A Conceptual Foundation for the Development of a Lifelong Learning Measurement Framework

Annex C. Suggested indicators and questions to explore within the framework

The framework provides a basis for proposing a set of exploratory indicators to quantify LLL dynamics and to capture interactions among stakeholders (learners, providers, policymakers), technologies (e.g. digital platforms), and data sources (surveys and administrative records). At this stage, we propose exploratory indicators that leverage existing data systems and Python-based analytics (e.g. Pandas for data processing and scikit-learn for clustering). These initiatives could be piloted through projects similar to Norway’s Learning Analytics project (Wasson et al., 2024[85]), which focuses on integrating new data sources to enhance learning. The proposals aim to address gaps in LLL measurement and foster resilient ecosystems, pending further validation and development.

Below is a concise overview of the indicators proposed for examination, organised by pillar, with explanations of their purpose, data sources, methods, potential insights, and relevance to the framework.

C.1 Core data and information for learning needs and demand

Building on the foundations outlined in Sections 3, which address multi-level learning needs (individual, organisational, and country-level), governance dynamics, and diverse assessment approaches from qualification gaps to well-being outcomes, this summary distils the principal data strategies for operationalising the demand pillar within the framework. Here, data acts as a central enabler, transforming raw information into actionable insights that link identified needs with provision and coordination mechanisms. To this end, data are organised by functional role: research for pattern recognition, policy design for targeted intervention, and coordination for systemic alignment.

Core data and Indicators

The following table (Table A C.1.) outlines key data types, their purposes, and examples, illustrating how they support the assessment of LLL needs.

Table A C.1. Key data and indicators focused on LLL needs

Purpose	Core Data/Indicators	Examples/Sources
Research	Skill gaps, motivations, barriers	PIAAC/OECD surveys on skills (Desjardins, 2020[86]); (Grotlüschen et al., 2016[46]); surveys focused on attitudinal and motivational studies (Boeren et al., 2012[87]) (Kalenda, Boeren and Kočvarová, 2023[59]); (Karger, Kalenda and Vaculíková, 2025[88]); (Mertens et al., 2025[79])
Policy Design	Barriers, expressed demands, transition dynamics	AES/LFS surveys; motivation studies; barriers (Roosmaa and Saar, 2017[63]); (Broek et al., 2024[89]); (Van Nieuwenhove and De Wever, 2023[90])
Coordination	Multi-level needs, conversion factors, mismatches	Administrative data (UNESCO, 2019[91]); longitudinal pathways (e.g. (Elias, Hogarth and Pierre, 2002[49]); (Bukodi and Goldthorpe, 2018[92]))

Addressing data gaps

Persistent gaps in data on LLL needs remain evident.

These include:

1. limited systematic measurement of learning needs across all societal levels;

2. unarticulated demands among low-skilled adults or marginalised groups;

3. insufficient monitoring of non-work-related needs, such as health and financial literacy, as well as civic engagement;

4. inadequate measurement of motivation and barriers, especially in relation to those not involved in organised learning activities;

5. underdeveloped measurement of motivation in relation to existing learning opportunities and interaction with them.

To address these challenges, we argue that policymakers should adopt more holistic data-ecosystem strategies. These should not only build on existing survey instruments (Sekmokas et al., 2024[93]) but also strengthen them through improved measures of learning needs, various types of barriers and motivations; enhanced data-collection techniques (for example, the longitudinal tracking of LLL at the national level, or supplement existing longitudinal projects by dedicated LLL modules); and integration with newly available data sources. These could include administrative records, real-time signals from digital platforms, and insights generated through community participation.

As an illustration, we propose a blue-sky approach to potential future integration of some of these data sources in the form of a composite Needs Ecosystem Index (see Section 5.4.4). This index is designed to capture the interaction of individual, community, and national factors, thereby supporting more responsive and equitable policies. It combines direct measures (such as skills assessments) with indirect proxies (for instance, AI-driven sentiment analysis of digital platform interactions) to identify shifts in motivation in real time, while ensuring fairness across diverse national contexts.

Although most international data frameworks and policy debates focus on adults aged 25–64, LLL must be understood as extending both earlier and later in the life course. Children and youth under 18, as well as older adults over 64, represent critical but under-examined groups whose learning trajectories have long-term implications for equity, well-being, and social cohesion.

For children and youth, learning occurs not only in formal schooling but also through non-formal and informal settings that include civic associations, cultural and sports clubs, digital platforms, and family environments. These spaces foster dispositions, social capital, and literacies that shape future participation in LLL. Yet access remains stratified, with disadvantaged youth often lacking structured opportunities. Emerging challenges around digital well-being and algorithmic curation also require systematic attention. Data collection must therefore move beyond PISA-style assessment to capture non-formal and civic learning, as well as online experiences that increasingly mediate early educational pathways.

For older adults, participation in LLL is frequently framed through a deficit lens of reskilling or labour-market withdrawal. A more holistic approach recognises their active role in maintaining cognitive resilience, digital inclusion, health literacy, and community vitality. Informal and community-based learning is especially prevalent, from senior universities and cultural associations to citizen science initiatives and intergenerational programmes. These activities are linked to positive health outcomes, reduced isolation, and stronger civic engagement, yet they remain largely invisible in current measurement systems. International surveys typically exclude those over 64, producing a systemic blind spot.

Addressing these blind spots requires both conceptual and methodological innovation. Conceptually, children and seniors must be fully integrated into the demand–supply–coordination framework, rather than treated as peripheral cases. Methodologically, data coverage must be widened: by embedding LLL modules in longitudinal youth and ageing studies; by extending international surveys to capture learning beyond the labour market; and by experimenting with new methods such as digital ethnography and participatory data logging. Doing so would produce a genuinely life-wide and lifelong evidence base

Innovative data solutions

Strengthening the assessment of LLL needs and demands requires innovative and inclusive approaches to data.

