness or scalability on their own.

• Findings may not transfer directly across contexts. Evidence from one institution, field, region or population does not establish that an intervention will produce the same effects elsewhere. Differences in governance, funding, institutional capacity, learner composition and labour markets may alter how it operates. Assessing wider applicability requires attention to the mechanisms and contextual conditions on which effects depend, the adaptations that can be made without weakening essential functions, and the capacities required for expansion.

Source: Skivington et al. (2021[1]), “A new framework for developing and evaluating complex interventions: Update of Medical Research Council guidance”, https://www.bmj.com/content/374/bmj.n2061; and Unlu, F. et al. (2021[2]), “Can quasi-experimental evaluations that rely on state longitudinal data systems replicate experimental results?”, https://doi.org/10.1002/pam.22295.

The evaluation method should match the policy question, intervention and context. Experimental, quasi‑experimental, longitudinal, comparative, statistical and qualitative approaches may be used separately or in combination. The appropriate choice depends on the nature and scale of the intervention, how participants are selected, the availability of baseline and comparison data, the timing of expected effects and variation in implementation across settings. No single method is appropriate for every policy question. The strength of the evidence depends on whether the selected design supports credible conclusions about the question being examined and makes its assumptions and limitations explicit.

The methodological challenges summarised in Box 3.1 reinforce the need to connect evaluation with diagnosis, implementation and policy decisions. They also show why isolated interventions may be insufficient where outcomes depend on complementary conditions, responsibilities are distributed across policy areas and effects emerge across several stages. A strategic approach can provide a framework for connecting these elements, aligning action with implementation capacity and supporting policy learning over time.

Why a strategy for advanced STEM skills may be needed

Advanced STEM skills are developed, demonstrated, recognised, used and renewed through connected education, research and employment systems. Constraints affecting these processes may originate in school preparation, higher education capacity, programme organisation, access to research and professional experience, recognition arrangements, recruitment practices, onboarding, work organisation, employment conditions or access to skills renewal. Responsibility for addressing them is consequently distributed across education, research, innovation, employment, industrial and regional policy and among public authorities, higher education institutions, research organisations, employers and professional bodies.

A strategy can provide value where these distributed responsibilities and interventions need to be connected around a shared diagnosis. It can establish a common understanding of the advanced STEM skills concerned, identify where consequential constraints arise and clarify how the differentiated responsibilities of the relevant actors should be aligned. It can also support continuity where developing institutional, educational, research or employer capacity requires action extending beyond individual programmes or short funding periods.

A separate STEM strategy is not necessarily required where existing education, research, employment and innovation strategies already provide a shared diagnosis, priorities, clear responsibilities, sustained implementation capacity and connected monitoring. The relevant question is whether existing strategic arrangements can address the identified constraints coherently across policy domains.

The purpose of diagnosis is not to demonstrate that every country has a general shortage of advanced STEM skills or that participation should increase across all STEM fields. It is to determine which disciplinary, interdisciplinary, transversal and professional competence is required; in which fields, qualification levels, occupations and locations; which populations and organisations are affected; and which mechanisms produce the observed constraint. Low entry, weak completion, limited progression, recruitment difficulties or low participation in continuing education do not by themselves establish their causes or the appropriate response.

A strategy should therefore begin with observed outcomes and examine competing explanations. Recruitment difficulties, for example, may indicate an insufficient availability of particular advanced STEM skills, but may also reflect narrow requirements, demand for directly comparable experience, location, pay and working conditions, weak onboarding or limited organisational capacity to use available competence. Low progression to advanced study may reflect financial, informational or institutional barriers, but may also result from attractive employment opportunities. Limited participation in skills renewal may arise from unsuitable provision, uncertain recognition, cost, time constraints, weak employer support or limited opportunities to apply the learning.

This diagnosis should be differentiated by field, level, population and location. Aggregate STEM indicators can support political direction and comparison, but cannot identify the particular competence required or the mechanism producing a constraint. The strategy should therefore combine evidence from education, research, employment, employers and workers and examine qualifications, experience, tasks, actual skills use, working conditions and progression over time.

A shared diagnosis provides the basis for selecting and sequencing priorities. It should identify which actors possess the relevant authority and instruments, what implementation capacity is required and how action at one point may affect another. Increasing entry without sufficient teaching and learner support may transfer the constraint to persistence and completion. Expanding advanced study without adequate supervision and research infrastructure may delay completion or weaken quality. Increasing graduate supply without suitable roles, effective onboarding and substantive work may transfer the constraint into employment and skills underutilisation.

A strategy should consequently be selective rather than comprehensive by default. It need not introduce a new intervention for every bottleneck or reproduce activity already functioning effectively under general education, employment or social policy. Its purpose is to identify the constraints that require connected or additional action, establish responsibilities and resources, and monitor whether advanced STEM skills are developed, recognised, used and continually renewed. Countries may adopt a dedicated STEM education strategy, incorporate advanced STEM skills into a broader skills, science, innovation or industrial strategy, or develop strategies for particular technologies and fields. These formats can coexist.

The gains and potential trade‑offs of a national strategy

A national strategy for advanced STEM skills can provide a common framework for policy areas and institutions that would otherwise act separately. It can connect the foundations developed in school with routes through VET and higher education, relate initial education to skills renewal, and bring research training into a wider account of advanced STEM skills. It can also establish links with research, innovation, industrial and regional policies that influence where advanced STEM skills are developed and used.

The first potential gain is a shared and differentiated diagnosis. A strategy can bring together evidence on participation, learning, progression, research capacity, educational and professional destinations, and the application of advanced STEM skills. This can help distinguish among constraints affecting different fields, qualification levels, regions and population groups. It can also enable governments to test whether an observed difficulty arises primarily from weak educational foundations, unequal access, insufficient institutional capacity, limited progression routes, conditions in research training, or constraints on the effective use of advanced STEM skills after education.

