Ireland Has a Viability Ceiling, Not a Planning Bottleneck
218 residential extension of duration applications in Q1 2026 — above the Covid peak. The acceleration started months earlier than anyone noticed.
The Viability Series · Part 1In Q1 2026, Irish developers filed 218 residential extension of duration applications — requests to keep an expiring planning permission alive because the holder cannot yet build. That is now above Q1 2021, when construction sites were physically closed under Covid restrictions.
These applications are a forward indicator. They capture viability stress roughly six to twelve months before it shows up in completions or prices. And the quarterly series shows this one started much earlier than the headline suggests: Q1 2025 produced 82 residential extensions. Q2: 84. Q3: 112. Q4: 190. By Q4 2025, extensions had already exceeded the Covid peak. Q1 2026 did not mark a spike. It confirmed a trend that had been compounding for three quarters.
When stalled developments are cross-referenced against spatial data, the cause is clear: 75% face at least one infrastructure or ground-condition cost burden — unsewered sites, pyrite/mica risk zones, or both. Fewer than 3% sit in a flood zone. Planning constraints are not what is stopping these homes from being built. Cost is.
What do extension of duration applications actually measure?
An extension of duration application is filed when a developer holds a valid planning permission but cannot activate it before it expires. To qualify, the developer must demonstrate that substantial works have not been completed. A developer filing one is formally stating: I have the permission. I cannot finish. Give me more time.
| Q1 2020 | 142 |
| Q1 2021 | 209 |
| Q1 2022 | 166 |
| Q1 2023 | 82 |
| Q1 2024 | 93 |
| Q1 2025 | 82 |
| Q1 2026 | 218 |
Three councils (DCC, Cork City, and Offaly) do not publish extension of duration applications. All totals here are lower bounds. But the same councils are missing in every year, so the trend is reliable even if the absolute numbers are conservative.
Four elevations in five years. Each one has a distinct cause. Q1 2021: Covid site closures. Q1 2022: Ukraine-driven construction cost inflation, which the SCSI recorded at 14% for the full year. Q1 2025-2026: a structural viability gap that has been building since 2022 and is now producing more extensions than a national lockdown.
The 2026 surge is also disproportionately residential. Historically, housing accounts for roughly 65% of all extension applications. In Q1 2026, that share jumped to 79%. Whatever is driving this, it is hitting housing harder than other development types.
When did the acceleration actually start?
The original version of this article showed only Q1 snapshots. The quarterly data tells a different story.
| Q1 '25 | 82 |
| Q2 '25 | 84 |
| Q3 '25 | 112 |
| Q4 '25 | 190 |
| Q1 '26 | 218 |
A wave of planning permissions granted in 2020 and 2021, during the Covid construction slowdown, are now reaching their five-year expiry dates. Developers who could not build during lockdown, and could not afford to build afterwards, are running out of clock. Meath makes this mechanism visible.
Why is Meath filing more extensions than any other council?
Meath filed 53 residential extension of duration applications in Q1 2026 — more than any other council, including Cork County (10) and Donegal (15). A year earlier, Meath filed 7.
All 53 come from distinct applicants across eight planning agents, with no batching or single-firm concentration. The spike began in Q4 2025 (51 extensions from 50 unique applicants) and held through Q1 2026. This is 53 separate households hitting the same constraint at the same time.
Eighty-three percent are new-build dwellings. Forty-six were filed in January — consistent with five-year permissions granted in January 2021 reaching expiry. Covid-era permissions — overwhelmingly one-off houses — hitting their statutory deadline.
Meath's situation is distinct because of the cost stack.
All extensions with available ground-condition data fall within the pyrite risk zone. Pyrite-resistant construction requires engineered fill and enhanced foundations, adding cost that did not exist in the original project budget. Seventy-four percent are on unsewered sites, meaning each needs its own septic tank and percolation area — an additional cost that varies with ground conditions.
These are houses in the commuter belt that people got permission to build and then could not afford to deliver. Ninety-two of Meath's 141 total extensions since January 2025 are in census areas classified as "rural with high urban influence" — the peri-urban ring where Dublin's housing demand meets rural construction economics.