First, updated and inclusive surveys. Large-scale instruments such as AES and PIAAC could be enhanced through AI-tailored questionnaires that reflect the realities of low-skilled and underrepresented groups. For example, a more feasible mode of data collection regarding the measurement of needs and demands. This would also help surface hidden needs – for example, barriers encountered by low-skilled learners – thereby improving inclusivity. These instruments would also benefit from additional pilot survey experiments, which would enable the development of measures to capture more nuanced dimensions of motivation (habitual vs. deliberate) and various perceived barriers (e.g. situational, dispositional, institutional, and time-related) among adults, including their interactions with motivational factors.

Second, extending the age scope of current surveys. Missing formative early experiences and later-life learning that influence well-being and social cohesion. To address this, we should: (a) extend longitudinal youth cohort studies with non-formal/civic modules, (b) incorporate LLL modules into ageing surveys such as SHARE (Survey of Health, Ageing and Retirement in Europe) and ELSA, and (c) pilot digital ethnography methods to capture learning on platforms where youth and seniors increasingly interact.

Third, cross-level data integration. Linking datasets across levels - individual (e.g. survey responses), group (e.g. community forum data), and national (e.g. policy impact metrics) – can provide a more comprehensive perspective. Cross-country observatories, proposed below (see Section 7), could play a central role in harmonising data and addressing institutional differences between countries.

Fourth, ethical use of big data and AI. Anonymised platform analytics – such as LinkedIn activity or YouTube engagement – combined with sentiment analysis, can help reveal unexpressed needs, including interest in non-job-related learning. The use of privacy-preserving synthetic data provides additional opportunities for sensitive analyses while protecting vulnerable groups.

Fifth, feedback loops. Establishing continuous cycles in which participation metrics directly inform policy design is crucial. For instance, mobile applications that monitor barriers in real time could enhance outreach to disadvantaged learners and better align interventions with equity objectives.

Key indicators, research questions, and policy questions

Key indicators and construction

As an illustration of advanced, “blue-sky” measurement within the demand pillar, we propose the Needs Ecosystem Index (NEI) - a synthetic composite of three indicators designed to assess learning needs:

1. Latent Needs Index: Captures unexpressed demands among marginalised groups by combining survey data on barriers (e.g. stigma scores with a 1–10 scale) with AI-analysed platform engagement (e.g. skill search queries). Results are normalised and weighted (60 per cent barriers, 40 per cent engagement) using pandas in Python.

2. Motivation-Barrier Ratio: Measures the balance between motivation and obstacles, drawing on enhanced AES or PIAAC module responses (motivation score ÷ barrier score). Ratios below 1 reveal high barriers, signalling inequities.

3. Transition Readiness Score: Evaluates preparedness for key life transitions, using longitudinal cohort data (e.g. from ELSA or the United States High School and Beyond Longitudinal Study of 2022 - HSBLS 2022). It compares skill gaps before and after transitions through regression modelling in statsmodels, incorporating demographic variables.

Machine learning techniques (e.g. principal component analysis in scikit-learn) are applied to standardise and weight these factors according to contextual relevance, giving greater weight to barriers in low-income countries. A threshold (e.g. an index score above 70) would indicate a robust ecosystem, while heatmaps could be used to visualise temporal shifts and identify where interventions are most urgently required.

Research questions

• How do socioeconomic factors shape latent learning needs across national contexts?

• What role do digital platforms play in converting unexpressed needs into participation?

• How do dispositional barriers evolve over the life course, and what predicts their impact on LLL engagement?

Policy questions

• How can subsidies be optimised to reduce barriers for low-skilled groups across welfare regimes?

• What interventions best address gender-based mismatches in learning needs?

• How should policies adapt to emerging needs driven by technological disruptions like AI?

C.2 Core data and information for learning opportunities

Building on Section 4, which traces the evolution of LLL provision from traditional triadic typologies to more adaptive frameworks reflecting flexibilisation, hybridisation, digitalisation, and institutional embeddedness, this summary consolidates data strategies for operationalising the supply pillar of the framework. The framework reconceptualises provision of LLL as a dynamic ecosystem, integrating diverse learning modalities across the formal education, labour market, welfare, and civic clusters.

Core data and indicators

Table A C.2. outlines key data types, their purposes, and examples, illustrating how they support the mapping of LLL provisions.

Table A C.2. Core Data and Indicators

Purpose	Core Data/Indicators	Examples/Sources
Research	Classify provision across various types of LLL, trends toward digitalisation, flexibilisation and hybridisation	AES (e.g. (Karger, Kalenda and Vaculíková, 2025[88])); (Desjardins and Kalenda, forthcoming[94]); administrative registries (e.g., (Manninen and Vuorikoski, 2024[95]); (Albrecht, van den Berg and Vroman, 2005[96])); InfoWeb for Continuing Education from Germany (IWWB, 2025[97]); platform logs
Policy Design	Participation and uptake - Accessibility metrics, equity outcomes	PIAAC, AES (e.g. (Desjardins, 2020[86]), (OECD, 2025[98])); administrative data
Co-ordination	Cross-cluster flows – Quantify movement between clusters	Longitudinal tracking; Linked admin IDs; case management systems

Addressing data gaps

Significant gaps persist in data on LLL provision. These include: (1) inadequate and often inappropriate monitoring of informal learning; (2) inconsistent measurement of hybrid forms that span formal and non-formal education, as well as the boundaries between non-formal and informal learning; (3) limited visibility of cross-cluster interactions, particularly in non-Western contexts where informal learning plays a predominant role; and (4) insufficient coverage of provision from direct providers. The less formalised and structured the learning activity, the more fragmented the data landscape becomes.

To address some of these challenges, we propose a blue-sky approach to data integration through a composite Provision Ecosystem Index. Conceived as a forward-looking data tool, the index is designed to help policymakers benchmark levels of digitalisation, monitor cluster-specific participation using enhanced datasets, and assess national accessibility standards. By consolidating fragmented data, the index would support more strategic decision-making, enabling systems to adapt provision to evolving societal and labour-market needs.

Innovative data solutions

To strengthen the mapping of LLL provision, several innovative strategies can be pursued.

First, enhancing existing survey modules: International surveys such as AES and PIAAC should be redesigned to better capture both informal learning and hybrid types of learning. Beyond that, informal and semi-formal types of learning, in particular, require new data-collection modalities capable of recording such activities in natural settings (e.g. such as mobile applications or diary-based tools that track learning throughout the day). This is especially important given the anticipated rise of informal learning and the growing role of digital platforms and conversational agents over the coming decade (Milana et al., 2024[99]).

