Improving children's outcomes through data
View this email in your browser

The Datamonger

Etazo Performance Data's newsletter

Dear <<First Name>>

Introduction and update

This month we're marking National Safeguarding Month by looking at recent data releases and publications linked to safeguarding children: through fostering, services to children with disabilities and provision of social work staff.
Please do get in touch if you have any suggestions for themes we could address in future issues, or if you'd like a conversation about how we could support you with focussed analysis, developing performance frameworks and meeting targets.
In the next issue we're planning to look at data related to child sexual exploitation, with a focus on how your organisation can use available data to monitor and plan services.
Best wishes
Jo and Georgia
In this issue
~  Data publications
~  Etazo researches:

      Children's social workforce 2016
      Fostering in England 2015/16
      Understanding the needs of
          disabled children

~  What makes a good data scientist
~  Useful links
Data publications

~ Outcomes for children looked after by LAs: 31 March 2016
23 March 2017 9:30am (confirmed)
~ National Survey for Wales, financial year ending March 2015: On-line safety for children
28 March 2017 9:30am (confirmed)
~ Children's Social Work Statistics 2015-16: Scotland
28 March 2017 9:30am (confirmed)
~ Children and Young Patients survey: 2016
April to May 2017 (provisional)
Recently published
~ Children's social work workforce 2016
16 February 2017
~ Children looked after in England including adoption: 2015 to 2016
28 February 2017
~ Children and Young People’s Health Services Monthly Statistics: Nov 2016
10 March 2017

Etazo researches   

Children's social work workforce 2016
DfE (February 2017)
DfE statistics
This publication reports on the data supplied by LAs in the Children’s Social Work Workforce return for 2015/16. In the past this return has been problematic for LAs to complete accurately and difficult to make sense of the published data, because the guidance has been interpreted differently in different LAs (something we talked about on the Google group when the 2014/15 data were published). This year the guidance was revised and the DfE thinks that the data are more accurate – we would still recommend caution when using the data locally, and there are a few obvious anomalies in the published information.
The report includes a new measure on the average caseload for each LA. As we understand it, this has been calculated by asking each LA for the number of open cases (which should equate to the number of children in need), the number of those which are allocated, the number of case-holding SWs and the number of agency SWs – then dividing the number of allocated children by these total SWs. We think there’s an problem here in that not all agency SWs will be caseholding, so the actual average caseload is likely to be higher than the reported figure of 16. Possibly partly because of this, the average caseload varies from 8 to 51, though the latter is an outlier – only ten LAs report average caseloads above 25 children.
A surprisingly high number of LAs, 28, report more than 80% of cases as not allocated. This may represent LAs which are under particular pressure and cannot allocate cases, or LAs with a high vacancy rate – taken as a group, these LAs do have a higher than average level of vacancies.
The percentage of staff who are agency workers varies considerably across the country. It is generally highest in London LAs. There are a handful of LAs with no agency workers at all, and two where more than half the staff are agency.
The report also includes sickness and retention data, which may be areas where your LA will want to compare itself with others.
Fostering in England 1 April 2015 to 31 March 2016
Ofsted (February 2017)
Ofsted fostering return

This data publication is the results of the Ofsted Fostering data collection for 2015/16. There is a report reviewing national trends, and it’s worth looking at the key points from that. Some of the changes noted are probably about improvements in data quality; for example, the increase in the proportion of children looked after who are disabled from 7% to 10%.
The increase in the proportion of care leavers who continue living with their foster carers at 18 is very striking, from 22% in 2014/15 to 54% in 2015/16.
There was a steep fall this year in the number of applications to be foster carers, from 16,920 in 2014/15 to 11,460. The report does not discuss why this might be. We wonder if some authorities have made cuts to their recruitment services.
This is the second year of collecting data on children going missing and on child sexual exploitation, and patterns are beginning to emerge. The most common reason for children looked after to go missing is related to contact with people, either wanting to see family and friends or wanting to avoid contact with family. There are still likely to be data accuracy issues: we find it hard to believe that ten LAs have no CLA at risk of or experiencing CSE, including one LA which as a coastal town fits the profile of areas where CSE is a problem.
Locally, we have used this return in conjunction with the CLA return to look at the number of LA foster care places in proportion to the number of children looked after in that authority. You can do this by working out the number of CLA as a percentage of the number of places. For a few LAs, the number of CLA is exactly the same as the number of places available, so the number of CLA is 100% of the number of places. In other authorities there are more or fewer children than foster care places available. So in Lambeth, for instance, there were 460 CLA at 31st March 2016, but only 165 approved places – the number of CLA is 279% of the number of places available. High numbers like this may represent pressure on places and more children in IFA foster care or residential care as a result. A small number of LAs are in the process of outsourcing foster care completely, and they will also show this sort of pattern. In a small number of LAs there are more places available than there were children looked after; in North Yorkshire, for instance, there were 410 children looked after and 555 approved foster care places, so the number of CLA as a % of places was 74%. For some LAs this may be a result of recent successful reductions in the number of children in care.
Understanding the needs of disabled children with complex needs or life-limiting conditions: What can we learn from national data?
Council for Disabled Children and the True Colours Trust (February 2017)