Even excluding Meath entirely, the rest of the country more than doubled: from 75 residential extensions in Q1 2025 to 165 in Q1 2026.
How wide is the viability gap outside Dublin?
Outside Dublin, the median new-build sale price is at or below the cost of building it. The SCSI has been reporting this. Developers have been citing it in planning documents. Nobody has measured it directly from transaction data, matching what buyers actually pay against the approximate cost of delivering.
The SCSI estimates the all-in cost of delivering a standard three-bedroom house outside Dublin (build cost, land, fees, VAT, and developer margin) at approximately €350,000 to €420,000 using conventional construction methods. Modular and timber-frame building may eventually move that floor, but the benchmark reflects what most developers are actually building with today. Below the lower end of that range, private delivery without subsidy is generally unviable. The Property Price Register shows what new homes are actually selling for.
Below this threshold, private residential delivery without subsidy is generally unviable. The benchmark covers build cost, land, fees, VAT, and developer margin for a standard 3-bed house.
In rural and small-town Ireland, the median new-build sale price is €321,586. The estimated delivery cost floor is approximately €350,000. The market price of a new home is lower than the cost to build it. That is not a delay problem. It is a non-delivery condition. When the sale price sits below the cost of delivery, supply does not slow. It stops.
Sixty-nine percent of new homes transacting in rural Ireland in 2025 sold below the approximate delivery cost threshold. Those homes were built, but the economics that made them viable in 2018 or 2019 are gone. Construction costs have risen approximately 39% since early 2020, and rural sale prices have not kept pace.
The threshold itself is not fragile. At €300,000, 35% of rural new builds are still below it. At €375,000, 77%. At €400,000, 85%. The regional ordering holds across the full range. The specific number matters less than the gradient: the further from Dublin, the wider the gap between what it costs to build and what buyers can pay.
Is the gap narrowing?
The first quarter of 2026 brought 2,255 new-build PPR transactions — enough to compare on a like-for-like basis with Q1 2025.
Prices are rising faster outside Dublin than inside it. The commuter belt median jumped from €374,449 in Q1 2025 to €405,286 in Q1 2026 — a gain of over €30,000 in a single year. Rural areas have gained €65,000 over three years (Q1 2023 to Q1 2026), from €265,000 to €330,396. Regional cities crossed the €350,000 median threshold for the first time in Q2 2025.
But the extension of duration data says the gap is not closing fast enough. Prices are moving toward the floor. The floor is also moving — construction costs have not stopped rising. The viability gap is narrowing from the demand side while potentially widening from the cost side, and the extension spike says the cost side is winning.
Are developers being taxed on land they cannot build on?
The Residential Zoned Land Tax was introduced to penalise landowners who sit on zoned, serviced land without developing it. The tax is 3% of site value, applied annually. The intent is to discourage land hoarding.
Ninety-one extension of duration applications filed since January 2025 sit on RZLT-liable land, representing 2,904 permitted housing units across 327.6 hectares (lower bound — coverage of zoning data varies by council).
DLR alone accounts for 656 units on 109.7 hectares. Cork has 447 units. Fingal has 242.
These developers obtained planning permission, could not make the numbers work, and are now paying an annual tax for the privilege of holding a permission they cannot activate. The RZLT was designed to unlock idle land. On these sites, the land is not idle by choice — many of the same sites face the infrastructure and ground-condition cost burdens documented below. The tax compounds a viability problem it was not designed to address. Part 3 of this series investigates where that land ends up.
Where do stalled developments actually cluster?
If planning constraints were the binding issue, stalled developments would cluster around flood zones, heritage designations, or ecological protections. They do not.
The extension of duration data, cross-referenced with 30 spatial layers, allows a direct test. Only 2.2% of extension applications sit in a mapped flood zone. Heritage designations (NIAH, SMR, Record of Protected Structures) appear at or below the baseline rate for all planning applications. Natura 2000 sites, protected species — none are overrepresented among stalled developments.