A second gain is the possibility of connecting action across the five bottlenecks. Measures intended to widen entry may have limited effects if learners encounter weak teaching or insufficient support after admission. Expansion at postgraduate level may move a constraint into completion if research funding, infrastructure and supervision do not expand accordingly. Short learning may increase participation in continuing education without enabling skills renewal if its level, recognition and relationship with larger qualifications remain unclear. A strategy can make these dependencies visible and support combinations of action that would be difficult to co‑ordinate through isolated programmes.

A third gain is clearer direction and responsibility. A strategy can identify a limited set of national priorities while allowing institutions to adapt their responses to particular fields and learners. It can clarify which public authorities are responsible for funding, regulation, quality assurance, qualifications, data and research, as well as which responsibilities rest with education and research institutions. It can also identify where co‑operation with employers and other actors is required without transferring their responsibilities to education providers.

Finally, a strategy can provide continuity and a basis for policy learning. Many changes to advanced STEM education require sustained commitments. Preparing qualified teachers and academic staff, building research environments, developing specialist infrastructure, renewing programmes and establishing new progression routes cannot normally be achieved through short interventions. A strategy can provide a longer time horizon, while periodic monitoring and review can allow priorities to change as evidence, technologies and societal needs develop. These gains are accompanied by two potential trade‑offs.

The first concerns the relationship between strategic breadth and sufficiently specific diagnosis. A broad strategy can give STEM political visibility and connect policy across educational stages and government portfolios. At the same time, it can encourage STEM to be treated as a sufficiently coherent category for policy action. Aggregate patterns may conceal major differences among disciplines, qualification levels, types of competence, regions and population groups. An apparent shortage in one specialised field does not establish a need for general expansion across STEM, just as weak participation in one educational route does not show whether the principal constraint lies in preparation, programme capacity, financial support or subsequent opportunities.

A broad strategy may therefore encourage expansion across STEM without identifying the location or cause of the constraint. Conversely, it may develop narrowly specialised provision around uncertain forecasts of technological or occupational demand. Both responses can reduce adaptability. General expansion may direct resources towards fields in which education supply is not the principal constraint, while premature specialisation may produce programmes dependent on a particular expectation of technological development or employer demand.

A strategy can manage this trade‑off by combining common direction with field- and level-specific diagnosis. Priorities should identify the mechanism they are intended to change and the evidence against which progress can be assessed. Investments in emerging fields should retain strong disciplinary and methodological foundations, enabling learners to move between applications and adjacent areas as technologies develop. Periodic review should revisit assumptions about demand rather than treating an initial forecast as a fixed basis for programme design or capacity expansion.

The second trade‑off concerns the relationship between political visibility and deliverable institutional commitments. A national strategy can convene actors, increase attention and establish ambitions that extend beyond individual programmes. However, political visibility, co‑ordination structures and stated objectives do not themselves create the staff, leadership, infrastructure, learner support, supervision, time and organisational authority required for implementation.

Strategies may add initiatives and expectations without addressing the funding, capability and capacity needed to deliver them. New programmes, partnerships and support commitments may be layered onto institutions already facing constraints, while temporary projects may increase activity without creating recurrent capacity. Expansion may then transfer pressure from entry to persistence, completion, progression or research training. Strategic ambition should therefore be matched by an assessment of the resources, expertise, infrastructure and time required. Strategies may need to sequence expansion, concentrate specialist provision, support collaboration or moderate the scale of intervention until sustainable institutional capacity has been developed.

Different formats for national strategies

Countries organise joined-up strategies for advanced STEM skills in different ways, reflecting their governance arrangements, policy priorities and the constraints they seek to address. Three broad formats can be distinguished: dedicated STEM education and skills strategies; broader skills, science, innovation or industrial strategies that include advanced STEM skills; and strategies focussed on particular fields or technologies. These formats may coexist and complement one another.

Dedicated STEM education and skills strategies

Dedicated strategies explicitly use STEM or an equivalent national term and usually seek to connect action across several stages of education. They can give visibility to relationships that are otherwise divided among school subjects, vocational programmes, higher education, adult learning and non-formal education. Their substantive contribution depends on whether this broad framework is used to select and connect action rather than mainly to assemble existing initiatives under a common label.

The European Commission’s STEM Education Strategic Plan provides a supranational example (European Commission, 2025[3]). Adopted as part of the Union of Skills, it places STEM education within a wider agenda concerned with competitiveness, preparedness and technological development. Its proposed actions include improved STEM skills intelligence, a competence framework, STEM education centres, joint programmes, micro-credentials, capacity building and closer relationships among education, research and industry. The plan seeks to provide shared direction and identify areas in which European co‑operation can complement national and regional responsibility.

Finland’s LUMA  Strategy illustrates a national strategy organised around a long-term vision (Ministry of Education and Culture, 2023[4]). LUMA refers to natural sciences and mathematics, while the term LUMATE, used in the accompanying action plan, makes the inclusion of technology and technical fields more explicit and corresponds more closely to the international use of STEM. The strategy connects competence in these fields with socially, ecologically and economically sustainable well-being and growth. It is complemented by a LUMATE action plan that identifies more concrete measures, timescales and responsible actors. Its activities are grouped around the development of teaching and education; monitoring, surveys and wider development; and communication and interest. This illustrates how a concise strategic framework can be followed by a more operational plan without requiring the strategy itself to specify every instrument.

Austria’s MI(N)Tmachen action plan covers the educational route from early childhood education through school, apprenticeship and higher education (BMBWF, 2023[5]). MINT is the German-language equivalent of STEM and stands for Mathematik, Informatik, Naturwissenschaften und Technik, meaning mathematics, computer science, natural sciences and technology. The title MI(N)Tmachen is a play on the German verb mitmachen, meaning to participate or join in. The plan brings together regional networks, information and guidance, teacher and programme development, support across educational transitions, research experience, inclusion, interdisciplinary higher education and opportunities for adults. It also situates its educational objectives within Austria’s wider research, technology and innovation strategy and higher education planning. It therefore provides an example of a dedicated STEM action plan connected with broader science and innovation objectives.

