F: Technical Annex on the Children and Young People’s Services (CYPS) Relative Needs Formula (RNF)
Updated 20 June 2025
Applies to England
The LG Future Final Evaluation Report provides the technical guide allowing the reader to access the detailed research underpinning the CYPS model. This note provides the updated parameters required to calculate need shares.
As explained in the consultation document, the CYPS model has been updated since the Final Evaluation Report with the latest available data, including Census 2021, the National Pupil Database and the Children in Need (CiN) and Children Looked After (CLA) datasets (previously based on 2015-16). This has resulted in changes such as the replacement of the ‘limited activity due to health’ variable collected in the 2011 census to ‘poor health’ from the 2021 Census. The CYPS RNF model was also trained on the full dataset, rather than the subset used in the Final Evaluation Report. Collectively, this has resulted in changes to the original model parameter estimates, although the main drivers underpinning the model remain deprivation levels and population numbers. Another smaller change has been the removal of ethnicity from the model. Paragraphs 12.2.16 to 12.2.18 of the consultation document provide the reason for this change.
The new parameter estimates and the resultant probability for an example child are presented in Table 1 and Table 2 below. These are, respectively, updated versions of Table 47 and Table 48 from the Final Evaluation Report. Note these parameters will change again when the model is updated with the new indices of deprivation measure (IDACI) when the data is expected to be published later this year.
The steps for calculating local authority CYPS RNF share remain the same as detailed in paragraphs 10.2.1 to 10.3.4 of the Final Evaluation Report.
Table 1: Model parameter estimates used to predict individual probabilities
Variable/Factor | CiN as at 31 March 2016 [CiN31March] | In Care during 2015-16 [InCare] | Care Ceased during 2015-16 [CeasedCare] |
---|---|---|---|
Intercept | -5.23 | -6.93 | -8.27 |
FSM Eligible (Reference=Not FSM Eligible) | 2.16 | 2.12 | 2.49 |
Female (Reference=Male) | -0.09 | -0.10 | -0.04 |
Age 8-9 (Reference = Age 6-7) | 0.03 | 0.07 | -0.29 |
Age 10-11 | 0.12 | 0.22 | -0.48 |
Age 12-13 | 0.31 | 0.51 | -0.22 |
Age 14-15 | 0.49 | 0.80 | 0.03 |
Age 16-17 | 0.73 | 1.14 | 0.49 |
LSOA: Proportion overcrowded households | -2.26 | -2.80 | -1.99 |
LSOA: Proportion children with parents with low quals | 1.01 | 1.68 | 0.77 |
LSOA: Proportion of children with poor health | 4.71 | 1.14 | -4.03 |
LSOA: 2019 IDACI Score | 3.85 | 5.50 | 3.77 |
LSOA: Travel Time to Town Centre (mins) | -0.01 | -0.02 | -0.01 |
LSOA: Population density (persons pe km2) | 0.00 | 0.00 | 0.00 |
Interaction: FSM * LSOA 2019 IDACI Score | -3.04 | -4.57 | -3.04 |
Interaction: FSM*Age 16-17 | 0.04 | 0.57 | 1.23 |
Table 2: Model parameter estimates used to predict individual probabilities
Variable | Factor/Description | Value | Model Parameter Estimate | Estimate Value × Model Parameter Estimate |
---|---|---|---|---|
Constant | Model Intercept | 1 | -5.231 | -5.2306 |
Sex | Male | 1 | 0.000 | 0.0000 |
Age | Age 14-15 | 1 | 0.491 | 0.4909 |
FSM | Yes | 1 | 2.163 | 2.1633 |
LSOA | 2019 IDACI score | 0.017 | 3.850 | 0.0654 |
LSOA | Proportion children with parents with low quals | 0.10478 | 1.013 | 0.1062 |
LSOA | Proportion of children with poor health | 0.01695 | 4.715 | 0.0799 |
LSOA | Proportion overcrowded households | 0.01802 | -2.262 | -0.0408 |
LSOA | Population density (persons pe km2) | 2442.33 | 0.000 | 0.0055 |
LSOA | Travel Time to Town Centre (mins) | 11.48 | -0.008 | -0.0918 |
Interactions | Interaction: FSM * Age16-17 | 0 | 0.035 | 0.0000 |
Interaction: FSM * LSOA 2019 IDACI Score | 0.017 | -3.036 | -0.0516 | |
Sum Product | -2.5036 | |||
Probability | 7.56% |