Several changes have been made to the methodology for calculating the statistics shown in this release.
The two changes that have had the most impact are:
- Updating the “weighting” methodology used to estimate regional and England “group-based provider” totals - in order to compare them with results from the 2025 Survey, estimates of the total and mean number of registered and booked group-based provider places, and total and mean number of paid group-based provider staff in 2024 have been revised down.
- “Imputing” missing data where providers report that they “don't know” the number of registered and booked places that they have - in order to compare them with results from the 2025 Survey, estimates of the total number of registered and booked places at school-based providers, group-based providers and childminders in 2024 have been revised up.
The sections below provide more detail of these changes and describe the scale of revisions made to 2024 figures.
A change has been made to the design of the survey in 2025. As a result, estimates of the average fees charged by “all providers” in 2025 are slightly higher than they would have been if the 2025 Survey had been carried out in the same way as the 2024 Survey. Comparisons of average fees for individual types of providers are unaffected.
Other changes described in the sections below include a change to how “average per child” calculations have been made (these are based on the number of registered children rather than registered childcare places that providers have) and changes to the wording of several questions on the Survey, which have improved the accuracy of the information collected but had an impact on some of the time series data.
Group-based provider weighting
Results from the Survey are weighted in order to be representative of all childcare providers in England. The weighting methodology is described in detail in a Technical Report available for download from “additional supporting files”.
Briefly, this methodology gives more weight to categories of providers that are under-represented in the sample of respondents to the survey, compared to the population of childcare providers as a whole, and less weight to categories of providers that are over-represented in the sample.
Analysis of results from the 2024 Survey identified that large group-based providers - those with an above-average number of registered places on Ofsted's Early Years Register of childcare providers - were over-represented in the sample. Specifically, group-based providers that responded to the 2024 Survey had approximately 5 per cent more registered places on the Early Years Register than average. The weighting methodology used to estimate totals for all group-based providers in England did not take account of this bias. This meant that some of the statistics produced from the 2024 Survey - such as the total number of registered group-based provider childcare places, total number of children booked to attend childcare at group-based providers and total number of paid staff employed by group-based providers - were over-estimated.
In the 2025 Survey, the weighting formula for group-based providers was updated to include a measure of “provider size”. Weights from the 2024 Survey have also been re-calculated, using the same formula used in the 2025 Survey, so that results from the two years can be directly compared.
Most “percentage” measures (such as the percentage of childcare places that are spare) and most “average” measures (such as the average number of weeks that childcare providers expect to be open during the year) are unaffected by this change. Detailed results from the Survey are shown in a set of tables, available for download from “additional supporting files” - where appropriate, these show if, and how, 2024 figures differ, depending on whether the "old” or “new” weights are used.
Updating the weighting methodology for the 2024 Survey does, however, have a more significant impact on “volume” measures. The table below shows:
- Total number of registered childcare places;
- Average number of registered childcare places;
- Total number of booked childcare places;
- Total number of paid childcare staff; and
- Average number of paid childcare staff
for group-based providers in England in 2024, based on the “old” and ”new” weighting methodologies.
Imputing missing data
Some providers either “don't know” or “prefer not to answer” certain questions on the Survey. Rather than risk them dropping out of the Survey altogether, providers are allowed to skip a question that they are unable or do not want to answer and move on to the next one.
Another important methodological improvement made in the 2025 Survey is to “impute” (or estimate) missing Survey data for important “volume” measures, such as the number of registered and booked childcare places. Not imputing these data led to England totals for previous Surveys being under-estimated.
The amount of missing data - and impact of estimating rather than excluding missing data - varies, depending on the measure and type of provider. Virtually all group-based and school-based providers, for instance, reported their "number of paid staff" so the very small amount of missing data for this measure has not been imputed.
It is more common, however, for providers to say they “don't know” or “prefer not to say” the number of registered and booked places that they have. Accordingly, estimates have been made of the number of:
- Registered nursery places;
- Before-school and after-school places for school-age children at group-based providers; and
- Children booked in full-day nursery care
in cases where it was missing. Specifically, where a provider told us they “didn’t know” or “preferred not to say” how many places they had, it was assumed, for the purpose of estimating regional and England totals, that they had the median number of places for providers of their type.