Germany’s MINT action plans connect formal education with non-formal learning environments and regional networks. The MINT-Aktionsplan 2.0 consolidated federal measures across educational stages and organised further action around co‑operation between formal and non-formal provision, quality, families, early learning and education research (BMBF, 2022[6]). The subsequent MINT+ framework, presented jointly by the federal ministries responsible for education and research in 2026, places additional emphasis on coherent education ecosystems, equal opportunities, talent in key technologies and shared responsibility among government, education, research, business and civil society. The “+” signals a broader approach that connects MINT subjects more strongly with creativity, societal questions and other capabilities relevant to innovation, including entrepreneurship. It also links MINT education more closely with Germany’s wider innovation and technology policy through the Hightech Agenda Deutschland (BMFTR and BMBFSFJ, 2026[7]).

These examples differ in scope and organisation. Finland combines a concise long-term vision with an implementation plan. Austria follows the education route and brings together several forms of provision and support. Germany places particular emphasis on regional and non-formal learning ecosystems, while the EU plan identifies common objectives and areas for co‑operation across national systems. In each case, the general gains and trade‑offs identified above remain relevant. A dedicated STEM framework needs to distinguish among fields and levels, select among possible actions and establish how institutions will acquire the capacity required for implementation. The evolution of the strategy in Germany illustrates how a national STEM framework can be periodically revised to reflect changing educational, technological and societal priorities rather than treated as a fixed statement of priorities.

Positioning advanced STEM skills within broader national strategies

Advanced STEM skills may also be addressed within a broader national strategy. This may be appropriate where the principal purpose is to connect education and research capacity with economic development, public services, technological transformation or regional specialisation. Relevant formats include national skills strategies, science and innovation strategies, industrial strategies, and strategies for digital or green transitions.

This format can strengthen the connection between advanced STEM skills and their effective use. It can bring educational priorities together with research funding, infrastructure, business investment, procurement, innovation support and regional development. It can also make clearer that skills alone do not produce innovation or productivity. Their application depends on complementary expertise, equipment, organisational capability and opportunities to undertake substantive work.

Ireland’s National Skills Strategy, for example, linked education and training with economic and social development, lifelong learning, inclusion, teaching quality and employers’ effective use of skills. It treats the relevance of learning to individuals, society and the economy as a shared concern while distinguishing the contributions expected from education providers, employers and public authorities. Not confined to STEM, this form of national skills strategy can accommodate advanced STEM skills within a broader account of participation, learning and use (Department of Education and Skills, 2016[8]).

Industrial strategies can integrate advanced STEM skills with investment and sectoral development. The United Kingdom’s Modern Industrial Strategy, for example, establishes a ten‑year framework for increasing business investment in eight growth-driving sectors, supported by dedicated sector plans and measures addressing research and development, infrastructure, skills and regional clusters. The eight sectors are advanced manufacturing, clean energy industries, creative industries, defence, digital and technologies, financial services, life sciences, and professional and business services. In this format, skills development forms one component of a wider account of the conditions required for sectors such as advanced manufacturing, clean energy, digital technologies and life sciences to develop (Department for Business and Trade, 2025[9]).

Embedding advanced STEM skills in a broader strategy can therefore provide a more complete account of effective use than an education strategy alone. It can connect programme and research capacity with environments in which advanced STEM skills may be applied. Education and science‑policy responsibilities should nevertheless remain explicit. Education institutions should not be treated primarily as suppliers of labour for selected sectors, and programme design should not be subordinated to short‑term occupational requirements. Where an industrial or economic strategy anticipates substantial expansion in a field, it should also consider the time and resources needed to build teaching, research and supervisory capacity.

Field- or technology-specific strategies: Lessons from national quantum strategies

A third format focusses on a particular scientific field or family of technologies. This may be appropriate where capability depends on specialised research infrastructure, concentrated expertise, long training horizons and close co‑ordination among research, education, innovation and technological uptake. National strategies for quantum technologies, artificial intelligence, semiconductors, biotechnology, cybersecurity, advanced materials and clean-energy technologies illustrate this approach. National quantum strategies provide a particularly useful example.

Field- or technology-specific strategies can identify competence and capacity requirements more precisely than a general STEM strategy. They can distinguish between researchers, engineers, technicians and professionals from adjacent fields, while connecting research funding, infrastructure, education and innovation to the maturity and expected development of the field. Their specificity does not remove uncertainty, however. Technologies may develop in directions that differ from initial expectations, and relevant competence may be distributed across established disciplines rather than embodied in a new qualification.

Quantum technologies illustrate these characteristics. Skills in quantum sensing, computing and communication draws on physics, mathematics, computer science and engineering, as well as technical expertise in areas such as fabrication, instrumentation and systems integration. Education and research training need to interact with specialist infrastructure, sustained research funding, knowledge transfer, standards, procurement and international co‑operation.

The OECD’s overview of national quantum strategies examines the timelines, motivations, scope, objectives, governance and assessment arrangements through which countries support the development and uptake of quantum technologies (OECD, 2025[10]). It identifies a portfolio of commonly used instruments, including institutional funding for public research, competitive project grants, grants for business research and innovation, public procurement and equity finance. This shows that national capability is not pursued through education or workforce measures in isolation.

The strategies reviewed by the OECD combine scientific and technological objectives with economic, societal and, in some countries, security considerations. Public research funding and specialist infrastructure remain foundational, while support for business research, innovation and uptake becomes more prominent as applications develop. Governance arrangements are used to connect public research organisations, higher education institutions, businesses and government authorities, although their composition and authority vary among countries.

Several lessons follow for advanced STEM skills policy. First, the relevant competence is not confined to graduates from programmes bearing a quantum label. Research and innovation draw on deep disciplinary competence developed in established fields, combined with increasing opportunities for interdisciplinary specialisation. School education contributes broad mathematical, scientific and computing foundations. VET may develop enabling technical competence as applications and infrastructure mature. Bachelor’s and master’s programmes can combine solid disciplinary foundations with specialised options, while doctoral and postdoctoral training remain central to research capability. Continuing education can enable researchers and professionals from adjacent fields to develop additional competence where their existing foundations are relevant.

