In England, a pilot born out of Covid urgency shows that in Stockport, housing homeless people in properties acquired by a social investor costs less than temporary accommodation. The Everyone In program mobilized private capital to purchase 528 homes in less than five years; at Stockport Homes, the model was estimated to cost more than 70% less than temporary accommodation. The final evaluation published on April 20, 2026 by PERU with partners including the Centre for Homelessness Impact documents this result, but also the fragility that accompanies it: leases expiring in 2030, with no permanence solution yet guaranteed.

The essentials

  • The pilot produced cost savings or cost neutrality in certain cases, notably in Stockport, without establishing a general rule about private capital. The evaluation identifies exit and refinancing risks that must be planned before 2030.
  • Everyone In mobilized an initial £50 million, then an additional £138 million, to acquire 528 homes by March 2025; in Stockport, the model costs between 12 and 22% less than the ordinary private rental market, and more than 70% less than temporary accommodation (PERU and partners, 2026).
  • One of the models relies on a social investor who purchases properties and leases them to an operator, with refinancing envisaged at the end of the lease. Mixed financing distributes risks, while leaving exit and continuity risks in place.
  • The 2030 deadline poses a real risk: without refinancing or renewal, expiring leases create a risk of disruption for affected tenants.
  • The challenge for European policymakers is to distinguish what constitutes genuine savings from what constitutes deferred savings.

Emergency accommodation is expensive, and it’s measurable

Emergency accommodation is an expense that renews itself without ever producing an asset. A night in a social hotel, boarding house, or shelter mobilizes staff, management contracts, administrative fees, and leaves people in an instability that worsens health, employment, and integration problems. Public costs accumulate without building capital.

This observation led the British government to launch Everyone In in 2020 in the urgency of the first lockdown. The initial objective was straightforward: get homeless people sheltered from the virus. But several actors saw in this window an opportunity to experiment with something different: buy properties with private capital, assign them to homeless people, and measure what it actually costs.

The evaluation conducted by PERU with notably the Centre for Homelessness Impact, whose final report is dated December 2025 and the executive summary March 2026, establishes precise results: For Stockport Homes, the evaluation estimates a cost 12 to 22% lower than comparable private rents. Compared to temporary accommodation, the gap exceeds 70%. This figure carries an idea formulated by the pilot: in the Stockport case studied, the model based on leases of properties acquired by an investor was estimated to be less costly than certain alternative accommodations.

How the structure actually works

The Everyone In mechanism deserves to be described precisely, because it explains both the performance and the vulnerability of the model.

A social investor, in this case a non-profit organization backed by impact funds, raises capital to purchase properties on the open market. These properties are then rented to homeless or severely precarious people, at a rent compatible with their resources. Rents and aid linked to complex needs finance costs and returns, while initial capital can be reimbursed through refinancing at the end of the lease in some models.

The initial financing, £50 million, came from a combination of public funds and impact investors. The pilot’s £50 million attracted an additional £138 million, and 528 homes had been acquired in England by March 2025. At each stage, property and operational risk is partly shared between the social investor and the housing provider, while the public sector provides legitimacy, approvals, and part of the co-financing.

This division of roles is the model’s strength. It is also its structural limit. Because the leases that organize this distribution have a fixed duration: most expire around 2030. At that date, two paths open. Either the investor refinances, seeking new capital in the markets, capitalizing on the accumulated portfolio, demonstrating that the model is economically viable over the long term.

Or the structure collapses, and in case of lease termination, alternative scenarios include portfolio transfer to a social investor or refinancing. Lease expiration creates a risk to the stability of certain households when housing is not permanent.

The lessons from Stockport and their limits

Stockport is a useful laboratory, not a directly exportable model. The municipality brings together several favorable conditions: a less tight property market than London, engaged local authorities, a tradition of partnership between the social sector and impact investors. These conditions are not universal.

The Centre for Homelessness Impact evaluation emphasizes that the 70% savings compared to temporary accommodation are calculated on a precise scope, at a precise moment. The comparison includes direct costs, rent, management, social support. Sustainability after 2030 depends on securing refinancing or a renewal mechanism, dependence on market conditions. The evaluation currently estimates neutral costs or savings, without establishing certain causal savings.

This point matters for policymakers looking at this model from other European countries. The Netherlands, Finland, and Austria have social housing stock far more developed than the United Kingdom, with long-term financing mechanisms already proven. Local coordination of social services remains a common challenge to most European systems, regardless of housing stock maturity. The Everyone In model interests these countries less for its financing architecture than for its ability to mobilize resources quickly where public stock is insufficient.

In France, Germany, Italy, and Spain, the shortage of affordable housing is severe in major cities, and public budgets remain under pressure. The social impact model, which brings in private capital with returns conditional on social outcomes, is an avenue several governments are beginning to examine seriously, without yet having resolved the question of long-term refinancing.

