The companion piece to this one, The arithmetic of not building, set out what the Irish State currently spends each year on housing solutions that produce no permanent public asset. Approximately €886 million in 2024 across emergency accommodation and HAP, flowing principally to private providers and private landlords. The piece named the gap between what the State spends on not-building and what it would have produced for the same money if it had been building.

This piece runs the working model on the question that follows. How quickly could the State, supported by State borrowing on currently-available terms, move every person currently in emergency accommodation into permanent local-authority housing, while continuing to fund emergency accommodation during the transition for as long as it takes for the new permanent capacity to come online? What would the model cost? What would the financing profile look like across the transition years and afterwards? And how does the financing-model total compare with the cost of continuing the status quo?

The model below is calibrated to the verified empirical figures used in the arithmetic-of-not-building piece, with current Irish sovereign-bond and EIB-HFA financing terms applied. It is intended to be defensible against pushback at every numerical step. The conclusions are read directly from the arithmetic. The political constraints on actually adopting the model are described in the closing section, because the model itself is operationally and financially feasible, and the reasons it has not happened are political rather than technical.

The givens

The empirical inputs to the model are all verified from published Government and EIB sources.

The population to be housed. The Department of Housing's March 2026 monthly homelessness report, published on 24 April 2026 and the most recent available at the time of writing, recorded 17,517 people in emergency accommodation: 11,946 adults and 5,571 children, distributed across 2,659 families and approximately 7,950 single adults. The total household figure is approximately 10,600, treating each family as one household and each single adult as one household. This is the figure the model uses for permanent-housing-unit demand. The actual figure may be slightly higher or lower depending on how shared-accommodation arrangements convert into permanent-housing-unit allocations. The headline figure has risen across each successive monthly report since at least mid-2024, with the March 2026 figure representing a 12.5 percent annual increase, child homelessness up 19 percent year-on-year, and family homelessness up more than 18 percent. The trend is continuing. The figure used in the model below should be read as a floor on demand rather than a ceiling.

The cost per unit. Department of Housing data on local-authority direct-build costs in 2024 indicate an all-in average cost of approximately €318,365 nationally and approximately €392,975 in the Greater Dublin Area. The 'all-in' figure includes site purchase, design fees, utility connections, surveys, and the broader development costs. Because the bulk of homelessness is concentrated in Dublin and other urban areas, the weighted average for housing the homeless population is closer to the GDA cost than the national average. The model uses €370,000 per unit as the working blended figure.

The current State spend on the status quo. Approximately €361 million on emergency accommodation in 2024 (with private operators receiving approximately 70 percent of that), plus €525 million on HAP in Budget 2024 (€482 million in Budget 2025). The two streams together come to approximately €886 million per year, with additional smaller streams (RAS, leasing, supplements) bringing the total flowing into the private rental and emergency-accommodation sectors to approximately €1 billion per year on a conservative central estimate.

The cost of borrowing. Ireland's 10-year sovereign bond yield in 2026 has been trading in the 3.0 to 3.4 percent range, with recent values around 3.25 percent. The European Investment Bank, in October 2025, launched a €400 million loan facility with the Housing Finance Agency to provide 40-year fixed-rate finance to Approved Housing Bodies and Local Authorities, supporting the delivery of social and affordable homes at rents at least 25 percent below market levels. The EIB Group has separately announced doubling its broader housing financing to €6 billion across Europe in 2026. The financing architecture for substantially expanded direct-build social housing in Ireland exists. The model uses 3.25 percent as a working blended cost-of-borrowing figure, reflecting Irish sovereign rates and the favourable EIB-HFA terms that should be available for this scale of programme.

The construction-sector capacity. Local-authority direct-build output peaked at approximately 9,000 units in 1975. The current Government target for total social-housing delivery is 10,000 units per year across all delivery mechanisms, of which only a fraction is direct local-authority build. As of Q3 2025, only 30 percent of the social-homes target had been met. Sinn Féin's published housing plan proposes 25,000 social-and-affordable homes per year. The realistic delivery rate for a substantially-expanded direct-build programme depends on the speed of capacity build-out (engineering offices, procurement processes, construction-sector labour availability, planning-and-land-acquisition pipelines).

The model

The model runs two tracks in parallel during the transition period.

Track A is the continuation of current emergency-accommodation and HAP expenditure, providing housing-of-last-resort to the homeless population for as long as it takes for permanent housing to come online. Track A spend declines progressively as households are moved into permanent housing.