Cost is overrepresented. Sixty percent of extension applications are on unsewered sites (versus 48.6% of all applications). Forty-four percent are in mica or pyrite risk zones (versus 32.4%). Only 25% face none of these cost burdens.
Stalled development clusters around cost, not regulation. Missing mains sewerage. Defective-blocks risk zones. The cumulative burden of building on sites where ground conditions add to an already unviable cost base.
The construction industry has faced significant challenges recently due to a convergence of inflation, material shortages, labour shortages, rising energy prices, and geopolitical conflicts. Inflation has driven up the costs of materials, making it difficult for companies to stay within budget and plan effectively, making it unfeasible to carry out some projects.
This is from a planning consultant's report for a 24-unit scheme in Dalkey, granted in 2019, still unbuilt five years later. The language is specific: not difficult, not challenging. Unfeasible. The planning system's own paper trail documents the viability crisis in the developers' own words.
Why can't planning reform fix this on its own?
Planning reform is producing measurable gains. Third-party submissions per application dropped from 0.34 per app in January 2024 to 0.27 in January 2025 (measured across all councils; seven major councils report zero submissions, so the true rate among reporting councils is higher). Average processing times for decided applications in Q1 have varied between 85 and 100 days since 2020, with the median consistently at 62-64 days — the mean is pulled up by applications that go through the further information process.
None of this is closing the viability gap.
The 69% of rural new-build transactions below the delivery cost floor is a market structure statistic, not a planning one. No amount of planning simplification changes the fact that a three-bedroom house in Roscommon costs approximately €350,000 to deliver and a buyer there typically cannot pay more than €320,000 for it.
Archa's data identifies approximately 70 residential schemes with at least 20 permitted units, no linked commencement notice, and planning expiry within 18 months, representing over 5,000 recorded units. Kildare alone accounts for 11 schemes and 1,679 units, a third of the total. The commuter belt holds 58% of all at-risk pipeline units. These sites have approval. They are waiting for the economics to work.
Planning reform has reduced friction, and faster processing does cut holding costs on developer finance. But it has not changed the underlying economics. A three-month faster decision does not close a €30,000 per-unit gap between sale price and delivery cost.
The Q1 2026 extension spike reached its current level before the Iran war, which began on 28 February, had any measurable effect on construction costs. Energy price shocks take six to twelve months to reach tender prices. Q3 2026 is the earliest point at which a war-driven cost spike would plausibly appear in extension volumes. The Q4 2025 baseline is 190 residential extensions. If Q2 or Q3 2026 exceeds that, the fourth shock has become the primary driver.
What does this mean before you submit a scheme?
The viability ceiling is not an abstract market observation. It is a project-level input that changes what you build, where, and whether you start.
Dublin metro is the one region where the numbers work. The median new-build sale price (€449,339 in 2025, €440,529 in Q1 2026) clears the delivery cost floor with margin. Extension of duration applications are low. The constraint here is land and permissions, not viability.
The commuter belt has improved but not uniformly. The median rose from €383,260 to €405,286 between 2025 and Q1 2026, and the share below €350,000 dropped from 31% to 21%. But Louth (median €348,018, 53% below threshold) functions more like a regional city than a Dublin commuter county. And Meath — where prices are above the floor — is filing more extensions than anywhere else, driven by ground conditions and infrastructure costs that the headline price does not capture.
Regional cities are crossing the threshold. Cork (€367,841), Galway (€378,855), and Limerick (€361,233) are all above or near the delivery cost floor. Waterford (€338,326, 62% below) is the exception. The gap here is closing, but the extension data says it has not closed yet.
Rural and small-town areas are where the gap remains structural. The median is €321,586 in 2025, rising to €330,396 in Q1 2026. Tipperary is the deepest large-sample gap: 91% of its 223 new-build sales were below €350,000. Without subsidy, cost reduction, or an above-market site, private delivery does not work at scale.
For policy advisers: The approximately 70 at-risk schemes represent over 5,000 units of permitted capacity that the market cannot currently activate. Ninety-one extension applications sit on RZLT-liable land — developers are paying tax on sites they have permission to build but cannot afford to deliver. Planning reform alone does not close this gap.
