The Abbotsford landslide in Dunedin in 1979 moved 18 hectares 48m in 15 minutes. Photo: Wikimedia Commons

Building housing on land that is vulnerable to instability has long been an issue in New Zealand. It’s time we face up to the problem, write Chris Nicoll and Martin Brook

The EQC evolved from the Earthquake and War Damage Act 1944, itself prompted by the local insurance market’s inability to provide adequate, national coverage for earthquakes following the Wairarapa earthquake sequence in 1942 which devastated Masterton and shook the corridors of power in Wellington.

How should the Act be applied to the damage caused by the extreme weather events we’ve witnessed in recent years, particularly 2023?

Preventing building where it’s not safe
Publicly-funded compo for climate retreat

The 1944 Act had one great advantage over the present EQC legislation, which meandered to its present form without benefit of any roadmap of principle. The former Act had a coherent philosophy which any member of the public could have understood. It was that the whole country should stand behind those people who suffer from unpredictable and unascertainable earthquake losses they cannot take any steps to avoid, encapsulated by the catchphrase “national disaster is the national care”.

The National Party opposition, in a commendable show of unity, supported the proposed legislation with only minor objections which were well-articulated and constructive. In short, the philosophy was as well-defined by informed debate, as one could expect in any democracy.

Over 70 years ago, the spectre of moral hazard was raised. That is, the lack of incentive to guard against risk by, for instance, trusting in public money being used to assist without the reciprocal obligation to keep a property safe or to mitigate loss when it is likely to occur. Three ‘moral hazard’ issues were raised and robustly debated.

First, the fact that the new regime only applied to people who had fire insurance. What if they didn’t? Should they be helped? The answer under the 1944 Act was ‘no’. Quite simply if you weren’t prepared to pay to be in a scheme you wouldn’t get its benefits.

Secondly, if you have a building that is likely to be badly affected because no reasonable earthquake resilience measures have been taken, should you pay a greater levy than those who have made reasonable attempts to mitigate? The answer was ‘no’ again, but only because then it was not practical to assess buildings individually.

Thirdly, a question of coverage was raised, what events should be covered? Why not cover all acts of God? Floods and cyclones were mentioned, hailstorms affecting horticulture and snowstorms affecting high country farmers; the Honourable Member for Otago observing there was little chance of earthquake in the South! The conclusion was that the market value of land was a partial reflection of its susceptibility to flooding and compensation whenever flooding occurred would be a pure gift from the state to those interested in obviously vulnerable property. By parity of reasoning, it could be said of the other risks that they were an expected cost of doing business, which likely loomed large in later debates, when it was decided to remove cover for commercial assets.

In short, the philosophy of the 1944 Act was to help those who had little or no control over an unpredictable natural disaster such as an earthquake. But now, with resilience to major storm events being tested, we need to take a more clearly defined two-pronged approach: one prong being EQC, and the other being a clear policy of early, equitable state intervention to prevent new building in vulnerable areas and to move existing residents to safety.

Building housing on land that is vulnerable to instability has long been an issue in New Zealand. Perhaps the most salient example is the Abbotsford landslide in Dunedin in 1979, where 18 hectares (nearly half residential) moved 48m in 15 minutes, as a “blockslide”. This left hundreds of people homeless, but there were no deaths. Since then, the area has been rezoned and includes trees and rugby grounds.

A less well-known example of building in an area vulnerable to land instability is in Muriwai, West Auckland. In August 1965, after two days of heavy rain, weak sands became saturated and flowed downslope across houses at Domain Crescent, as “debris avalanches”, at velocities up to 90 km/hr, a total distance of 250m. Two landslides on consecutive days destroyed two houses and killed two occupants. While the 1965 landslide runout zone was left vacant for the 1970s, by the end of the 1980s, houses were being rebuilt on the site of the 1965 landslides. Domain Crescent was unfortunately affected again, by the recent Cyclone Gabrielle event.

Having analysed SPOT satellite imagery (1.5m resolution) taken two days ago, comparing last week’s Motutara Rd fatal landslide visible on the satellite photo with the 2016 LiDAR DEM (explained below) it seems last week’s fatal landslide occurred on a site of three existing landslides. This suggests people chose to build on unstable ground, the ground where this year’s landslide killed two firefighters responding to an emergency.

Myriad approaches exist to deal with such land instability issues, approaches that can be categorised as either passive or active.

Active approaches are focused on slope engineering, which is using engineering methods to make a slope more stable. Effectively though, this puts the ambulance at the bottom of the cliff, if one were to look at a hierarchy of risk controls. Moreover, slope engineering can be extremely costly. Some form of managed retreat might be the more economic option, a concept which is not new to New Zealand – the Crown bought out 8000 properties when the area around the Avon River was red-zoned following the Christchurch earthquakes.

Passive approaches are concerned with planning. Unfortunately, the Local Government Amendment Act (No 2) 1981 s 641A allowed local authorities to permit relocatable building on unstable land, and not be under any civil liability to any person having an interest in the building. The certificate of title was required to say the permit had been issued under these circumstances. Such properties were not covered by EQC. It is unsurprising the site of the 1965 landslides at Domain Crescent, Muriwai, had several houses built on it after 1981.

We now have the technology that would identify land instability hazards. This is a region-wide mapping approach, based on a high-resolution LiDAR digital elevation model (DEM). This is where a whole range of geological, geomorphological, and landcover/land use parameters are layered onto the DEM in a GIS (geographical information system), and statistically modelled. A susceptibility map is then produced and in theory, this can then inform future planning decisions.

Once unstable hazardous slopes have been identified, slope monitoring approaches such as InSAR (satellite radar) can monitor slope movement as a landslide early warning system.

The third ‘passive’ approach is to then apply adequate setback distances above and below slopes, to remove houses from high-risk areas. There are a range of methods to calculate setback distances.

Clearly, we will need to focus on the recovery stage following Cyclone Gabrielle but given extreme weather events are likely to become more frequent, we need a robust and national debate on how we can best protect people from such events, and to ensure people are actively discouraged (or prevented) from being placed in areas of calculable risk.

Chris Nicoll, Department of Commercial Law, Faculty of Business and Economics, University of Auckland, is co-author of 'Colinvaux' Law of Insurance in NZ'.

Leave a comment