We Can Fix Emergency Aid with Disaster Insurance – Stefan Dercon and Theo Talbot

Summary

The CGD Podcast: We Can Fix Emergency Aid with Disaster Insurance

Guests: Stefan Dercon (Chief Economist, DFID & Oxford Professor), Theo Talbot (Senior Analyst, CGD Europe)
Host: Rajesh Merchandani (CGD)
Date: August 10, 2016

---

Episode Overview

This episode explores how the international community could radically improve disaster relief by moving from reactive fundraising to proactive financing—specifically through disaster insurance. Stefan Dercon and Theo Talbot argue that insurance and risk pooling can incentivize better preparedness, provide reliable funding for disasters, and strengthen government capacity, all while reducing delays and inefficiencies plaguing the current aid model.

---

Key Discussion Points & Insights

The Problem with Current Disaster Relief Funding

• Reactive Approach: Current model relies on donor appeals and public fundraising after disasters strike.
  • Unpredictable funding—often, only about half of pledged donations arrive.
    • Stefan Dercon [01:23]: “Whenever we do these appeals, maybe about half of these appeals get actually paid up. Now, you may have made some kind of plan, but if half the money only comes in, how are you going to use it?”
• Incentive Failures: Scramble for resources fosters poor planning, fragmented and uncoordinated responses, and competition among NGOs to “plant the flag.”
• Capacity Depletion: Delays and uncertainty can erode government capacity, leaving people without timely support.
• Case Example: Haiti earthquake, where slow donor responses meant the first meaningful payout came from an insurance pool—not from aid organizations.

Insurance as a Solution and Its Benefits

• Thinking in Terms of Liability: Insurance requires acknowledging what’s at risk and planning for specific trigger events.
  • Shifts mindset to one of preparedness and deliberate financial planning.
  • Encourages governments to consider: What do we want to protect? Who do we protect? Under what circumstances?
• Clarity and Predictability: By knowing coverage and triggers in advance, governments and donors can act rapidly and effectively when disaster occurs.
• Multiple Financial Tools: Insurance is one instrument—others include contingency funds and contingent credit lines, all of which require planning and clear criteria for use.
• Incentive to Reduce Risk: Insurance premiums provide motivation to invest in risk mitigation, as lowering risk means lower premiums.

How Disaster Insurance Can Work

• Risk Pools: Regional pools already exist (e.g., African Risk Capacity [ARC], Caribbean Catastrophe Risk Insurance Facility [CCRIF], PCRAFI in the Pacific).
• Parametric Insurance: Payouts are triggered by measurable parameters (e.g., wind speed, earthquake magnitude) rather than subjective assessments.
  • Theo Talbot [07:10]: “Parametric insurance kind of takes the reliability of the insured person, the insured party, out of the equation...because these are contracts written on things like wind speed or the violence of an earthquake. These are geophysical data that can’t be affected by the insured or the insurer.”
• Real-World Example: Haiti’s first earthquake payout came from a Caribbean risk pool using parametric contracts, paving the way for fast and direct government relief.

Strengthening Government Systems & Response

• Integration with Social Protection: Kenya links their ARC insurance to social protection programs, allowing rapid scale-up to more people in case of drought.
  • Stefan Dercon [11:00]: “They’ve linked the ARC...to actually a social protection system that exists all the time. But the insurance policy allows the system to scale up during a drought to reach more people...That’s the kind of way that we want to do.”
• System Incentives: Participation in risk pools incentivizes governments to set up effective response mechanisms in advance, contrasting with improvised post-disaster systems.

Financial and Political Challenges

• Affordability and Buy-In:

  • Difficulty convincing cash-strapped governments and populations to pay insurance premiums for uncertain benefits.
  • Poorer countries have competing spending needs, making insurance a tough political sell.
    • Rajesh Merchandani [12:25]: “I certainly don’t like having to write a check to the insurance company...How do you get cash strapped governments...to decide to spend money on insurance?”
  • Political incentives often favor visible, post-disaster spending over less tangible, preventive investment.
    • Stefan Dercon [15:50]: “There’s plenty of evidence...that governments or people in power...can get electoral benefit from being seen to be the benefactor after a disaster.”

• Encouraging Long-Term Thinking: Policymakers and donors should champion preparedness as being in the long-term interest, despite political headwinds.

  • Stefan Dercon [15:50]: “It’s actually up to those of us who are involved in preparing policy...to bring that long term dimension.”

