All posts
data-analysis

The Pivot: Ireland's Social Housing Is Now Majority Bought, Not Built

DHLGH's own reports show the state stopped being a primarily-builder of social housing at some point between 2018 and 2020. By Q3 2025, 57% of what is planned and 73% of what is completed is acquired from private developers. Nobody announced this.

The Viability Series · Part 10
Archa Intelligence··9 min read

DHLGH's own quarterly reports have been recording two versions of the same shift for years. The completed-units share crossed into turnkey-majority by 2018. The planned-pipeline share crossed in Q3 2020. By Q3 2025, 57% of what is planned and 73% of what is completed is turnkey: private developer construction acquired by an Approved Housing Body or local authority after the fact. Nobody announced this.

The state has not stopped delivering social housing. The Q3 2025 pipeline covers 44,613 units across 2,654 schemes, the biggest on record. What has stopped is the state building it. A turnkey unit is financed by a state capital programme (SHIP, CALF, CAS), designed and constructed by a private developer on the developer's own balance sheet, then acquired by the state on completion. The state is the buyer. The builder is someone else.

How is social housing being delivered now?

DHLGH's Q3 2025 Construction Status Report carries the full pipeline split.

Delivery categoryUnitsShare
Turnkey (CALF + SHIP + CAS)25,46857.1%
Direct-build (council + AHB construction)17,72439.7%
Renewal / regeneration1,4213.2%

Of the 20,538 units marked "Completed" in that same Q3 2025 register, 73.1% are turnkey. Of the 12,070 units marked "On Site," turnkey leads again. Pre-construction stages (Capital Appraisal, Pre-Planning, Pre-Tender Design) hold 7,546 direct-build units and zero turnkey units. That is structural, not a snapshot coincidence. Turnkey schemes skip the state's own design and tender stages because the developer has already done that work privately before the state commits.

When did the state stop being a builder?

The pipeline flipped in Q3 2020. Turnkey units rose from 14,029 (Q2 2020) to 15,556, an 11% quarter-on-quarter jump. Direct-build moved from 12,292 to 12,755 over the same window, a 4% rise. Both grew. Turnkey grew faster, and the gap between them widened. Q3 2020 was the first quarterly report in which turnkey crossed majority share of the planned pipeline (51.0% against 41.8% direct-build). Of the 21 quarterly reports since, turnkey has led direct-build in 17. The four exceptions are every quarter of 2022, during the CALF collapse covered below.

The crossover was not driven by the state pulling back from construction. Between Q3 2019 and Q3 2020, direct-build units grew 10% (11,607 to 12,755), slow but positive. Over the same year, turnkey grew 82% (8,544 to 15,556). This is turnkey acceleration, not direct-build contraction. Private developers and AHBs stepped in faster than the state stepped back.

The completed-units share tells a longer story. By Q4 2018, the first quarterly report with meaningful completion volumes, turnkey was already 54.8% of completed units. It rose to 63.4% by Q4 2020, 68.9% by Q4 2022, and 73.1% by Q3 2025. The pipeline crossover in Q3 2020 is the quarter the structural shift became visible in PLANNED deliveries. Completed deliveries had been turnkey-majority earlier.

Over the full 8.5-year series, turnkey share roughly doubled (28.6% in Q1 2017 to 57.1% in Q3 2025) while direct-build share fell by 24 percentage points (63.6% to 39.7%). The total pipeline more than tripled in absolute size. The shift is not about the state building fewer units. It is about the composition of what is being delivered.

How broad is the shift across the AHB sector?

The pattern is not concentrated in one dominant AHB. It is the sector.

AHBQ4 2018 turnkey shareQ3 2025 turnkey share
Clúid Housing68.2%89.1%
Respond83.0%96.8%
Tuath Housing Association80.4%94.0%
Co-Operative Housing Ireland65.7%99.0%
Oaklee Housing61.1%95.3%
Circle Voluntary Housing15.9%86.4%
The Iveagh Trust70.8%100%

Seven AHBs with pipelines of 50 or more units in Q3 2025 operate in the 90-100% turnkey band. Those seven represent 87.5% of all AHB units in the register. No significant AHB still functions primarily as a builder. Circle Voluntary Housing is the most striking conversion: 15.9% turnkey in 2018 to 86.4% in 2025, a 70-point shift in seven years. When every significant actor in a sector converges on the same strategy, that is not market positioning. That is the structure of the system.

What did the 2022 shock reveal?

2022 tested whether the shift was cyclical or structural. Russia's invasion of Ukraine in February reset European construction cost inflation. ECB rate hikes began in July. Between Q4 2021 and Q4 2022, every social housing delivery channel contracted.

ProgrammeQ4 2021 unitsQ4 2022 unitsChange
CALF Turnkey (AHB)15,0169,495−37%
SHIP Turnkey (council)3,4181,925−44%
Direct-build14,85311,938−20%
Total pipeline36,68128,037−24%

Every channel fell. Turnkey took the biggest proportional hit because AHB and council turnkey prices are set by the private market, and private developers could not hold their bid prices as construction costs rose 23% over 2021-2023 (per SCSI index readings). Direct-build contracted less because council capital budgets absorbed some of the shock. But the 20% direct-build drop rules out any reading that direct-build was insulated.

