21 Nov 2024
Alison Wallace, Emeritus Professor, School of Business and Society at the University of York introduces findings from the first UK examination of housing risk-profiling technologies
Over the last decade, welfare reform, increased marketisation of social housing finance and practices and rising social housing ‘affordable’ rents have, in combination, given rise to social landlords undertaking stringent affordability assessments as part of their routine pre-tenancy checks. Our Code Encounters Nuffield Foundation- funded research looked at the use of digital risk profiling tools that mediate access to housing across the market.
The study included an exploration of how social landlords approached affordability assessments, what data they used to verify tenant circumstances, how they calculated affordability and the degree to which letting decisions were algorithmically informed.
We found a variegated approach not least in data used, how affordability thresholds were set and how expenditure was calculated, but around the sense of change and uncertainty about greater automation. Landlords nonetheless felt an imperative to digitalise and automate to save time and expense. Critically, within this, the role of affordability assessments was unclear, often framed by participants as a triage for access to tenant support and sustainability, but still resulting in exclusion for some, often young people, without a clear alternative path to housing.
Following the pandemic and the cost of living crisis, providers are wrestling with how to reconcile social purpose with business risks arising from external economic and policy change.
‘It's okay, I think, and quite right if people are wanting to find out background information about someone in that context, but if it's being used to simply exclude someone from social housing provision, then I think we're missing the point of what social housing is for, and the types of people that are more naturally going to be needing social housing.’ (Stakeholder 13)
Following detailed financial appraisal, some applicants were routed towards debt advice or undertook successful income maximisation checks, and indeed individuals and communities gained significant sums in previously unclaimed benefits. Tenants welcomed this support.
There were disparate approaches to affordability assessments, however, particularly around estimating expenditures, which would be critical to housing access at the margins. Providers set affordability thresholds in different ways, and these points of entry had often been lowered in recent years to fail fewer tenant applicants and ease pressure on service use. This raises questions about the utility of the assessments at the outset.
Despite efforts to reframe or drop the practice of detailed affordability assessments, especially post-pandemic, interviews indicated that providers still used these assessments to refuse tenancies. Participants regularly cited that the practice was impacting young people in particular, and others for whom the benefit system is now inadequate.
‘So, we don't sometimes use it as a blocker to getting housing. It's more of an advice before they get a property. Obviously, there's always going to be some cases where it's just really not affordable, and we need them to go away, and come back at a later date, when they've evidenced those changes [reducing expenditures].This depends on things like what their circumstances are like, and what their housing needs are like, and how quickly they need to move, and how drastic that a lack of affordability is, as well.’ ’ (Social landlord 7)
Many providers currently adopted digital platforms or were looking to draw in additional data and to automate what are often, although not exclusively, currently analogue systems. Of interest to some was Open Banking data, where people give consent to third parties to access detailed banking transaction data to assess income, expenditure, verify circumstances and reveal new insights into debt, gambling or child support for example.
Before adopting greater digitisation and paying for platforms, landlords must first (re)consider the purpose of the assessments, the data regulation implications of proportional data collection to inform decisions and find ways to protect tenants’ privacy and autonomy. Landlords also need to evidence the link between initial assessments and tenancy performance.
‘So, we're doing research at the moment to see if we can get better insight into what is happening to people at those income brackets. If they start getting into arrears, when is that? Is that uniform across, or is there no correlation? [£0-£50 per week spare income] We don't know yet, because we haven't done the research. What we do know is, those that demonstrate that they will have, or we assume, because they've got low income, will have difficulties affording that property or sustaining that property, we're getting in there from the get-go to offer whatever help we can.’ (Social Landlord 9)
Greater data and automation may create administrative efficiencies, and tenants like the convenience of some lettings automation, but staff and tenants value human interaction and relationships. Although tenants were uncomfortable with banking data being used as the insights were considered intrusive and went beyond confirming ability to meet rent commitments.
But critically algorithmic assessment is not the answer to policy-driven challenges accepted by or created within the sector, like affordable rents, or welfare reform. While systems have value to enhance administration and tenant circumstances - as technology can maximise income or even to help tenants retain ownership of data and in control of assessments - the adoption of new data resources and automation needs a clear purpose that accords with sector purpose and values.
Affordability assessments contribute to tension between local authorities and providers returning nominations of homeless households, which evidence suggests has reduced only marginally. If landlords are using the tools to triage for support, then there should be a distinction between proportionate mandatory data collection that materially influences letting decisions, and additional data relating to support services that may be discretionary that tenants can choose to volunteer and benefit from.
If data is primarily collected to inform letting decisions, then providers should consider whether this task is necessary, how it aligns with core values, and critically, what other offers, such as furnished tenancies or foyers for young people, are offered, individually or collaboratively in the district, to serve excluded groups.
With a new government prioritising social housing standards and supply, there is an opportunity to reassess the role of affordability assessments and automation, balancing support and inclusion with data privacy and efficiency.
The full report on data, automation and access to social housing can be obtained from the Code Encounters website along with other outputs of the project. https://www.york.ac.uk/chp/housing-markets/code-encounters/
Emeritus Professor Alison Wallace, School of Business and Society, University of York.
Alison is now an Emeritus Professor in Social Policy and Housing and has been part of the Centre for Housing Policy since 2001, now the School of Business and Society at the University of York. Alison was formerly a social housing practitioner and then undertook teaching and research on low cost home ownership, private renting and shared ownership housing. Recently she developed an interest in digital technologies and housing working on damp mould and Internet of Things sensors and now a Nuffield Funded Project on digital risk profiling and access to housing. alison.wallace@york.ac.uk