By Eric Crampton 22/02/2018


Seventh anniversary of the Christchurch earthquakes.

GNS figures there’s about a 1/120 annual risk of something of similar size hitting Wellington: 0.833%.

Recovery here will be much harder than in Christchurch. There are few local options for firms to relocate to while downtown is torn down and possibly rebuilt. There are few transit routes in and out. The airport could easily be out of service. The Port is only now just talking about getting floating wharf facilities that could be more quickly serviceable after an earthquake. And while there are lots of ways of getting from anywhere in Christchurch to the Hospital, or at least close enough to Hagley Park to be walked over, lots of Wellington suburbs will be cut off from the hospital. Wellington has reticulated gas. The fire service will have a very difficult time accessing some suburbs. The immediate calamity will be worse, and recovery from it will also be harder.

Post-earthquake recovery policy in Christchurch was its own separate disaster. The morass of agencies and regulations contributed to what I’ve called a confusopoly: nobody knew what they were allowed to do because they needed to get permissions from too many outfits, none of which really knew what they were doing.

In Christchurch, that meant business fled from downtown to the suburbs.

In Wellington, it could easily mean the death of the city.

I’m not speaking hyperbolically here. There is reasonable risk that, after a Mercalli VIII event, Wellington will cease to exist as a city.

Here’s the path to failure. It doesn’t have to play out this way, and I sure hope it doesn’t. But it’s more likely than I’d like.

Government replays Christchurch, with a poor governance setup for a recovery agency. Everybody knows Christchurch didn’t work well; nobody’s prepared for anything else so they go with the current off-the-shelf option, which is a do-over.

Nobody has put sensors into key buildings before the earthquake that would let us know how far beyond design spec key parts of the buildings moved, so engineering assessments will take a long time. Downtown gets cordoned off for months while they figure out which buildings are too risky. The Lambton/Featherston core is under cordon for at least half a year, maybe more than a year, during engineering assessments. Safe buildings cannot be reoccupied because everyone’s worried about the other ones that could fall on the safe ones and so it’s army barricades again.

Little priority is put on regime certainty: government announces a review of building standards following the earthquake, and no buildings over two stories are allowed pending that review. So even those places that could be safe to build on aren’t built on. Essential government functions will have to leave quickly to be able to get on with things, and there won’t be enough available safe commercial space here to move people into.

The business core will relocate, possibly to Auckland, possibly elsewhere. The tech core and Weta don’t need to be here – they could relocate anywhere. There’s a coordination problem in figuring out where the tech sector would go, but a Schelling point might just be to follow wherever Xero goes.

A lot of the core public service will be gone, perhaps with a promise to return when facilities are restored. Civil servants will follow government to wherever it goes.

The University will collapse as students flee even more quickly than they did in Christchurch: recall that the area around the University of Canterbury was relatively undamaged.

After an initial property jump for relatively undamaged homes, there will be a long and terrible decline as people come to realise that the city is finished. Many homeowners will cash-settle rather than repair, and will flee to try to rebuild their lives elsewhere.

And the city will be left with a terrible mess: a rather smaller population of those with the fewest exit options, a big pile of cash-settled unrepaired broken houses, little tax base, and few prospects.

Christchurch makes sense as a city: it’s a regional service centre for a broad agricultural periphery with a good university and an important port. If Christchurch didn’t exist, you’d have to invent it. So even though government made a complete hash of the downtown recovery, it was going to come back because there was a reason for it to come back.

It’s harder to see that for Wellington, even if you can’t beat it on a good day.

ew Zealand government “Beehive” and the Parliament Buildings. Wikimedia Commons.

All of that means that government simply cannot afford to screw up a Wellington post-quake scenario the way it screwed up Christchurch. It may be the difference between the life and death of the city. Restoring regime certainty will be critical.

I hope that the current government puts some priority on it. We’re still playing Russian Roulette. The revolver’s barrel has 120 chambers, one of them has a bullet in it, and we pull the trigger every year. We have the chance to build a Kevlar helmet to blunt the bullet’s effect if we get an unlucky draw. Kevlar ain’t perfect, but it helps reduce the risk.

A bit of early planning could be rather helpful. Bryce and I put together a few options for forward planning that would make recovery easier.

If government wishes to return to Wellington, it will have to move incredibly quickly to demolish its own broken buildings and start building new. No multiyear processes for figuring out the optimal design of some new government precinct, just start building. That on its own would do a lot to avoid a death-spiral.

I also expect that we need a liability regime around dangerous buildings impeding downtown access. If the risk imposed by one office tower means that adjacent office towers cannot be occupied, the owners of the adjacent towers should be due compensation from the risky tower. That compensation should fully reflect the costs they face after some minimally-reasonable period after the earthquake. At the margin, this would mean more buildings are demolished rather than held for extensive rehabilitation periods during which they continue to impose risk on neighbours. Getting this regime right would mean a much shorter cordon period.
As postscript, I worry a lot about uninsurable risk. The odds each year are small, but if it happens, there are losses against which it is currently impossible to insure. Homeowner insurance and income protection insurance don’t cut it. There are no Case-Schiller indices that could let you buy options that would act as insurance. There isn’t any obvious stock portfolio that would track.

And there are no retail-level parametric insurance products that take even the simple form “Pays $x if a Mercalli VIII event hits Wellington in 2018; pays $0 otherwise.” The fair price of that contract is 0.833% of the contract’s price; I’m more than happy to pay a reasonable margin over it to lay off that risk. But contracting costs are too large for anyone to offer this contract to just me, and too few people would be willing to pay say $8300 per year for a contract offering $500,000 in case of earthquake. Not that I’ve been offered a quote – that’s just a doubling of the fair price to reflect a cost of risk transfer. Absent any substantial demonstration of people-other-than-me being interested in buying it, it’s worth nobody’s time to even offer a quote.

Even Paddy Power declined to quote me a price for a bet on it; bets are the same thing as insurance.

I’m not worried enough to flip to being a renter rather than a home-owner, but I do wish we were closer to Arrow-Debreu worlds. Hopefully, blockchain parametric insurance options will get us there.

Image: Christchurch City (New Zealand) from the Port Hills. Wikimedia commons.