A monitoring system's function is to tell whether measures are reaching children, and to allow a change of course when they are not. The countries that built systems that do this sequenced the build: governance first, then a results chain, then a phased indicator set, and treated every reporting cycle as a chance to learn. Romania created a sequenced roadmap and a dashboard concept that tracks key indicators and progressively links them to financial monitoring; Luxembourg made its data constraints explicit and put its statistical office in charge of the data; Ireland linked a national outcomes framework to practical measurement and learning tools. The recurring trap mostly goes unnoticed: an indicator set chosen for what is easy to collect rather than what is meaningful to track.
Build the system that learns
Three applied activities from the same evidence base: map the twelve components of a working M&E system, sequence your own build against the country that started where you are, and lead five decisions over 36 months without taking the plausible shortcut.
The problem
A biennial progress report goes to the European Commission on time, full of indicators. Asked in a committee whether any specific group of children is measurably better off than two years ago, the honest answer is that no one can say. The indicators were chosen because the data happened to exist, not because they track what the measures were meant to improve. The numbers that would answer the question sit in health, education and social-protection systems that do not communicate with one another. The report is complete. It demonstrates little to the country.
Three structural reasons keep recurring.
First, indicators are driven by what can be collected. When indicator sets are built from what is easy to collect rather than from a theory of how measures produce outcomes, the system measures activity, not results. The logic of intervention has to be reconstructed later, at much greater cost.
Second, data are fragmented and non-interoperable. Sectoral administrative data, with no interoperability across ministries and levels of government, break the monitoring chain: even good indicators cannot be compiled or compared if the underlying data never connect.
Third, M&E is built as a reporting afterthought, separate from planning and from decisions. When monitoring is designed in isolation from the plan and disconnected from how decisions are made, evidence rarely leads back into revising measures.
The rest of this brief addresses one question: what does a monitoring system that actually changes decisions require, and in what sequence?
“Countries that make progress do so not because they have perfect systems from the outset, but because they establish sound governance foundations, invest in a deep understanding of the situation of children, develop their indicator frameworks iteratively, and treat each reporting cycle as an opportunity to learn and improve.”
What the working systems share
Across the documented countries, the most functional systems share one characteristic: two governance tiers and a clear data owner. A political-strategic level, a high-level inter-institutional group that holds cross-ministerial accountability; and a technical-operational level, a permanent expert working group including statistical, research and civil-society actors that runs the day-to-day monitoring, data collection and reporting. Bulgaria’s two-level model is the reference: a high-level group provides overall coordination and monitoring, while a permanent expert group develops two-year operational plans and progress reports. The other critical component is naming a data owner: Luxembourg formally mandated its statistical office, STATEC, as the central data-governance institution for the whole monitoring framework, putting standardisation and quality assurance with the institution best equipped to hold them. None of this requires new legislation; it can be established through mandates, terms of reference and clear reporting cycles.
Diagnose first, standardise everything, then design the dashboard
A roadmap built from a mapping of what already existed, and fiches for over 200 indicators.
Romania began with a diagnostic mapping of its M&E components, data flows, overlaps and gaps against the ECG Monitoring Framework, then built a roadmap. A first initiative produced standardised fiches for 146 indicators; a second, TSI-funded, extended the scope to over 200 indicators assessed against a nine-criteria matrix (data routine, availability, periodicity, definition clarity, formula clarity, biennial-reporting compliance, disaggregation, territorial coverage, interoperability). It mapped data flows across child protection, health and education, and proposed interim data-sharing protocols, feeding a curated shortlist into a dashboard concept.
What transfers: where data systems are fragmented, a country can start from ECG-aligned indicators and progressively integrate national indicators, disaggregation needs and financial linkages, supporting decision-making, accountability and the EU reporting cycles.
The catch: Romania’s documentation openly acknowledges gaps in state-budgeted M&E capacity. The process was possible with TSI and UNICEF support, and durability depends on internalising it into the national budget.
Logic before data, and the statistical office in charge
Theories of change first, then an indicator set that is transparent about its own gaps.
Luxembourg built theories of change for each domain before selecting indicators, then classified the whole set into three groups: indicators available with the minimum required breakdowns, indicators available but lacking the required breakdowns, and placeholders that are not yet available but are considered critical. It mandated its statistical office, STATEC, to ensure data management, and paired a high-level Steering Committee with Technical Working Groups spanning ministries, local authorities, observatories, research institutes and civil society.
