Managing retreat

Dr Eric Crampton
28 February, 2023

Managed retreat from floods and rising sea levels seems like a difficult problem. How to decide who might be allowed to live where, and what risks are acceptable? It sounds impossible. Any decision would be horribly fraught.

But some problems are only impossible if you look at them from the wrong end.

The typical maze in a kid’s book is a lot easier if you start from the end of the maze and work back to the beginning.

That’s a trivial example. But think about another difficult problem that has already been solved.

Imagine someone asked you to design a system that made sure that, with very high reliability, enough food made it to your city – and in the varieties that suited local tastes.

Where would you even start?

You’d have to know who liked what and how much, where everyone lived, who is good at growing and making different kinds of food, how to transport it, and when to deliver it. You’d need to figure out where to grow which crops and in what quantities. It would require information that would be impossible for anyone to acquire.

And yet Auckland is fed. Nobody tried to ‘manage’ the entire problem. Instead, people and businesses made their own plans and decisions, guided by basic ground rules penalising fraud, enforcing contracts, and preventing adulterated products.

And it all just works – or at least very tolerably well, given the scale of the problem! Think about how terrible the ‘How do we feed Wellington’ problem would sound if you put it that way. Then marvel at the solution that’s been right in front of us all along. We all plan our own shopping trips; grocers, supermarkets, vegetable markets and butchers compete for customers; and growers and processors compete to supply those markets. If there’s a bad apple harvest, apple prices rise. Shoppers adjust their trollies as best suits, without anyone telling them to do so.

We’ve similarly been looking at managed retreat the wrong way round. Let’s try it the other way, by starting from the underlying problem.

Right now, a lot of people live in risky places – and other people wind up picking up the tab if things go wrong. And the riskiness of risky places seems to be increasing.

The Earthquake Commission charges the same amount to insure a $1m property on the safest ground in the country as it does to insure a $1m property that takes land slips every year. And a million-dollar unreinforced brick house sitting right on the Thorndon fault would face the same EQC premium as a million-dollar weatherboard home in the safest part of Karori, or Hamilton.

Motu’s study of the data from 2000 to 2017 found that higher-income neighbourhoods benefit most from this coverage.

So we have a government-provided insurance programme in EQC that has helped ensure broad insurance coverage, but at a cost of encouraging more building in risky places than otherwise would have happened. And the benefits of this flat premium structure have disproportionately gone to richer households.

Council infrastructure can also be a lot more expensive to provide and maintain in places subject to frequent flooding and landslips. But if rates and water charges do not vary with the cost of providing services, councils can be left in a difficult spot. If they allow a lot more building in riskier places, they could be stuck with a very costly bill down the track.

When building in risky places is subsidised through flat fee structures for insurance and council services, there will be more of it than there otherwise would be.

Seeing the problem that way makes the solution rather obvious.

The Earthquake Commission has decades of claims history and rather decent expectations of where future claims will fall. EQC should be allowed to set premiums that reflect differences in risk. A shift to a fully risk-based premium structure could be spread out over time so the immediate sticker shock is not as large.

While private insurers are not forbidden from setting very different premiums in the riskiest places, they may feel politically constrained against doing so. Insurance is a highly regulated industry. Private insurers might worry, whether reasonably or unreasonably, that increasing premiums in risky places would lead to populist backlash and punishment through regulatory channels.

If EQC were allowed to set premiums that varied with underlying risk, private insurers might find it safer to follow. Premiums would drop in safer places and rise in riskier places.  

That, all on its own, would start encouraging different choices. You can’t get a mortgage without insurance, and most of us can’t buy a house without getting a mortgage. So land in safer places would be safer bets for developers – and especially if safer places were zoned to allow more housing.

There’s another side to insurance. Insurance premiums are generally sensitive to measures that owners might take to reduce risk. If you install a sprinkler system, you’ll likely pay less for fire insurance. Smokers pay more for life insurance. As insurance premiums in risky places start rising because of that risk, owners in those neighbourhoods might want to work together to put in flood protection works, or to stabilise risky slopes, to help keep their insurance bills down.

But right now, there’s no good way for owners to coordinate the funding and financing to make it work. Councils at their debt limits cannot reasonably fund them, and a special rate only on protected properties cannot finance them.

So councils, and potentially others, need to be able to take on long-term debt that is separate from council main balance sheets, backed by dedicated revenue streams like special ratings districts, or user fees, or betterment levies.

Finally, councils need better ways of withdrawing from or applying higher service fees to places that become too expensive because of natural hazards. Currently, councils can ‘stop’ local roads when they are no longer cost-effective for councils to supply. But that is harder for other works.

Councils could offer existing residents a choice. Council could maintain service while setting higher rates. Or it could transfer local infrastructure to a special-purpose local board that would levy serviced properties for the infrastructure – with lower council rates to reflect the reduction in council-provided services.

People could then decide on their own ways of adapting, while bearing their own costs.

If a family were happy to get a few more decades’ out of a beachside bach likely to be underwater by 2100, or from a home with excellent clifftop views but subject to erosion, there would be no need for council or anyone else to draw a line on a map saying they could not. They could enjoy their homes at their own risk and at their own cost.

Higher insurance premiums in risky places would encourage more building in safer places, without anyone having to issue edicts. Getting rid of bans on building more housing in some of our safer places would make the job easier too.

Risk-sensitive premiums would encourage building protective works against those risks, without anyone having to compel it.

It would vastly simplify the problem of adapting to climate change and managing retreat from risky places. It wouldn’t avoid the problem of planning. It would instead enable all of us, both individually and as communities, to do more of the planning ourselves – rather than having planning be done to us.

Flipping the problem, approaching it from the right end, makes it far easier to solve.

Read more

Stay in the loop: Subscribe to updates