Second, cross-national data harmonisation: The creation of organisational units tasked with integrating administrative records from diverse providers of formal and non-formal education, based on successful national examples (e.g. IWWB, 2025), developing classification methodologies, and combining these with platform-based data would enable meaningful comparisons across varied LLL systems.

Third, AI-assisted metadata analytics: Advanced analytical tools applied to privacy-compliant platform data (e.g. course completions, engagement timestamps) can provide estimates of informal and hybrid participation, with synthetic data used where direct measurement is not feasible.

Fourth, equity-driven feedback loops: Real-time accessibility indices should guide adjustments to provision to ensure that learning opportunities remain aligned with goals of inclusivity and equity.

Key indicators, research questions, and policy questions

Key indicators and construction

As an example of a potential “experimental” solution within the supply pillar, we propose the Provision Ecosystem Index (PEI) - a composite indicator designed to assess the breadth and responsiveness of learning supply. It comprises three core dimensions:

1. Hybridisation Ratio: This indicator captures the balance between formal, non-formal, and informal provision. It is constructed by calculating proportional shares (e.g. hybrid activities/total provision) using survey data such as AES, with clustering algorithms in SciPy applied to classify provision types according to structural and digitalisation attributes. The index is defined as: MH Iᵢ = w_f Fᵢ + w_n Nᵢ + w_i Iᵢ + w_d Dᵢ, where F, N, I represent the shares of formal, non-formal, and informal provision, and D denotes the digitalisation score. All values are normalised to the interval [0,1]. Weights (w) are determined using a combined expert-Principal component analysis (PCA) approach, and the reported index is scaled to a 0–100 range.

2. Accessibility Index: This dimension evaluates the extent to which provision is equitably available across various demographic groups, identified by expert-policy groups. It draws on administrative databases, aggregating uptake rates (e.g. older vs. younger, urban vs. rural) and weighting them against barriers such as cost, distance, time, or broadband access. The metric is expressed as: Access_g = k∑ v_k U_gk − b∑ λ_b B_gb, where U denotes uptake by channel k and B represents barrier metrics for group g.

3. Provision Alignment Score: This regression-based indicator (scaled 0–100) compares provision profiles with proxies of demand, such as employer needs or community survey data. It incorporates fairness diagnostics to identify disparities within subgroups.

All components of the PEI are z-score normalised within each country and weighted using the combined expert-PCA method. To facilitate interpretation, a tiered classification (P25, P50, P75) identifies levels of ecosystem maturity, with robustness checks conducted via ±20% weight variation.

Research questions

• How do digitalisation trends influence the hybridisation of learning provisions across institutional clusters?

• What factors explain variations in informal learning access between welfare regimes?

• How do economic shifts like automation affect the evolution of labour-market provisions?

Policy questions

• How can funding models be adapted to increase accessibility in underrepresented regions?

• What incentives best encourage employers to integrate hybrid forms of non-formal education and informal learning in workplaces?

• How should civic provisions be scaled to address equity gaps in diverse national contexts?

C.3 Core data and information for coordinating lifelong learning

Building on Section 2, which examines the coordination pillar through educational biographies, life-course transitions, skill mismatches, and policy instruments, this summary outlines the data strategies required to align the evolving needs of citizens and organisations with supportive structures within the framework. The framework treats LLL as a dynamic ecosystem. Here, data is the cornerstone: it turns raw metrics into evidence that reduces coordination failures - such as information asymmetries or misaligned incentives - while widening equitable access. The approach prioritises reusable, AI-enhanced datasets to close gaps, such as fragmented tracking of transitions and under-measured societal benefits (e.g. civic resilience). The synthesis reinforces the coordination pillar’s analytical rigour, and ushers in the subsequent focus on data frameworks to support inclusive, adaptive LLL ecosystems across global contexts.

Core data and indicators

The following Table A C.3. outlines key data types, their purposes, and examples, illustrating how they support LLL coordination.

Table A C.3. Key data and indicators focused on LLL coordination

Purpose	Core Data/Indicators	Examples/Sources
Research	Pathway sequences, transition dynamics, and capability stocks	Longitudinal cohorts (e.g., (Marcotte et al., 2005[32]); (Bukodi and Goldthorpe, 2018[92]); (Steindórsdóttir et al., 2023[100]); (dos Santos et al., 2024[101])) matched with administrative data; sequence analysis
Policy Design	Instrument impacts, incentive effects, equity outcomes	ALMP evaluations (Albrecht, van den Berg and Vroman, 2005[96]); (OECD, 2020[102]); (Bonoli, Emmenegger and Felder‐Stindt, 2025[103]); combination of ALMP data and PIAAC/LFS data (e.g., (Dämmrich, Kosyakova and Blossfeld, 2015[104]); (Desjardins and Ioannidou, 2020[105]); (Cabus, Ilieva-Trichkova and Štefánik, 2020[64]))
Coordination	Mismatch indices, externality metrics, governance alignments	Administrative registers (UNESCO, 2019[91]); welfare-labour metrics

Addressing data gaps

Major gaps remain in the data on the coordination of LLL. These include: (1) fragmented coverage of non-linear transitions, such as those linked to family or retirement; (2) limited measurement of non-economic externalities, including social cohesion; and (3) weak integration across levels, from the individual to the organisational and national.

To address some of these shortcomings, we propose an experimental, forward-looking Coordination Dynamics Index. Grounded in complexity theory, this index combines direct measures - such as transition tracking - with indirect proxies, including AI-based analysis of platform interactions. This approach enables real-time monitoring of system alignment, while embedding equity as a core design principle.

Innovative data solutions

To strengthen coordination in LLL, several innovative data strategies can be pursued.

First, enhanced longitudinal tracking. Existing longitudinal datasets, such as ELSA, BCS70 NEPS, should be augmented with AI-driven modules to capture non-linear transitions (e.g. entry into parenthood or retirement). This would make visible the informal contributions to learning that occur across diverse contexts.

