In part 1 of the series, Sophie Gawryla, one of our new Research Impact and Engagement Managers, shares her favourite five posts from the UK Data Service Impact blog – highlighting the power that data can have in giving a voice to those who are often unheard.
In social sciences, we often say that data tells a story. But whose story gets told? Who decides how it’s told – and who gets left out altogether? We may assume that, in a world saturated with information, everyone’s voice is heard. However, it is often the quiet, complex and marginalised that get lost in the noise.
The UK Data Service, through its support for researchers and its stewardship of key datasets, helps change that. By facilitating access to rigorous, secure and ethically governed data, we make it possible to hear what is often muted – whether that’s the challenges faced by low-paid workers, the journeys of children with neurodiversity, or the realities of homelessness.
In this post, I revisit my top five posts from the UK Data Service Impact blog that speak to this idea – each one a vivid example of how data, in the right hands and with the right support, can give voice to those who might otherwise remain unheard.
Together, they follow a thread from governance to policy change – a full cycle of impact.
Permission to listen
Before data can speak, someone must decide who gets to listen – and who gets to ask the question in the first place. This is the space of data governance. Often dismissed as admin, it is the foundation of ethical, inclusive research.
Karen Mansfield, in “Data governance in research: science or ‘admin’?”, reflects on her journey through cognitive training studies to large-scale linkage across health, education and social care datasets. In doing so, she highlights a key insight:
“Between the daily problem-solving and discussions with information compliance, information security, ethical, legal and contracts teams, I often asked myself: is this science or ‘admin’?”
Her answer is clear: governance frameworks – far from being dry bureaucracy – are the necessary, moral scaffolding of good research. We can’t listen to those who do not want to be heard – and for those who do want a voice, we need to look after their data in a secure and safe way.
This idea that the decisions about what data is needed, and from whom, should not be neutral. They reflect value judgements about which communities and issues matter, which resonates with the rest of my blog post selections.
Take the ECHILD project as an example, which links health, education, and social care data for children in England. This kind of research, which allows us to understand how a child’s experiences intersect across systems, is only possible because of the ability to anonymise and protect the people behind the voices. Karen points out:
“The richer the data are in demographic or contextual information […] the higher the risk is that the data subjects could be identified”
It’s a tension between data and voice: the more nuanced and powerful the data, the greater the responsibility to protect it. The success of ECHILD relies on processes that are sophisticated enough to allow for rich, linked insights – without comprising privacy. As Karen concludes, that’s not admin, it’s science.
We see a similar thread in the Living Wage Foundation’s research on in-work poverty. Their research, which uses the Labour Force Survey and Family Resources Survey from the UK Data Service, was a strong precursor to the introduction of the National Living Wage.
This kind of impact story didn’t start at the publication of their research, but with a governance structure that permits sensitive, safeguarded, financial and household data to be used in a secure way.
For example, to access safeguarded and controlled data through the UK Data Service, researchers need to register and apply (where applicable) for access. Fortunately, we offer Safe Researcher Training courses to help users navigate this space.
Ultimately, if we want data to illuminate the lives of those who are underrepresented, we must start by rethinking how access is managed, and whose priorities drive the process – even if this means choosing different questions, datasets and methods of analysis.
When stories collide
How can we make the most of the data once we have accessed it? Matthew Jay tells us about ECHILD in his post “Introducing ECHILD: the role and power of routine data linkage”. This is a powerful example of how linking datasets with administrative data can weave together pieces of a child’s experience into a more complete narrative. As Matthew puts it:
“Initiatives […] have begun to realise the enormous potential that data linkage across different sectors can have. Given the nature of administrative data, these linkages enable new research that would previously have been impossible.”
What does this mean for voice? It means we can start to see children as people whose mental and physical health, schooling and care are intricately interconnected.
ECHILD catches the subtle connections previously unexplored; for example, Matthew discusses a study on gestational age at birth and primary school results. He revealed that children born just a few weeks earlier than full term were statistically less likely to reach expected educational standards by age 10 or 11. Katie Harron notes that without ECHILD:
“[I]t is unlikely we would have had a large enough sample to carry out these analyses […] ECHILD is not subject to attrition in the same way that traditional study designs are.”
