Why Official Statistics producers are vital to administrative data research
As a new report from the Office for Statistics Regulation (OSR) examines how Official Statistics producers can unlock the public value of their data through onward sharing, Dr Emma Gordon, Director of ADR UK, explains why these producers are so important to administrative data research.
Today (8 October 2020), the OSR published their report on ‘Unlocking the value of data through onward sharing’. As Director of a partnership that exists to do just that, I wholeheartedly welcome this new guidance. The report makes clear that the principles in the Code of Practice for Statistics – which ensures they are high quality and have public value – extend beyond statistics production to data sharing and access.
From our perspective, statistics producers are in an enviable position: if there is data of sufficient quality to support decision making, they will generally have access to it. Statistics producers will also have spent the time needed to understand the data’s quality issues, and how it should be curated to support research and analysis.
This means statisticians are exactly the people who should feel empowered to help facilitate administrative and survey data being made accessible to external researchers, through appropriate routes. As the report articulates, this includes the full spectrum from publication as open data, through to using secure research facilities such as those offered by ADR UK and other ESRC investments such as the UK Data Service.
Of the data standards elements presented, the two that are core to the vision of ADR UK are that data should be linkable and curated.
By linking administrative data sources, it is possible to reach across traditional departmental boundaries to more fully understand the impact of policies on society. As is becoming clear in the management of the Covid-19 pandemic, it is not enough to have good data about the health of the population if we don’t also understand other elements of people’s lives such as their caring responsibilities, job security, income, living conditions and ethnicity. All these factors interact to determine how different sectors of society will be affected, in terms of both health and other elements of wellbeing. It is only by linking this data and making it available to researchers, following the principles of the Five Safes, that we can properly understand the impact of the pandemic on society.
For data owning organisations to engage in the creation of new datasets for research, they need to know that the effort they put in is justifiable. This is why curated data is so important. Anyone who has been involved in setting up a new data sharing agreement knows that this is an understandably detailed and lengthy process. Knowing there is a commitment to continued curation of the data means the research value can be maximised, and the initial resource needed to make it accessible can reap rewards for years to come.
Moving forwards, we are aiming for datasets created as part of the ADR UK investment to be trackable. Public money is used to fund our programme, and we need to be able to show the public, data owning organisations and government the research their investment is facilitating. Building our published case study collection will not only make it easier for decision makers to find policy-relevant research, but will also help reassure data owning organisations that time invested in working with us is well spent.
At ESRC and across the ADR UK partnership, we are also excited about the potential for synthetic datasets to improve researchers’ ability to use data. These would help researchers develop their proposals, and could also play a vital role in training the next generation of researchers to use administrative datasets effectively. Like a flight simulator, they could enable rigorous and realistic training to be delivered, without requiring direct access to sensitive linked datasets which is rightly very tightly controlled.
ADR UK brings external researchers closer to policymakers, to support evidence-based policymaking. Statisticians and others involved in the production of Official Statistics are key to us forming this bridge, which is why the three Chief Statisticians from the devolved administrations, as well as a representative from the Office for National Statistics, all sit on our Leadership Committee. We look forward to supporting the OSR and statistics producers across government to deliver on the ambitions of this report.
This blog has also been published on the Office for Statistics Regulation website.