Effective policy design, implementation and evaluation require a clear understanding of the different types of data that inform decisions across education, employment and skills systems. Multiple sources of data exist, each offering distinct strengths and responding to different analytical needs. Table 1 summarises these sources, their main uses, and their key strengths and limitations. Figure 2 complements this overview by presenting a heat map that visually compares the relative strengths of each data source across key analytical dimensions, including coverage, timeliness, granularity and capacity for causal inference.
A challenge that can arise in policymaking is the expectation that available datasets can answer a wider range of policy questions than their design allows. In practice, the ability to draw valid conclusions depends on both the analytical capacity of decision makers and the structure and quality of the underlying data. Some sources are appropriate for describing trends, diagnosing disparities and informing strategic planning, while others are required for assessing long-term trajectories or determining whether a policy or programme caused an observed change. Misinterpretation of data – such as drawing causal conclusions from descriptive indicators or extrapolating results beyond the population studied – can lead to weak or misleading policy choices.
For this reason, policymakers need a clear understanding of the kinds of questions that different data sources can answer reliably, and those for which they are not suitable. This includes knowing which can support system-level monitoring and allocation decisions, which can credibly inform programme evaluation, and which require complementary evidence to avoid incorrect inference. Recognising the analytical boundaries of each dataset is as important as understanding its potential.
Ten data sources are relevant for education and skills policy, organised into five broad categories that reflect distinct analytical logics and policy functions: population and household surveys, administrative and assessment data, programme‑focussed data, evaluative and qualitative methods, and emerging data sources. While these categories are not mutually exclusive, recognising the distinct logic of each helps policymakers identify which sources are fit for purpose for a given analytical need, and which combinations of evidence are required to support robust decisions. Table 1 summarises these ten sources within their respective categories, outlining their main uses for policymaking and their key strengths and limitations. Figure 2 complements this overview with a heat map comparing the relative strengths of each source across key analytical dimensions, including coverage, timeliness, granularity and capacity for causal inference. Annex A provides a detailed profile of each data source, including illustrative examples, and the types of conclusions that can legitimately be drawn from each.
