At the 2023 Spring Budget, the then-chancellor, Jeremy Hunt, announced a series of reforms to the childcare system to enable more parents to work and, in turn, boost growth. This included a commitment to 30 hours of funded childcare for most working parents, a policy Labour has retained since coming to power last year. Realising there was a big gap in evidence on why and where parents were struggling – or managing – to access services, Ofsted’s Early Years Data and Analysis team launched a project to quantify neighbourhood-level access to childcare in England.
Two-and-a-half years on, the resulting project, Understanding barriers for working parents: where are childcare deserts and oases?, is feeding into No.10 delivery dashboards and informing the Department for Education’s evaluation of its £14bn investment in expanded childcare – and has received plenty of plaudits.
Last December, the team won the Evaluation and Analysis Award at the 2024 Civil Service Awards. It followed this up with two prizes at the Analysis in Government awards in January, before winning the Royal Statistical Society’s Campion Award for Excellence in Official Statistics (awarded in partnership with the UK Statistics Authority and CSW) in July. Civil Service Awards judges said the team “transformed disparate cross-government and commercial data into novel, real-world insights on the everyday experiences of parents seeking nurseries/childminders”. The RSS’s judges praised it as a “strong example of statisticians working to fill a significant gap in our data” and said it has delivered “essential insights for both policymakers and the public”. And the government analysis function said the project enabled targeted interventions; supported the government’s growth and opportunity missions; and demonstrated cost-effective innovation by leveraging existing data and tools.
CSW spoke to the team’s leader, Anita Patel, to find out how they achieved such acclaim.
From anecdote to hyperlocal evidence
While the team knew anecdotally that local access to childcare was one of the biggest barriers to parents returning to work, there was very little evidence to substantiate and quantify this.
Existing models “made a lot of assumptions about local authority or aggregate data that do not reflect what happens in reality, where parents are crossing multiple areas because of their work – or whatever their situation is – to access childcare,” Patel says. In London, for example, some working parents might make a 15-minute tube journey to access childcare, spanning multiple boroughs. Traditional methods could not measure this or look at how childcare access varies within a local authority.
So the team decided to “dig deep” into hyperlocal statistics, looking at 180,000 neighbourhoods in England, and used a two-step floating catchment model utilised by the University of South Wales “to account for the location of childcare providers; different transport modes, such as driving or public transport; and the number of children aged zero to seven in those areas accessing the childcare”.
Patel says this meant working with “extremely large, complex datasets at a really granular level”. Helpfully, the team managed to get funding from the Cabinet Office and Treasury in January 2024 to help accelerate the work and put the ideas into practice.
Getting help
Another big challenge was how to communicate these granular findings in a visual and interactive format. “We were a little bit stuck when it came to using our GOV.UK publishing platform,” Patel recalls. The team liaised with existing contacts in the analysis hub at the Office for National Statistics to see if there was an opportunity to collaborate. The ONS helped in two ways: it added in some of its own socioeconomic data to enhance the insights; and its data journalists helped the Ofsted team to create “semi-automated robo-journalism” to showcase the interactive, neighbourhood-level statistics. To operate the tool, the user simply types in a postcode and instantly gets an automated story on that neighbourhood.
This kind of collaboration was “really central” to making the project a success, Patel says.
Within government, the team worked with policy colleagues at Ofsted, DfE officials, and staff within the No.10 data science team – as well as the aforementioned support from the Cabinet Office and HMT.
They also worked closely with international experts from academia. Initially, the team got in touch with academics in Melbourne who had done similar work, to understand how it could be deployed in England. The Australian team had come up with the idea of childcare deserts and oases, and Patel says the discussions were really useful to understand some of the challenges she and the team might face.
“Obviously, geographically, Australia is very different to England,” she says. “But there are similar challenges in terms of urban and rural access to childcare, so there were a lot of parallels and things to learn from them about how to build the model.”
Then, looking a bit closer to home, the team established a partnership with computational geographers at the University of South Wales who had established the “multi-model” two-step floating catchment method, which looks at how far a childcare provider or service is from the centre of a given neighbourhood – which comprised 40 to 250 households – and then at the children who are competing to access those places. This work was based on the Welsh model of 15 hours’ free childcare and it was, therefore, relatively straightforward to adapt the methodology to Ofsted’s registered places in England. The academics in Melbourne also helped the team to understand how similar metrics had been developed across Australian cities.
Patel’s team also worked with data scientists at the University of Liverpool who specialise in population data, and who shared their experience of sequence analysis and mixed effects modelling. “They were vital to our quality assurance of findings ahead of publication,” Patel says. Working with academics from across the world allowed the officials to turn data into real-world accessibility metrics that were easy to understand and could inform wider government work.
Adapting to a new administration
A general election in July 2024 and a change of government provided another challenge. The disruption delayed some of the timescales for delivery of the project, and the new administration’s policy to open many more school-based nurseries across the country has necessitated an expansion of the project’s focus. SBNs currently make up about 20% of provision in the country, and Labour has pledged to open some 3,000 more, taking the proportion to 27%. Around 200 SBNs are being launched in September, with another 100 joining them later this year.
“The [new government’s] focus is very much on expanding early years in schools and that was something that we didn’t include initially,” Patel says. Schools that have nursery provision for two-to-four-year-olds were not initially included in the project because they do not have to register with Ofsted. The data on them is collected by DfE and it doesn’t entirely align with Ofsted’s data. But Patel says the team built enough flexibility into the project’s design to pivot and include data from school-based nurseries.
“That’s where our partnership with DfE works really well,” Patel adds. “And we are working closely to the point that we just received some data from the department which will enable us to look at the whole early years [system]. We’re working hard to incorporate that. That’s one of our immediate next steps.”
The key to success
What advice does Patel have for officials across government who are approaching similar tasks?
“Collaboration is key,” she says, advising officials to use “any contacts you have across government departments, whether that’s in the analyst community or whatever you’re specialising in, but also academics”.
Patel adds that having a dedicated team with a broad range of specialisms was also vital to the project’s success, and to support this it was important to have “really clear milestones and a shared vision” that meant the multi-disciplinary team could still “measure progress and keep things on track”. She says it is also important to look for opportunities to be innovative and to focus on having a tangible impact. “I think what is great about this is it has delivered real-world impact based on a policy that’s been set out and we’ve been able to do quantitative analysis to support that as the rollout happens,” she explains. That impact has been significant. Patel says the project has generated “substantial insight” around childcare access and helped to monitor policy implementation and mitigate potential risk around childcare policy.
The work has also had sustained interest across government – and beyond – and “has really helped for any future decisions in the childcare sector”, she adds. “We’ve done so many presentations, I think I’ve lost count. Within government, but also externally – to think tanks, academics.”
Winning the RSS’s Campion Award, to add to a well-stocked cabinet, has been “the icing on the cake”, Patel says. “It was a true honour”.