The actors advancing the model

Everyone In worked because several actors accepted to share risk differently.

The Centre for Homelessness Impact plays the role of rigorous third-party evaluator: it measures results, publishes data, identifies selection biases, formulates limitations. This transparency about limitations is precisely what makes the model credible for subsequent investors. A pilot that publishes only its successes does not attract serious capital.

Local authorities, Stockport in the lead, provided what no pure private investor can produce alone: detailed knowledge of homeless people on the territory, a network of social workers, legitimacy to allocate housing according to vulnerability criteria rather than solvency. Without this local infrastructure, the model becomes an ordinary real estate operation dressed up in social action.

Impact funds accepted a below-market return over the lease term, in exchange for partial security on portfolio value and visibility of social results. This is an investment thesis based on model stability and long-term value of assets acquired in neighborhoods undergoing rehabilitation.

An enforceable municipal debt illustrates similar logic: making local public commitments verifiable and binding to create the conditions of trust that private capital seeks before entering a long-term project.

The 2030 bifurcation: genuine savings or deferred savings

The Everyone In model worked, the figures confirm it. The question concerns the sustainability of the savings realized: are they permanent or temporary?

If leases expire in 2030 without structured refinancing, two effects compound. The 2030 deadline risks creating a stability break for affected tenants. The evaluation identifies a continuity disruption risk for tenants, without quantifying or predicting added pressure on local accommodation. For Resonance, portfolio transfer to a responsible investor is the documented exit scenario. For SASC, refinancing is envisaged after approximately ten years.

The evaluation establishes favorable or neutral public cost results for the pilot, but leaves uncertainties about its sustainability after 2030.

This risk is documented in the Centre for Homelessness Impact evaluation. It outlines the condition for the model’s success, without invalidating it. For the savings to be real and not deferred, several elements must be in place before 2030.

The first is a pre-established refinancing mechanism, not improvised at expiration. This assumes that second-generation investors, pension funds, insurers, sovereign wealth funds, already have a legal and tax framework to enter into this type of asset with guarantees on lease duration and the social nature of occupation. Several European countries are working on “social impact bonds” that could play this role, but none yet has a sufficiently liquid and deep market to absorb portfolios of several hundred homes.

The second is a policy of portfolio appreciation. Properties acquired in 2020–2025 in popular neighborhoods have an asset value that can evolve. If this value increases, refinancing becomes easier. But price increases also make homes harder to maintain at social rent. The social investor must then choose between profitable exit and social mission, and this arbitrage, in current structures, is not always contractually framed.

The third is clear public commitment over duration. Local authorities that co-financed Everyone In need to know their partner will continue housing vulnerable people after 2030. A framework agreement with tenure protection clauses protects tenants without the local authority bearing the risk alone.

These three conditions can be met. They will not be met spontaneously. They require decisions today, on contracts, guarantees, regulations, so that 2030 is a transition, not a rupture.

Measures available to European governments

Several European countries are examining the Everyone In model with a view to scaling it up. The question is whether it can be replicated in tighter metropolises, with more expensive property markets and larger homeless populations.

The answer depends less on pilot size than on rule quality. A social investor who knows it can refinance in ten years on stable terms is prepared to enter a more expensive property market by accepting a lower initial return. An investor who does not know what will happen at lease termination will demand either higher returns or explicit public guarantees, which amounts to recreating through another channel the public spending one wanted to reduce.

The right level of regulation is one that sets refinancing rules in advance, without guaranteeing returns. It creates predictability without socializing losses. This is precisely what “social impact bond” type contractualizations seek to do, with results still limited in volume, but principles that several Northern European countries (Finland, the Netherlands, Denmark) are beginning to codify in their affordable housing legislation.

For governments wanting to draw inspiration from Everyone In, the evaluation recommends securing exit mechanisms before 2030, among other conditions for proper implementation. The pilot aimed for modest returns for investors and showed that some structures could work, without demonstrating short-term general profitability. Consolidating it over twenty or thirty years requires a contractual architecture that no one builds in urgency, and that everyone has interest in building before urgency returns.

Social innovation at local scale often follows this path: a pilot that proves the concept, then a decade during which the rule of the game stabilizes, before scaling up becomes possible. Everyone In is today at the hinge between pilot and rule of the game. What the United Kingdom decides between 2025 and 2030, on leases, refinancing, public guarantees, will determine whether this model becomes a European reference or remains a well-documented test whose main takeaway is its limitations.

Sources

  1. PERU, with partners including the Centre for Homelessness Impact, Everyone In Social Investment Pilot: Final Evaluation, December 2025 (executive summary March 2026), https://www.homelessnessimpact.org/publication/everyone-in-social-investment-pilot-final-evaluation