Second, cross-level data harmonisation. Dedicated observatories, proposed in the conclusion of this report, could integrate individual data (e.g. responses from cross-sectional surveys like PISA and PIAAC), organisational data (e.g. training logs), and national records (e.g. administrative datasets). Such integration would align provision with need and allow for institutional variation to be more systematically accounted for.

Third, ethical AI and blockchain analytics. Anonymised platform data (e.g. career-service recommendations) combined with blockchain-verified records would enable transitions and externalities to be tracked more reliably. Where direct evidence is limited, synthetic data could be deployed to enable robust yet ethically sensitive analysis, particularly for vulnerable groups.

Fourth, feedback loops. Real-time indicators (e.g. mismatch indices) should feed back into governance processes to support ongoing adjustment. For example, mobile applications that track transition readiness could provide timely support to disadvantaged learners, thereby advancing equity objectives.

Key indicators, research questions, and policy questions

Key indicators and construction

As an illustration of advanced, “blue-sky” measurement within the coordination pillar, we propose the Coordination Dynamic Index (CDI) - a synthetic composite of three indicators designed to assess coordination of LLL.

The framework yields key indicators for coordinating LLL:

1. Transition Dynamics Index: Measures readiness for life shifts, aggregating longitudinal survey data (e.g. skill readiness scores, 1-10) with platform engagement, normalised and weighted (60% survey, 40% engagement) using pandas for processing.

2. Mismatch Resolution Ratio: Quantifies alignment of skills with needs, built from PIAAC surveys, calculating ratios (e.g. resolved mismatches/total), with thresholds (<1 indicating persistent gaps) to flag coordination failures.

3. Externality Impact Score: Evaluates societal benefits like resilience, derived from administrative data, scoring impacts (0-100) via regression in statsmodels, incorporating demographic and policy variables.

Using machine learning - such as Bayesian hierarchical models in Python - it normalises and weights factors by context (e.g. higher weights for transitions in volatile markets). A threshold (e.g. index > 70) signals effective coordination. Network visualisations track stakeholder interactions and help target misalignments, including access barriers for marginalised groups.

Research questions

• How do early educational experiences shape long-term coordination of LLL pathways across contexts?

• What role do digital platforms play in mitigating information asymmetries in LLL coordination?

• How do governance structures influence the alignment of multi-level needs with provisions?

Policy questions

• How can tripartite agreements optimise coordination for low-skilled groups across welfare regimes?

• What interventions best address skill mismatches in volatile labour markets?

• How should policies leverage data to enhance societal benefits like civic engagement?

C.4 Summary boxes

The framework provides a basis for proposing a set of indicators to quantify LLL dynamics and to capture interactions among stakeholders (learners, providers, policymakers), technologies (e.g. digital platforms), and data sources (surveys and administrative records). At this stage, we propose exploratory indicators that leverage existing data systems and Python-based analytics (e.g. Pandas for data processing and scikit-learn for clustering). These initiatives could be piloted through projects similar to Norway’s Learning Analytics project (Wasson et al., 2024[85]), which focuses on integrating new data sources to enhance learning. The proposals aim to address gaps in LLL measurement and foster resilient ecosystems, pending further validation and development.

Below is a concise overview of the indicators proposed for exploration organised by pillar, with explanations of their purpose, data sources, methods, potential insights, and relevance to the framework.

C.4.1 Demand: measuring capability gaps

Proposed indicators in Box A C.1. aim to quantify learning needs as capability gaps, supporting the Demand pillar’s focus on identifying expressed and latent needs across individual, organisational, and national levels (Section 3).

Box A C.1. Demand Indicators

Latent needs index

Description: Proposed to measure unexpressed demands among marginalised groups (e.g. low-skilled adults).

Data: Survey responses on barriers (e.g., AES/PIAAC) and platform engagement metrics (e.g. YouTube skill searches).

Method: Normalises and weights data (60% barriers, 40% engagement) using pandas, potentially piloted through community-driven data logging.

Potential insights: Could reveal hidden needs, such as dispositional barriers.

Why in framework: Aims to quantify demand gaps to inform targeted interventions, supporting bounded agency and RA3’s inclusive data strategies.

Motivation-barrier ratio

Description: Proposed to assess the balance between learner motivations and barriers to highlight potential inequities.

Data: AES/PIAAC responses on motivation and barrier scores.

Method: Calculates ratio (motivation ÷ barrier), with <1 indicating high barriers, using pandas for aggregation.

Potential insights: This could reveal why needs go unmet, for example, subsidies in some countries may reduce financial barriers more effectively than in others.

Why in framework: Seeks to inform policies to reduce barriers, aligning with the demand pillar’s focus on capability gaps and RA2’s equity focus.

Transition readiness score

Description: Proposed to evaluate preparedness for life transitions (e.g. career shifts, retirement).

Data: Longitudinal cohorts (e.g. BCS70, NEPS, PIAAC longitudinal study in Germany) and job platform analytics (e.g. Indeed job searches).

Method: Regression modelling in statsmodels, incorporating demographic variables.

Potential Insights: Could highlight skill gaps affecting transitions.

Why in Framework: Aims to link demand to coordination, supporting policies for smooth transitions and RA3’s longitudinal tracking.

C.4.2 Supply: Mapping Diverse Provisions

Proposed indicators in Box A C.2. map learning opportunities across various types of learning, aligning with the supply pillar’s emphasis on provision diversity and accessibility (Section 4).

Box A C.2. Supply Indicators

Hybridisation ratio

Description: Proposed to measure the blend of formal, non-formal, and informal provisions, along digital types of learning.

Data: AES survey data on learning formats; Coursera logs for micro-credentials.

Method: Calculates proportions (hybrid activities ÷ total) using SciPy clustering to categorise forms by structure and digitalisation.

Potential insights: Could capture hybridisation, flexibilisation and digitalisation trends.

Why in framework: Seeks to map provision diversity, informing policies on emerging forms like micro-credentials (Section 6.2) and supporting RA1’s taxonomy.

Accessibility index

Description: Proposed to evaluate provision availability across demographics (e.g. older vs. young learners, high vs. low occupational status).

Data: Administrative registries from providers; LinkedIn analytics on course access.

Method: Aggregates weighted rates, normalised via pandas.