That level of detail gives voice to groups of children that previous studies were unable to, whether it be early- or late-term births, children with rarer health conditions, or learners with complex needs. It’s the small, quieter patterns that suddenly speak volumes.
This technique isn’t isolated. As James O’Toole explains in “Unlocking energy data for public interest research”, the Smart Energy Research Lab (SERL) collects “half-hourly electricity and gas smart-meter data […] from 13,000 consenting GB homes”, stretching back to 2019.
But the innovation isn’t just in the access to these energy readings – it’s in stitching those readings together with Energy Performance Certificate (EPC) ratings, weather data, and participant survey details such as household demographics and building characteristics.
That linkage transforms a home’s daily routine into data-driven decision making. What does this insight sound like? Homes deemed less efficient by EPC models are using less energy than expected and advice on how to change your energy use at home. That nuance is only visible when half-hour granularity meets contextual data – and you can only hear it because James and the team linked those datasets.
In an echo to the previous section, this work partly exists because of the ethical and structural governance frameworks provided by the UK Data Service. That trust was important to develop given the scepticism from both the public and media on using smart meters.
The voice given by the data
In “How can we tackle in-work poverty in the UK”, Emma Zimmerman tells the story of the Living Wage Foundation’s (LWF’s) campaign. It was supported by datasets accessed through the UK Data Service and confronts a persistent myth: that work alone is enough to protect against poverty. Joe Richardson, the LWF’s Research Manager, writes:
“Without this access to data, our labour market research would be significantly more difficult.”
But this isn’t just about income brackets and payslips – it’s about how the data has given a voice to those who worry if they’ll have enough hours next work, or who earns less than the cost of living in the city they serve. By showing how these patterns disproportionately affect disabled people, ethnic minorities and certain regions, the data takes on an advocacy role.
And it works! Through campaigns like Living Hours, the Living Wage Foundation has driven policy, local government engagement and resulted in real wage increases – at the time of writing, Emma records the movement, and subsequent National Living Wage introduced in 2016, has put “over £1.6bn into the pockets of low paid workers, and secured pay rises for almost 300,000” more.
The data, and the research using it, has amplified the lived experiences of those who would otherwise be left out of the economic decisions that directly affect them.
We are proud to highlight these narratives here at the UK Data Service; indeed, one of our own Data Impact Fellows, James Cockett, reflects on his experience sharing evidence with the Low Pay Commission and we have documented the work of the Institute for Employment Studies on this topic in a case study.
Housing and homelessness data – taken from sources like the English Housing Survey, UK Household Longitudinal Survey and British Cohort Study – offer similarly powerful insights.
Professors Glen Bramley and Suzanne Fitspatrick discuss the development of the Homelessness Monitor in “How have the data helped us understand and improve the UK housing and homelessness situation?”. They used linked, longitudinal data to press the government on problematic statistics. Their findings compelled the UK Statistics Authority to investigate, leading to official changes and recommendations to how homelessness is measured.
This monitor turned marginalised life experiences – crowded houses, hidden homelessness and unsuitable temporary accommodation – into a voice given through the data. As Glen and Suzanne report:
“The findings of this work have been endorsed by a wide range of housing organisations and seek to reinforce a policy shift […] there is compelling evidence that England needs at least 90,000 net additional social rent homes a year.”
These narratives show that data – with context – becomes a voice for those who may not otherwise be heard. It’s not just information, or part of the noise, it’s advocacy for those living quietly in the data – whether that is a result of lack of representation in surveys, in-work poverty, the cost-of-living crisis, or homelessness. You can read more about this work in our case study.
However, data is only one part of the puzzle. The voice that data facilitates needs to be combined with real, lived experience so that data-driven decisions – that result in impactful policy changes – really make a difference.
Eve Little dives into the impact of lived experience in part two of this series on our top ten blog posts.