Second, education and research capacity cannot be separated. Increasing the number of specialised learners without educators, supervisors, research positions, infrastructure and viable career routes would provide a limited response. Access to research facilities and active research environments affects both the quality of training and the competence that learners can develop. The long time required to educate researchers and build specialist communities also means that capacity decisions need to extend beyond short-term demand signals.

Third, the skills question cannot be reduced to estimating a required number of graduates. National capability depends on the balance of disciplinary backgrounds, the interaction of research and technical roles, and movement between higher education, public research and industry. It also depends on whether organisations possess the infrastructure, complementary expertise and absorptive capacity required to use specialised competence. A strategy focussed only on education supply would overlook much of the system through which that competence produces scientific, economic or societal value.

Finally, quantum illustrates the need for periodic strategic review. The relative maturity of sensing, communication and computing applications differs, and future trajectories remain uncertain. Strategies need to support current research and capability while preserving the capacity to adjust priorities. Strong disciplinary foundations, interdisciplinary routes and mobility between research and professional settings can reduce the risks associated with dependence on a single technological forecast. The OECD’s wider quantum work similarly emphasises anticipatory governance, international co‑operation and the need to address the opportunities and risks of quantum technologies as they develop.

National quantum strategies thus demonstrate the advantages of field-specific co‑ordination but also reinforce the two trade‑offs identified at the beginning of this section. Specificity can direct attention to distinctive research and infrastructure needs, yet it should not become excessive dependence on uncertain forecasts. Political ambition can mobilise resources and institutions, yet its achievement depends on sustained teaching, research, infrastructure and supervisory capacity.

The formats of a strategy discussed above differ in scope, governance and degree of specialisation, but they face a common implementation challenge: translating strategic direction into priorities that address diagnosed constraints, assign clear responsibilities and are supported by sufficient capacity. This requires action across the policy portfolios and actors that shape how advanced STEM skills are developed, demonstrated, applied and renewed. The eight areas for policy action provide a framework for identifying and combining such action.

Eight areas for policy action to guide strategies for advanced STEM skills

The development and effective use of advanced STEM skills is a cross-portfolio policy concern. In many systems, responsibilities for higher education and for science, technology and innovation are distributed across different ministries, agencies or policy portfolios. Effective co-ordination can connect the higher education framework with the national science, technology and innovation ecosystem and strengthen coherence across the routes through which advanced STEM skills are developed, demonstrated, applied and renewed.

The eight areas for policy action translate this cross-portfolio perspective into strategic directions that countries can select, combine and adapt to their institutional arrangements, policy priorities and diagnosed constraints. They are intended to inform the design, implementation and evaluation of strategies for advanced STEM skills. Their application should account for differences in system size and capacity, including the particular challenges smaller higher education and research systems may face in sustaining specialised programmes and research capacity, attracting and retaining talent, and balancing national provision with cross-border collaboration.

Learning from existing practice can be an important element in this, provided that examples are interpreted in relation to the constraints, institutional arrangements and implementation capacity of each system. Table 3.2 summarises the principal approaches and examples discussed in Chapter 2 and shows how they relate to the eight areas.

Table 3.2. Eight areas for policy action: implementation approaches and examples

Area for policy action	Principal approaches discussed in Chapter 2
1. Enabling learners with the interest and potential to succeed to enter or return to high-quality STEM higher education	Strengthen mathematical, scientific and digital foundations; widen access to substantive STEM experiences; strengthen the relevance of STEM and learners’ sense of belonging; improve information, guidance and transitions; and use diagnostic assessment, preparatory provision and other routes to support entry or return.
2. Making learning, persistence and completion central to the quality of STEM programmes while preserving academic standards	Diagnose differences in learning, persistence and completion; connect evidence with co-ordinated academic, practical and psychosocial support and institutional improvement; strengthen teaching, assessment, feedback, supervision and programme coherence; redesign first-year experiences and programme structures where required; enable flexible continuation, interruption and return; and address financial and other constraints on participation.
3. Establishing transparent and usable routes between levels, fields, employment and return	Strengthen recognition of prior learning, credit transfer and competence frameworks; establish transparent progression between qualification levels; enable movement across disciplines and fields; recognise relevant prior learning and experience; provide bridge, conversion and preparatory routes; and support movement between education, employment and return to study.
4. Widening equitable access to substantive research and professional experience	Embed research and professional experience within programmes; fund participation where unpaid or additional activity would restrict access; strengthen access to research environments, infrastructure, partnerships and mobility; connect doctoral education with cross-sector research and professional settings; and expand access to paid, substantive and appropriately supported professional experience.
5. Renewing programmes around profiles of disciplinary, interdisciplinary, transversal and professional competence	Define the competence profiles required in established and emerging fields; preserve strong disciplinary and methodological foundations; integrate interdisciplinary, transversal and professional competence; strengthen connections among STEM fields and with relevant non-STEM fields; involve external partners in programme design and delivery while preserving institutional responsibility for educational quality; and use evidence on technological, scientific, societal and professional developments to support continuing programme renewal.
6. Improving career entry and the effective use and continued development of advanced STEM skills	Build shared routes between higher education, research and employment; support entry into suitable first professional roles; improve employers’ interpretation and recognition of advanced STEM skills; connect recruitment with effective onboarding, substantive work and continued development; strengthen work organisation and management practices that enable advanced STEM skills to be applied; and improve employment conditions, career development and advancement where these affect skills use and retention.
7. Building coherent routes for skills renewal across working life	Develop provision for updating and deepening existing competence, moving into adjacent fields and making more substantial professional transitions; enable adults to return to higher education through differentiated entry, recognition of prior learning, diagnostic assessment and bridge or refresher provision; support flexible participation; clarify the level, recognition and relationship of shorter learning to larger qualifications; and enable learning to accumulate into coherent and usable progression routes.
8. Building sustainable pedagogical, organisational and resource capacity in higher education institutions and strengthening their role in policy learning	Sustain qualified academic, technical, professional and support staff; strengthen curriculum design, teaching, assessment, feedback, supervision and programme leadership; develop connected learner support, recognition, interruption and re-entry arrangements; provide suitable teaching and research infrastructure and environments; strengthen institutional data, analytical and research capacity; support collaboration and shared provision where appropriate; secure sustainable finance and implementation time; and enable higher education institutions to contribute to monitoring, evaluation and continuing policy learning.

