Faye Gracey is the Head of Data Acquisition and Linking for Research for the ADR Wales team in Welsh Government. In a new blog, she describes her experience presenting at the European Survey Research Association 2025 Conference hosted by Utrecht University in Utrecht, The Netherlands.
Last week, I had the privilege of opening a lively debate. I proposed that “Administrative data research (ADR) will replace the vast majority of surveys—if not all—in time”—also the title of my presentation. A bold claim, admittedly, but grounded in the evolving reality: in many cases, administrative data already delivers research insights faster, more reliably, more ethically, and more cost-effectively than traditional surveys.
Standing at the front of Megaron Theatre, I began to wonder if my words were too bold. A colleague had jokingly warned me I might be trying to “sell a plant-based diet to farmers”—a playful dig at the challenge of promoting administrative data research to survey purists. Fortunately, the reaction wasn’t quite so dramatic.
The debate was designed to provoke, and it did. Comments and questions from the audience were thoughtful and reflective of genuine concerns. It was clear that while the UK has made significant strides in opening up administrative data for secure, de-identified research, some countries are still catching up. Some attendees were unfamiliar with Trusted Research Environments (TREs), or with examples of administrative data research driving real-world impact.
Yet what struck me most was how many researchers at a conference focused on surveys are already doing work that mirrors what we do as part of ADR UK. (I would be less surprised at an International Population Data Linkage Network event to find so many). Lisa Ziemba from Statistics Austria presented developments and plans for their Administrative Data Collection for the Social Sciences (ADCOL), which sounded strikingly like our work in Wales. Throughout the five-day conference, administrative data was a recurring theme—how it is used to shape survey sample frames, validate findings, and increasingly, to link directly to survey responses.
A smaller, but growing number of presentations also explored TREs in detail. I was particularly interested to learn more about the EOSC-ENTRUST project, which aims to create a European network of TREs and to drive interoperability, which seems to be following a path like that of our partner SAIL Databank working with DARE UK.
I’m more convinced than ever that the use of linked administrative data will expand in the coming years. That said, some surveys will still have a vital place in capturing nuance and direct perspectives that can be hard to replicate. When surveys are used, let’s help managers to routinely consider the feasibility of ingestion into a TRE to open up further research opportunities.