Potential insights: Could identify access inequities, e.g. older learners may face different barriers than younger learners in different regimes.

Why in framework: Aims to inform equitable provision policies, aligning with the Supply pillar (Section 4) and RA2’s national variations.

Provision alignment score

Description: Proposed to assess alignment between learning opportunities and needs.

Data: PIAAC or employer data on training needs; funding levels from administrative records.

Method: Regression in statsmodels, scoring 0–100, with fairness diagnostics.

Potential Insights: Could highlight mismatches, e.g. labour-focused provisions may neglect civic needs in fragmented systems.

Why in Framework: Seeks to link supply to coordination, supporting policies for effective provisions (Section 2) and RA3’s data integration.

C.4.3 Coordination: Aligning pathways and policies

The proposed indicators in Box A C.3. and Box A C.4. aim to align demand and supply, supporting the coordination pillar’s role in resolving asymmetries and facilitating transitions (Section 2).

Box A C.3. Coordination Indicators

Mismatch resolution ratio

Description: Proposed to quantify alignment between skills and needs.

Data: PIAAC/employer surveys; labour registries.

Method: Calculates resolved mismatches ÷ total, with thresholds <1 signalling gaps, using statsmodels.

Potential insights: Could reveal persistent mismatches.

Why in framework: Aims to identify coordination barriers, informing transition policies (Section 2.3) and RA2’s variation analysis.

Externality impact score

Description: Proposed to evaluate societal benefits like resilience and cohesion.

Data: Administrative data on programme outcomes; social media analytics.

Method: Regression in statsmodels, scoring 0–100, incorporating demographic variables.

Potential insights: Could capture wider outcomes, e.g. civic programs may boost community cohesion.

Why in framework: Seeks to support policies for social outcomes (Section 3.3), aligning with RA3’s focus on externalities.

Governance alignment score

Description: Proposed to quantify policy-stakeholder synergy.

Data: AES/PIAAC responses; LinkedIn analytics on engagement.

Method: Principal component analysis (PCA) in scikit-learn, scoring 0–100 (>80 indicates strong alignment).

Potential insights: Could flag governance failures, e.g. fragmented systems may show lower alignment.

Why in framework: Aims to support stakeholder-driven policies for resilient systems (Section 2.5), advancing RA2’s governance focus.

Box A C.4. Resilient Data Ecosystems Indicator

Description: Proposed to synthesise indicators to link demand, supply, and coordination, measuring system interdependencies.

Data: Combined survey (PIAAC, AES), administrative (UOE, INES), and big data (LinkedIn, Coursera logs).

Method: Dynamic analytics (e.g. clustering in scikit-learn, regression in statsmodels).

Potential Insights: Could reveal interdependencies, e.g. demand gaps may reduce supply uptake in uncoordinated systems.

Why in Framework: Seeks to unify pillars to foster adaptive ecosystems, supporting RA1–RA3 by integrating diverse data for policy and coordination (Sections 2–4).

References

[96] Albrecht, J., G. van den Berg and S. Vroman (2005), “The Knowledge Lift: The Swedish Adult Education Program that Aimed to Eliminate Low Worker Skill Levels”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.673516.

[31] Arrow, K. (1973), “Higher education as a filter”, Journal of Public Economics, Vol. 2/3, pp. 193-216, https://doi.org/10.1016/0047-2727(73)90013-3.

[30] Becker, G. (1964), Human capital: A theoretical and empirical analysis with special reference to education, University of Chicago Press.

[57] Billett, S. (2002), “Critiquing workplace learning discourses: Participation and continuity at work”, Studies in the Education of Adults, Vol. 34/1, pp. 56-67.

[6] Boeren, E. (2016), Lifelong Learning Participation in a Changing Policy Context. An Interdisciplinary Theory, London: Palgrave Macmillan.

[87] Boeren, E. et al. (2012), “Why do adults learn? Developing a motivational typology across 12 European countries”, Globalisation, Societies and Education, Vol. 10/2, pp. 247-269, https://doi.org/10.1080/14767724.2012.678764.

[103] Bonoli, G., P. Emmenegger and A. Felder‐Stindt (2025), “Re‐Skilling in the Age of Skill Shortage: Adult Education Rather Than Active Labor Market Policy”, Regulation & Governance, Vol. 20/2, pp. 482-494, https://doi.org/10.1111/rego.70065.

[48] Borgonovi, F. and J. Suarez-Alvarez (2025), “How can adult skills assessments best meet the demands of the 21st century?”, OECD Social, Employment and Migration Working Papers, No. 319, OECD Publishing, Paris, https://doi.org/10.1787/853db37b-en.

[61] Boshier, R. (1971), “Motivational Orientations of Adult Education Participants: a Factor Analytic Exploration of Houle’s Typology”, Adult Education, Vol. 21/2, pp. 3-26, https://doi.org/10.1177/074171367102100201.

[55] Bourdieu, P. (1986), The forms of capital, In J. G. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp. 241–258). Greenwood Press.

[35] Bourdieu, P. and J. Passeron (1977), Reproduction in education, society and culture, Sage, https://uk.sagepub.com/en-gb/eur/reproduction-in-education-society-and-culture/book203162.

[3] Boyadjieva, P. and P. Ilieva-Trichkova (2021), Adult Education as Empowerment. Re-imagining Lifelong Learning through the Capability Approach, Recognition Theory and Common Goods Perspective, Palgrave Macmillan.

[70] Broek, S. et al. (2025), “Exploring card-sorting potential to uncover the interplay between adult learners’ motivations and barriers to (start to) learn”, International Journal of Lifelong Education, Vol. 44/3, pp. 330-346, https://doi.org/10.1080/02601370.2025.2462163.

[89] Broek, S. et al. (2024), “Conditions for successful adult learning systems at local level: creating a conducive socio-spatial environment for adults to engage in learning”, International Journal of Lifelong Education, Vol. 43/2-3, pp. 200-223, https://doi.org/10.1080/02601370.2024.2338366.

[7] Bronfenbrenner, U. (1979), The ecology of human development: Experiments by nature and design, Harvard University Press.

