The Scale Trap: Too Small to Finance, Too Big to Survive

Roughly 400 residential developers receive planning permission for schemes of ten or more units in Ireland each year. Three-quarters of them are first-time grantees. They average 43 units per scheme. The commencement sweet spot, identified in Part 7, starts at 61 units.

The gap between 43 and 61 is eighteen units. It is also the gap between what an inexperienced builder can self-finance and what the system reliably delivers. There is no financing product at current interest rates that bridges it. The developer who builds 30 houses in Louth cannot grow to 70 without institutional backing, and institutional capital has retreated from Irish residential since the ECB moved to 4.5%.

Above 100 units, commencement falls sharply: 9-33% outside Dublin (Part 7). Nine out of nine large SHD grants from 2021 sit uncommenced — including a 183-unit Cairn Homes scheme in Kilkenny and a 253-unit mixed-tenure LDA scheme on state-owned land in Limerick. The LDA site carries protected-structure and school-displacement conditions, and other LDA schemes are progressing. But the pattern across all nine 2021 grants is the same: well-capitalised developers and the state itself stalling at scale.

The market is boxed in. A floor below which you can't finance, a ceiling above which you can't build, and a geography that punishes you for being in the only zone where the economics work.

How many developers does Ireland have?

Fewer than you think. Between 2018 and 2024, an average of 560 unique developers per year submitted residential applications for ten or more units. Roughly 400 received grants. Of those, 275-340 were receiving their first ever grant. Only 80-110 were returning developers — the experienced core of the entire industry.

Each year, 300 new entrants attempt residential scheme development with no track record, no prior relationship with a council planning department, and no portfolio of alternative sites. They hold one permission. They build at 56%. These commencement rates are lower bounds — building control data matching is incomplete, particularly in Dublin. But the same method applies to every developer, so the gaps between groups are real even if the absolute levels are understated.

The returning 100 developers carry the institutional knowledge: which councils process fastest, where the infrastructure works, which scheme sizes and types convert. They build at 62% overall, and at much higher rates on the sites they select. But they are only 100 firms in a country that needs 33,000 new homes per year.

The floor: why developers can't scale up

The full residential pipeline tells the story of what enters the system each year.

Roughly 4,000 one-off houses are granted — single dwellings on individual sites. Another 620 small schemes of two to nine units. And 400 scheme grants of ten or more units. The one-off house is the volume product. The scheme grant is the unit product.

One-off houses commence at 70.5%, face ABP involvement on 2.5% of applications, and process in 112 days. Scheme housing of ten or more units commences at 57%, faces ABP on 30%, and processes in 150 days. The one-off builder and the scheme developer operate in different systems sharing a name.

The scheme developer who wants to scale faces a financing gap. At current ECB rates, development finance for a 70-unit housing estate requires capital reserves and a track record that a first-time builder with a 30-unit permission does not have. The builder knows the sweet spot exists. They can see it in the commencement data: 68-76% for 61-100 units versus 54% for 10-30 units. Getting there requires a step change in capitalisation that the market does not provide.

Build-to-rent was the institutional bridge. In the 2019-2021 era, BTR schemes commenced at roughly half their grants. For recent grants the commencement rate is near zero. Only 5-15 BTR schemes are identifiable per year (Part 7 methodology), and the recent cohorts trend toward the lower end of that range. The bridge is gone.

The ceiling: why developers can't scale through

Above 100 units, the economics change. Construction complexity increases non-linearly with apartment height and scheme size. Part V at 20% bites harder in absolute terms. Financing quantum rises. And the data shows the result: commencement rates of 9-33% for schemes over 100 units outside Dublin.

The apartment penalty is structural at most scales. Apartments commence at 46% versus 64% for houses in the raw aggregate. Controlling for scheme size, the gap narrows to 10-15 percentage points but persists in every band. The exception is mid-size apartments of 61-100 units in the commuter belt, which reach 75% — near parity with houses. The penalty concentrates in small apartments (10-30 units) and in Dublin, where apartment economics are most exposed to financing costs. The gap was modest in 2018 and widened to over 20 points by 2022 as interest rates rose. Apartments require more institutional financing per unit, and that financing now costs 4.5%.

New apartments sell at a median of €383,000 nationally in 2025. Industry estimates for apartment delivery costs in Dublin range from €450,000 to over €500,000 per unit depending on height and specification. The gap is not a margin problem. It is a product that cannot be sold for what it costs to build.

On this evidence, the ceiling is not primarily about planning. Every large scheme in the 2021 graveyard had planning permission. The constraint is the economics of delivering at scale in the current interest rate and construction cost environment.

The geography: the only viable zone is the hardest to build in

Part 7 identified the viable band: 61-100 units, across all three regions, 68-76% commencement. But the viability ceiling cuts across that band. The commuter belt is the only non-Dublin zone where 2025 median new-build prices (€383,000) sit clearly above the €350,000 SCSI delivery cost floor. Regional cities (€362,000) clear it by approximately €12,000. Rural and small-town Ireland (€322,000) sits €28,000 below.

But the commuter belt has the worst infrastructure profile of any zone. Fifteen percent of scheme sites are served by a failing wastewater treatment plant. The average site carries 80 water supply advisories. Twenty-three percent of scheme sites are not connected to public sewerage at all.

For the schemes that do commence on failing infrastructure, the worst quartile wait 729 days — two full years — from grant to first brick. On a five-year permission, that is 40% of the clock burned before construction starts.

An unsewered commuter belt scheme actually commences more reliably (65%) than a sewered one (57%). Developers who build their own treatment system bypass the bottleneck. The ones who depend on Irish Water to upgrade a failing plant wait for a timeline the state does not provide.