In the following, the eight areas for policy action are discussed in terms of their rationale, scope and implementation considerations.

Area for policy action 1. Enabling learners with the interest and potential to succeed to enter or return to high-quality STEM higher education

Advanced STEM skills cannot be developed across the broadest possible talent base when avoidable gaps in preparation, information, institutional capacity and accessible provision determine who can enter or return to STEM higher education.

Effective action requires clear roles across entry and re‑entry routes. Public policy can address wider barriers relating to prior educational opportunities, finance, accessibility and location. Higher education policy can provide funding, admission, recognition, quality-assurance and information frameworks that support entry and return. Higher education institutions remain responsible for admission and recognition decisions, preparatory provision, programme information, transition support, and the quality and capacity of provision. Schools, vocational providers, employers and other organisations can contribute preparation, guidance, occupational information, substantive STEM experiences and access to professional communities.

Policy monitoring should follow routes into higher education, including access to prerequisite subjects and the preparation they provide, participation in guidance and preparatory provision, application, admission, enrolment and initial progress. It should examine who benefits from these opportunities, where qualified and interested learners cease to progress, and whether interventions reduce or reproduce differences across population groups, institutions and locations.

Area for policy action 2. Making learning, persistence and completion central to the quality of STEM programmes, while preserving academic standards

Entry or return to higher education is only the first step. Learners must subsequently have a genuine opportunity to develop the intended advanced STEM skills and complete their programmes. Learning, persistence and completion should therefore be treated as central dimensions of programme quality, without lowering academic standards.

Constraints may arise at several levels. Individual-level constraints include gaps in preparation and financial or practical pressures. Programme‑level constraints include weaknesses in curriculum design, teaching and assessment, limited practice and feedback, excessive workloads and weak belonging. Institutional-level constraints include inadequate learner support, inflexible study arrangements and insufficient programme leadership. These constraints may interact and be concentrated in particular courses, transitions or learner groups. They may be especially consequential in introductory and gateway courses during the early undergraduate years, where avoidable weaknesses in teaching, assessment and programme culture can lead to the loss of students who entered with strong academic preparation and the potential to succeed. Withdrawal, transfer, interruption or delayed completion does not by itself reveal the cause or indicate learner failure.

Effective action requires clear roles. Public policy can address financial, practical and accessibility barriers and ensure sufficient teaching and learner-support capacity. Higher education policy can reflect learning, persistence, completion and practical accessibility in funding and quality assurance. Higher education institutions remain responsible for curriculum, teaching, effective teaching, including appropriately designed active and interactive approaches, assessment, programme coherence, learner support, flexible continuation and programme culture. They also need the records, information systems, designated responsibilities and academic guidance required to preserve completed learning, maintain proportionate contact during interruption and support planned re‑entry. Programme leaders play a central role and require appropriate authority, time, support and recognition.

Policy monitoring should examine learning and competence development, course performance, credit accumulation, workload, access to practical learning and support, transfer, interruption, re‑entry, delayed completion and completion. A unique student identifier that remains with the learner across institutions, education sectors and periods of study can support secure linkage of enrolment, credit, transfer and completion records and help distinguish permanent withdrawal from transfer, interruption and later return. Its use requires consistent reporting, interoperable systems and clear safeguards governing access, linkage, privacy and permissible uses. Monitoring should identify where difficulties are concentrated, distinguish changes in educational route from displacement by avoidable barriers, and assess whether learners develop the intended disciplinary, interdisciplinary, transversal and professional competence.

Area for policy action 3. Establishing transparent and usable routes between levels, fields, employment and return

Formally available routes do not provide genuine opportunities when learners cannot understand their requirements, obtain recognition for relevant previous learning, or access the preparation and support needed to reach their intended destination. These routes include return after an interruption, re‑entry after employment and entry to a different level or field for skills renewal.

Progression may be constrained by unclear admission requirements, inconsistent recognition of previous learning, unnecessary repetition, missing disciplinary or methodological foundations, limited bridge provision, unequal access to information and research experience, financial and geographical barriers, and selection points that learners could not reasonably anticipate. At advanced levels, limited funding, supervision and research infrastructure may restrict progression even where qualified and interested learners are available. Low immediate progression does not by itself indicate a barrier, since entry into suitable employment may be an informed and valuable outcome.

Making progression routes usable requires connected responsibilities. Public policy can support coherent qualification frameworks and address financial, geographical and administrative barriers across education, research and employment. Higher education policy can establish admission, credit, recognition and funding frameworks that support movement between institutions, levels and fields. Higher education institutions remain responsible for admission and recognition decisions, transparent programme requirements, curriculum mapping, bridge provision and ensuring that learners admitted through different routes possess or can develop the necessary foundations. Institutions also need sufficient organisational capacity to retrieve and interpret previous records, provide timely recognition decisions and develop an individual route connecting previous learning with current programme requirements. Vocational providers, research organisations, employers, professional bodies and qualification authorities can help connect educational and occupational routes and clarify their requirements.

Policy monitoring should follow application, admission and recognition decisions, recognised credit, additional preparation, progression, remaining duration and cost, completion and subsequent education or employment. Where learners move between institutions or return after an interruption, linked longitudinal records can show whether previous learning is preserved and whether formal re‑entry leads to continued progression and completion. Monitoring should distinguish voluntary entry into employment from progression prevented by avoidable barriers and assess whether formal progression rights translate into usable routes across institutions, fields, educational backgrounds and learner groups.

Area for policy action 4. Widening equitable access to substantive research and professional experience

Research and professional experience help learners develop advanced STEM skills and provide evidence that institutions and employers use when assessing readiness for further study and work. Unequal access can therefore affect persistence, progression and career entry.

