icon picker
An Update on the SAVVI Projects

by Andrew Humphreys, Published 19 March 2024

Wigan Council: supporting vulnerable people in a flood

We have completed our work with Wigan Council to establish data reuse propositions for the purpose of finding people who are vulnerable in an emergency and the council has now established the necessary data feeds and deployed a technical solution to reuse the data in an emergency situation. Wigan Council are now in a much stronger position to find vulnerable people should an emergency situation occur in their area.
We have also developed a set of precise definitions for this data reuse pattern that any local authority with ‘Category 1’ civil contingencies responsibilities can use to replicate the data reuse that Wigan Council has established. These definitions will be the first full set of data reuse definitions for a vulnerability initiative that are published in the SAVVI catalogue of repeatable vulnerability data reuse propositions.
We are also finalising a full case study that will set out all of the SAVVI definitions that have been developed and are now in use by Wigan Council. This represents a significant milestone for SAVVI and we will soon be launching this case study at a showcase event so look out for the invite to that.

Rural and Island Child Poverty in Scotland: finding vulnerable families

We are working with the and three councils in Scotland (Argyll and Bute,
Angus and Inverclyde Councils) to apply SAVVI to finding and supporting families that are vulnerable because of child poverty. Rural and Island locations are of particular concern because of their higher cost of living. The geography of these communities also means that more traditional geographic analysis, i.e. targeting ‘deprived communities’ does not help find those in need.
As set out in the , we started by agreeing a purpose definition based on powers and duties set out in law. In this case, we identified an ‘’ that local authorities in Scotland have to address child poverty through income maximisation measures, defined in the . The purpose definition allowed the local authorities to define the vulnerability that they are seeking to address and the needs and outcomes that their interventions will meet and achieve, using our definition of vulnerability: ‘An Increased Risk of a Poor Outcome’.
Once the needs and outcomes were defined, we were able to support the local authorities to agree on the risk factors needed for them to prioritise the families most likely to be vulnerable and the data sources from which these could be derived. There were three groups of risk factors needed for the risk model:

Risk factors indicating families are known to local services, e.g. income maximisation
Risk factors indicating families are receiving nationally administered income-related benefits
Risk factors that increase the likelihood of child poverty or its impact, e.g. location or household composition

The risk model was designed to prioritise families:
Known as at risk locally but not receiving nationally administered benefits
Receiving nationally administered benefits but not known locally
At higher risk of poor outcomes because of child poverty
So that they could be prioritised for income maximisation interventions needs assessments.
The risk factors in group 1. that the three councils agreed upon as an initial risk model were:
Families who have applied for a Crisis Grant in the last twelve months
Families who are receiving Council Tax Reduction
Families whose Council Tax accounts are in arrears

Risk factors in group 2. were:
Families who have a Universal Credit claim
Families who are receiving welfare benefits via Scottish Child Payment

Risk factors in group 3. were:
Families with a child under one year
Families with a pregnant mother
Single parent families
Families with three or more dependent children
Families living in rural or remote locations
Families living on islands

The proposed data sources are a combination of internal datasets (Council Tax and Crisis Grants), externally held data from DWP in the UK Government and the Social Security Directorate of the Scottish Government and geographic data, which is not personal in nature.
From the risk model, with the support of the Improvement Service, we collated a body of evidence for why these risk factors are effective in prioritising those who are more vulnerable and have used this in data reuse propositions for the internally sourced data and data reuse requests for externally sourced data. These data reuse propositions and requests articulate a lawful basis for the new purpose and a legal gateway, which is the point(s) in law that give powers or duties to the local authorities to reuse the data defined in the risk model.
So far, the three councils have agreed on a legal gateway for one of the two internal sources of data and are looking at using the public service delivery power and multiple disadvantages objective within the Digital Economy Act 2017 as the legal gateway for the other. We are also researching legal gateways for the external data requests and mapping the SAVVI process to the protocols for requesting the data..
We have also had some dialogue with the Social Security Directorate of the Scottish Government on sharing the Scottish Child Payment data with the local authorities. Our plan is to develop all four data reuse propositions and data requests in the coming weeks.

Greater Manchester Combined Authority: reusing data to deliver a supporting families programme

GMCA has continued to develop its data mesh project which initially enables data sharing and use for the Supporting Families programme by GMCA and four of the ten local authorities in Greater Manchester. We recently met with GMCA who explained that they have now brought new data feeds online for the project and we agreed that we will review their information about these to reflect them in SAVVI terminology with a view to them being reused in other projects, which we are now doing.

Want to print your doc?
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
) instead.