by: Elliot Maassen, Senior Associate
22 April 2026
Your email:
The information in these articles is general information only, is provided free of charge and does not constitute legal or other professional advice. We try to keep the information up to date. However, to the fullest extent permitted by law, we disclaim all warranties, express or implied, in relation to this article – including (without limitation) warranties as to accuracy, completeness and fitness for any particular purpose. Please seek independent advice before acting on any information in this article.
In December 2025, the Parliamentary Commissioner for the Environment (PCE) published a detailed review of the Land Use Capability (LUC) system and how it is used across a range of regulatory instruments. The headline finding is that the LUC system has real limitations that make it a poor fit for the regulatory work it is increasingly being asked to do. However, the report’s implications reach well beyond the LUC system itself and are worth paying attention to as New Zealand’s resource management framework is reshaped.
The RMA reform programme is built around a proposition of greater national-level direction and standardisation. The PCE’s report is a prescient reminder that national level policy and rules will only be as effective as the information upon which they are built. The quality of the underlying data and classification systems matters just as much as the quality of the standards themselves. This is not a novel observation, but as the PCE’s review of the LUC system demonstrates, it is a point that deserves attention from those shaping environmental law reform.
The LUC system has been in use in New Zealand since the 1950s. It was originally designed to assess the capability of land to support long-term primary production. The classification works through three components – LUC class, subclass and unit – that collectively identify how suitable land is for different uses, the dominant limitation on that use, and the specific characteristics of that limitation.
National-scale mapping began in the early 1970s, eventually producing a national dataset, the New Zealand Land Resource Inventory (NZLRI), which has remained largely unchanged since. That national coverage has made it an appealing dataset for policy-makers.
Over time, the LUC system has become embedded in several significant regulatory instruments. The National Policy Statement for Highly Productive Land (NPS-HPL) uses the national LUC dataset as the default basis for identifying highly productive land. The National Environmental Standards for Commercial Forestry rely on a derivative classification, the Erosion Susceptibility Classification, for regulating forestry activities. Most recently, amendments to the Climate Change Response Act use LUC classifications to restrict forestry conversions registrable under the Emissions Trading Scheme (ETS).
The most fundamental limitations with the NZLRI dataset come from its age. The mapping was paper-based and involved a significant degree of expert judgement, which inevitably introduced subjectivity. New Zealand’s landscape has changed considerably in the intervening half century – through irrigation and drainage management practices, among other things – but the NZLRI has not kept pace with those changes.
The regional coordination of the original mapping exercise also produced local inconsistencies that persist in the national dataset today. The NZLRI soil factor has well-documented limitations.
A more detailed soil map known as the S-Map, produced by Bioeconomy Science Institute, is available in some areas of the country. Comparisons between the S-Map and the NZLRI maps illustrate the practical significance that different datasets can have on LUC mapping.
In Canterbury, a comparison of the two datasets to identify LUC 1 and 2 land showed an overlap of only 24% for the area analysed. A similar comparison in Hawke’s Bay for LUC 1–3 land produced a 41% overlap. To put it simply, the two best available methods for identifying highly productive land disagree about where that land is the majority of the time.
When laws or regulations rely on a tool or a dataset, the limitations of that tool or dataset flow straight through to the law or regulation. That is precisely what is playing out here.
A 2024 Environment Court decision (Bluegrass Limited v Dunedin City Council [2024] NZEnvC 83) confirmed that during the NPS-HPL transitional period, landowners cannot challenge their LUC classification in consenting processes – even where more detailed, site investigations reach a different conclusion. Land mapped as LUC 1, 2 or 3 in the NZLRI is, in the PCE’s phrase, “stuck” as highly productive land until regional councils complete their own mapping. Resource management practitioners will be all too familiar with the roadblock that the NPS-HPL can present for land that is mapped LUC 1, 2 or 3 but which, on closer inspection, plainly has limited productive capacity.
The same dynamic is at work with the ETS. In an effort to curb land use change from pastoral farming to forestry, recent amendments to the CCRA use the LUC classification and the accompanying NZLRI LUC dataset to restrict the number of exotic forestry conversions that can be registered to the NZ ETS. The restrictions make the boundary between LUC 6 and 7 legally significant. The trouble here is that the NZLRI dataset is far too coarse for property-level distinction. Disputes about where that boundary falls between LUCs are an entirely foreseeable consequence. Landowners who disagree with their classification may correct it through a property-scale assessment, but at their own cost.
As the PCE rightly observes, there is real difficulty with justifying regulatory restrictions that limit access to a potentially significant revenue stream, but which require landowners to fund the correction of inadequacies in the Government’s own dataset. Where the information base is known to be unfit for the purpose it is being put to, it is perverse that the cost of remedying the dataset should fall on the people subject to the restriction.
The PCE’s report highlights two pitfalls: regulation that gets ahead of the data it relies upon, and regulation that repurposes tools never designed for regulatory use. The LUC story is a case study in both. The NZLRI was convenient and had national coverage, but it was never validated for property-level decisions – and the LUC system’s productivist origins make it a poor fit for the environmental management questions that the NPS-HPL and NES-CF now ask of it.
None of that is a reason to abandon standardisation and sometimes decisions simply must be made on imperfect information. But when an entirely new system is being designed from the ground up, there is a rare opportunity to build the information base alongside the regulatory framework – commissioning fit-for-purpose datasets, aligning commencement dates with realistic data milestones, and investing where regulatory stakes are highest.
The reform process has articulated a clear vision of what good legislative architecture looks like. The PCE’s report is a timely reminder that architecture is only as strong as its foundations. If the information underneath the new framework is not being built with the same care as the framework itself, then regulation will need to be designed around the limitations of that information.
WWW.WYNNWILLIAMS.CO.NZ
