I admire SA independent senator Nick Xenophon hugely. He’s a rare combination of brains, enterprise and principle. I knew him at Adelaide University; he had all those qualities then, and he seems to have kept them intact over the quarter-century since.
But I have wondered for weeks why he is so keen to have the states insure against natural disasters.
Today I may have found the answer to my puzzle. Xenophon is being advised by former insurance industry executive John Tsouroutis. When Malcolm Farr profiled Tsouroutis in a recent article for The Punch, the picture emerged of a bloke who thinks outside private insurance is always good.
This seems to me exactly wrong. At a government level, outside private insurance is almost always bad.
Before we go any further, it pays to remember that insurance only spreads risk rather than eliminating it. Remember also that private insurers need to make profits. (I make these obvious points here only because they seem to have disappeared in the debate over state insurance.) Insurance is likely to cost you more than not being insured. You don’t insure to save money. You insure to avoid unlikely episodes that would cause you too much stress to fund yourself when they happened. For instance, you insure your family against the chances of its principal breadwinner being killed or disabled, because your savings won’t cover the costs of this sort of unlikely but potentially devastating blow.
So I was startled to hear that many state governments currently take out natural disaster insurance. As a general principle, governments should not insure against Bad Things, be they floods, bushfires or earthquakes – or, for that matter, recessions. Not only should Queensland not take out such insurance, but the other states should stop doing so.
And here’s why:
- Insurance works best when it covers events which would overwhelm you if they happened. For a government, a $7 billion disaster like Yasi is big, but hardly overwhelming. Queensland is a $250 billion economy, and the state government brings in revenue of more than $37 billion a year. Pre-disasters, the Queensland government was projecting deficits of more than $3 billion for 2010-11 and 2011-12, due to the impact of the economic slowdown. Borrowing to pay $7 billion for Yasi would add another $0.5 billion a year to that figure at most – large, but manageable at less than 2 per cent of overall expenses. If you want Queensland to insure against huge blows to its economy, forget cyclones – get the state to buy insurance against recessions.
- Insurance works best when it covers unlikely events. In Australia, natural disasters do not fall into this category. For the people being hit by them, cyclones and bushfires are frequently once-in-a-lifetime events. But for governments, over the course of a decade or so, some sort of disaster is statistically likely. We’re a nation of droughts and flooding rains, not to mention bushfires and cyclones. We know this stuff happens. If we’re smart, we’ll have our governments put aside some money for that rainy day.
And of course, sometimes disaster is not even completely surprising to the people it’s happening to. The north Queensland coast, for instance, attracts cyclones like a summer evening barbecue attracts mozzies. That’s why when Cyclone Yasi hit, the locals kept comparing it to the recent Cyclone Larry. - Insurance works best when it covers events that are out of human beings’ control. For individuals, catastrophic health problems are like that. Cyclone, flood and bushfire damage are not always avoidable, but their costs can be minimised, for instance by not building flood-vulnerable buildings in flood plains and by building for high wind resistance. Insurance acts as encouragement not to worry about the things you insured against – a problem known as “moral hazard”. A government insured against flood losses will be more likely to unwisely allow flood-vulnerable building in a flood plain, and to ignore or (more likely) under-fund flood mitigation efforts.
Private insurers deal with this moral hazard by jacking up their premiums. A government insuring against floods will likely end up paying a private sector premium to escape the consequences of its own expected inaction and short-term lack of will. - Governments possess vast abilities to tax and borrow. If they won’t put aside money to cover the costs of a natural disaster, their next best course is to raise the money after the event, not to insure. Their funding costs are probably going to be lower than the insurer’s borrowing costs.
(Note that on these arguments, smaller governments may have a case for outside insurance. This might help explain the perspective of John Tsouroutis, Xenophon’s adviser; he most recently ran the Northern Territory government-owned insurer, TIO.)
For most governments, the best insurance against floods, bushfires, earthquakes and recessions is four-fold:
- Implement good policy. Minimise building in floodplains where floods would destroy the buildings’ value. Create anti-cyclical policies which lessen the boom-and-bust cycle. And so on.
- Build government rainy-day funds. You can call your fund a Future Fund, a sovereign wealth fund or just a really big budget surplus. At a state level, it will guard against bushfires, floods, cyclones, earthquakes and other blows. At a national level, it will guard against the day when the banks threaten to collapse, the Chinese stop buying steel, or the commercial property sector suddenly loses faith.
- If a big natural disaster comes along at a time when a government needs to be tightening fiscal policy anyway, a disaster levy makes sense, even though it is somewhat opportunistic. Right now, a levy makes sense. If the Queensland floods had hit two years ago, in contrast, the best reaction would have been government borrowing.
- Don’t pay for private insurance.
Who stops this happening? You could blame a lot of people – governments, self-interested private insurers, and so on. And Senator Nick Xenophon and his adviser sure aren’t helping matters.
But the real blame lies with people who constantly protest the sight of a government Budget in surplus, and people who forget that over the course of years, bad things always happen and need to be paid for. In short, the people who most need to take a longer-term view are voters.
I disagree. Private insurance provides substantial advantages over self-insurance for political reasons.
Firstly, it is more resistant to the future Parliament problem. Parliament cannot bind its future self. No future fund, no rainy day balance, is immune from being appropriated by an Act of Parliament at any time for any reason, such as a looming election. When money is given to the insurer, it is now out of the reach of the bright young things in Treasury.
