Across England, the question of who gets a social tenancy — an affordable flat — is quietly being reshaped by software, spreadsheets and, in a small but growing number of cases, open banking feeds. What used to be a personal conversation with a housing officer is increasingly an automated data exercise: income in, benefits out, debts tallied, risk scores generated.

On the surface this change all looks reasonable. Landlords are under pressure to manage arrears and sustain tenancies in the middle of a long-running welfare squeeze. Checking whether someone can actually afford the rent is common sense. But when you listen to what tenants, landlords, and software firms say about how these systems work in practice, a much more complicated picture emerges.

To start, many landlords ask for hard copies of statements, which is burdensome for many in an increasingly digital world. Take one tenant my colleagues and I spoke with who applied for a flat and was asked for three months of paper bank statements. She had no printer, so she trekked to the local library where her council’s housing hub is based, logged into online banking on a public computer, and printed everything off. Then she logged into her investment account and did the same. By the time she had assembled the paperwork, the fear was not just that her finances might be judged wanting, but that the flat would simply have gone to someone who could move faster.

‍Figure 1. St Peter's Estate in Bethnal Green, London (Photo by Miles Glendinning, Creative CommonsAttribution 4.0 International)

Another tenant, who did everything online, had to ask their bank to post hard copies of the statements out. While they waited, they knew the landlord would not even look at their application. Every day lost felt like a day in which dozens of other people could be getting ahead of them in the queue—and given competition for below-market rents, probably were.

More and more landlords, though, as asking for direct, digital access, aided by software companies. Landlords and software providers point the above snafus as bureaucratic friction that are exactly what their tools are meant to remove. Why chase paper when a click of a button can give a two– or three-month view of income and spending straight from the bank? Why rely on tenants’ self-reported budgets when open banking can categorize every transaction, generate neat charts, and feed straight into an affordability model?

Yet from the tenant’s side of the desk, the offer of a “frictionless” process can look more like a demand for radical transparency. Bank statements do not only show wages and utility bills. They display every takeaway, every gift, every pub trip, every subscription, and a lot of one’s web traffic. One interviewee put it very simply: “They knew everything about me, because it’s all there in my bank statement.” Another talked about carefully crossing out adult content subscriptions before handing over statements because “it would look a bit weird.” A third joked grimly that their transactions probably showed they “spend way too much on beer, buy too much make up and like underwear.”

The discomfort here is not just about embarrassment. It is about how these traces are read. Tenants know that someone else might look at the data and draw conclusions about responsibility, self-control, or “living within one’s means.” What for one person is a small luxury that keeps them going may for an unseen caseworker or algorithms look like reckless spending.

Credit scores, by contrast, can feel almost reassuring. People understand that a history of missed payments will count against them and that a clean file is an asset. Many tenants now track their scores through apps and expect them to be checked when they apply for a home. Credit files, for all their problems, with their cold, hard numbers at least feel loosely tethered to repayment histories rather than to the intimate details of everyday life.

Open banking pushes the relationship between tenant and landlord into new territory. It promises speed and a fuller picture, but at the cost of exposing far more than many people are comfortable with and stripping away the limited control they have when handing over selected paper statements. One tenant summed it up: they would reluctantly print statements because they could “take things off” if they really wanted to, but the idea of giving a third party a direct feed into their account was “a bit too nosy.”

Landlords are not unaware of these problems. Social housing is not a simple testing ground for fintech ideas because there is always the question of social purpose hanging over the table. Some housing providers are wary of giving third party firms deep access to sensitive tenant data. Others worry about being swamped with detail they do not have the capacity to interpret. Even software companies are feeling their way around how to use the data. Some see open banking as useful context rather than the core of a decision.

Social purpose is also what makes affordability checks so fraught. Landlords know that if they get the decision wrong and let someone into a tenancy that is doomed from the outset, there are real human costs. Nobody wants to set a household up to fail, particularly in a system where moving into arrears can reverberate for years. This is one reason some providers talk about their tools as a way of breaking cycles of debt. If the model shows that someone is living off credit to cover basic bills, the response may be to insist on support and budgeting advice rather than simply to reject them.

“Open banking and affordability models do not just allocate tenancies, they harden the line between those who can inhabit subsidized space and those who are kept at the threshold.”

