The strength of a National Action Plan is directly related to its evidence chain: the traceable path from who is vulnerable, where, and why, to the measures planned and the indicators that follow. Greece grounded its plan in commissioned research, a deep-dive analysis and a second study, plus a national stakeholder survey identifying barriers, priorities and children's perspectives, then gave the ECG Recommendation the force of law. Portugal reached comparable coherence with no externally commissioned research, by structuring the entire plan around an in-depth analysis of nine access barriers and replicating the same diagnosis down to municipal level. UNICEF's review of the EU plans found that plans anchored in data, with dedicated evidence generation built in from the start, had more coherent logical chains, more grounded targeting and more actionable monitoring.
Build the plan that reaches its children
Three applied activities from the same evidence base: follow an evidence chain until it breaks, choose how to build evidence under a real deadline, and pick the governance model that survives your own institutional conflict.
The problem
In any country, the authorities responsible for the Child Guarantee can develop a National Action Plan that includes hundreds of measures, covers all five policy areas, and references the right data. Months into implementation, asked in a committee which specific groups of children in need the plan will actually reach and where, and how change will be measured, the honest answer is often that no one can trace it. The plan catalogues what the country already does and fails to respond to the specific needs the Child Guarantee aims to address.
Three structural reasons keep recurring.
First, sub-national and demographic disaggregation is missing. National averages are not an accurate picture of the most vulnerable children. Nationally or regionally collected and aggregated indicators and existing programme data conceal territorial, ethnic, gender and age-specific concentrations of deprivation. Across all 27 EU plans, most of the evidence base stopped at national level; gender disaggregation was almost absent and differences between children and adolescents were rarely distinguished, so the evidence could not fully identify the children the plan was meant to prioritise.
Second, consolidation is mistaken for analysis. Assembling measures already planned across sectoral strategies into one thematic document is fast, and the result looks complete on paper. It rarely turns into a comprehensive and coherent plan that targets specific vulnerable groups against identified barriers.
Third, the chain is rarely made explicit or traceable. Even where good evidence exists, the connection from analysis to target group to measure to indicator is often undocumented. That removes accountability for the choices made, and makes each subsequent planning cycle start from scratch.
The rest of this brief covers what it takes to build and trace that evidence chain, and what a country needs to make it happen.
“We aimed for a plan that is child-centred and grounded in lived experience, not only in statistical analysis. Previous national measures were fragmented and lacking coordination.”
Different routes to a situation analysis
Documented countries reached this in different ways, and varied within each. Bulgaria, Greece and Lithuania each commissioned dedicated deep dives with UNICEF technical support during the Child Guarantee testing phase, some based on a single study and others on several. Czech Republic synthesised more than 120 existing sources without commissioning new research. Portugal combined a barrier analysis with a review of existing data. The approach was not the deciding factor. What mattered was whether the evidence triangulated multiple sources (administrative data, primary data collection, for example on barriers to accessing services, and families’ and children’s own perspectives) and whether findings were translated into the identification of target groups, the planning of specific measures and the development of dedicated indicators, rather than left as a report.
An independent entity as coordinator anchored in law, and evidence from every direction
Deep dive, national survey, school questionnaire and focus groups, triangulated into one plan.
The National Centre for Social Solidarity (EKKA), an independent public-law entity rather than a ministerial unit, was made National Coordinator by law. That secured political neutrality across ministries, operational flexibility, and continuity independent of political cycles. The evidence combined a UNICEF-supported deep dive, a second study by the IOBE economic-research foundation, a field survey drawing around 350 responses from 173 organisations across all 13 regions and all 332 municipalities, a national school questionnaire, and 8 focus groups reaching Roma children, children with disabilities, unaccompanied minors, LGBTIQ+ children and children in institutions. The plan set 8 target groups, 25 quantitative targets and a 132-indicator M&E annex.
What transfers: coordinating through an existing independent public-law entity; triangulating two analytical studies with a national stakeholder survey and a students’ questionnaire; ensuring representativity for vulnerable children.
The catch: the deep dive was funded and staffed through the ECG testing phase. Before an integrated data platform existed, data stayed fragmented and non-interoperable, which risked leaving some areas out.
Nine barriers as the spine of the whole plan
Comparable coherence with no new external research, pushed down to municipal level.