Access may be constrained by unpaid or poorly paid participation, informal selection and networks, institutional prestige, geographical mobility, additional accommodation, transport and care costs, and limited availability of suitable opportunities. Experiences may also provide little educational value where learners receive routine or peripheral tasks, insufficient supervision or feedback, or few opportunities to connect experience with academic learning. Participation alone does not establish that advanced STEM skills have been developed.

Research and professional experience may also depend on access to research environments, specialist facilities, data, equipment and complementary expertise that individual institutions or employers cannot provide alone. Shared facilities, collaborative research environments and partnerships among higher education institutions, research organisations, employers and public bodies can widen access, particularly for learners and organisations in less research-intensive institutions or regions. Such arrangements should define responsibilities for access, supervision, safety, intellectual property, data use, assessment and recognition and should ensure that participants undertake substantive work rather than observe or provide routine assistance.

Higher education institutions remain responsible for transparent selection, academic oversight, learning objectives, assessment and connections with the curriculum. Research organisations and employers are responsible for substantive tasks, prepared supervisors, appropriate feedback, accessible working environments and opportunities to exercise technical and professional judgement. Where specialist capacity is distributed across organisations, public authorities, funders and intermediary organisations can support shared facilities, collaborative research environments and access arrangements that enable learners and early-career professionals to participate in substantive research and innovation activity.

Policy monitoring should examine who benefits; how participants are selected; whether participation is financially and practically feasible; the tasks and responsibilities assigned; and the quality of supervision and feedback. It should also follow subsequent progression into further study or suitable employment and assess whether the learning and evidence produced remain portable beyond the participating institution, employer or sector.

Area for policy action 5. Renewing programmes around profiles of disciplinary, interdisciplinary, transversal and professional competence

Advanced STEM education must respond to changes in knowledge, technology and professional practice while preserving the durable foundations needed for future learning and responsible application.

Programme renewal may be constrained by unclear learning outcomes, weak connections among courses, insufficient assessment of intended competence, superficial integration across disciplines, limited professional and societal application, and curricula organised around short-lived tools or immediate employer requirements. Programme titles and competence frameworks may not reveal the depth of disciplinary learning, the form of integration achieved or the educational and occupational routes supported. Adopting a competence profile does not by itself establish that the intended learning is developed or assessed.

Programme renewal is principally an institutional responsibility, supported by appropriate public frameworks and informed by other actors. Public policy can support national or sectoral competence frameworks, skills intelligence and collaboration where common terminology and shared evidence provide wider benefits. Higher education policy can provide funding, quality-assurance expectations and educational-development capacity while preserving institutional responsibility for academic standards. Higher education institutions remain responsible for curriculum, teaching, assessment and the coherence of disciplinary, interdisciplinary, transversal and professional competence. Educators and programme leaders translate competence profiles into sequenced learning and credible assessment. Employers, professional bodies, research organisations and societal partners can clarify changing problems, tasks and contexts, but should not narrow programmes around the immediate requirements of particular technologies, occupations or organisations.

Skills intelligence can identify changes in occupations, tasks, technologies and competence requirements and help institutions assess the need for adaptation, new provision or collaboration. It should remain one input to institutional judgement, not a direct curriculum specification. Available evidence may reflect broad categories, stated recruitment requirements, temporary or local demand, and employer recruitment practices rather than the advanced STEM skills actually needed and used. Institutions should interpret it in light of programme purpose, level, disciplinary foundations and intended destinations, distinguishing needs for programme renewal from optional specialisation, continuing education, work-based learning or employer-provided onboarding.

Policy monitoring should examine whether competence profiles and relevant evidence are translated into coherent curricula, teaching, substantive learning opportunities and credible assessment. It should assess whether learners develop durable disciplinary and methodological foundations, whether interdisciplinary provision achieves genuine integration, and whether learners can communicate, demonstrate and apply their competence in further education, research and employment.

Area for policy action 6. Improving career entry and the effective use and continued development of advanced STEM skills

Developing advanced STEM skills will produce limited economic and societal value if individuals cannot obtain suitable work or if employers cannot recognise, use and continue developing the competence available.

Career entry and effective skills use may be constrained on both the individual and employer sides. Individual-facing constraints include unequal access to professional experience, occupational information and professional networks, as well as financial, geographical and personal barriers to suitable work. Employer-side constraints include difficulty interpreting and validating advanced STEM skills, narrow recruitment criteria, demands for directly comparable experience or organisation-specific knowledge, insufficient onboarding, restricted task allocation, limited autonomy and feedback, weak development opportunities, and working conditions or advancement structures that undermine progression and retention. Recruitment difficulty does not by itself establish that relevant advanced STEM skills are unavailable.

Employers remain primarily responsible for recruitment, onboarding, task allocation, work organisation, working conditions and continued development. Their involvement during education can improve routes into STEM careers, but does not substitute for realistic entry requirements, effective onboarding or opportunities to use and further develop advanced STEM skills after recruitment. Employer organisations, professional bodies and intermediaries can provide shared information, frameworks and support, particularly for employers with limited internal capacity.

Effective use also depends on whether organisations have suitable demand and application capacity. Employing people with advanced STEM skills will have limited effects where organisations lack research and innovation activity, access to appropriate technologies and infrastructure, complementary technical and managerial expertise, or the organisational capacity to define substantive problems and incorporate specialist evidence into decisions. These constraints may be particularly important for SMEs and public organisations that cannot maintain extensive internal research, technical or innovation functions.

Responsibilities differ across actors. Public policy can address wider barriers to suitable work and retention. Higher education policy and institutions can make assessed advanced STEM skills clearer, provide career guidance and connect learners with substantive professional experience. Employers can contribute to the planning, co-design and delivery of STEM programmes by identifying durable changes in professional practice, contributing authentic problems and specialist expertise, and providing funding, traineeships, internships, placements and other forms of work-based learning. Such involvement can strengthen the relevance of programmes and help learners develop, apply and demonstrate advanced STEM skills in professional settings. Higher education institutions nevertheless remain responsible for curriculum coherence, assessment, academic independence and qualification standards (see Area for policy action 5).

