ESCoE conference 2026 – Measuring an economy in transition

By Rebecca Riley

Following ESCoE’s May 2026 conference at King’s Business School, ESCoE Director Rebecca Riley reflects on the event and economic measurement in a changing economy.

ESCoE’s 2026 conference brought together policymakers, academics and statisticians to confront a shared reality: economic measurement is being reshaped by rapid technological change, evolving data sources and shifting economic structures.

What struck me across the three days was not simply the scale of this change, but the clear consensus around what must come next. Ensuring economic statistics remain relevant, trusted and useful will require continued collaboration, innovation and long-term investment.

Building trust in economic statistics 

The conference opened with a welcome address from James Benford (Director-General for Surveys and Economic and Social Statistics, UK Office for National Statistics). He highlighted the importance of the constructive dialogue and rigorous debate that we aim to foster at this conference – across different parts of the statistical system and in wider public discourse. This matters not only for improving the quality and relevance of economic statistics, but also for building and maintaining trust in them.

I think many of us found it very interesting to hear from André Loranger (Chief Statistician of Canada) about the common challenges faced by statistical agencies worldwide and how Statistics Canada is dealing with the “insatiable” demand for bigger, better, faster and cheaper data. Meeting this demand, while maintaining quality and trust, is a shared challenge across countries. Find out more in André’s keynote interview.

These issues were revisited throughout the programme and brought together in the final panel session that reflected on the progress made in the recovery of the UK’s official economic statistics over the last year, chaired by Darren Morgan (Engagement Director, ESCoE). Panellists James Benford, Rob Kent-Smith (Deputy Head, Office for Statistics Regulation), Alison Kilburn (Director of Analysis and Chief Economist, Department for Science, Innovation and Technology) and Jill Leyland emphasised the importance of economic statistics for policymaking, and their crucial role in shaping decisions on budgets, priorities and evaluating impact. There was agreement that maintaining trust requires statistical producers to be transparent, including being open about the uncertainty associated with the estimates that economic statistics ultimately are, as well as continued engagement with all users.

AI and economic measurement

Not long ago a colleague of mine suggested that I checked some of their algebra as they had been working with Claude. I asked “Who’s Claude?”, wondering what might be the matter with them. So, it felt very timely and important to listen to advice from Michael McMahon (University of Oxford) in his keynote on how we can use AI to enhance economic data and research. Find out more in Michael’s keynote interview.

Thursday’s keynote from Carol Corrado (Georgetown University) provided very careful and inspiring thinking around how National Accounts might evolve to capture the many changes that AI might bring. She argued that current statistical frameworks struggle to make the economic impact of AI visible, to fully capture rapid improvements in quality, and to account for low or zero-priced digital services. Find out more in Carol’s keynote interview.

We also heard from a panel of statistical agencies represented by Jennifer Withington (Statistics Canada), Chris Sibley (Central Statistics Office Ireland) and Grant Fitzner (Office for National Statistics), and chaired by Diane Coyle (ESCoE and University of Cambridge), who discussed what developments in AI might mean for economic statistics production and the distinct constraints faced by statistical agencies.

New data and methods for economic measurement

Beyond AI, sessions explored new data sources and methods that are transforming economic measurement. They showed how measurement is moving beyond traditional surveys towards a hybrid model, combining survey, administrative and alternative data sources.

In a panel on nowcasting – the prediction of the present, the very near future, and the recent past – chaired by Stuart McIntyre (ESCoE and the University of Strathclyde), Raffaella Giacomini (UCL), Michele Lenza (European Central Bank) and Giulia Mantoan (Bank of England) explored how real-time data can improve the timeliness of economic measurement. Read more about ESCoE’s work on nowcasting.

This theme continued in special sessions on firm‑to‑firm transactions data and industry and regional classifications for sectoral and local economic analysis. Here, researchers demonstrated how detailed transactions data can track activity in near real-time, map supply chains and production networks, and identify key dependencies and vulnerabilities. Reflecting on work on firm-to-firm transactions data, Dan Mawson (Department for Business and Trade) emphasised how progress in this area has involved many people and organisations, spanning more than a decade. Those investments are now enabling us to map the nature of supply chains at a time when this capability is particularly important, demonstrating the value of long-term investment and successful, sustained collaboration.

Firms, productivity and business dynamism

A special session on business dynamism chaired by Jakob Schneebacher (ESCoE and Institute for Fiscal Studies) explored how firm behaviour is changing over time and what that means for aggregate economic developments. Evidence presented by Javier Miranda (Halle Institute for Economic Research, CompNet and ESCoE), Sophie Piton (Bank of England) and Sabien Dobbelaere (Vrije Universiteit Amsterdam and Tinbergen Institute) suggested that while headline rates of firm creation remain relatively stable, deeper analysis shows that many new firms are small, short‑lived and contribute little to productivity or employment growth. Read about ESCoE work on business dynamism and market structures. 

Labour markets, households and distribution

Another strand of the programme focused on labour markets and households, including a special session on measuring labour market activity using Linked Employer-Employee Data (LEED), chaired by Grant Fitzner (Chief Economist, Office for National Statistics). 

This highlighted yet another example of what successful collaboration can look like, showcasing the development of new linked employer employee data by my ESCoE colleagues John Forth (Bayes Business School) and Alex Bryson (UCL) together with Tom Wickersham’s Economic Microdata Transformation team at the Office for National Statistics. Their work provides a tool for cross checking labour market surveys and for creating new and important information on business dynamism and labour market flows. 

What’s next for economic measurement?

This year’s discussions highlighted both the scale of the challenges ahead and the opportunities created by new data and technologies. Across the conference, a shared direction of travel emerged. Improving economic measurement will require integrating new data sources with established frameworks; investing in skills, infrastructure and collaboration; developing new approaches to capturing AI and intangible activity; and maintaining transparency and trust in an increasingly complex system.

Thank you

A record number of attendees reinforced that the ESCoE Conference continues to provide a vital forum for collaboration and debate across the economic measurement community. 

Thank you to King’s Business School, Kings College London and the Office for National Statistics for their support for this conference, and to all speakers, organisers and participants for another insightful event.

Catch up online

Slides and recordings from selected sessions are available on the ESCoE website. 

What’s next?

The next ESCoE Conference will take place in Bristol in Spring 2027. It will be our 10th anniversary, and we look forward to seeing many of you there. If you can’t wait until then, please join us in Sydney, Australia for our conference with the University of New South Wales on 7-8 December 2026. The call for papers will open later in June 2026.

ESCoE blogs are published to further debate.  Any views expressed are solely those of the author(s) and so cannot be taken to represent those of ESCoE, its partner institutions or the Office for National Statistics.