Track B is the direct-build local-authority programme, financed by State borrowing, producing permanent public-asset housing units at approximately €370,000 per unit. Track B units come online progressively as construction completes.

The model assumes a realistic ramped delivery profile, calibrated to the construction-sector capacity discussion above:

  • Year 1: 2,000 units delivered (programme launch and capacity build-out)
  • Year 2: 4,000 units delivered (programme reaches mature delivery rate)
  • Year 3: 6,000 units delivered (full capacity, with construction-sector adjustment)
  • Year 4: 5,000 units delivered (continuing to address waiting-list and replacement demand)
  • Year 5+: 4,000-5,000 units per year, sustained

Cumulative units delivered across the transition: 12,000 by end of Year 3, sufficient to house the entire current emergency-accommodation population (approximately 10,600 households) with capacity to spare for the wider social-housing waiting list.

The capital cost

Capital cost across the four-year transition window:

  • Year 1: 2,000 units × €370,000 = €740 million
  • Year 2: 4,000 units × €370,000 = €1.48 billion
  • Year 3: 6,000 units × €370,000 = €2.22 billion
  • Year 4: 5,000 units × €370,000 = €1.85 billion

Cumulative capital expenditure: approximately €6.3 billion across the four years, financed by State borrowing.

This is a substantial capital programme. It is not, by international comparative standards, an unprecedented one. Ireland's National Development Plan annual capital budget runs at approximately €13-15 billion per year across all sectors. The €6.3 billion housing-transition capital, spread across four years, represents approximately 10-12 percent of total annual national capital expenditure. It is a reorientation of priorities, not a doubling of the State's overall capital programme.

The capital is borrowed, not paid out of current Exchequer revenue. The borrowing creates an asset on the State's balance sheet (the housing units, in public ownership) that substantially offsets the new liability. International accounting conventions for public-housing capital expenditure are well-established and the structure is not novel.

The financing profile

At a blended cost-of-borrowing of 3.25 percent and a 40-year amortisation schedule (matching the EIB-HFA term structure), the annual debt service on €6.3 billion is approximately €265 million per year for 40 years.

Rental income from the new units offsets a substantial portion of the debt service. Local-authority differential-rent arrangements typically produce average rents in the €350 to €500 per month range, depending on tenant income and household composition. At an average of €400 per month per unit across 10,000+ social-housing units in the new stock, gross annual rental income is approximately €48 million per year, before management and maintenance costs. After operating costs (typically 25-30 percent of gross rental income for local-authority stock), net annual rental income is approximately €33-35 million per year.

Net annual cost of the financing model after rental income is therefore approximately €230 million per year for 40 years. This is the net annual cost of the programme once it is fully operational.

The transition profile, year by year

The combined annual spend across both tracks during the transition:

Year 1. Track A spend continues at full €886 million (no households yet moved). Track B spend incremental debt service approximately €25 million (year-one debt only). Combined operational annual cost approximately €911 million. Capital expenditure of €740 million is borrowed.

Year 2. Track A spend declines to approximately €705 million as 2,000 households are moved into permanent housing (~20 percent reduction). Track B incremental debt service rises to approximately €85 million (cumulative debt). Combined operational annual cost approximately €790 million. Capital expenditure of €1.48 billion is borrowed.

Year 3. Track A spend declines to approximately €355 million as 6,000 households are moved (~60 percent reduction). Track B incremental debt service rises to approximately €175 million. Combined operational annual cost approximately €530 million. Capital expenditure of €2.22 billion is borrowed.

Year 4. Track A spend declines to approximately €88 million for residual emergency-accommodation needs (~10 percent of original level). Track B debt service approximately €265 million. Combined operational annual cost approximately €355 million. Capital expenditure of €1.85 billion is borrowed.

Year 5 onwards. Track A residual approximately €88 million per year. Track B debt service €265 million per year for the remaining 36 years of the amortisation, partially offset by approximately €33 million net rental income. Combined operational annual cost approximately €320 million per year for 36 years, declining toward Track A residual only after year 40.

The comparison

Status quo (Track A only): €886 million per year, indefinitely. No permanent public assets created. Homeless population persists at current levels or grows. Cumulative spend across 40 years at status-quo level: approximately €35 billion. End-of-period assets: zero.

Financing model (Track A + Track B): peak combined annual cost in Year 1 approximately €911 million (essentially the same as status quo). Annual cost declines progressively through the transition. From Year 5 onwards, annual operating cost is approximately €320 million, sustained for 36 years thereafter. Cumulative spend across 40 years: approximately €13-14 billion (€2.5 billion across the transition, plus €265 million per year × 36 years debt service, less rental income). End-of-period assets: 12,000+ permanent local-authority housing units, valued conservatively at €4-5 billion in 2026 prices and likely substantially more in 2066 terms after appreciation.

