Given that it is increasingly difficult to get affordable earthquake insurance, we may have to radically revise the system.

Suppose you are exposed to a risk that you cannot protect yourself from. You might set aside a capital sum which would cover the loss if the event occurred. But if the loss was your dwelling from an earthquake, you’d have to put aside its entire value, effectively doubling the cost of the house even though the probability of a total wipeout is low (or so you hope).

Instead we use insurance to cover the contingency. Broadly there are two kinds of insurance that are sometimes traced back to two instances (illustrative, but not entirely historically accurate).

One involved a group of Swiss farmers who agreed that if any of them suffered a loss of a cow, the loss would be shared among all of them. Observe that such social sharing can only work if the shock does not affect a large part of the group.

The second kind of insurance is said to have led to the development of Lloyds of London. It involves paying a fixed sum to someone to take the risk of an event happening. The counter-party covers a lot of similar events and gambles that only some will happen, with the expectation that the total claims are less than the fixed sums paid to them; if it does not, they make a loss. Typically they, often a corporation, take on a portfolio of such gambles in the likelihood that an earthquake in Wellington won’t occur at the much the same time as ones in San Francisco and Tokyo. It is possible the insurance company will not be able to cover its losses and goes bankrupt. (AMI had state help to prevent it going under when it was too exposed to the Canterbury earthquakes. The world’s largest insurer, AIG had to be nationalised by the US during the GFC.) In which case the insurer would lose their cover.

It is much more complex than this, but the important thing here is that there are these three responses: self, share (or social) and shift insurance. Of these ‘share’ and ‘shift’ are the significant policy options (although if there is an excess to one’s insurance that is ‘self’).

In 1993 Earthquake Commission Act New Zealand made a dramatic change to its earthquake insurance changing the balance between share and shift. I focus on dwellings – there were also changes to commercial buildings – and I simplify.

Before, the essence of a dwelling’s earthquake recovery protection was a share scheme in which the risk was covered by the New Zealand government. Insurance levies were paid to the Earthquake and War Damage Commission; if there was damage, the EWDC paid for it. If the EWDC had insufficient reserves, the government (taxpayer) covered the deficit. So the residual individual earthquake risk was shared across the entire New Zealand community.

After 1993 only the first $100,000 of damage was covered by the renamed Earthquake Commission (EQC) to which insurance was paid. (The government has recently hiked the level to $300,000. The $100,000 in 1993 would purchase about $200,000 of building construction today.) The rest of the earthquake protection has to be purchased from private insurers. Basically, since 1993 there has been a rebalancing from sharing to shifting – a privatisation of earthquake protection.

(EQC insurance also covers other major catastrophes such as natural landslips, volcanic eruptions, hydrothermal activity, tsunamis and the impact of storms and floods on residential land.)

There were numerous reasons justifying the change. The minister in charge of the bill was Ruth Richardson so that there would have been a neoliberal-minimalist state element among them. (The change was discussed by the Rogernomics Government.)

The winding down of the sharing state was happening in other policy areas. The Richardson-Shipley ‘reform of the welfare state’ was a reversion to the charitable aid which had preceded the 1898 Pensions Act, the first major social insurance in New Zealand.

The notion of social insurance, where certain burdens are shared across the nation, hardly appears in today’s rhetoric of even the Labour Party. When did it last use Nash’s ‘the young, the old and the sick will have the first call upon the state’? An exception is the proposed social insurance for the unemployed who have been made redundant. One is struck how the dominant criticism is from minimalists who see the levies as a tax, rather than a means of sharing the burden of adjustment across all workers. (I assume they would not object if private insurance did this, but it wont.)

Another justification for the 1993 earthquake legislation was the Treasury trying to minimise its exposure to risk because it felt the taxpayer was overcommitted. It worried that it might not be able to fund other costs associated with a major disaster, such as social welfare payments and the restoration of government facilities.

Following the Canterbury earthquakes the Key-English Government chose to fund its reconstruction out of current revenue rather than imposing a special levy, say on income tax, justified by the entire nation sharing the burden. That would have been the approach of the traditional welfare state. (The consequence was that public spending was squeezed; we are paying the cost of the social damage it caused through to this today.)

My view is that a role of the government is to do things which the private market cannot do effectively. There were private insurance companies willing to sell earthquake insurance – take the risk – so there was a case for changing the balance from share to shift. Note that the insurance companies do not take all the risk but shift some on by reinsuring with others offshore, as does the EQC.

Thirty years later the system is not working well. The Canterbury and Kaikoura earthquakes have indicated that seriously damaging earthquakes are more likely than we once thought. Apparently a period of earthquake quiescence has come to an end; there is increasing uncertainty about the future risks of these catastrophes and reinsurers are becoming more wary. (Climate change is another source of increased uncertainty.)

The situation is further complicated by there being only three significant private insurers, which means the market is not fully competitive and may work less than smoothly.

In any case, private earthquake insurance is becoming increasingly expensive. For instance Wellington apartment buildings of a certain age and height are suffering extraordinary increases. In one case I saw, private earthquake insurance costs have increased since the Kaikoura earthquakes by over eightfold.

It is not the private insurers’ fault; they cannot find cheap reinsurers. The risks since the Canterbury earthquakes have become much more uncertain. They may be able to find a reinsurer but not at an acceptable price while the local domestic insurer may concerned about being too concentrated – as the AMI experience demonstrated. The best approach in such circumstances is prevention, but retrospective fitting of older buildings can be very expensive.

Perhaps we should revisit the 1993 decision. Should we change the balance, going back to greater use of sharing the risk across the nation rather than depending on shifting all of it offshore?

Insurance is a complex area. We have had numerous inquiries into the operation of the EQC but not a single one as looked at the self-share-shift balance. Hopefully, we will get around to it before the next great earthquake.