In a recent assessment of the success or otherwise of pooled service delivery initiatives in the Pacific island region, Matthew Dornan and I did not consider the World Bank-led Pacific Catastrophe Risk Insurance pilot as we felt it was too recently established to provide sufficient information for us to evaluate. Being a pilot it has not had a great deal of exposure. However, it has generated some recent discussion (unfortunately this item credits the Asian Development Bank with having established the scheme) further to its recent payout of US$1.9 million to the government of Vanuatu in the aftermath of Cyclone Pam. This is the second time the scheme has made a payment to one of the pilot member countries, the first being US$1.27 million paid to Tonga further to the passage of Cyclone Ian during January 2014.
Less positive was the experience of Solomon Islands, who withdrew from the scheme after experiencing two events (the Santa Cruz earthquake and flooding in March 2014), neither of which triggered a payment from the scheme despite the fact that they had very severe impacts on the country socially and economically. This has, it would appear, led to the consideration of increasing the range of responses available to better suit members’ needs. Which is what pilot projects are for – to learn.
In 2014, the Forum Economic Ministers’ Meeting (FEMM) received a paper [pdf] that documented the experiences of the member countries that had participated in the scheme up to that point and captured the lessons learned from the pilot project. It highlights the need to develop a wider range of responses with particular reference to the experience of Solomon Islands highlighted above. Further to consideration of this information, the FEMM decided that the scheme should continue and requested that the World Bank examine ways in which this could be effected, including options for future financing. To this end a workshop was convened in Suva in March of this year.
The $1.9 million payout to Vanuatu will, it is true, barely scratch the surface in terms of what is required to finance relief, recovery and rebuilding in Vanuatu further to the impacts of Cyclone Pam. This insurance scheme is not intended to meet all costs, rather it is designed to be able to respond quickly to assist governments with immediate/short-term costs associated with the disruption of critical service delivery. Given the enormous impact of Cyclone Pam it will go only a small way to doing that. But it constitutes one of only a handful of direct cash injections into government coffers since the cyclone (the others being release of funds by the Reserve Bank and cash donations from Papua New Guinea and Solomon Islands). In addition, the promptness of response will add to its credibility not only in Vanuatu but elsewhere in the region and we can expect that this mechanism will form part of longer-term discussions about disaster preparedness.
More generally, how does this example of pooling stack up in terms of success or otherwise? Here are some snapshot answers to the questions we used in 2013 as the basis of our assessment of 20 pooling initiatives:
Was the pooling initiative ever implemented? – as a pilot, yes and we are still waiting to see if and how it will be expanded.
Has the pooling initiative been sustained over a period of time or did it cease? – to be confirmed.
Has it remedied a service provision deficit at national or sub-national level? – yes, it has provided access to a service that was previously unavailable to Pacific island governments and has utilised bulk purchasing power to achieve costs savings of approximately 50% to each country.
Has it delivered a service or good other than capacity building? – yes.
It is true that there has been a significant amount of donor support for this initiative with the World Bank providing technical support and the government of Japan playing a large part in subsidising the costs of members’ premiums (other than for Cook Islands who paid for their premium for this season). But each of the members of the pilot has made a nominal contribution (US$20,000) in each of the seasons that has been covered reflecting a degree of high-level political support that we struggled to find in relation to some of the other activities we examined.
Tess Newton Cain is the Founder of the Devpacific Thinknet and a Visiting Fellow of the Development Policy Centre. She lives in Vanuatu.
I agree that insurance is an underutilized part of the disaster risk management toolkit in the Pacific, and that PCRAFI is a useful experiment to look at the specific risk of emergency cash. It’s not unheard of, however, for governments to consider blanket insurance coverage or protection schemes – although it is less prevalent in developing countries (Turkey has an earthquake insurance scheme; advanced countries like New Zealand’s EQC and Japan’s earthquake insurance scheme are also good examples where countries negotiate bulk re-insurance deals for blanket coverage).
That said, that type of insurance is likely to be prohibitively expensive for small island developing economies, especially those in the Pacific, without some form of pooling, reinforcing your post’s basic premise.
Thanks for this and especially for the examples of governments that do purchase insurance.
Hi Tess. If the cost of the cyclone is something like $300 million, as Matt suggests in his blog today, it does make you wonder why the payout is only $1.9 million. Doesn’t really seem like an insurance program, maybe a rapid early-payment scheme?
Thanks for this comment. Matt’s assessment is in relation to economic output not the value of damage & loss sustained – we won’t have a dollar figure for that until we get the PDNA report on the 23rd. The payout is not designed to cover the cost of damage and loss, it is designed to provide a rapid injection of cash to governments to assist with the maintenance/restoration of critical services in a post-disaster scenario. And, unlike a lot of other things, the payout does not come from donors, it comes from the private sector (the global insurance market), leaving more of their funds available for other things. Governments generally self-insure when it comes to physical assets such as buildings so it would be odd for the government of Vanuatu to try and take out the sort of insurance cover you might have for your house where you would expect to receive a greater proportion of the value and/or cost of repair. Even if they did take this out (assuming it existed) it would not respond as promptly as this scheme has. As we know it is a pilot scheme so there are opportunities to refine its design and operation, including the modelling on which it is based if that is necessary. Or donors could set up a MDTF (or something similar) that can respond as quickly and to an equal or greater quantum if that is preferable, but it means they are carrying the risk that they may need to pay out to more than one country in a given period – under a scheme such as this the risk is transferred to the insurance market. For what it’s worth I was just at a press conference at which the Minister of Finance made specific reference to this payout with approval and confirmed that Vanuatu would be maintaining its participation.