Unlocking new frontiers in linked data with ODISSEI

This blog offers a preview of the keynote by Tom Emery at the ADR UK Conference 2025. Drawing on over a decade of experience with ODISSEI—the Dutch national infrastructure for social science—Tom will reflect on the promises and challenges of using administrative data to answer complex social questions. He’ll share lessons from building ODISSEI and explore the possibilities that lie ahead for linked data and public policy. 

For the last 10 years, I have been working in the Netherlands, helping to build the Dutch national infrastructure for social science, known as ODISSEI (Open Data Infrastructure for Social Science and Economic Innovations). When I first started working with administrative data, I was struck by two things: its potential for uncovering insights about society, and the myriad hurdles—access, linkage, and quality—that make research with administrative data tricky. 

The obstacles in using administrative data often obscure its potential. Given that many of these challenges still exist, nobody truly knows how useful this data might be in understanding and improving society. Are we on the cusp of personalised policy and super-intelligent policy design? Or are we heading toward an endless proliferation of ineffective and unjust algorithms? We don’t have the answer yet. 

In this keynote, I want to share some of the many things I’ve learned about the challenges of building administrative data infrastructure, the questions I still have about the potential of administrative data, and where we might end up in the coming decades. 

Supporting administrative data analysis with ODISSEI 

At ODISSEI, we have been working to remove obstacles and constraints on research so that we can gradually answer these big questions. Historically, accessing administrative data—such as tax records, employment histories, or health encounters—meant navigating a labyrinth of approvals, privacy protocols, and technical silos. ODISSEI has tackled these barriers head-on by: 

• Supporting access through a single data portal, access grants, subsidies, and community collaboration initiatives 

• Promoting best practices through shared training materials, workshops, and a network of secure computing environments 

• Fostering trust between government agencies and academic researchers by co-designing development frameworks. 

The result? What used to be an intimidating and cumbersome process, often taking months of back-and-forth, can now be initiated within weeks—without compromising privacy or security. 

But access is only the first step. To address complex social questions—such as how job changes affect health outcomes, or how neighbourhood characteristics shape educational attainment—you need to link data across sources and add meaningful context. ODISSEI facilitates this enrichment by: 

• Supporting record linkage services that connect disparate datasets—including surveys, digital trace data, and genotypes—with administrative records while preserving anonymity 

• Providing harmonised metadata so researchers know exactly what each variable means, how it was collected, and how it compares across sources 

• Offering value-added tools such as scientific use files with clear provenance to their administrative base, supporting complex, high-quality downstream research. 

This enriched infrastructure has opened the door to new kinds of inquiry—ones that simply weren’t possible when data lived in isolated silos. 

From access to insights 

To see ODISSEI’s impact in action, consider two recent projects that showcase its potential. 

The first is the PreFer Data Challenge, launched by Gert Stulp at the University of Groningen. The challenge invited researchers to predict who would have a child within the next three years using survey and administrative data. Thanks to ODISSEI’s secure linkage infrastructure and harmonised guidelines, teams from around the world analysed data on millions of individuals while maintaining confidentiality. Their innovative modelling approaches not only set new predictive benchmarks but also highlighted potential policy levers to support greater agency in realising childbearing intentions. 

The second example is PopNet, an initiative building “family trees” and social networks at the population level by connecting family members, colleagues, neighbours, and classmates across decades. Through ODISSEI’s metadata services and secure computing layers, PopNet researchers have traced how family ties and neighbourhood networks evolve—and how these dynamics influence outcomes such as educational attainment and health trajectories. This work is shedding light on social processes that span generations and administrative boundaries, offering fresh perspectives on inequality and social mobility. 

Outstanding questions 

Despite these successes, there is still much we don’t know about the full potential of administrative data. That’s why, at this year’s conference, my keynote will explore some of the unanswered questions and emerging frontiers, including: 

• How can we conduct cross-national administrative data analysis? 

• How can we responsibly integrate new data types (such as digital footprints or sensor data) with traditional registers? 

• How much does additional computing power help in modelling complex social processes? 

• How do we ensure that the insights we generate translate into fair and effective policies? 

I’m excited to share stories from the development of ODISSEI and to hear perspectives on how we can harness administrative data to enhance lives. If you’re curious about the challenges ahead and the opportunities waiting to be unlocked, I hope you’ll join us in Cardiff this September. 

Explore the full ADR UK Conference 2025 programme 

To plan your visit and discover the full range of keynotes, workshops, and panels: 

View the full conference programme