Research, innovation, industrial, regional and public-sector policy can complement these education–employer connections by strengthening the environments in which advanced STEM skills are applied. Depending on the diagnosed constraint, relevant instruments may include research and innovation funding, technology-diffusion and extension services, access to shared laboratories, testing facilities, data and computing infrastructure, collaborative research and demonstration projects, and innovation-oriented public procurement. Support for management and organisational development can help organisations identify suitable applications, integrate new technologies with existing processes, establish multidisciplinary teams and give technical expertise an effective role in organisational decisions.

For SMEs and public organisations with limited internal capacity, intermediary and shared arrangements may be especially important. Higher education institutions, public research organisations, technology centres, professional bodies and regional innovation organisations can provide access to facilities, specialist advice, collaborative research and professional networks that individual organisations could not sustain alone. Public support should strengthen organisations’ capacity to undertake research, adopt and adapt technologies and use advanced STEM skills, rather than subsidise isolated recruitment without the infrastructure, complementary expertise or organisational changes needed for effective application.

Policy monitoring should follow individuals beyond initial employment and examine access to suitable roles, recruitment outcomes, onboarding, actual skills use, task complexity, autonomy, professional judgement, feedback, continued development, advancement, working conditions and retention. It should also examine whether organisations increase their capacity to undertake research, adopt or adapt technologies, access suitable infrastructure and complementary expertise, and organise work so that advanced STEM skills influence products, services, processes and decisions. For publicly supported measures, monitoring should establish whether organisations develop lasting internal or shared capacity, whether benefits extend beyond a single recruitment episode, and whether effects differ by organisational size, sector and location.

Area for policy action 7. Building coherent routes for skills renewal across working life

Changes in knowledge, technology and work create learning needs throughout working life that cannot be met through initial education or disconnected short courses alone. Skills renewal may be constrained by provision that is unavailable, fragmented, difficult to navigate or poorly recognised; unclear prerequisites and learning outcomes; missing foundations; inconsistent recognition of previous learning; unsuitable cost, scheduling or delivery; limited employer support; insufficient access to specialist educators, infrastructure and learner services; and few opportunities to apply the competence developed.

Effective action for skills renewal requires clear roles and coherent routes. Public and higher education policy can provide finance, information, quality-assurance and recognition frameworks. Higher education institutions remain responsible for accessible entry, recognition, coherent provision and credible assessment, while employers should enable participation and opportunities to apply the competence developed. Institutions also need the organisational capacity to support people returning after employment, interruption or a longer absence. This includes accessible information and guidance, reliable records of previous learning, timely recognition decisions, diagnostic assessment, bridge or refresher provision and clearly assigned responsibility for helping learners identify an appropriate route.

Based Routes should clarify prerequisites, learning outcomes, assessment, recognition and further opportunities, with diagnostic assessment and bridge provision addressing essential gaps. Flexible provision should retain access to the educators, infrastructure, practical learning and support required for advanced learning. Professional bodies and other providers can meet specialised needs, while higher education has a distinctive role where skills renewal requires advanced disciplinary and methodological learning, research-informed content, specialist infrastructure or a recognised qualification. Short courses and micro-credentials may support focussed learning or contribute to larger routes but should not substitute for missing foundations or the coherent learning required for substantial occupational change.

Policy monitoring should extend beyond enrolment and completion to the competence developed, the credibility and portability of recognition, application in work or further education, progression towards the intended destination and longer-term employment outcomes. It should examine who participates and benefits, who receives employer support, whether short courses and micro-credentials form coherent routes where accumulation is intended, and whether learners can understand likely recognition and progression outcomes before committing substantial time and resources. It should also examine whether returning learners receive timely recognition and guidance, whether previous learning is preserved where it remains relevant, whether identified gaps are addressed through proportionate preparation and whether learners continue towards their intended educational or occupational destination.

Area for policy action 8. Building sustainable pedagogical, organisational and resource capacity in higher education institutions, and strengthen their role in policy learning

Higher education institutions need sustainable capacity not only to expand provision, but also to maintain its quality, support different educational routes, provide substantive practical and research learning, renew programmes and respond to skills-renewal needs.

Pedagogical capacity includes sufficient qualified educators, supervisors and technical staff, with adequate time and support to develop and use effective teaching approaches, including active and interactive pedagogy, and to undertake assessment, supervision, curriculum development and educational improvement. Organisational capacity includes effective programme leadership, manageable workloads, co‑ordination across courses and organisational units, functioning learner-support arrangements, recognition and guidance for returning learners, processes for interruption and planned re‑entry, and the capacity to collect, connect and use data on learning, progression, competence development and outcomes. Resource capacity includes stable and sustained finance, specialist facilities, laboratories, equipment, digital environments, data systems and access to suitable research and professional settings. Where learners transfer, interrupt or return after several years, compatible records and, where permitted, a unique student identifier can support continuity across institutions, education sectors and periods of study. These forms of capacity may differ across fields, qualification levels, institutions and locations.

Building sustainable capacity requires co‑ordinated action and continuing capacity building. Public policy can align ambitions for advanced STEM skills with decisions on education, research, infrastructure, regional development and public finance. Higher education policy can provide sustained funding, quality-assurance and data frameworks that reflect the differing costs and requirements of high-quality STEM provision. Higher education institutions remain responsible for staffing, workload management, infrastructure, programme leadership, organisational co‑ordination, learner support, data governance and the quality of provision. Research organisations, employers and other partners can contribute specialist expertise, facilities, data and professional environments, with clear responsibilities for teaching, supervision, assessment, accessibility, support and appropriate data use.

Capacity-building measures should strengthen educators, supervisors, technical staff, programme leaders and institutional teams, rather than focus only on additional places or infrastructure. Collaboration across higher education institutions can support shared provision and facilities, exchange effective approaches, develop common evidence and extend access to specialised expertise. Such collaboration should support mutual learning while preserving institutional responsibility for academic standards, programme quality and learner support.