Missing data from the 2025 Survey have been imputed and the imputed figures used in estimates of England totals. Missing data from the 2024 survey have also been estimated and revised estimates of England totals produced. These are shown in the table below. This shows that, proportionally, estimates of registered childminder places have seen the biggest impact.
The reason that the mean number of registered school-based provider and group-based provider places per provider have fallen, despite the total number of registered places increasing, is that, where missing data have been estimated, “per provider” calculations are based on bigger numbers of providers.
Childcare fees for “all providers”
Prior to 2025, there were two versions of the Survey. The “main” version asked about all of the topics covered by the Survey while a “short” version asked a “booster” sample of providers questions about fees and funding rates, so that these could be estimated at local authority level. Almost all of the providers sent the short survey were childminders.
In order to boost responses to the main survey, in 2025 all providers were sent this version of the Survey and the short version was dropped. To compensate for the short version of the Survey being dropped, instead of questions about fees only being included on certain “variants” of the main survey, in 2025 they were made “core” questions i.e. every provider was asked to answer them.
This change to the design of the survey means that average fees for “all providers” that responded to the 2024 Survey are not directly comparable to those for “all providers” that responded to the 2025 Survey. Specifically, because, in 2025, all group-based providers (rather than only group-based providers sent certain variants of the “main” survey) were asked about the average fees that they charged, and the “short” version (which was primarily sent to childminders) was dropped:
- Group-based providers account for 42 per cent of “all providers” in the 2025 sample (compared to 32 per cent in 2024); and
- Childminders account for 45 per cent of “all providers” in 2025 (compared to 52 per cent in 2024).
Calculations of the average fee charged by group-based providers, average fee charged by school-based providers and average fee charged by childminders are unaffected.
However, because group-based providers (who charge, on average, higher average fees) make up a bigger proportion of “all providers” in 2025 than in 2024 and childminders (who charge, on average, lower average fees) a smaller proportion, the average fee for “all providers” in 2025 is higher than it would have been under the old survey design. It is estimated that mean fees for 3 to 4 year olds are approximately 13p per hour higher than they would have been, mean fees for 2 year olds approximately 17p per hour higher and mean fees for under-2s approximately 20p per hour higher.
So, although comparisons are made between mean and median fees for “all providers” returning the 2024 and 2025 Surveys, it should be noted that in 2025 this calculation gives more weight to fees charged by group-based providers and less weight to fees charged by childminders. Accordingly, in the tables accompanying this release, the statistical significance of differences between mean fees fess charged by “all providers” in 2024 and 2025 has not been calculated.
Mean fees calculated for “all providers” in 2025 are considered a better measure of the fee charged by the “average” provider because in the 2025 Survey all childcare providers are asked about average fees, whereas in 2024 mean fees for “all providers” were skewed by almost all providers that were sent the “short” survey being childminders.
Unlike fees, the average funding rates received by different types of providers for “funded entitlement” hours are very similar, so the change to the design of the survey has not had an impact on average “all provider” funding rates.
Tables on childcare staff joining and leaving providers
Although the questions are designed to be consistent, school-based providers and group-based providers reported a significantly bigger number of “net joiners” (number of paid staff who joined in the last 12 months minus the number of paid staff who left) than their total number of paid staff increased by. Group-based providers, for instance, reported approximately 20,000 “net joiners” in the 12 months prior to the 2025 Survey but that their total number of paid staff had only increased by a fraction of this (approximately 2,500).
Intuitively, it is easier for providers to report how many staff they currently have than to remember how many staff have joined and left over course of the last year. Accordingly, tables showing the total number of paid staff employed by providers have been retained in the publication while tables showing the number of staff joining and leaving providers have been dropped while the discrepancy between “net joiners” and “total paid staff” is further investigated.
Tables showing staff turnover rates continue to be shown as although there is some uncertainty over the total number of staff leaving providers used in these calculations these tables show a clear and consistent difference between provider types i.e. higher “staff turnover” rates for group-based providers than school-based providers.