Portugal structured both its analysis and its measures around 9 access barriers grouped in 5 categories (financial, capacity, administrative, geographical, cultural), rather than commissioning primary research. A four-level governance model runs from ministerial coordination down to Local Child Guarantee Units embedded in existing municipal social-action councils, each mandated to carry out social and barrier diagnosis locally. The plan carries 5 target groups, 76 measures and EUR 340.09 million of ESF+.
What transfers: the barrier systematisation is a low-cost scaffold that forces coherence between diagnosis and measures; embedding local units in pre-existing municipal bodies avoids creating and funding new institutions.
The catch: consultation methods are thinly documented, so participation’s influence on the plan cannot be verified; local replication depends on uneven municipal capacity; and the barrier analysis is only as good as the sectoral data feeding it.
“From the very beginning, a lot of importance was given to the construction of a diagnosis as detailed as possible. In this sense, in the preparation of the Action Plan, a process of consultations and study was carried out, looking for evidence, data and information of different natures, with the aim of identifying all the available information, but also identifying the information gaps to be worked on at the level of the implementation process of the Action Plan.”
What makes it work
- Dedicated evidence generates better plans. Countries that commissioned a purpose-built analysis of child poverty, barriers to services and structural drivers of exclusion produced plans with more coherent logical chains, more grounded targeting choices, and more actionable monitoring frameworks.
- Governance architecture determines the quality of the process. The institutional placement of the National Coordinator, the composition and mandate of the working groups, and the degree of cross-sectoral ownership built from the beginning directly shape who contributes what to the analysis, whose knowledge is incorporated, and whose commitments are included in the plan.
- Leadership anchored at the right institutional level. The coordinator needs both the political authority to convene sectors and the operational continuity to sustain momentum across a multi-year cycle. Greece secured it through statutory mandate, Bulgaria through a two-tier political-plus-technical structure.
- Civil society and children in the analysis, not only the consultation. Direct service providers can corroborate or correct institutional assumptions about who is being missed and why, while children’s and families’ voices increase relevance and anchor a rights-based approach.
- Enough time and support for evidence generation. The deep-dive countries succeeded partly because the testing phase gave them a protected evidence window. Candidate countries should budget the evidence phase explicitly rather than assume it can be compressed into the drafting timeline; external technical support is a significant multiplier of plan quality.
What gets in the way
Named and specific, drawn from the documented plans rather than from principle.
- Consolidation produces a compliant document that fails on targeting. Planning by collecting existing measures under the Child Guarantee umbrella satisfies the reporting obligation but misses the Recommendation’s purpose: reaching specific vulnerable children.
- No sub-national or demographic disaggregation. Most analyses stopped at national level. However well-intentioned the measures, targeting cannot be equitable when the evidence cannot identify which children are most at risk or where they are.
- Financing beyond ESF+ is largely unaddressed. Most plans lean on ESF+ 2021-2027 without naming national financing past it. Only a few, such as Romania, committed state budget for 2028-2030.
- Measures listed without costs. Actions appear with no attached budget or cost estimate, so feasibility cannot be judged and the plan cannot be sequenced or prioritised.
- A plan with no revision cycle. A single document written to run from adoption to 2030 with no scheduled review cannot absorb new evidence. Bulgaria’s biennial operational cycle is the exception across the documented countries.
What transfers, and how
The planning process needs mandate and time more than it needs special infrastructure. Commissioning or synthesising a situation analysis, creating an intersectoral working group, building in structured consultation, ensuring children’s participation, and linking indicators to the evidence are governance choices available in most contexts. Bulgaria, Greece, Lithuania, Czech Republic and Portugal made them from very different starting points.
Adaptation beats replication. The same planning logic produced different models: Lithuania commissioned two studies, Greece paired two analyses with a national survey, Bulgaria added child-led research, Portugal used a barrier analysis. Copy the logic, not the exact methods and instruments.
Coordinator placement is a design decision with long-term consequences. Whether the coordinator sits in an independent body (Greece), an existing line function (Lithuania, Bulgaria) or a dedicated unit (Portugal) shapes neutrality and continuity far more than day-to-day administration. Attaching the role to an existing institution avoids additional costs.
Transferability tags
Sources and depth
- ECG NAP Comparative Synthesis Report (2026)EN