[54] Brown, P., H. Lauder and S. Cheung (2020), The Death of Human Capital? Its Failed Promise and How to Renew It in Age of Disruption, Oxford University Press.

[92] Bukodi, E. and J. Goldthorpe (2018), Social Origins, Ability and Educational Attainment: Is there a Wastage of Talent?, In: E. Bukodi & J. H. Goldthorpe. Social Mobility and Education in Britain: Research, Politics and Policy (pp. 108-127) Cambridge University Press.

[64] Cabus, S., P. Ilieva-Trichkova and M. Štefánik (2020), “Multi-layered perspective on the barriers to learning participation of disadvantaged adults, Mehrebenen-Perspektive auf die Barriere für die Weiterbildungsteilnahme benachteiligter Erwachsener”, Zeitschrift für Weiterbildungsforschung, Vol. 43/2, pp. 169-196, https://doi.org/10.1007/s40955-020-00162-3.

[37] Cedefop (2025), European skills index. Skills matching. European Centre for the Development of Vocational Training, https://www.cedefop.europa.eu/en/tools/european-skills-index/skills-matching.

[72] Cedefop (2016), Leaving education early: putting vocational education and training centre stage, Publications Office, https://doi.org/10.2801/893397.

[51] Cedefop (2010), The skill matching challenge: Analysing skill mismatch and policy implications, https://www.cedefop.europa.eu/files/3056_en.pdf.

[22] Chetty, R. et al. (2024), Changing Opportunity: Sociological Mechanisms Underlying Growing Class Gaps and Shrinking Race Gaps in Economic Mobility. NBER WORKING PAPER NO. 32697, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w32697.

[15] Clarke, A. (2005), Situational Analysis, SAGE Publications, Inc., 2455 Teller Road, Thousand Oaks California 91320 United States of America , https://doi.org/10.4135/9781412985833.

[62] Cross, K. (1981), Adults as learners: Increasing participation and facilitating learning, Jossey-Bass.

[104] Dämmrich, J., Y. Kosyakova and H. Blossfeld (2015), “Gender and job-related non-formal training: A comparison of 20 countries”, International Journal of Comparative Sociology, Vol. 56/6, pp. 433-459, https://doi.org/10.1177/0020715215626769.

[86] Desjardins, R. (2020), “PIAAC thematic review on adult learning”, OECD Publishing.

[23] Desjardins, R. (2017), “Political economy of adult learning systems: Comparative analysis of strategies, policies and constraints”, Bloomsbury Academic.

[52] Desjardins, R. (2014), “Rewards to skill supply, skill demand and skill match–Mismatch using direct measures of skills. Studies using the adult literacy and lifeskills survey”, Lund economic studies no 176, Lund University.

[105] Desjardins, R. and A. Ioannidou (2020), “The political economy of adult learning systems—some institutional features that promote adult learning participation, Die politische Ökonomie von Weiterbildungssystemen – einige institutionelle Merkmale, die Weiterbildungsbeteiligung fördern”, Zeitschrift für Weiterbildungsforschung, Vol. 43/2, pp. 143-168, https://doi.org/10.1007/s40955-020-00159-y.

[94] Desjardins, R. and J. Kalenda (forthcoming), Worlds of Adult Learning Systems.

[42] Desjardins, R. and A. Warnke (2012), “Ageing and Skills: A Review and Analysis of Skill Gain and Skill Loss Over the Lifespan and Over Time”, OECD Education Working Papers, No. 72, OECD Publishing, Paris, https://doi.org/10.1787/5k9csvw87ckh-en.

[101] dos Santos, S. et al. (2024), “Does training beget training over the life course? Cumulative advantage in work-related non-formal training participation in Germany and the UK”, European Sociological Review, Vol. 40/3, pp. 464-478, https://doi.org/10.1093/esr/jcae022.

[5] Elder, J. (1994), “Time, Human Agency, and Social Change: Perspectives on the Life Course”, Social Psychology Quarterly, Vol. 57/1, pp. 4-15, https://doi.org/10.2307/2786971.

[49] Elias, P., T. Hogarth and G. Pierre (2002), “The wider benefits of education and training: a comparative longitudinal study”, (No. 178). Norwich. Warwick Institute for Employment Research, Department for Work and Pensions, https://warwick.ac.uk/fac/soc/ier/publications/2002/elias_et_al_2002_.

[71] Eraut *, M. (2004), “Informal learning in the workplace”, Studies in Continuing Education, Vol. 26/2, pp. 247-273, https://doi.org/10.1080/158037042000225245.

[76] Erola, J., M. Mills and H. Solga (2023), “Beyond Education and Training – How Can We Adapt to Future Needs of Local Labour Markets?”, Population and Policy 42.

[18] Evans, K. (2007), “Concepts of bounded agency in education, work, and the personal lives of young adults”, International Journal of Psychology, Vol. 42/2, pp. 85-93, https://doi.org/10.1080/00207590600991237.

[25] Field, J. (2011), “Researching the benefits of learning: the persuasive power of longitudinal studies”, London Review of Education. Advance online publication, https://doi.org/10.1080/14748460.2011.61632.

[28] Folke, C. (2006), “Resilience: The emergence of a perspective for social–ecological systems analyses”, Global Environmental Change, Vol. 16/3, pp. 253-267, https://doi.org/10.1016/j.gloenvcha.2006.04.002.

[9] Giddens, A. (1984), The constitution of society: Outline of the theory of structuration, University of California Press.

[82] Grimshaw, D. (2022), “Informal and precarious work: Persistent inequalities and new challenges”, United Nations Department of Economic and Social Affairs (UNDESA), https://www.un.org/development/desa/dspd/wp-content/uploads/sites/22/2022/06/Damian-Grimshaw-Paper.pdf.

[84] Grotlüschen, A. et al. (2023), “Adult learning and education within the framework of lifelong learning”, International Perspectives in Adult Education, No. 81, https://www.dvv-international.de/en/adult-education-and-development/international-perspectives-in-adult-education.

[46] Grotlüschen, A. et al. (2016), “Adults with Low Proficiency in Literacy or Numeracy”, OECD Education Working Papers, No. 131, OECD Publishing, Paris, https://doi.org/10.1787/5jm0v44bnmnx-en.