Ireland is no longer constrained by how many homes it can approve. It is constrained by how many it can afford to build.
Part 2: The Silent Lapse — at most one in twenty developers with expiring permissions files an extension. The rest let the permission lapse in silence. The 2021 record cohort is hitting its five-year mark.
Part 3: The RZLT Paradox — 85% of recycled RZLT sites change hands. The tax designed to unlock land is transferring it from developers who cannot build to institutional buyers who can afford to wait.
Part 4: Permitted but Caught Out — six policy changes hit 170,000 permitted homes mid-flight. Lapse rates exceed 80% regardless of developer size, but one-site operators lose everything.
Every figure in this article can be queried for any specific council, scheme type, or geographic area. A viability assessment for a specific site — cross-referencing local sale prices, delivery cost benchmarks, ground conditions, infrastructure constraints, and extension of duration rates — takes minutes to produce.
Methodology
Extension of duration (updated April 2026): 28 of 31 councils. DCC, Cork City, and Offaly do not publish this application sub-type — all EOD totals are lower bounds. Q1 2026 total of 218 reflects the full quarter (January through March), up from 210 in the original March 22 query which captured applications received before that date. Contributing councils: 25 in Q1 2026, within the 21-25 range observed across all years. This is a council publishing decision, not a platform limitation. Quarterly series based on received_date, residential dev_type only. Three applications in Cavan and Kilkenny are filed as "Permission" but describe extensions of duration in their proposal text — not included in the 218 total.
Viability gap: 2025 PPR new-build transactions (property_type = 'new'), split by county into four geographic categories: Dublin Metro (Dublin) / Commuter Belt (Meath, Kildare, Wicklow, Louth) / Regional Cities (Cork, Galway, Limerick, Waterford) / Rural and Small Town (all remaining). The approx. €350k delivery cost floor is the SCSI's published all-in benchmark for a standard 3-bed outside Dublin and varies by scheme, location, and specification. PPR covers 26 of 31 county areas. Claims are robust to threshold variation: the regional ordering and narrative hold from €300k to €400k. Excluding not-full-market-price transactions pushes medians up €3,500-€7,600 and does not change the narrative.
2026 Q1 PPR: 2,255 new-build transactions (January 2 to March 27, 2026). All four regions have 483+ sales, comparable to Q1 2025 sample sizes. March data may be slightly incomplete due to PPR registration lag.
Meath analysis: 53 residential EODs verified against 53 unique applicant names and 8+ distinct planning agents. Dev_action breakdown: 44 new-build, 6 alteration, 2 demolition, 1 extension. Monthly breakdown: 46 in January, 3 in February, 4 in March. Pyrite zone: 46 of 53 have defective_blocks_risk data, all 46 in pyrite zone (7 have NULL — not geocoded or at data boundary). Unsewered: 39 of 53 (73.6%). Urban-rural classification from CSO Census 2022 small areas joined via small_area_geogid. The "92 of 141" peri-urban figure covers all Meath EODs since January 2025, not just Q1 2026.
RZLT analysis: 91 EODs on rzlt_liable = true land since January 2025. Unit count (2,904) and site area (327.6 hectares) from num_units and rzlt_site_area fields. Coverage: num_units is 59% populated, rzlt_site_area depends on GeoHive zoning coverage (34% of applications nationally). Both figures are lower bounds.
Constraint analysis: Unsewered rate (60%) from is_sewered = false on EOD applications received since January 2025 versus all applications in the same period. Defective blocks rate (44%) from defective_blocks_risk IS NOT NULL. Environmental constraints (flood zone, Natura 2000, heritage) measured against the same baseline. Flood zone data from OPW CFRAM zones A and B.