What transfers: indicator design from a theory of change; a classification that keeps gaps transparent and stakeholders engaged while they are worked on; data governance assigned to the national statistical authority.
The catch: it relies on a capable, well-resourced statistical office and structured technical facilitation. Luxembourg’s own lessons point at what to do earlier: clearer technical profiles for the indicator working groups, structured input from children on priority domains, and interim dissemination products that demonstrate value between milestones.
Permanent units and an evidence habit built over a decade
Not a project: standing research infrastructure funded from the national budget.
Ireland’s Department of Children (DCEDIY) sustains two permanent internal units: a Research and Evaluation Unit for standards, guidance and training, and a Data and Analytics Unit for integration, analysis and reporting. The research infrastructure supports them: the Growing Up in Ireland longitudinal study, and the What Works Evidence Hub curating programmes with evidence and cost ratings, provide the monitoring and evidence foundations for implementing and tracking the NAP. Capacity is funded from domestic sources alongside EU support, and the outcomes framework dates back to the mid-2010s.
What transfers: survey-based and longitudinal evidence assets that complement administrative data and support policy learning over time; an evidence hub for continuous learning; dashboard-style reporting that improves the accessibility and comparability of indicators.
The catch: this is a long-term trajectory backed by sustained domestic funding, dedicated in-house analytical capacity, and prior institutional maturity.
What makes it work
- A robust, inclusive governance architecture. A high-level political coordination body plus a permanent technical working group responsible for M&E operations, data governance and capacity-building, with all ECG domains represented in both tiers.
- A formal, active role for the statistical authority. Placing definitions, quality assurance and interoperability with the national statistical office gives standardisation a durability that a ministry unit alone rarely sustains.
- Capacity built in two directions at once. Systems need both the ability to produce rigorous, disaggregated data and the ability to use it: revising plans, designing new interventions, and reallocating resources toward what works.
- Plan and monitor together, from the start. M&E cannot be designed in isolation from planning: each measure should trace back to a diagnostic finding and forward to an indicator that captures the change it is meant to produce.
- Start from the ECG Monitoring Framework and what is already collected with rigour. Develop the indicator architecture with top-down logic, derived from the theory of change and the results chain, combined with bottom-up knowledge of what is already routinely collected at national and local level.
What gets in the way
Named and specific, drawn from the documented systems rather than from principle.
- Indicators chosen for availability, not meaning. When the set is built from what is easy to collect rather than from a theory of change, the system measures activity, and the intervention logic must be reconstructed later.
- A weak indicator framework undermines attribution. If indicators are inconsistent, monitoring data cannot be compared across institutions, evaluation cannot attribute findings to measures, and reporting to the Commission cannot rest on evidence of progress.
- Relying on projects and external funds. Most documented systems still depend on time-limited external support (TSI, UNICEF). Without internalised, state-budgeted units, the system cannot become sustainable.
- Monitoring built apart from planning and from decisions. Where M&E is a separate reporting track, evidence rarely reaches the people revising the plan.
What transfers, and how
Two-tier governance needs mandate, not legislation. A political-strategic layer and a permanent technical-operational layer can be established through ministerial orders, terms of reference and a fixed reporting cycle: Bulgaria, Luxembourg and Spain built theirs from very different starting points. A constant across Member States is that statistical institutes and research bodies anchor the technical layer, most explicitly in Luxembourg’s formal mandate for STATEC as the data-governance hub of the entire monitoring framework.
The three-group indicator classification allows for improvement. Organising indicators by immediate availability, partial availability and placeholder makes gaps transparent and delivers early results while methodological work continues. Luxembourg’s practice indicates it is highly replicable: a foundation for iterative improvement that keeps stakeholders engaged.
Standardised fiches reduce ambiguity across institutions. Documenting every indicator to a common template forces methodological consensus exactly where definitions diverge between institutions.
Evaluability assessment comes before evaluation. Checking that a measure has a clear results chain, measurable indicators and available data before commissioning an evaluation, as Croatia and Greece sequenced explicitly, is what makes evaluation findings usable.
Transferability tags
Sources and depth
- ECG M&E Systems Synthesis (2026)EN