[47] Hanushek, E. et al. (2025), “Age and cognitive skills: Use it or lose it”, Science Advances, Vol. 11/10, https://doi.org/10.1126/sciadv.ads1560.

[33] Hanushek, E. and L. Woessmann (2015), The Knowledge Capital of Nations: Education and the Economics of Growth, Cambridge, MA: MIT Press.

[53] Hartog, J. (2000), “Over-education and earnings: where are we, where should we go?”, Economics of Education Review, Vol. 19/2, pp. 131-147, https://doi.org/10.1016/s0272-7757(99)00050-3.

[40] Hodge, S. et al. (2025), “The perfect (skills) match? Tracking a chimera of lifelong learning policy”, International Journal of Lifelong Education, Vol. 44/4, pp. 351-357, https://doi.org/10.1080/02601370.2025.2528002.

[77] Hornberg, C., J. Heisig and H. Solga (2024), “Explaining the training disadvantage of less-educated workers: The role of labor market allocation in international comparison”, Socio-Economic Review, Vol. 22/1, pp. 195–222.

[60] Houle, C. (1961), The Inquiring Mind, Madison: The University of Wisconsin Press.

[34] Illich, I. (1971), Deschooling society, Harper & Row.

[97] IWWB (2025), InfoWeb Weiterbildung [Website], https://www.iwwb.de/information/UEber-das-IWWB-weiterbildung-20.html.

[16] Jarvis, P. (2004), Adult education and lifelong learning, Routledge.

[59] Kalenda, J., E. Boeren and I. Kočvarová (2023), “Exploring attitudes towards adult learning and education: Group patterns among participants and non-participants”, Studies in Continuing Education, Vol. 46/3, pp. 33–50.

[27] Kalenda, J. and R. Desjardins (2025), Trends and patterns in the development of adult learning systems, In R. Desjardins & J. Kalenda (Eds.), A Modern Guide to Adult Learning Systems (pp. 382–408). Edward Elgar.

[68] Kalenda, J. and I. Kočvarová (2022), “Why do not they participate? Reasons for non-participation in adult learning and education from the viewpoint of self-determination theory”, (RELA) Journal of European Research of learning and education of Adults, Kalenda, J., & Kočvarová, I. (2022). Why do not they participate? Reasons for non-participation in adult learning and education from the viewpoint of self-determination theory. (RELA) Journal of European Research of learning and education of Adults, 13(1), 1–16., pp. 1–16.

[78] Kalenda, J., J. Vaculíková and I. Kočvarová (2024), “Gender inequality in adult education: a comparative study of four adult learning systems”, Cogent Education, Vol. 11/1, https://doi.org/10.1080/2331186x.2024.2390688.

[58] Kalenda, J., J. Vaculíková and I. Kočvarová (2022), “Barriers to the participation of low-educated workers in non-formal education”, Journal of Education and Work, Vol. 35/5, pp. 455-469, https://doi.org/10.1080/13639080.2022.2091118.

[69] Karger, E., J. Kalenda and I. Kočvarová (2022), “Legitimisation of non-participation in adult education and training”, International Journal of Lifelong Education, Vol. 41/3, pp. 307-325.

[88] Karger, T., J. Kalenda and J. Vaculíková (2025), “Participation reimagined: beyond the one-dimensional approach to participation in adult learning and education”, International Journal of Lifelong Education, Vol. 44/3, pp. 291-305, https://doi.org/10.1080/02601370.2025.2496654.

[73] Kluge, A. and J. Schilling (2003), “Organizational learning and learning organizations: Theory and empirical findings.”, The Psychologist-Manager Journal, Vol. 6/1, pp. 31-50, https://doi.org/10.1037/h0095917.

[75] Lancaster, A. (2020), Driving Performance Through Learning, Kogan Page Limited.

[10] Latour, B. (2005), Reassembling the social: An introduction to actor-network theory, Oxford University Press.

[95] Manninen, J. and K. Vuorikoski (2024), “Policy vs. practice. An analysis of the course provision of Finnish study centres between 2016 and 2021”, International Journal of Lifelong Education, Vol. 45/1, pp. 49-60, https://doi.org/10.1080/02601370.2024.2442625.

[32] Marcotte, D. et al. (2005), “The Returns of a Community College Education: Evidence From the National Education Longitudinal Survey”, Educational Evaluation and Policy Analysis, Vol. 27/2, pp. 157-175, https://doi.org/10.3102/01623737027002157.

[45] Massing, N. et al. (2025), “A first look at low literates in PIAAC Germany 2023”, International Journal of Lifelong Education, Vol. 44/5, pp. 497-521, https://doi.org/10.1080/02601370.2025.2520491.

[79] Mertens, B. et al. (2025), “Unraveling low-educated adults’ motivation for second-chance education: A multidimensional perspective”, Learning and Instruction, Vol. 97, p. 102075, https://doi.org/10.1016/j.learninstruc.2024.102075.

[74] Messmann, G., M. Segers and F. Dochy (2018), Informal Learning at Work. Triggers, Antecedents, and Consequences, Routledge.

[67] Michie, S., M. van Stralen and R. West (2011), “The behaviour change wheel: A new method for characterising and designing behaviour change interventions”, Implementation Science, Vol. 6/1, https://doi.org/10.1186/1748-5908-6-42.

[41] Milana, M. et al. (2025), “Skills, skills, skills! The European Commission’s endorsement of a skills-first approach and the future of adult education and lifelong learning”, International Journal of Lifelong Education, Vol. 44/2, pp. 127–130.

[99] Milana, M. et al. (2024), “Artificial intelligence (AI), conversational agents, and generative AI: implications for adult education practice and research”, International Journal of Lifelong Education, Vol. 43/1, pp. 1-7, https://doi.org/10.1080/02601370.2024.2310448.

[66] Neidhöfer, G. et al. (2021), “Social mobility and economic development”, CEDLAS Working Paper No. 286. Universidad Nacional de La Plata.

[2] Nussbaum, M. (2011), Creating capabilities: The human development approach, Harvard University Press.

[98] OECD (2025), Trends in Adult Learning: New Data from the 2023 Survey of Adult Skills, Getting Skills Right, OECD Publishing, Paris, https://doi.org/10.1787/ec0624a6-en.