Policy monitoring should examine staffing, workloads, educator and supervisor development, infrastructure, programme leadership, learner support, organisational co‑ordination, data capacity, and the sustainability and geographical distribution of resources. It should assess whether institutions can support high-quality learning, progression, substantive research and professional experience, programme renewal and skills renewal, and whether learners can access this capacity in practice.

Higher education institutions have a core role in policy learning because they generate and interpret evidence on implementation, learning and outcomes and can identify how common policies operate across fields, programmes and learner groups. Monitoring arrangements should support institutional improvement, capacity building and structured learning across institutions and between institutions and policymakers, rather than serve only reporting and accountability purposes. Where provision expands, monitoring should establish whether constraints are resolved or transferred to later stages.

Conclusion

An effective strategy for advanced STEM skills begins with a sufficiently specific diagnosis. It should connect the development, recognition, effective use and renewal of advanced STEM skills rather than address individual educational or employment outcomes in isolation. Strategies may be dedicated to STEM, incorporated into broader education, skills, research, innovation or industrial strategies, or focused on a particular field or technology. Their value depends less on their formal designation than on whether they connect responsibilities and interventions around shared priorities and diagnosed constraints.

International mobility and collaboration should form part of this strategic perspective, connecting the attraction of learners, researchers and professionals with the recognition of qualifications and experience, integration into education, research, employment and wider society, suitable conditions for retention and continued development, opportunities for return and collaboration, and consideration of effects on origin and destination countries. A strategic focus on advanced STEM skills does not imply prioritising STEM over non-STEM fields. Sustainable growth and responsible innovation depend on contributions from both and on their effective integration. Attention to career entry and effective skills use also moves the discussion beyond the number of STEM graduates produced to whether their competence is recognised, applied and continually developed at work.

System size and capacity affect how a strategy can be implemented. Smaller higher education and research systems may face particular challenges in sustaining specialised programmes and infrastructure, maintaining research capacity, attracting and retaining talent, and providing opportunities across all fields nationally. Shared provision, access to cross-border infrastructure, institutional collaboration, mobility and international research and education partnerships can complement national capacity. Such arrangements should preserve clear responsibilities for academic quality, recognition, learner support, supervision and the continuity of educational routes.

The eight areas for policy action are intended to inform the design, implementation and evaluation of strategies for advanced STEM skills and support co-ordinated action across policy portfolios. They are strategic directions that countries and other actors can select, combine and adapt to their diagnosed constraints, institutional arrangements, implementation capacity and policy priorities, rather than a standard set of measures for every system. Under conditions of technological and labour market uncertainty, their application should combine continuity in the development of durable foundations and long-term capacity with the ability to review and adjust priorities as evidence, technologies and competence requirements develop.

Higher education institutions are strategic partners in implementing these directions and supporting policy learning. Their contribution depends on sustainable pedagogical, organisational and resource capacity, including qualified academic and technical staff, programme leadership, learner support, specialist infrastructure, research environments, staff time, stable finance, and the data and analytical capacity required for institutional improvement and policy learning.

Ultimately, the effectiveness of a strategy should be assessed by whether people from a sufficiently broad talent base can develop advanced STEM skills, whether those skills are recognised and used effectively, and whether they can continue to develop and be renewed across working life.

References

[2] Barnow, B. (ed.) (2021), “Can quasi‐experimental evaluations that rely on state longitudinal data systems replicate experimental results?”, Journal of Policy Analysis and Management, Vol. 40/2, pp. 572-613, https://doi.org/10.1002/pam.22295.

[6] BMBF (2022), MINT-Aktionsplan 2.0 [STEM Action Plan 2.0], Federal Ministry of Education and Research, Berlin, https://www.bildung-forschung.digital/digitalezukunft/de/bildung/mint/mint_node.html (accessed on 12 March 2026).

[5] BMBWF (2023), Aktionsplan MI(N)Tmachen [Join in STEM Action Plan], Federal Ministry of Education, Science and Research, Vienna, https://www.bmfwf.gv.at/dam/jcr:611422de-e98e-4638-9047-98e3191d9175/Aktionsplan%20MI(N)Tmachen%20Endfassung%20Englisch.pdf (accessed on 12 March 2026).

[7] BMFTR and BMBFSFJ (2026), Der Aktionsplan MINT+: Entdecken, entwickeln, verändern [The STEM+ Action Plan: Discover, Develop, Transform], Federal Ministry of Research, Technology and Space and Federal Ministry of Education, Family Affairs, Senior Citizens, Women and Youth, Berlin, https://www.publikationen-bundesregierung.de/pp-de/publikationssuche/aktionsplan-mint-2438556 (accessed on 8 August 2026).

[9] Department for Business and Trade (2025), Industrial Strategy Sector Plans, https://www.gov.uk/government/publications/industrial-strategy-sector-plans (accessed on 12 March 2026).

[8] Department of Education and Skills (2016), Ireland’s National Skills Strategy 2025: Ireland’s Future, https://hdl.handle.net/20.500.14765/90371. (accessed on 12 March 2026).

[3] European Commission (2025), STEM Education Strategic Plan: Skills for Competitiveness and Innovation, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2025) 89 final, European Commission, Brussels, https://education.ec.europa.eu/sites/default/files/2025-03/STEM_Education_Strategic_Plan_COM_2025_89_1_EN_0.pdf (accessed on 3 June  2026).

[4] Ministry of Education and Culture (2023), Finnish National STEM Strategy and Action Plan: Experts in Natural Sciences, Technology and Mathematics in Support of Society’s Welfare and Growth, Publications of the Ministry of Education and Culture, Finland, No. 2023:22, Ministry of Education and Culture, Helsinki, https://urn.fi/URN:ISBN:978-952-263-733-8 (accessed on 12 March 2026).

[10] OECD (2025), “An overview of national strategies and policies for quantum technologies”, OECD Digital Economy Papers, No. 379, OECD Publishing, Paris, https://doi.org/10.1787/5e55e7ab-en.

[1] Skivington, K. et al. (2021), “A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance”, BMJ, p. n2061, https://doi.org/10.1136/bmj.n2061.