Robust and integrated skills data are essential for effective policymaking, particularly in improving the allocation of public resources across education and training. Yet many governments continue to face significant information gaps that constrain their ability to assess programme effectiveness, prioritise spending and target investments towards interventions that yield the strongest outcomes.
This paper examines how stronger skills data can support more evidence‑based and efficient financing decisions in education and training systems. It begins by identifying the main data gaps that currently hinder strategic decision making, including fragmentation of data systems across ministries and levels of government, limited longitudinal linkages between education and labour market records, uneven coverage across the skills system – particularly in early childhood and adult learning – and persistent issues with timeliness, accessibility, definitional consistency and data quality. These gaps leave policymakers without a consolidated view of how public resources translate into results, limiting their capacity to respond to evolving labour market needs.
The paper then reviews ten data sources relevant for education, employment and skills policy, outlining for each their practical applications, strengths and limitations, and the types of conclusions that can legitimately be drawn. This helps decision makers select appropriate evidence for different policy questions and avoid common misinterpretation.
The paper also examines what is at stake. Better data can enable sharper financial decision making – supporting cost-benefit analysis, revealing fiscal inefficiencies and improving the targeting of resources towards high-return programmes – while reinforcing the already well-established economic and social returns to skills investment. At the same time, governments are not the sole actors in skills development: employers, private providers and individuals invest significantly in training, and better public data should complement rather than substitute for these contributions.
To understand what prevents progress, the paper introduces a framework grouping the main obstacles into four categories: institutional, governance and financing barriers (such as fragmented mandates and unstable funding for data infrastructure); human-capital and analytical-capacity gaps (shortages of specialised data talent and limited data literacy among decision makers); legal and regulatory constraints (particularly around data protection and cross-agency sharing); and technical and interoperability challenges (legacy systems, inconsistent standards and cybersecurity risks). Examples from across the OECD illustrate how governments are addressing these barriers through integrated data systems, centralised analytical units and legal reforms that enable secure data linkage.
The paper concludes that improving data systems is not primarily about collecting more information. In many countries, substantial relevant data already exist but remain underutilised because they are fragmented, stored in incompatible systems or inaccessible to decision makers. The priority is to unlock and integrate existing data assets through better co‑ordination, stronger analytical capacity and governance arrangements that encourage evidence use. While the costs of modernising data infrastructure are real, the potential gains – in more effective resource allocation, improved policy outcomes and stronger public trust – are considerable. Digital technologies, including artificial intelligence, can act as multipliers of these returns, but only when supported by strong data governance and high-quality underlying data.