[39] OECD (2024), Do Adults Have the Skills They Need to Thrive in a Changing World?: Survey of Adult Skills 2023, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/b263dc5d-en.

[102] OECD (2020), Strengthening the Governance of Skills Systems: Lessons from Six OECD Countries, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/3a4bb6ea-en.

[56] OECD (2019), Skills Matter: Additional Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/1f029d8f-en.

[83] OECD (2007), Understanding the Social Outcomes of Learning, OECD Publishing, Paris, https://doi.org/10.1787/9789264034181-en.

[12] Poell, T., D. Nieborg and J. van Dijck (2019), “Platformisation”, Internet Policy Review, Vol. 8/4, https://doi.org/10.14763/2019.4.1425.

[8] Prigogine, I. and I. Stengers (1984), Order out of chaos: Man’s new dialogue with nature., Bantam Books.

[4] Putnam, R. (2000), Bowling alone: The collapse and revival of American community, Simon & Schuster.

[44] Rammstedt, B. et al. (2024), PIAAC 2023. Grundlegende Kompetenzen Erwachsener im internationalen Vergleich, Waxmann Verlag GmbH, https://doi.org/10.31244/9783830999652.

[63] Roosmaa, E. and E. Saar (2017), “Adults who do not want to participate in learning: a cross-national European analysis of their perceived barriers”, International Journal of Lifelong Education, Vol. 36/3, pp. 254-277, https://doi.org/10.1080/02601370.2016.1246485.

[26] Rothstein, B. and D. Stolle (2008), “The State and Social Capital: An Institutional Theory of Generalized Trust”, The State and Social Capital: An Institutional Theory of Generalized Trust. Comparative Politics, Vol. 40/4, pp. 441–459, http://www.jstor.org/stable/20434095.

[65] Rubenson, K. (2018), Conceptualising participation in adult learning and education: Equity issues, In M. Milana, S. Webb, J. Holford, R. Waller, & P. Jarvis (Eds.), The Palgrave international handbook on adult and lifelong education and learning (pp. 337–357). Palgrave Macmillan.

[19] Rubenson, K. and R. Desjardins (2009), “The Impact of Welfare State Regimes on Barriers to Participation in Adult Education”, Adult Education Quarterly, Vol. 59/3, pp. 187-207, https://doi.org/10.1177/0741713609331548.

[17] Salling Olesen, H. (2001), “Professional identity as learning processes in life histories”, Journal of Workplace Learning, Vol. 13/7/8, pp. 290-298, https://doi.org/10.1108/13665620110411076.

[24] Schuller, T. et al. (2004), The benefits of learning: The impact of education on health, family life and social capital, Routledge.

[50] Schultheiss, T. and U. Backes‐Gellner (2023), “Different degrees of skill obsolescence across hard and soft skills and the role of lifelong learning for labor market outcomes”, Industrial Relations: A Journal of Economy and Society, Vol. 62/3, pp. 257-287, https://doi.org/10.1111/irel.12325.

[36] Schultz, T. (1959), “Investment in Man: An Economist’s View”, Social Service Review, Vol. 33/2, pp. 109-117, https://doi.org/10.1086/640656.

[14] Scott, A. (2024), The Longevity Imperative, Basic.

[93] Sekmokas, M. et al. (2024), “Updated Framework for Monitoring Adult Learning: Enhancing Data Identification and Indicator Selection”, OECD Education Working Paper No. 317.

[1] Sen, A. (1999), Development as freedom, Oxford University Press.

[29] Spence, M. (1973), “Job Market Signaling”, The Quarterly Journal of Economics, Vol. 87/3, pp. 355-374, https://doi.org/10.2307/1882010.

[100] Steindórsdóttir, B. et al. (2023), “Career transitions and career success from a lifespan developmental perspective: A 15-year longitudinal study”, Journal of Vocational Behavior, Vol. 140, p. 103809, https://doi.org/10.1016/j.jvb.2022.103809.

[81] Stiglitz, J. and B. Greenwald (2014), Creating a learning society: A new approach to growth, development, and social progress, Columbia University Press.

[13] Suleyman, M. and M. Bhaskar (2024), The Coming Wave. AI, Power and our Future, Vintage.

[91] UNESCO (2019), Global report on adult learning and education, UNESCO Institute for Lifelong Learning.

[38] UNESCO (1996), “Learning: The treasure within”, Report to UNESCO of the International Commission on Education for the Twenty-first Century. UNESCO, https://unesdoc.unesco.org/ark:/48223/pf0000109590.

[43] van der Velden, R. and I. Bijlsma (2019), “Effective skill: A new theoretical perspective on the relation between skills, skill use, mismatches, and wages”, Effective skill: A new theoretical perspective on the relation between skills, skill use, mismatches, and wages. Oxford Economic Papers, Vol. 71/1, pp. 145–165.

[11] van Dijck, J., T. Poell and M. De Waal (2018), The Platform Society: Public Values in a Connective World, Oxford University Press.

[90] Van Nieuwenhove, L. and B. De Wever (2023), “Psychosocial Barriers to Adult Learning and the Role of Prior Learning Experiences: A Comparison Based on Educational Level”, Adult Education Quarterly, Vol. 74/1, pp. 62-87, https://doi.org/10.1177/07417136231147491.

[80] Vignoles, A., F. Galindo‐Rueda and L. Feinstein (2004), “The Labour Market Impact of Adult Education and Training: A Cohort Analysis”, Scottish Journal of Political Economy, Vol. 51/2, pp. 266-280, https://doi.org/10.1111/j.0036-9292.2004.00306.x.

[85] Wasson, B. et al. (2024), “Implementing Learning Analytics in Norway”, Journal of Learning Analytics, Vol. 11/2, pp. 268-280, https://doi.org/10.18608/jla.2024.8241.

[20] West, R. and S. Michie (2020), A brief introduction to the COM-B model of behaviour and the PRIME Theory of motivation, Qeios.

[21] World Bank (2016), Snapshot. Investing in the early years for growth and productivity, Washington DC. World Bank Group, https://www.worlddata.info/life-expectancy.php.