Figure 2. Stafford: Corporation Street flats, Rydal House (photo by Jonathan Hutchins, CC-BY-SA-2.0)

Some organizations lean into this more supportive version of affordability. Landlords describe routing applicants towards income maximization services, discretionary housing payments, or debt advice. They stress that very few people are flatly turned away. In one case, a large housing association using a major credit scoring tool reported refusing only three social rent applicants in two and a half years, in all cases because they would not engage with support, not because they had poor credit histories.

But that is not the whole story. In many cases the numbers simply do not add up. Young single people trying to move out of the family home face a double bind: they are stuck with lower benefit rates for under-25s and hit by the bedroom tax if they take a two-bedroom property because there are too few one-bedroom homes available. When an affordability model sums this up, it often concludes that the tenancy is unsustainable before it even begins. As one landlord put it, with evident frustration, “you cannot sustain that tenancy.”

In those moments, social landlords become the point at which the contradictions of the welfare state are enforced. On the one hand, the benefits system is supposed to provide a basic income for people who cannot secure enough from work. On the other hand, those same benefit levels are treated as insufficient to qualify for a supposedly residual welfare tenure. Digital affordability tools do not create that contradiction, but they do make it highly visible and harder to fudge. The system is effectively saying: “Your officially calculated income is not enough to live on here.”

Zooming out, these practices are part of what Marion Fourcade and Kieran Healy call an ordinal society, where people are continually sorted and ranked across multiple domains. Tenants become what we might call “ordinal social tenants”: data profiles that travel through credit bureaux, tenant referencing platforms, income calculators and experimental open banking tools. It is their “eigencapital” — the composite of their transactions, debts, benefits and payment histories — that moves across systems, not their stories.

Social housing still differs from the private rented sector and mortgage markets in important ways. Affordability checks are less automated, probationary tenancies and support services exist, staff assert the right to “let against the data” in borderline cases. But the direction of travel is similar. The same kinds of data are being pulled in, the same risk logics applied. Pre-tenancy assessment becomes an obligatory passage point, a digital gate that applicants must pass before they can put the key in the door.

Figure 3. Portobello Way: new-build social housing (Photo by Lydia Shingbrightly CC BY 2

“Affordability tools could be used to document the structural impossibility of sustaining even subsidized rents on current benefit levels.”

So, what would it mean to do this differently?

One option is simply to slow down and insist that human judgement remains central. That does not mean throwing away every tool, but it does mean treating scores and model outputs as prompts for conversation rather than verdicts. If a model says a tenancy is marginal, the question should be “what support or changes would make this workable?” not “how do we refuse this application?” That kind of stance requires resources, political backing, and a willingness to accept calculated risks.

Another is to turn the gaze of these systems upstream. Affordability tools could be used to document the structural impossibility of sustaining even subsidized rents on current benefit levels. Aggregated data could show, with painful clarity, how many young single people are locked out of housing by policy design, how often the bedroom tax collides with local stock profiles, how thin the margin really is between survival and failure. Instead of using data to ration scarcity, social landlords could use it to support collective claims for better welfare and rent setting.

There are also basic governance questions that cannot be dodged. Tenants should know what data are collected, how long they are kept, who they are shared with, and when decisions are influenced by algorithmic tools. They should have clear routes to challenge decisions and correct errors. Regulators need to pay attention not only to overt discrimination but also to the quieter ways in which past inequalities are baked into models that treat historical arrears as neutral indicators of individual risk.

Social housing is not just another market. If the sector drifts fully into the same scoring logics that govern credit cards and private letting platforms, the “ordinal social tenant” will become just another line in a portfolio spreadsheet. The promise of social housing has always been that it can offer some insulation from risk, not simply subject people to a more legible and data-intensive version of it.

The technologies are not going away. Open banking feeds, credit models, and affordability calculators will almost certainly become more common, not less. The real choice is whether they are used to further narrow the gates of social housing or to expose and contest the conditions that make those gates so tight in the first place.

Author’s note: For a more detailed analysis of the research summarized here, see Alison Wallace, David Beer, Roger Burrows, Alexandra Ciocănel and James Cussens, “Data and Automation in Pre-tenancy Affordability Checks In Social Housing” (Centre for Housing Policy, University of York, 2024).

Citation

Roger Burrows, “Living by the Score in Social Housing,” PLATFORM, December 15, 2025.