Then came the recovery. By Q4 2023, CALF Turnkey was back to 14,109 units. By Q4 2024, it exceeded its pre-shock peak at 18,805 units. By Q3 2025 it reached 20,655. The state paid what the market asked. Turnkey prices reset higher, AHB and council procurement resumed, and the pipeline returned. There is no reading of the 2022 data where the state tested whether to build directly at scale and decided not to. The state tested whether to pause acquisition. The answer was no.

Why this matters

Policy and media discourse still talks about social housing delivery in construction language: units completed, targets met, pace of build. DHLGH's data has been telling a different story for at least five years. The state is not primarily a builder of social housing. It is a buyer. The mechanism that determines how many social homes arrive next year is not public-sector construction capacity. It is private developer balance-sheet capacity, private financing conditions, and the rate at which AHBs can close acquisitions financed by HFA loans and CALF drawdowns.

Two implications follow. First, the price-setting power for social housing sits with the private market. The 2022 CALF collapse and its 2023-24 recovery showed this cleanly: when private delivery prices rose, state acquisition paused, then resumed at the new prices. At scale, the state is a price-taker.

Second, direct-build capacity has grown in absolute terms (6,335 units in Q1 2017 to 17,724 in Q3 2025) but has fallen from 64% to 40% of the pipeline mix. Reversing the composition would require substantial direct-build expansion, turnkey contraction, or both. Neither shows in the current data.

What this means before you submit

If you are an AHB executive, AHB pipeline volume now depends on private developer delivery capacity. The 2022 shock is a worked example of how quickly the turnkey pipeline stalls when financing conditions shift, and of how long recovery takes (four quarters to return, six to exceed the prior peak).

If you are a council housing officer, SHIP Turnkey is now the dominant delivery programme. SHIP Construction is a minority of council delivery even at council scale. Framework design decisions take on different weight when the framework is the acquisition channel, not the construction channel.

If you are a planning consultant, a residential scheme with plausible AHB off-take sits in a different financing environment than a scheme selling only to the open market. The DHLGH register publishes the scheme-level pipeline quarterly. Your client's scheme either sits in it or does not.

If you are reading national supply numbers, state delivery and private market delivery are not separate tracks in this data. They are the same track, with different end-buyers. Any supply-volume forecast depends on private developer capacity and private financing conditions whether the end-buyer is an individual household or an AHB.

The 32-quarter DHLGH series analysed here is now queryable on Archa. For any residential site or scheme, we can tell you whether DHLGH has it in the pipeline, which AHB is acquiring, what stage it is at, and how the surrounding LA's turnkey share has moved over time. The pattern documented in this article is specific evidence. The same read is available for any site in minutes.

Methodology

The data source is the DHLGH Social Housing Construction Status Report, published quarterly as CC-BY-SA-4.0 open data at data.gov.ie. All 32 publicly available quarterly reports from Q1 2017 through Q3 2025 have been ingested and exposed through Archa's DHLGH schemes dataset. Q2 2017 is absent from the open data catalogue.

Q3 2025 figures are post-deduplication. The raw CSV contains roughly 10% phase-row duplicates that collapse by the unique composite key (report quarter, funding programme, LA, scheme name). Integrated figures: 2,654 schemes, 44,613 units. Turnkey classification matches "Turnkey" in the Funding Programme column (CALF, SHIP, CAS variants). Direct-build matches "Construction" excluding Turnkey and Renewal.

AHB normalisation uses uppercase matching with legal-suffix stripping and accent-stripping. A small number of Q3 2025 rows have mangled Unicode in the source CSV which produces a second normalised Tuath name variant; the Tuath figures combine both.

Per-scheme completion data is point-in-time: the cumulative-completions figure in each quarterly report includes schemes that reached "Completed" in any prior quarter and are still in the current register. Schemes fully handed over may not appear in later reports, which means year-over-year "share of completions" readings are comparable but annual flow figures are not directly recoverable from this source.

Sources

  • DHLGH Social Housing Construction Status Report Q3 2025 data.gov.ie, published 19 December 2025, CC-BY-SA-4.0
  • Historical CSR reports Q1 2017 through Q2 2025 DHLGH open data catalogue
  • Archa DHLGH schemes dataset public MCP-accessible via Archa's query_data tool; refreshed monthly
  • SCSI construction cost index Society of Chartered Surveyors Ireland, H1 2023 report (for 2021-2023 construction cost inflation context)

Figures deliberately excluded

This article focuses on the state's delivery mechanism. What is not in this article, specifically who the private developers are on the other side of these transactions, where the land comes from, and who owns the finance behind it, is the subject of Parts 11 through 15 of this series. Each of those layers has enough specific evidence to stand as its own article. Forcing them into this one would dilute every finding.


The Viability Series — data-driven investigation into why Ireland's housing delivery system is structurally broken.

  1. The Viability Ceiling — the economics are broken outside Dublin
  2. The Silent Lapse — permissions are dying, fewer than one in twenty developers file to extend
  3. The RZLT Paradox — the state taxes the problem, 85% of recycled sites change hands
  4. Permitted but Caught Out — six policy changes hit 170,000 permitted homes mid-flight
  5. Five Systems, No Feedback — five state systems, one site, zero information flow
  6. The Optionality Gap — the options trader and the committed builder
  7. The Eye of the Needle — what Ireland can still build
  8. The Scale Trap — too small to finance, too big to survive
  9. Where Ireland Actually Builds — the geography of what gets through
  10. The Pivot — Ireland's social housing is now majority bought, not built (this article)