At-risk pipeline: Approximately 70 schemes / 5,100 units (April 2026). Filters: decision_outcome = 'granted', num_units >= 20, dev_type = 'residential', building_control_linked = false, expiry within 18 months, excluding extension-of-duration applications. The original March 2026 figure of 74 schemes / 5,026 units reflected the data at that date; the small change is attributable to new commencement notices filed and data re-scraping, not methodological differences. Both decision_outcome (61-95% by council) and num_units (59% populated) limit this figure — treat as a lower bound.
Processing days: Average of (decision_date minus received_date) for applications with status_category = 'decided' in Q1 of each year, no outlier cap. The Q1 2026 figure (85 days mean) is based on 1,724 decided applications. The median is 64 days — the mean is consistently pulled up 25-35 days by applications that go through the further information process.
Submissions per application: January of each year, all councils. Seven major councils (DCC, DLR, Fingal, South Dublin, Cork County, Wexford, Cork City) report zero third-party submissions, diluting the per-app rate. Values for 2020-2025 are reliable. The January 2026 figure is excluded from this update due to likely scraper lag (submissions for recent applications may not yet be linked in the register data).
Document extract: Verbatim from Archa document ID 763720, applicant's planning consultant report, DLR ref D19A/0244/E (39 Castle Park Road, Dalkey — 24 units), December 2024. This is the applicant's own submission, not the council's assessment.
Sources
Archa Planning Intelligence (primary)
Extension of duration figures, at-risk pipeline, processing times, and spatial constraint analysis from Archa's database — 28 of 31 councils for planning data. Originally queried 22 March 2026; updated 10 April 2026 with full Q1 2026 data, 2026 PPR transactions, and new spatial overlays (census demographics, zoning, RZLT, environmental constraints). Available at archa.ie.
Property Price Register
New-build sale prices by area type from 776,000+ PPR transactions. 2025 calendar year (14,428 new-build sales) and Q1 2026 (2,255 new-build sales), split by county into Dublin Metro / Commuter Belt (Meath, Kildare, Wicklow, Louth) / Regional Cities (Cork, Galway, Limerick, Waterford) / Rural and Small Town (all remaining). Published at propertypriceregister.ie.
CSO Census 2022
Urban-rural classification and small area demographics from Census 2022 (18,919 small areas). Used for the Meath urban-rural breakdown and development trend analysis. Published at cso.ie.
Construction costs
SCSI Tender Price Index, February 2026 (2.5% annual inflation in 2025; cumulative approx. 39% above 2020 baseline). Delivery cost floor (approx. €350k) from SCSI published benchmarks. 14% peak in 2022 from SCSI press release. All at scsi.ie.
Spatial data sources
OPW CFRAM flood zones (Zone A: 1:100yr, Zone B: 1:1000yr). GeoHive MyPlan zoning (82K polygons, 34% application coverage). RZLT map via GeoHive. GSI groundwater vulnerability. EPA sewerage agglomeration boundaries. Defective concrete blocks risk zones (Statutory Instruments under Defective Concrete Blocks Act 2022). NIAH, SMR, and RPS heritage designations. NPWS Natura 2000 sites and qualifying interests.
Planning document extract
Verbatim text from applicant's planning consultant report, Archa document ID 763720, DLR ref D19A/0244/E (39 Castle Park Road, Dalkey — 24 units), December 2024. Verified via Archa document corpus. Applicant's own submission, not the council's assessment.
Figures deliberately excluded
- Commencement divergence chart (removed in April update): The NBCO building control database contains one row per building, not per notice. A 100-unit scheme produces approximately 100 rows. Raw record counts between Q1 2025 and Q1 2026 tripled nationally, but distinct planning permission numbers rose only 30%, indicating a structural change in CSV methodology rather than a tripling of construction. Several councils showing apparent "collapses" (step-function drops from hundreds of records to single digits) are consistent with delayed BCMS uploads rather than cessation of construction activity. The chart was removed because the data quality does not support the council-level comparison it attempted.
- DCC, Cork City, and Offaly extension of duration: These councils do not publish this application sub-type in their open data. All EOD totals in this article are lower bounds.
- Submissions per application, January 2026: Excluded due to likely scraper lag. Recent applications have not yet had their submissions linked in council register data. The 2020-2025 series is reliable.