Back in January 2019, the National Audit Office published a report which claimed that 44 per cent of the variance in Local Authority Child Protection Plans could be explained by ‘local authority characteristics’ other than deprivation (which could explain 15 per cent). Although they never claimed this, the finding was largely interpreted as evidence that local authority policy and practice was more important than structural disadvantage - by some magnitude. However, an analysis by myself and colleagues of the Child Welfare Inequalities Project recently published in Children and Youth Services Review found that, once deprivation and other structural factors are accounted for at the neighbourhood level, differences between local authorities shrink dramatically and in some instances practically disappear.
The Nuffield Foundation funded a modest extension to our original project to use some more complex modelling techniques on the data we had collected from 18 systematically chosen local authorities. These local authorities were chosen to be a representative sample of all local authorities in England, reflecting many different regions and varying levels of deprivation and ethnic makeup. For this analysis we aggregated Childen in Need, Child Protection Plan and Children Looked After records up to the Lower Super Output Area level and joined several LSOA measures onto this data. An LSOA can be thought of as a large ‘neighbourhood’, with around 1,500 people living in it. This contrasted to the NAO’s approach which, although not made transparent, seemed to use far less granular data.
Our paper focused mostly on the magnitude of structural inequalities and the size of the Inverse Intervention Law and the Inequalities Intervention Law, but the technical appendices also contained measures of variance linked to neighbourhood local authority membership and how these changed after controlling for different structural factors. It’s these measures that I want to expand upon here, because I believe they are important for the ongoing ‘Policy and Practice’ versus ‘Structural Disadvantage’ narrative that appears so frequently.