The €33,000 margin between the commuter belt median sale price and the delivery cost floor must absorb: development finance at 4.5%, RZLT at 3% of site value, professional fees, site maintenance, insurance, and the carrying cost of 264 to 729 days waiting. For the committed builder on a failing WWTP, the margin is gone before the first foundation is poured.

The one-off paradox

While scheme developers are trapped between a floor and a ceiling in the commuter belt, the one-off house builder faces a different constraint: the system won't let them build where they want to.

In Kildare, one-off house refusal rates exceed 50% (629 refusals from 1,222 decided applications). In Fingal, 45% (n=338). In Meath, 26% (n=959). These are commuter belt counties where demand is highest and prices are strongest. The National Planning Framework encourages compact growth and discourages one-off rural housing.

In Tipperary, 3% are refused. In Monaghan, 5%. In Mayo, 9%. One-off houses are the most reliable housing product in Ireland — 70.5% commencement — and the system grants them freely where prices are lowest and restricts them where prices could support them.

The result in Tipperary tells the story from both ends: one-off houses commence at 74%, scheme housing at 42%. The same county, the same planning authority. The one-off builder operates below the friction. The scheme developer operates inside it.

What this means before you submit

If you are a first-time scheme developer, the data says you are one of approximately 300 new entrants this year. Your average scheme size (43 units) is below the sweet spot (61-100). Your commencement probability is 56% overall, but ranges from 35% to 87% depending on council and friction profile. Before committing capital, the infrastructure serving your site may matter more than the planning permission on it.

If you are considering scaling from a small scheme to a mid-size one, the financing gap is the constraint. The sweet spot is visible but the bridge to reach it does not exist in the current rate environment. The builders who operate at 61-100 units are predominantly experienced multi-grant developers, not first-timers who scaled up.

If you are a one-off house applicant in the commuter belt, the refusal rate in your council is the first number to check. In Kildare it is 52%; in Tipperary, 3%. The gap between councils is larger than the gap between any other variable in this series.

Methodology

Developer population. Unique applicant names on residential permission applications of 10 or more units, received 2017-2025, across all 31 local authorities. Annual averages over the period: 560 unique developers apply, roughly 400 receive grants, 275-340 of those grantees are first-time, and 80-110 are returning multi-grant developers. The "three-quarters first-time grantee" figure uses the midpoint of the 275-340 range (~308 of 400). "First-time" means no prior grant in the dataset; "returning" means at least one prior grant. SPV contamination means the true number of independent developers is lower — approximately 14% of one-grant developers share a planning agent with multi-grant developers.

Pipeline volumes. One-off houses identified via the rural one-off flag (AI-extracted, 21% of applications populated, only true values reliable). Small schemes: residential applications with 2-9 units. Scheme housing: permission-type applications with 10 or more units. All figures are permission-type applications, excluding EODs, compliance, and retention.

Infrastructure data. WWTP compliance (Pass/Fail) from EPA monitoring, mapped to the treatment plant serving each site. Water supply advisories from Uisce Eireann, aggregated by supply zone. The sewered indicator reflects whether the site falls within an EPA Urban Wastewater Treatment Plant agglomeration boundary. 23% of commuter belt scheme sites are outside these boundaries and rely on private treatment.

Grant-to-commencement timing. Measured as commencement date minus decision date for commenced schemes only. Survivorship bias applies: schemes that abandoned due to infrastructure delays do not appear. The P75 figure (729 days for failing WWTP sites in the commuter belt) represents the slowest quartile of successful schemes, not the full population including failures.

One-off refusal rates. Based on decision outcomes for one-off house applications (rural one-off flag) by council, 2018-2022. Kildare's 52% refusal rate is calculated from 593 grants and 629 refusals (n=1,222 decided applications).

PPR prices. 2025 calendar-year new-build sales (14,428 transactions), grouped using the same 4-zone schema as Part 1: Dublin Metro = the four Dublin local authorities; Commuter Belt = Meath, Kildare, Wicklow, Louth; Regional Cities = Cork, Galway, Limerick, Waterford; Rural and Small Town = all remaining counties. Zone medians: Dublin €449,339, Commuter Belt €383,260, Regional Cities €362,411, Rural and Small Town €321,586. Median prices are zone-level, not site-level — individual scheme viability depends on local micro-market conditions. Q1 2026 figures (commuter belt €405,286, rural €330,396) are also available but the 2025 full-year medians are used here for consistency with the developer behaviour data, which extends only through 2025.

The delivery cost floor (~€350,000) is from SCSI benchmarking, calibrated to standard scheme housing. It likely overstates costs for small regional builders and understates costs for apartment delivery in Dublin. The commuter belt margin of €33,000 is the gap between the 2025 zone-level PPR median (€383,000) and SCSI floor — not a site-specific calculation.

Social housing. New-build only from DHLGH social housing delivery statistics (LA, AHB, Part V construction). 604 units in 2016 to 9,089 in 2025. Excludes acquisitions and leasing (~6,000-8,000 additional units per year).

Figures deliberately excluded. The "activation cliff" (65% to 49%) was dropped after critical review found it compared mature cohorts with immature data. At equivalent maturity, the decline is approximately 5-8pp. The "2pp to 21pp apartment gap" was dropped because the 2018 baseline is classification-sensitive; the robust finding is a 10-15pp gap present since 2019. BTR year-on-year percentages are not cited due to samples of 5-15 per year; the direction (roughly half to near-zero) is used instead.

Sources