The financing-model total is approximately 60 percent lower than status-quo total spend across the same 40-year period. The financing-model end-of-period asset position is approximately €4-5 billion higher than the status-quo end-of-period asset position (which is zero).

The financing model housed approximately 10,600 households permanently. The status quo housed nobody permanently.

The financing model produced ongoing rental income to the State of approximately €33 million per year, indefinitely, after the debt is repaid. The status quo produces no rental income to the State.

The financing model ended modern Irish homelessness for the cohort currently in emergency accommodation by approximately end of Year 3 of the programme. The status quo does not end modern Irish homelessness on any timescale visible in the published Government forecasts.

The capacity constraints

The model assumes the construction sector and the local-authority engineering offices can scale to the delivery rates described. This is the principal practical constraint. Three observations on the constraint.

First, the construction-sector capacity question is, on the empirical record, addressable through deliberate policy. Ireland's construction sector is currently operating at near-capacity for private housing. Adding 6,000 social-housing units per year on top of current activity would require either expanding the construction-sector workforce (through training pipelines, immigration of construction trades, and improved working conditions to attract and retain workers) or reallocating existing construction-sector capacity from private housing to social housing. Both approaches are operationally available. The reallocation approach is, on the available record, more politically straightforward in the short term because it does not require the longer-lead-time workforce expansion. The expansion approach is, on the same record, more sustainable in the medium term because the underlying capacity question affects all Irish housing delivery, not just social housing.

Second, the local-authority engineering and architectural offices have substantially atrophied since the 1970s peak. Restoring direct-build capacity at the local-authority level requires re-staffing, re-training, and procurement-process redesign. This is a multi-year project rather than a single-year project. The Year 1 delivery figure of 2,000 units in the model accounts for this build-out time. The 6,000-units-per-year figure in Year 3 assumes that local-authority capacity has been substantially restored by that point.

Third, the planning-and-land-acquisition pipeline requires Government attention to compress. Current planning timelines for large social-housing projects can run to two or three years for site identification, planning permission, design completion, and construction-readiness. The 1970s programmes operated under faster timelines partly because the State had substantial existing land banks and partly because the planning architecture of that period was less constrained by the post-1990s regulatory layers. A revived direct-build programme would benefit from statutory measures to accelerate planning and site-acquisition for social-housing projects specifically. Such measures are technically straightforward and have been implemented in comparable jurisdictions, including the post-2008 Vienna and post-2010 Helsinki social-housing programmes.

The capacity constraints are real. They are not insurmountable. The model assumes a realistic ramp that accommodates the constraints rather than ignoring them.

Why this is not happening

The model above is operationally and financially feasible. The capital is available at favourable rates from the EIB-HFA partnership and from sovereign borrowing. The construction-sector capacity question is addressable. The institutional knowledge for direct-build local-authority housing exists in residual form and can be deliberately restored. The end-of-period public asset position is substantially better than the status quo. The cumulative cost across the relevant time horizon is substantially lower than the status quo.

The model is not happening because the political-coalition pressure for it has not built to the level the existing political settlement would have to feel before adopting it. The constituencies that benefit from the status quo (the construction-industry federation, the institutional-landlord sector, the private-developer-led delivery model, the Departments and Civil Service teams calibrated to the existing architecture) have substantially more access to the political coalition than the constituencies that bear the cost (households in emergency accommodation, families on the social-housing waiting list, communities affected by the broader housing crisis).

This is the same pattern the rest of the political-literacy series on this site has been documenting. The model exists. The financing exists. The capacity is recoverable. The political will is the constraint.

The question that produced this analysis was: how quickly could the State, supported by State borrowings, move the homeless to housed while financing both tracks?

The arithmetic answers: approximately three years to house everyone currently in emergency accommodation in permanent local-authority housing, financed by approximately €6.3 billion in State borrowing over four years, with the cumulative-40-year cost of the programme substantially lower than the cumulative-40-year cost of the status quo. The political answer is: the State could do this within the next Government term if the political coalition decided to. It has not. The reasons it has not are not in the arithmetic. The reasons are visible in the lobbying-register data, in the Ireland.Inc framing, and in the structural pattern across multiple domains of Irish public policy that the rest of the series has been mapping.

The arithmetic is the easy part. The political work follows.

Related on this site

Plus the full Political Literacy archive.