From Data Stewardship to Policy Insight: Data Collaboratives for Synthetic Microdata and Policy Microsimulation.
Conference
Format: CPS Abstract - IAOS 2026
Keywords: datastewardship, synthetic data
Session: Official statistics for policy making
Thursday 14 May 9 a.m. - 10:30 a.m. (Europe/Vilnius)
Abstract
National Statistical Offices (NSOs) are increasingly expected to underpin evidence-informed policy making with integrated, timely insights, while safeguarding trust, privacy, and methodological integrity. However, delivering on this promise now depends on advanced technical capabilities, including privacy-preserving approaches such as synthetic data, that many NSOs, especially smaller ones, cannot realistically develop and sustain on their own.
At Statbel, we address this challenge through a stewardship model centered on data integration, structured as a Data Integration Value Chain. The model acts as an operating logic to move beyond one-shot microdata projects and turn recurring needs into domain-focused data collaboratives that progressively stabilize and scale integrated data assets.
Our academic partner, the Center for Applied Public Economics (CAPE, UCLouvain), has developed BEAMM, a synthetic-data platform that uses Statbel’ administrative and survey microdata as its backbone (https://beamm.brussels/about/). Using advanced statistical matching and AI-supported synthetic data generation, BEAMM produces privacy-preserving, analysis-ready microdata and supports nowcasting/forecasting and tax-benefit microsimulation for policy evaluation.
In our paper and presentation, we first introduce Statbel’s collaboration-centred stewardship approach and its operational translation through the Data Integration Value Chain. We then present CAPE’s BEAMM platform as the technical counterpart, explaining how it implements synthetic microdata generation, nowcasting/forecasting, and tax-benefit microsimulation.
Together, we showcase an end-to-end setup that links an NSO’s stewardship and governance strengths to synthetic-data-enabled analytical infrastructure, turning integrated data into decision-ready evidence for policy design and evaluation. We argue that this partnership model offers a credible way forward for NSOs to strengthen their position in a changing data landscape, particularly in support of evidence-informed policy making, by enabling advanced microsimulation for policy evaluation.