From September 2025, the full rollout of the current model of funded hours for accessing childcare provision in England will be complete. This will mean that children in families where all parents work at least 16 hours per week will have access to 30 hours of funded childcare from nine months to four years old. All three- to four-year-olds will have at least 15 funded hours of access, regardless of parental working status, as will two-year-olds from low-income households or with additional support needs. Families on universal credit and in work will also be supported to meet the cost of childcare.
However, this patchwork of policies will fail to ensure childcare is affordable for many families eligible for the full offer of support; accessible to low-income families who are excluded from this full offer; or of good enough quality across the country. The education secretary has acknowledged that the current system lacks coherence and that the government plans to pursue more comprehensive reform:
“There is a big question about how we simplify the system… it’s not as easy to navigate for parents as it should be. It’s a system that’s had bits added on to it at different times — I think there is a need to bring it all together. We will do that, and I’ll set out plans later this year around an early years strategy.”
This briefing proposes an alternative system, developed by NEF in collaboration with Pregnant Then Screwed and the Joseph Rowntree Foundation. Our proposal is significantly simpler and more progressive than the current system, and it would help families afford more hours of childcare. It is built around two elements:
- Universal core provision of 15 hours a week, free for all children from nine months to four years old, focused on ensuring equitable access to high-quality early years education.
- For working families, a cap on childcare costs as a percentage of family earnings (we propose 5% but we also model for 7.5% to demonstrate the impact on costs) for hours purchased above this core provision to enable them to work.
In contrast, the current system is complex, regressive and does not support enough use to help people work or support sufficient high-quality provision across the country. Our analysis finds:
- Low- and middle-income families in full-time work will still need to spend a high proportion of their earnings on childcare once the current system is fully rolled out. Families with multiple children and single parents will be hit particularly hard. A working family on £34k pa (at the 30th percentile on the household income distribution) would have to spend over 11% of their gross earnings to have one child in full-time childcare (40 hours per week). A family on £49k (the 50th percentile) would need to spend 8% of their earnings, while a family on £124k (the 90th percentile) would only need to spend 3.1%. As such, lower income families face stronger disincentives to work full-time, because the proportional costs of paying for childcare are so much higher.
- By contrast, a system with a cap on costs at a percentage of earnings would be inherently progressive – supporting the families who would benefit most from access to childcare. This would incentivise them to purchase more hours than under the current system, by ensuring only modest increases in their childcare costs as their earnings and hours rise – a stark contrast to the current system. On average, those in the bottom 60% of the earnings distribution would be better off by £1,081-£1,259 a year from our proposed system.
- Across the whole earnings distribution, almost 60% of families with pre-school children would be better off under our proposed system, based on our modelling of the most likely usage scenario, with 66% being no worse off. Those faring worse would primarily be families spending less than 5% of their earnings on additional hours above the 15 universal hours, and therefore not benefit from the cap. Further work could explore how to mitigate this impact, but it is a knock-on consequence of a system that is actively encouraging a higher level of usage to support labour market outcomes.
- At current hours of usage our proposed system would initially be cost-neutral, but as usage expanded it would present a net additional cost to the government of ~£3 – 3.4bn (based on current rates paid to providers, which we believe need to be reviewed regardless of the system in place) compared to the fully expanded funded hours system. However, these costs decline by as much as 11% once a labour market response of increased working hours are accounted for, and would fall further if we were to incorporate the wider economic benefits of this labour market response, such as higher tax revenues.
- Our system would also be much easier for families to understand – directly meeting the education secretary’s demand for a simplification of the current system. It would also be easy to communicate, with the universal hours pitched as an entitlement to support child development and the percentage cap pitched as support to families to help them get into work. We believe this could have potential wider behavioural benefits which it is not possible to model for here, but which we plan to explore in future.
Our proposed system would also help to increase the supply and quality of provision, by creating more demand, and therefore incentives for providers, in lower-income parts of the country. However, the government will also need to address the question of what level of payments to providers are required to resource high quality provision, responding to rising costs providers have faced in recent years. We have treated the level of payments required as a separate question from the design of our proposed system, to be able to make a side-by-side comparison to the current system. However, we model for a 10% uplift on current rates paid to providers as a demonstrative example of the impact on overall costs in our system. We fully support calls for a more comprehensive review of rates paid to providers.
Realising the full benefits of our proposed system, and ensuring fewer families lose out compared to the current system, would require additional investment. However, our analysis suggests that, unlike in the current system, the costs would decrease in response to positive labour market outcomes. These outcomes would also result in higher tax revenues and long-term savings from improved prospects for children from low-income households, which, although not modelled for here, would further offset the cost of our proposal.
Finally, we set out how a transition to our proposed system might be achieved. We acknowledge challenges that would need to be addressed but argue that these, along with the additional investment our proposal would require, need to be considered in the context of the failings of the current system. Our model also offers significant flexibility, meaning it could be adapted to fit different funding envelopes and meet different prioritisations of policy objectives.UFCP