Secondly, private insurance premiums provide an incentive against introduce retarded policies such as building in flood zones. They also provide independent discipline to develop better asset tracking, management and planning.
Thirdly, raising taxes immediately after a disaster is neither popular nor very helpful to the economy.
Fourthly, any money spent on disaster recovery must compete with other Treasury items. It is possible and indeed likely that various sacred cows being unslaughterable will lead to flatly irrational outcomes. Insurance, being an outside source of funds, is immune to this problem.
Does insurance for state governments have anything to do with AAA credit ratings?
I think the response Jacques is that future governments are also likely to stop paying insurance premiums for precisely the same reasons you are saying they will dip into any fund. The end point in both cases being that when a disaster or recession strikes, we will go into deficit because we either failed to accumulate a fund to deal with it, or we stopped paying insurance premiums. Of course, viewed that way, is an occasional deficit under those circumstances a bad thing compared to the alternative of diverting tax money into these funds or into insurance premiums before the fact?
At last someone out there realises a state government has a capacity for self insurance that a household or small business doesn’t. Insurance by anyone big enough to carry the risk just donates money to insurers. Does BHP insure its cars?
Paul Montgomery, ratings agencies can do these sums as well as anyone. They look at the long term. They know that insurance is expensive and not best practice. If anything, the decision to insure should weigh against a government when the rating is being assessed. Insurance is a source of costs, not a source of funds.
Paul
I’m not an expert on insurance law or the basis on which they price their product. However, as a matter of first principles I think insurers fix premiums (irrespective of whether the insured is a government, corporation or individual) on the basis of their assessment of the risk against which they’re insuring, not the creditworthiness of the insured. An insurer is not lending money to the insured but insuring against the occurrence of a specified risk. Non-payment of the premium isn’t a risk for the insurer, because the insurer as a matter of contract either gets paid the premium annually, monthly or weekly in advance, and if the insured doesn’t pay on the due date he/she/it is no longer insured.
Ken, if your reply is to me (too many Pauls) I can’t see your point. Why would a government want to give away money to an insurance company when it’s big enough to cover its own risks? There’s no point. And risk to the insurer is irrelevant to me.
No I was answering Paul Montgomery at #2. I agree with David Walker. It’s generally stupid, wasteful and pointless for a government to insure risk (other than by funding the risk by maintaining a surplus or future fund).
The biggest problem with government self-insurance over largish items (like multi-billion dollar expenses) is item [2] on David”s list above. It is hard for governments to save large chests of funds for a “rainy day”.
What would these funds be denominated in? An Australian dollar is nothing more than an IOU note signed by government… so for a government to keep a big chest full of IOU notes promising something from itself to itself is a bit pointless. I mean, I could write IOU notes to myself as well… really won’t help when my house burns down because then I’ll be wanting help from someone other than just myself.
The government could stockpile gold and silver (universal currency), but that would freak people out. They could stockpile foreign currency on the basis that if they got into trouble they could pay foreigners to help, a suitably chosen basket of foreign currency would solve the denomination problem.
However the next problem is preventing the next party at next election from making promises all over the place to spend off the “rainy day” chest and porkbarrel all the marginal seats. You know they would do this in a flash. They really can’t be trusted and I must admit that if I was in a marginal seat, I’d vote for the pork, because why shouldn’t I?
Private insurance contracts are very much like a PPP but for the purpose of government “rainy day” saving. Like any PPP, the contract details should be public so we know what we are getting, but because there is actually a contract we know what we are getting.
I agree with the post in general, but there are two reasons I could see for why government might choose to buy insurance.
Firstly, I would note that insurers have a comparative advantage over government in assessing the risks of natural disasters. Picking up on Tel’s language and assuming that the market for insurance is reasonably competitive, then purchasing insurance is probably a cost-effective way of working out how much to put aside for a (very) rainy day. There are better things for Treasury-esque folk to be doing then working out future liabilites for natural disasters.
Secondly, governments might choose to purchase insurance because it is cheap. The premiums reported for Victoria and New South Wales seem like pretty good value. So it may not be a case that the governments think they should buy insurance, but were offered a good deal and took advantage of it.
Possibly the question is not the best one.
I suspect that there is a level of damage above which the State and Federal governments would struggle to fund a recovery.
The question is thus, not whether or not insurance is advisable, but at what level and for what purposes.
Bushfire seems to me to be a lumpy cash flow, but still the kind of thing for which a community, through its government, can manage recovery. Major earthquake (thankfully rare in Australia, hence premiums should be low)… perhaps. Cyclones are an annual event in Qld but truly massive or frequent ones could be beyond the capacity to self-fund recovery.
So, blanket disaster insurance is not justified and is a waste of money. Targetted disaster insurance with a suitable excess, say $5B in the case of Qld’s cyclones, is another matter.
Finally, the commonwealth government already underwrites the States against losses of this nature, so disaster insurance via the for-profit insurance industry is not justified at State level. Read my second sentence above. If insurance is to be obtained, it should be obtained at national level, and then only for those risks and at that level for which the combined strengths and resilience of all levels of government will be inadequate.
I haven’t seen any such in the past 60 years, so the issue is probably a dead letter.
Senator X, nice man and smart though he undoubtedly is, has this one wrong.