By John Forth, Alex Bryson and Christina Palmou
Quality data that links employers and employees is key for many research questions – yet is insufficiently developed in the UK.
Linked employer-employee data (LEED) are a valuable component of any country’s data infrastructure, because they provide a more comprehensive understanding of the dynamics of the economy and labour market than studying organisations and households in isolation.
The ability to track workers in firms over time allows the development of a range of indicators of labour market dynamics, such as hiring rates, separation rates and job-to-job flows. LEED also allows for a better understanding of the sources of earnings inequality and volatility and can provide valuable insights into the “human side” of productivity.
The limited scope of the existing sources of LEED means that the UK currently has a significant gap in its statistical and research infrastructure. Economic and social policymaking is either based on UK LEED with significant limitations, or it is based on insights developed from LEED in different economies and labour markets, namely the US and Europe.
A strategic opportunity
Our ESCoE Technical Report argues that the UK now has a strategic opportunity to build a new LEED infrastructure around the data from the HM Revenue and Customs (HMRC) Pay-As-You-Earn Real Time Information (PAYE RTI) system.
These data have the potential to overcome many of the limitations of existing sources of LEED in the UK:
The report proposes that the new LEED infrastructure should have the following components and features:
Policy relevance
A new LEED infrastructure with these features could provide timely and authoritative information on a wide range of policy-relevant issues. Our report outlines several illustrative use cases, covering issues of direct relevance to government, such as business dynamism, productivity, wage growth, inactivity and job quality. A new LEED infrastructure would also provide the basis for further economic and social science research that would enhance our understanding of the economy and labour market.
Curating a core infrastructure
In principle, there are already opportunities for accredited researchers to use the PAYE RTI data in the HMRC Datalab and Integrated Data Service (IDS), with permission from relevant data owners. However, our report proposes that the Office for National Statistics (ONS), together with HMRC, ESCoE and Administrative Data Research UK (ADR UK), take the lead in constructing, documenting and maintaining a canonical LEED infrastructure, built around PAYE RTI. This coordinated approach would help prevent the emergence of multiple LEEDs that lack transparency, contain errors in data linkages, are difficult to replicate, or fall short of best practice in the design of such a vital data asset. This is why the report emphasises the importance of data curation.
The development of this LEED infrastructure is technically feasible, given the support of relevant stakeholders, and the potential value is significant.
A roadmap for new LEED
The report sets out a roadmap for the development of a new LEED infrastructure, including three related stages:
To move forward, the report proposes that ESCoE, ONS and HMRC should collaborate over the next twelve months to develop a prototype for a new LEED infrastructure, consisting of a LEED spine linked to a small number of additional datasets. The ONS, as the UK’s National Statistical Institute, is uniquely positioned to lead on this, drawing on its deep expertise in data linkage and its existing work with PAYE RTI data. Its established relationship with HMRC, who own the PAYE RTI data, would also be a key asset. ESCoE would bring academic insight into how best to design and analyse LEED, and the report highlights the value of involving other partners across government and academia who have experience working with PAYE RTI.
However, the ultimate ambition goes beyond a prototype. The aim is to build a sustainable LEED infrastructure that exists for many years to come, similar to the Longitudinal Employer-Household Dynamics (LEHD) infrastructure in the United States or the Linked Employer-Employee Data infrastructure (LIAB) in Germany. That means thinking from the outset about how to turn the prototype into a long-term data asset, owned and updated by the ONS, linkable to other survey and administrative datasets, and ready to address policy challenges, both now and into the future.
The need for collaboration
Making this vision a reality will require strong support from across government. Departments will play an important role in enabling the data linkages needed to build the LEED spine. ADR UK will also be central to this effort as it enters its next phase of funding from April 2026. Since the LEED spine will cover all tax-paying employees in the UK, it could also serve as a powerful sampling frame for new surveys commissioned by government or the ESRC. Additional surveys will be necessary, both for economic measurement and policy purposes, to enrich the administrative data with data items that more comprehensively capture the employer-employee relationship. Policymakers and researchers will also have a vital role to play, championing the case for this new data infrastructure and shaping its development to answer critical policy questions and fill evidence-gaps.
The opportunity is clear, and now is the moment for all parties to come together and build a LEED infrastructure that delivers lasting public value for the UK.
“ADR UK is committed to maintaining and enhancing our existing Flagship linked datasets and opening up research access to new ones, where there is evidenced need from both government and researchers. We know a Linked Employer and Employee Dataset (LEED) is a key data gap for the UK, and are keen to work with data owners and researchers to fill it. We welcome this report as a valuable contribution to the evidence base, which will help us to move forward with this vitally important work.”
Emma Gordon (Director of the ADR UK Strategic Hub and Director of ADR England)
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 the ESCoE, its partner institutions or the Office for National Statistics.