The RZLT Paradox: A Tax on Small Developers That Accelerates Land Consolidation

A Senior Executive Planner in Cork County Council assessed an extension of duration for 42 apartments on RZLT-liable land at the former Loreto Convent in Youghal. The planner concluded the apartments could not be delivered. The extension was refused. The land remains on the RZLT map, and the 3% annual charge on its site value continues to accrue.

In Kilternan, the applicant for 16 houses on RZLT-liable land wrote that it is "not possible to raise finance for this scheme until access is afforded to the public secondary road, taken in charge after five years of asking." The council had not taken the road in charge. That extension was also refused.

The planning system's own documents describe sites where construction is blocked by cost, by infrastructure, or by the state's own inaction. The RZLT charges 3% of site value annually on the same land. Neither system references the other.

When these developers eventually sell — and the data shows 85% of recycled RZLT sites change hands — the buyers are not other small builders. They are institutional acquirers with balance sheets large enough to carry the tax while they wait for the economics to turn. The RZLT was designed to unlock land for housing. On this evidence, it is transferring land from developers who cannot build to developers who can afford not to.

Does RZLT appear in planning documents?

Not once. Across 140 extension of duration applications on RZLT-liable residential land, and a broader pool of 500 residential EODs, no document mentions "RZLT" or "Residential Zoned Land Tax" by name (889 documents searched, 86% extraction coverage).

This is legally expected. The RZLT has no bearing on the statutory test for an extension of duration. Planning consultants have no reason to cite it. But it means the planning system grants, refuses, and extends permissions on RZLT-liable land without knowing the site carries a tax liability. Revenue charges the tax without knowing why the site is undeveloped.

Two state systems operating on the same land. No feedback loop between them.

When is the state itself the bottleneck?

The RZLT assumes non-development is a landowner choice. In several documented cases, state agencies are the constraint.

Kilternan, Dublin (D18A/0137/E, 16 units). The access road has not been taken in charge by the council after five years. The applicant cannot raise development finance without public road access. The EOD was refused. The site remains RZLT-liable.

Cobh, Cork (24/6184, 76 units). Irish Water's service infrastructure prevented the developer from finalising service agreements required by the planning conditions. The applicant could not "secure funding or commence." The land is zoned R1 — new residential — and taxed as if development should happen.

Ballinteer, Dublin (D18A/1096/E, 6 units). The applicant here is not a developer. Cabhru Housing Association, a social housing provider, faced "cost increases" severe enough to require reviewing the specification. They were waiting on the state's own Capital Assistance Scheme funding. A social housing association waiting on CAS funding is not hoarding land.

These are individual cases, not proof that all RZLT-liable non-commencement is state-caused. But they show what the aggregate statistics cannot: specific sites, specific obstacles, and specific state agencies in the way. And in every case, the planning documents that describe these obstacles make no reference to the RZLT liability on the same land. The planner who processed Cabhru's extension had no mechanism to flag that a social housing association was being taxed for not building homes the state hadn't funded.

What do developers and planners say about costs on RZLT land?

The planning documents on RZLT-liable EOD applications use consistent language about viability, even when the specific circumstances differ.

The Blackrock and Murphystown language is near-identical, likely from the same consultant or a template that circulates among DLR extension applicants. The Youghal quote is different. That one was written by the planner, not the applicant.

In at least three of these cases the extension was refused. The applicants cited cost escalation, infrastructure delays, and funding constraints. The planning authority denied more time. The RZLT continues regardless.

How many RZLT-liable permissions are expiring?

An estimated 2,791 residential new-build grants on RZLT-liable land reach their five-year expiry between 2025 and 2027. Of these, 1,849 have no linked commencement notice — the permissions that appear to have expired or be expiring unused. The remainder may have commenced but lack a building control record, or may have been superseded. 731 RZLT-liable grants expire within the next twelve months, representing 5,741 units approaching the cliff edge. Q3 2025 is the largest quarter by total grants (459), reflecting Covid-era 2020 permissions hitting their five-year mark.