Changes to survey question wording
In some cases, the wording of questions asked on the Survey changed in 2025. The main changes made, and reasons for making them, are described in more detail in the Technical Report. Briefly, changes were made to make questions:
- More clear - On the 2024 Survey, childminders were asked “How much childminding income do you currently personally earn per year?” In the 2025 Survey, clarification was added to this question ("By this we mean, the income you take from the business in the form of a salary") in order to distinguish “personal” childminding income from fee income (which, in addition to paying the childminder's salary, pays for materials, equipment, the salaries of any assistants that they employ etc.);
- More consistent - On the 2024 Survey, providers were asked “How many paid childcare staff do you currently employ on a temporary basis?” In the 2025 Survey, this was changed to “How many temporary paid childcare staff are involved in the delivery of your childcare provision?” This was changed to be consistent with other questions on the survey, which ask about staff delivering childcare, as opposed to any ancillary staff such as accountants and administrators that the provider might employ.
- More precise - On the 2024 Survey, school-based providers were asked “How many children in your nursery have Special Educational Needs or Disabilities (SEND)?” In some case, schools reported a very high number of children having SEND (more than the total number of children that they had registered in their nursery provision), suggesting that they were counting children in Key Stage 1 and Key Stage 2, as well as children attending nursery provision, that had SEND. In the 2025 Survey, therefore, this question was changed to “How many of the […] under school-aged children registered with your nursery provision have Special Educational Needs or Disabilities?”
Although, in the long-term, changes such as this improve the quality of information collected on the Survey, in the short-term they may have an impact on time series.
The proportion of total childminder costs that are accounted for by staffing (including the “salary” that childminders pay themselves), for instance, fell from 77 per cent in the 2024 Survey to 70 per cent in 2025; the number of temporary paid staff at group-based providers and school-based providers fell by over 20 per cent, from 25,600 in the 2024 Survey to 20,200 in 2025; and the proportion of children registered with school-based providers that have SEND fell from 16 per cent in the 2024 Survey to 14 per cent in 2025 (although this fall is not statistically significant, it is different to the trend for group-based providers, where the proportion of children with SEND saw a statistically significant increase between 2024 and 2025).
Where changes to question wording may have impacted on figures in tables, this is highlighted in footnotes to tables.
“Percentage of children” measures
There are several tables in this release showing the “percentage of children” who have special educational needs and disabilities (SEND), have an Education, Health and Care (EHC) plan and have fees paid using Tax Free Childcare. In 2024 and earlier years, these tables showed the number of children who, for instance, had SEND as a proportion of the number of registered childcare places at the provider. In 2025, these calculations have been updated to show the number of children who have SEND as a proportion of the number of registered children at the provider. This is a more appropriate measure of the prevalence of SEND. Figures for 2024 have been recalculated so that they can be directly compared with those for 2025.
The Survey asks group-based providers whether they are privately owned (and run “for profit”) or owned by voluntary, community or charitable organisations. The average “private” group-based provider has different characteristics to the average “voluntary” group-based provider - on average, “private” providers have more children registered, charge parents slightly higher fees and are significantly more likely than “voluntary ” providers to be part of a “chain” of childcare providers.
A slightly higher proportion of group-based providers that responded to the 2025 Survey reported that they were "voluntary" compared to the 2024 Survey. The weighting formula for group-based providers includes a variable relating to “ownership type”. This variable, however, separates providers into those owned by “individuals” and those owned by “organisations". This means that weighting does not correct for changes between one year and the next in the proportion of group-based providers responding to the survey that are “private” and the proportion that are “voluntary”. As a result, the breakdown of group-based provider figures between “private” and “voluntary” providers shows year-on-year fluctuation.
In this year's release, “volume” measures - such as the total number of registered group-based provider places and total number of paid group-based provider staff - only show a “total” figure for group-based providers rather than a split between private and voluntary group-based providers. In most cases, however, “percentage” and “average” measures continue to show a breakdown between private and voluntary group-based providers as these are largely unaffected by differences in response rates.
For the 2026 Survey, we will look to update the group-based provider weighting formula so that “ownership type” reflects whether providers are “private” or “voluntary”, as opposed to whether they are owned by “individuals” or “organisations”.
Feedback
The Department is proposing to continue to use the new methodologies described in this section for the next publication and explore further improvements.
Users are invited to provide feedback on these changes. Please send any feedback to EY.AnalysisANDResearch@education.gov.uk by 31st March 2026.