Periodic Reporting for period 1 - EU-ALMPO (EU-ALMPO -EU Active Labor Market Policies Observatory)
Reporting period: 2025-01-01 to 2025-12-31
European labour markets are undergoing profound transformation driven by digitalisation, the green transition, demographic ageing and geopolitical instability. At the same time, many Member States face a persistent paradox: unemployment coexists with skills shortages and unfilled vacancies. Active Labour Market Policies (ALMPs) are the main public instruments addressing these mismatches, yet across the EU they remain fragmented, unevenly evaluated, weakly coordinated with skills intelligence and insufficiently supported by advanced data analytics.
EU-ALMPO addresses this gap by establishing the EU Active Labour Market Policies Observatory — a structured, AI-supported knowledge and decision-support infrastructure designed to improve the effectiveness, efficiency and accessibility of ALMPs. The Observatory integrates: a validated analytical framework on ALMP effectiveness (WP1), comparative mapping of national skills ecosystems (WP2), an AI-powered repository of structured policy evidence (WP3), data-informed labour market analysis tools (WP4–WP5), participatory experimentation and evaluation (WP6–WP7), structured dissemination (WP8), and robust governance and quality assurance (WP9).
Its objective is to strengthen evidence-based ALMP design by combining policy analysis, labour economics and AI-enabled knowledge structuring. The pathway to impact includes consolidating evidence on determinants of effective ALMPs, transforming heterogeneous documentation into machine-readable datasets, developing AI-assisted decision-support tools, testing approaches through experimentation, and promoting EU-level learning.
EU-ALMPO addresses this gap by establishing the EU Active Labour Market Policies Observatory — a structured, AI-supported knowledge and decision-support infrastructure designed to improve the effectiveness, efficiency and accessibility of ALMPs. The Observatory integrates: a validated analytical framework on ALMP effectiveness (WP1), comparative mapping of national skills ecosystems (WP2), an AI-powered repository of structured policy evidence (WP3), data-informed labour market analysis tools (WP4–WP5), participatory experimentation and evaluation (WP6–WP7), structured dissemination (WP8), and robust governance and quality assurance (WP9).
Its objective is to strengthen evidence-based ALMP design by combining policy analysis, labour economics and AI-enabled knowledge structuring. The pathway to impact includes consolidating evidence on determinants of effective ALMPs, transforming heterogeneous documentation into machine-readable datasets, developing AI-assisted decision-support tools, testing approaches through experimentation, and promoting EU-level learning.
During RP1 (M01–M12), EU-ALMPO established the analytical backbone of the Observatory and built its core knowledge infrastructure.
WP1 – Analytical Framework
WP1 delivered the conceptual foundation through structured definitions of ALMPs and skills mismatch, a systematic review of labour market trends, and a meta-evaluation of 140+ ALMP studies identifying determinants of effectiveness and unintended effects (displacement, deadweight, substitution, creaming). These activities culminated in Deliverable D1.1 — the Analytical Framework for Designing and Implementing ALMPs — validated in an international workshop with 46 participants. The framework now serves as the methodological reference for subsequent technical work.
WP2 – Skills Ecosystem Mapping
WP2 mapped skills ecosystems in five Member States, identifying governance models, institutional arrangements and skills intelligence mechanisms. Ten in-depth case studies were completed using a Context–Mechanism–Outcome (CMO) approach, generating 37 structured configurations highlighting recurring governance and effectiveness patterns.
WP3 – AI Knowledge Infrastructure
WP3 delivered the AI-assisted Document Annotation Tool (D3.1) an ALMP-specific taxonomy and a human-in-the-loop validation workflow, enabling transformation of policy documents into machine-interpretable datasets and preparing for Retrieval-Augmented Generation (RAG) integration.
WP4–WP7 – Foundations
Conceptual alignment between labour market variables and AI architecture was achieved; modelling principles were defined; evaluation baselines established; and pilot coordination initiated.
All RP1 scientific objectives were achieved. Deliverable sequencing adjustments were procedural and did not affect progress.
WP1 – Analytical Framework
WP1 delivered the conceptual foundation through structured definitions of ALMPs and skills mismatch, a systematic review of labour market trends, and a meta-evaluation of 140+ ALMP studies identifying determinants of effectiveness and unintended effects (displacement, deadweight, substitution, creaming). These activities culminated in Deliverable D1.1 — the Analytical Framework for Designing and Implementing ALMPs — validated in an international workshop with 46 participants. The framework now serves as the methodological reference for subsequent technical work.
WP2 – Skills Ecosystem Mapping
WP2 mapped skills ecosystems in five Member States, identifying governance models, institutional arrangements and skills intelligence mechanisms. Ten in-depth case studies were completed using a Context–Mechanism–Outcome (CMO) approach, generating 37 structured configurations highlighting recurring governance and effectiveness patterns.
WP3 – AI Knowledge Infrastructure
WP3 delivered the AI-assisted Document Annotation Tool (D3.1) an ALMP-specific taxonomy and a human-in-the-loop validation workflow, enabling transformation of policy documents into machine-interpretable datasets and preparing for Retrieval-Augmented Generation (RAG) integration.
WP4–WP7 – Foundations
Conceptual alignment between labour market variables and AI architecture was achieved; modelling principles were defined; evaluation baselines established; and pilot coordination initiated.
All RP1 scientific objectives were achieved. Deliverable sequencing adjustments were procedural and did not affect progress.
EU-ALMPO advances the state of the art in three domains:
(i) ALMP Evaluation Methodology
A multi-dimensional framework linking policy design variables to contextual conditions and measurable outcomes, enabling structured gap analysis.
(ii) Comparative Skills Ecosystem Intelligence
Mechanism-based cross-country analysis moving beyond descriptive benchmarking toward structured comparative learning.
(iii) AI-Enabled Policy Intelligence
Development of a structured ALMP taxonomy, AI-assisted annotation infrastructure, and foundations for an ALMP Design Wizard. Unlike traditional observatories, EU-ALMPO establishes a policy intelligence architecture supporting AI-assisted retrieval and synthesis.
Potential impacts include improved evidence-based ALMP design, reduced fragmentation, enhanced responsiveness to digital and green transitions, and stronger targeting of vulnerable groups. The architecture is scalable and designed for long-term EU integration.
Key needs for uptake include completion of analytics tools, piloting of innovation experiments, institutional embedding with PES and EU bodies, sustained data access, compliance with the EU AI Act, long-term governance arrangements, alignment with ESCO/ISCO/NACE standards, and further research integrating predictive modelling.
(i) ALMP Evaluation Methodology
A multi-dimensional framework linking policy design variables to contextual conditions and measurable outcomes, enabling structured gap analysis.
(ii) Comparative Skills Ecosystem Intelligence
Mechanism-based cross-country analysis moving beyond descriptive benchmarking toward structured comparative learning.
(iii) AI-Enabled Policy Intelligence
Development of a structured ALMP taxonomy, AI-assisted annotation infrastructure, and foundations for an ALMP Design Wizard. Unlike traditional observatories, EU-ALMPO establishes a policy intelligence architecture supporting AI-assisted retrieval and synthesis.
Potential impacts include improved evidence-based ALMP design, reduced fragmentation, enhanced responsiveness to digital and green transitions, and stronger targeting of vulnerable groups. The architecture is scalable and designed for long-term EU integration.
Key needs for uptake include completion of analytics tools, piloting of innovation experiments, institutional embedding with PES and EU bodies, sustained data access, compliance with the EU AI Act, long-term governance arrangements, alignment with ESCO/ISCO/NACE standards, and further research integrating predictive modelling.