In late September and early October last year Hurricane Matthew smashed through the Caribbean. Winds of up to 145 miles per hour killed hundreds of people and caused extensive damage, devastating huge swaths of the region. It was the worst storm to hit the area in a decade.
As ever when this sort of storm hits, appeals for help were not far behind. But many of the worst hit countries were part of an insurance scheme that paid out within weeks, rather than the months that it sometimes takes for aid to arrive.
By mid-October, the Caribbean Catastrophe Risk Insurance Facility had paid US$23m to Haiti, one of the worst hit by the hurricane. The policy also paid out to Barbados, Saint Lucia and St Vincent & the Grenadines.
The Caribbean scheme is not the only one of its type. Across the world, similar initiatives are proliferating, from insurance against pandemic outbreaks to cover for the cost of droughts and floods. While insuring for natural disasters is not new, the use of insurance as a form of aid to developing countries is gathering pace.
Supporters say that using insurance to respond to natural disasters can help to save millions of lives in the world’s poorest countries, by putting in place a quicker and more predictable source of funds when the worst happens.
In 2015 the G7 promised to deliver climate risk insurance to 400m of the world’s most vulnerable people by 2020. Germany, Japan and the UK are leading the charge to put the promise into action.
“We are on the cusp of something really exciting,” Lord Bates, the UK’s international development minister, told a conference last month on the issue in London. Jo Scheuer, director of climate change and disaster risk reduction at the UN Development Programme, recently called it “a game changer” for the developing world.
Taking out insurance can leave African states “better equipped” to manage climate change, says Mohamed Béavogui, director-general of African Risk Capacity, an insurer set up by the African Union to help countries finance disaster response. Otherwise, “by the time the response to a crisis arrives, much of the damage has already been done”.
The insurance industry has assembled a host of executives, alongside the UN and the World Bank, into a group called the Insurance Development Forum, which aims to extend the use of insurance and risk management techniques to “build greater resilience and protection” in the developing world.
“For the first time ever we’ve had [insurance] carriers and brokers at CEO level sat round a table and working together. I’ve never seen so many alphas in the insurance industry working together,” Stephen Catlin, the IDF’s chair, told the London conference.
Michael Bennett, head of derivatives and structured finance at the World Bank, says that “between 2007 and 2017 there was $1.5bn of risk transfer [by the bank]. But there has been $1bn” of transactions in the past two months. He adds: “There is a lot of momentum now.”
There is plenty of potential for more of the same. According to Swiss Re, natural catastrophes around the world caused US$166bn of economic losses in 2016. Only $46bn of that was covered by insurance. Much of the rest, especially in the developing world, is left to humanitarian aid.
“The current humanitarian system works really badly,” says Owen Barder, a vice-president at the Center for Global Development, a think-tank.
He says there are two big problems. The first is the often late arrival of aid. Countries only ask for help when problems become apparent, and that help can take months to arrive. By that time the problem has become much worse, and much more expensive. Second, aid is unpredictable. Governments that ask for help have no idea what they are going to receive, and so cannot make firm plans for how to spend it.
Stefan Dercon, chief economist at the UK government’s Department for International Development (DfID) and an Oxford professor of economic policy, says: “When there are appeals, often less than half of what is appealed for is actually collected.”
So-called parametric insurance, say its backers, can provide solutions to both problems. Here, the insurance policy pays out not when there is firm evidence of a loss, but as soon as the first sign of trouble emerges. That can be when a storm happens, for example, or in the early stages of a drought.
The policyholders, in this case the governments of the countries affected, know exactly what they are going to get and when they are going to get it. They can then spend the money in the worst affected areas.
“If you give people money straight away, they can stay put. Otherwise they leave their land and sell their assets in a fire sale, which turns something bad into an absolute catastrophe,” says Mr Barder. Prof Dercon adds: “Faster payouts really help with slow onset disasters such as droughts.”
A recent US$322m pandemic insurance bond sold by the World Bank shows how this should work. Governments should receive money in the early stages of an outbreak, when there is still time to contain it. According to the DfID, it would have cost $5m to contain the 2014 Ebola outbreak when it was first detected in Guinea. Eight months later the cost had increased to $1bn. The pandemic ended up killing more than 11,000 people.