The Role of the Private Sector and Remaining Gaps

• Industry Engagement: Insurance industry is increasingly interested in developing these new markets, seeing benefits for risk diversification and business growth.
  • Theo Talbot [16:25]: “The insurance industry is increasingly...ready to work on this platform.”
• Limits of Insurance: Not all risks are easily insurable or contractible—so a robust humanitarian system remains necessary as a “last resort” insurer:
  • Stefan Dercon [17:12]: “We’ll still need a humanitarian system that actually can bail the whole system out...and cover those risks that are hard to plan for.”

Policy & Political Dimensions (UK/Donor Perspective)

• UK as a Leader: UK’s large humanitarian spending, insurance expertise, and political momentum make it well placed to lead innovation in disaster insurance.
  • Stefan Dercon [18:05]: “As the UK we spent billions on humanitarian aid...getting these incentives in place...is one part of making [the] humanitarian system function better.”
  • Leveraging London’s status as a global insurance hub supports both development and national interests.

---

Notable Quotes & Memorable Moments

• Stefan Dercon [01:23]: “If we wait until the disaster strikes and then we are going to scramble for resources...it's already quite late. And the problem is by doing this so late, we create very poor incentives to actually prepare ourselves properly for these responses.”
• Theo Talbot [02:59]: “When we insure something, we acknowledge that we have that liability… Preparedness creates government capacity.”
• Theo Talbot [07:10]: “Insurance doesn't set most people's heart racing, but Stefan and I...are grinning about the possibility of this.”
• Theo Talbot [09:02]: “The first payout [after Haiti’s earthquake] came from the Caribbean Sovereign Risk pool...That one little pinprick of light went so right.”
• Stefan Dercon [15:50]: “There’s plenty of evidence that...money that [would have been] spent for preparedness doesn't really give you electoral gain.”
• Stefan Dercon [18:05]: “The UK is the world center of insurance. Surely we won’t have that much objection...to actually harness some of the expertise and the capital that is present in London.”

---

Timestamps for Major Segments

• 00:05 — Introduction of guests and topic
• 01:16–02:46 — Why reactive disaster aid is problematic
• 02:46–04:10 — How insurance changes incentives, enhances preparedness
• 04:10–06:37 — Designing insurance contracts for countries (clarity, triggers)
• 07:10–09:44 — The appeal of parametric insurance; Haiti earthquake example
• 09:44–12:25 — Effectiveness of existing risk pools (ARC, CCRIF); integration with social protection
• 12:25–15:45 — Political and financial challenges for governments and populations
• 15:45–16:25 — Importance of long-term policy perspective
• 16:25–17:43 — Private sector’s increasing interest; need for humanitarian backup
• 17:43–19:49 — UK government’s commitment; donor policy implications

---

Further Reading & Resources

• Book: Dull: How Planning Ahead Will Make a Difference by Stefan Dercon & Daniel Clarke (World Bank)
• CGD Working Group: Led by Theo Talbot and Owen Barder — policy paper titled Payouts for What? Why Disaster Aid Is Broken and How Catastrophe Insurance Can Help to Fix It. (cgdev.org)

---

This episode is a comprehensive look at why international disaster aid needs reform and how innovative financial instruments like insurance can offer a more effective, coordinated, and proactive future for global humanitarian response.

Transcript

[Stefan Durkan]

Foreign.

[Rajesh Merchandani]

Hello, I'm Rajesh Merchandani. Thanks very much for joining me for this edition of the CGD podcast. Now, we've all seen the commercials on television. When a disaster strikes, we are urged to dig deep and send money. And we do. People are really generous. Generally when disasters like earthquakes or cyclones or famines happen, we can't stop natural disasters happening. And despite our best efforts, man made disasters like conflict and the human suffering they unleash are likely to continue. But is there a better way to fund the relief effort? The model of waiting till disaster strikes and then asking for money, could we do better than that? My guests today believe insurance can help in such circumstances. Stefan Durkhon is the Chief economist at the UK's Department for International Development and also professor of Economics Policy at the University of Oxford and co author of a new book called Dull How Planning Ahead Will Make a Difference. And Theo Talbot is my colleague based at CGD Europe in London, a senior analyst for cgd and he's also co running a new working group there on catastrophe insurance for humanitarian and emergency assistance. Stefan and Theo, welcome to the podcast.

[Theo Talbot]

Thank you, Vajesh.

[Stefan Durkan]

Thank you, Stefan.

[Rajesh Merchandani]

Let's start with the problem. Why is the way the world responds to disaster bad?

[Stefan Durkan]

If we wait until the disaster strikes and then we are going to scramble for resources, whether it's local governments having to go to national governments, or national governments having to go to international community or NGOs having to go and raise money from the public, it actually is already quite late. And the problem is by doing this so late is that we create very poor incentives to actually prepare ourselves properly for these responses. So when we do these responses on the basis of money raised afterwards, well, we usually don't know beforehand how much we're going to get. Whenever we do these appeals, maybe about half of these appeals get actually paid up. Now, you may have made some kind of plan, but if half the money only comes in, how are you going to use it?