Not all of the 1,849 uncommenced sites are "lost." Roughly 35% already have a newer granted permission within 50 metres — the site was re-permitted, not abandoned. The net figure after accounting for re-permitting is closer to 1,200 uncommenced grants with no replacement in the pipeline.

Are RZLT sites being sold rather than built?

Outside Dublin — where spatial matching is most reliable — 12.1% of expired RZLT sites have a new residential application within 50 metres, compared to 10.1% for non-RZLT sites. A modest difference. Consistent with the tax creating some pressure to sell, but not large enough to prove it.

What is more striking is who files the new application. Of the 48 recycled RZLT sites outside Dublin with clean spatial matches, 41 (85%) have a different applicant name on the new application. These are not the same developers coming back to try again.

The data cannot prove RZLT is causing these sales. But the pattern is visible in the planning register: permitted RZLT-liable land changing hands after permission expiry.

Who absorbs the pressure and who absorbs the land?

The RZLT charges the same 3% rate regardless of who holds the land. But 3% of one site's value is existential for a single-site developer and a line item for an institutional buyer holding twenty.

A small developer or SPV holding one permitted site that can't be built faces the tax annually with no revenue to offset it. The options narrow to: extend the permission (if eligible), sell, or absorb the charge until it becomes untenable. The 85% name-change rate on recycled sites shows which option most choose.

The names that replace them tell the rest of the story. In the broader dataset, SPV-held sites are acquired by listed housebuilders operating across multiple counties. Large sites fragment into clusters of smaller applications by individual names — consistent with subdivision and selloff. The pattern is institutional acquirers targeting RZLT-pressured land at discount.

The tax was designed to penalise landowners who sit on zoned land. On this evidence, it penalises small holders who cannot build and accelerates the transfer of permitted land to larger entities who can afford to wait for the economics to improve. Whether that is a policy success or a policy failure depends on whether the goal was to unlock construction or to consolidate ownership. The data shows the latter is happening. The former, at these sites, is not.

What does the data show about RZLT commencement rates?

The aggregate numbers are misleading and this article will not use them.

In raw terms, 33.8% of RZLT-liable residential grants from 2020-2021 have a linked commencement notice, versus 62.1% for non-RZLT grants. That gap collapses under proper controls. RZLT land is disproportionately in Dublin, where building control data linkage is poor (23% in Dublin versus 64% elsewhere). RZLT schemes are larger on average (13.4 units versus 4.2), and larger schemes have lower commencement rates everywhere.

When controlled for both geography and scheme size, the RZLT commencement gap disappears for most categories. Two-to-ten-unit schemes outside Dublin: 53.0% RZLT versus 53.1% non-RZLT. Eleven-to-fifty-unit schemes outside Dublin: 61.8% versus 62.9%. The only band with a meaningful gap is 50+ unit schemes outside Dublin (57.6% versus 68.7%), but this is based on 33 RZLT grants — too small a sample for structural claims.

The planning data cannot tell us whether RZLT-liable land is less likely to be built on than comparable non-RZLT land. What it can tell us, through the documents, is that specific RZLT-liable sites face specific obstacles that planners and developers describe in their own words.

What does this mean if you hold RZLT-liable land?

If you hold a 2020-2021 permission approaching expiry on RZLT-liable land, the extension data from the Viability Series applies with an additional dimension: the 3% annual charge continues regardless of whether the permission is extended, expires, or is replaced. The planning system and the tax system impose independent obligations on the same site.

If you are acquiring sites from RZLT-pressured sellers, the planning register shows a pattern of applicant name changes on recycled RZLT sites (85% outside Dublin). Permitted sites that have expired or are approaching expiry on RZLT-liable land are entering the market. The zoning and infrastructure status from the original permission is a starting point for due diligence — not a guarantee that the same scheme is still viable.

If you are advising on RZLT exposure, the planning documents from extension applications on RZLT land contain site-level evidence of viability constraints: planner reports acknowledging infeasibility, infrastructure blockages, cost escalation. All of it is in the planning record and all of it predates the RZLT liability.