Kenya has introduced an insurance scheme to help farmers deal with climate change, whereby the government pays premiums for up to five cows in arid areas in the north and north-east. At times of drought, the payouts are designed to provide food and other necessities to keep the animals alive, not to compensate for dead cows.
“It’s all about prevention rather than replacement,” says Richard Kyuma, who runs the Kenya Livestock Insurance Programme. “It’s a new idea and there’s a lot of research going on to make it even better.”
Despite those potential benefits, not everyone is convinced that insurance is the answer to disaster response in developing countries. “There is a lot of hoopla from the insurance sector and from donors,” says Debbie Hillier, senior humanitarian policy adviser at Oxfam. “The rhetoric has massively gone beyond reality. They say that insurance can make everything fine but that’s patently not true.”
One of the big problems, say critics of the idea, is called basis risk. This is when the insurance policy does not pay out because the specific nature of a particular disaster does not meet the conditions laid down in the policy.
A recent drought in Malawi showed basis risk in action. According to a report from Action Aid, an NGO, the country paid $5m for an insurance policy from African Risk Capacity, which was supposed to cover drought.
But when the rains failed last year and 6.7m people needed help, says Action Aid, the policy did not pay out. That was largely because the types of maize seeds used by the farmers were not the same ones envisaged by the policy, so the damage caused by the drought was worse than the insurer’s model predicted. After adjusting its model, ARC agreed to pay out $8m but Action Aid says the total cost of the drought response was $395m.
In its report, Action Aid argues that: “The G7, World Bank, Insurance Development Forum, ARC and others promoting the expansion of climate risk insurance markets for the poor and vulnerable should pause and reconsider this quest in the face of a lack of evidence of its equity and effectiveness, and indications that it may be exacerbating inequality and vulnerability.”
ARC responded that some of the NGO’s claims were flawed and its recommendations were misguided. Thomas Kwesi Quartey, a Ghanaian diplomat who is deputy chair of the African Union commission, says the ARC has paid out $34m over the past three years to countries affected by drought, including Senegal, Niger and Mauritania. “This is solidarity combined with innovation in practice,” he says.
Rowan Douglas, chair of the Insurance Development Forum’s implementation group, says the experience in Malawi shows there will always be some basis risk with this type of insurance.
“The question has got to be: if these instruments work effectively say 90 per cent of the time, is that better than not having them at all and losing all the benefits of this important innovation?” Mr Douglas asks. He says there are many situations where schemes such as the ARC have worked well: Haiti’s rapid payout for Hurricane Matthew is an example.
One of the solutions to this problem is better modelling. Risks in Europe and the US are extensively modelled, partly because there is a well developed insurance industry. Insurers will know, sometimes down to a few metres, the flood risks for a specific property.
That is not the case in the developing world. “Only around half the world’s population is covered by mainstream insurance risk models at the moment,” says Mr Douglas. “Although the expertise has been around for three decades.”
Experts say climate change, which could make weather patterns less predictable, is compounding the modelling problem. But even if the models can be more finely tuned and the basis risk cut down, there are still obstacles to the wider use of insurance in response to natural catastrophes. One is the question of who pays the premium for the policy. In some of the schemes operating at the moment, donor governments pay. In others, it is the recipient countries.
According to Mr Barder, there are reasons why both donor countries and recipients are reluctant to pay up. “Politicians have short-time horizons,” he says. “You could pay a premium while in office and not get a payout. Why would you spend your scarce resources on an insurance policy?”
Likewise, he says, some donor countries hold back. “It is very rare for donor countries to pay the premium . . . there is a real reluctance,” he says, partly because some donors prefer to wait until the aftermath of a disaster to decide how much they want to give. There is also, he adds, some suspicion of the private sector.
Finally, some NGOs argue that the focus on insurance detracts from more important work in building resilience to natural disasters.
“The insurance debate has crowded out everything else. There is a place for insurance, but it is a small place,” says Ms Hillier. “It is crowding out disaster risk reduction and social protection, where there are massive gaps.”
Richard Ewbank, global climate adviser at Christian Aid, says: “Having a more efficient system to shovel money out of the door is important but we need to focus on early action and early warning to bring the costs down to a more manageable place.”
Even supporters of the idea argue that there is a long way to go. “We are in the 21st century when it comes to lending to the developing world, but only in the 12th century when it comes to risk,” says Prof Dercon.
“We are just trying things out. If something doesn’t work well, then we’ll learn the lessons.”