Council for Disabled Children
This report focusses on data available on children with complex disabilities, including the School Census and the Children in Need Census.
The School Census figures show a rise of 48% since 2004 in the number of children with complex conditions who are in state schools (defined as children with EHCPs or Statements whose primary needs suggest complex conditions [Severe Learning Disability, Profound and Multiple Learning Disabilities or Multi-Sensory Impairments], or who have ASD and are educated in special schools).
CiN Census data are more difficult to interpret. The report authors note that
It is clear from the data analysed that thresholds for support from social services for disabled children and their families are high and vary widely from area to area. This may reflect tight local authority budgets and more positively, the roll-out of Early Support.
They note the range in the rate of children in need with disabilities across different LAs, and that most CiN only have one or two disabilities recorded. We think they are wrong to make the assumption from the latter data that these CiN are less likely to have complex disabilities, or they would have more disabilities recorded; in our view this is more likely to be about recording practice in LAs. We also think the report may make some incorrect assumptions about recording of need codes of children looked after who are disabled (too detailed to go into here, but get in touch with us if you’re interested).
The report also reviews data on children with learning disabilities or ASD who were in-patients for mental health. Under 18s had much higher levels of “adverse experiences and restrictive measures” (self-harm, accidents, assault; restraint or seclusion) than older age groups. More than two thirds had experienced either adverse experiences or restriction in the previous three months.
And they look also at CAMHS data on children with a learning disability or autism: at June 2016 over forty thousand were waiting for mental health services. As the report authors say, this shows high levels of unmet need.
The report makes it clear that not enough information is available on children with complex disabilities. The SEN data are the only available data showing the rise in this group, and these have not previously been analysed in this way and the trend highlighted. There are also gaps in this information for independent schools and for children or young people who are not of school age. Commissioners, LAs and LSCBs particularly struggle to unpick local data and generally have poor data about the needs of children with complex disabilities.
The report authors recommend that the DoH introduces incentives and penalties to support better reporting of data to the Children and Young People’s Health Dataset, that the DfE extends School Census EHCP / Statement reporting requirements to the independent sector, that information of disability is added to the CLA data return, and that an Early Help dataset is developed. Data colleagues in LAs are likely to be heavily involved in some of these changes if they are taken forward.
They also recommend making better use of existing data, such as matching between datasets and increasing the focus of analysis of returns such as the CiN Census on children with complex needs.
The report includes sizable and useful appendices analysing disability data from the School Census (Annex B), the CLA return (Annex C), the CiN Census (Annex D) and the Mental Health Services dataset (Annex E).
See also our previous issue on data on children with a disability:
The Datamonger, February 2016.

What makes a good data scientist – “don’t worship the machine”!

“At the M.I.T. conference, Ms. Schutt [Rachel Schutt, a senior statistician from Google] was asked what makes a good data scientist. Obviously, she replied, the requirements include computer science and math skills, but you also want someone who has a deep, wide-ranging curiosity, is innovative and is guided by experience as well as data.”
From “Sure, Big Data Is Great. But So Is Intuition”, New York Times
Useful links
Definitions and data about asylum seekers and refugees in the UK, from Full Fact -
“Lies, damned lies, etc: Why reporters must handle data with care”, from the Royal Statistical Society -
“The 38 best tools for data visualization” -

Children's Social Care Data Google Group
We administer the email group for local authorities to discuss children's social care data. Membership is recommended by the DfE and Ofsted as a good resource for performance staff to discuss interpretation of data, definition of indicators and year-end returns, and is of course completely free.

One group member recently told us that the group has been "a real help" with tips and pointers, such as problems with the DfE website and experiences of inspections. The group member also said, referring to end of year returns, "It’s been reassuring to know that as I plough through 903 errors, others are going through the same process".

If you or your performance staff would like to join, please ask them to email us or to go to the group's homepage at
Forward this newsletter
Copyright © 2017 Etazo Performance Data, All rights reserved.

Want to change how you receive these emails?
You can
update your preferences or unsubscribe from this list

Email Marketing Powered by Mailchimp

Georgia Corrick
07789 993 904

Jo Price
07790 181 539

Find us on Linkedin