[Rajesh Merchandani]

And that's what we're seeing more and more of.

[Stefan Durkan]

And we see that more and more. And we see all the time then on the ground those who've raised down the money saying, well, there's clearly not enough money around, let's keep it close to my organization, make sure we plant our flag and show that this is organization X, NGO Y or government Z that actually is providing the support. And then actually we also end up with a situation where the response is very uncoordinated.

[Rajesh Merchandani]

Theo we can't prevent natural disasters, but the point that you're making in the work that you're going to be doing in the working group is that we can better prepare. Why do you think insurance is one way to do this?

[Theo Talbot]

So I think Stefan has set out really eloquently, as always. Stefan is why we do things after the fact and why that's bad. It's costly. It leaves people left out of the process of rehabilitation, of rebuilding. It deletes government capacity. So we know that we can do better. And we have programs that are already in place in the Pacific, in the Caribbean and in Sub Saharan Africa, demonstrating how some of this can work at sovereign scale. When we insure something, we acknowledge that we have that liability. That's very different from the way we typically, as the global public sector, as donors, that's royal. We deal with catastrophe now, which is to wait until something's gone wrong, wait until someone says, I need this amount of money to tackle it, and then decide how much money to chip in. As a fragmented approach that doesn't acknowledge that we have the cost to bear. When we insure something, we do. And the second we have a cost to bear, we investigate ways to lower that cost. Preparedness creates government capacity. It means that we can invest money where it's most effective. And it means that we know how much we're going to have brought to bear because we insured ourselves rather than having to go to the international community and. And hope for the best, resulting in delayed, fractured and mismanaged funding.

[Rajesh Merchandani]

So the kind of disaster insurance you're talking about and the risk pools that exist at the moment, I mean, you talked about acknowledging our liability. When we insure something, we acknowledge it has value. That's a kind of tangible thing. But then if you're looking at the country, how do you ensure, again, what do you insure against and how do you ensure a country against disaster? How would you design those contracts?

[Stefan Durkan]

Well, I would actually make it very clear that insurance has a role to play in this. But let's not start with the product. The key here is to actually get the government to think, like setting up an insurance, like being very clear about what is it that it will want to protect, who it is that it will want to protect, under what circumstances, that you have very clear triggers. Insurance contracts have these nice features. It's very clear what is being protected. There's very clear circumstances in which this will happen. But so how will a country do this? Actually, once we have a sense of what we want to protect, who and under what circumstances, we'll actually go through very careful financial planning. Insurance is then one instrument that actually can take the risk, that can transfer that risk of the government's balance sheet and then actually in this kind of circumstance will get a payout. Governments can do other things as well. And it's actually important that it makes carefully these choices. It can also keep the risks on the balance sheet by, for example, setting up its own contingency funds. It can actually negotiate lines of contingent credit, meaning credit it gets on the circumstance, circumstances. So it can actually go through a whole range of financial instruments of which insurance is one. But they all have the same feature. There's a clarity about what is being protected, under what circumstances are we going to use it. And if we then manage to do it via an insurance style instrument, we'll actually transfer the risk on the balance sheet and we'll get these incentives that Theo talks about. Actually it clarifies how much it will be costing to actually get someone else to cover this risk. Meaning also that it makes it clear to us what money alternatively could spend of trying to, for example, reduce the risk and to get the insurance premium down and to get the cost of all these things at a more reasonable level.

[Rajesh Merchandani]

So using the kind of insurance framework changes the mindset of government to understand that the country is valuable and these threats are going to happen. But Theo, how would those contracts actually work? And how do you kind of like, you know, take into consideration the sort of problems that might arise, for example, when do they pay out? How do developing countries get access to the capital markets to be able to fund these? You can design for all of those.

[Theo Talbot]

Right? And I think this kind of links to Stefan's point that when we talk about insurance, we talk about insurance thinking about than insurance contracts. But insurance contracts, specific flavors of insurance contracts, can address all the problems you set out. So one of the big innovations in the last couple of decades in insurance specifically, we get quite excited about this. Insurance doesn't set most people's heart racing, but Stefan and I, for those who can't see us, are grinning about the possibility of this. Get excited about this thing called parametric insurance. Parameters are fancy words for variables. And parametric insurance kind of takes the reliability of the insured person, the insured party, out of the equation. Because these are contracts written on things like wind speed or the violence of an earthquake. These are geophysical data that can't be affected by the insured or the insurer. And that makes them transparent. It means that they pay out when the contract is triggered. It means that we can do this without reflecting on who we're insuring, because what we're ensuring is what really matters. And I think it's really worth, you know, in the suite of options that Stephane laid out, I get quite excited about the particular applicability of insurance models because we can very clearly contrast some experiences of vulnerable countries from the old way of doing things. What we think should, should be, should be limited to the most extreme cases, the uninsured risks and the new way in Haiti, in sort of post earthquake Haiti, we had this huge, fragmented, urgent response. Every donor was called. The prime minister of Haiti at the time said there'd be a donor conference. The donor conference happened three months late. The delays cost lives. A cholera epidemic eventually took hold. So we know that the old way of doing things is quite bad. The share of public resources that went to help rebuild Haiti, a vanishingly small share of those went through government systems. So they also deleted government capacity. But what most people seem to forget about is that the first payout before the World bank, before the imf, before the US Government responded, the first payout to Haiti came from the Caribbean Sovereign Risk pool, which used a parametric contract, realized that an earthquake had been triggered based on transparent geologic seismic data and paid directly into the government coffers. Now, of course, Stephane is right. Many other things went wrong, but we should be really proud that that one little piece, that one little pinprick of light went so right. And I think the question that Stephane will be talking about in the context of a working group that CGD has set up is how do we make that pinprick of light much more visible, much more accessible to other countries facing this kind of vulnerability?

[Rajesh Merchandani]

There are these risk pools that exist already. The African risk capacity P CRAFI is the Pacific version, and this Caribbean one as well. Is there evidence that they're working, they.

[Stefan Durkan]

Are doing what they set out to do, and they're one part of the jigsaw that actually we want to have in place. When in Haiti there was a payout, we're not necessarily saying that actually the money would have been used very well. So for us, it's very important to actually see this as one part of an effective system. It's an insurance like system where the financing on standby via the risk pool is there. Where we do need these other parts, very crucially, we need to get the investments in the systems on the ground that actually translated the transfer of the money into actually a clear response that benefits people. Now, the important thing here is that Being part of such a risk pool in principle could give you these incentives. It could force you to actually do this. And we see this for example in the context of ark. There's much more talk about that government's Ministry of Finance in Africa that are members of this pool have to have a clear system of how they actually will then disperse the money that they're getting.

[Rajesh Merchandani]

ARC is African risk capacity.

[Stefan Durkan]

Yes, the Africa risk capacity and that we're getting there. And so for example there is in the way in Kenya they've done it is that they've linked the arc, the Africa risk capacity product. They've linked it to actually a social protection system that exists all the time. But the insurance policy allows the system to scale up during a drought to reach more people and to have bigger payouts for the people affected. Now that's the kind of way that we want to do. Now these kind of things were not in place in Haiti, but actually membership of these pools could do this. So if you're asking how effective is it? It's very hard to do in this particular case, in the case of arc, for example, say well these are the number of life saves and whatever. But the one thing we really see, it's creating the incentives to actually getting the response capacity in place so that actually the systems are in place so that actually when something happens, responses are taking place and it actually reaches the people that we want to do, which is far better than the alternative. That if a drought is happening and then we start scrambling for setting up the whole system and I have to start thinking about who we'll target and then have to see shall we do it from government, shall we do it in parallel system and lots of valuable time gets lost as we see all the time in disasters that actually just making up the rules and probably losing the effectiveness.

[Rajesh Merchandani]

So you talked about incentives, but you know, few people like paying out for insurance. I know you guys get very excited about insurance. I'm less excited about insurance. And I certainly don't like having to write a check to the insurance company for anything frankly. Because you see the money going out and you wonder am I ever going to get anything back from that money? In a way, I hope I don't translate that dilemma that we have in the rich world to people in poor countries who can't afford insurance because they have to put food on the table. How do you get around that? And then how do you get cash strapped governments, both in rich and less well off countries to decide to spend money on insurance when they have Other spending necessities, how do you get around the politics of that? Or how do you get the buy in from people and governments?

[Stefan Durkan]

So you have, in a part of Kenya you have a lot of people covered by a social protection system, a cash transfer system all the time. However, when something bad happens, the system is designed to actually scale up to cover more people and to give bigger transfers. That scale up is financed by an insurance policy with the Africa risk capacity. And so as a result, actually the system can actually cater for more people, but in a very clear, well defined way beforehand that actually will be covered on top of that the better of people. They are clearly told, look, when the drought hits, you will not get a peel but actually payout. But actually what we can do is to offer you a product, a very well designed insurance policy that will actually pay out with the same triggers to you as well if you want to do. But then you have to participate. So you get a situation that in a very poor country at least you make it very clear to people who is being covered and who is not being covered and what are the alternative options for the other people, people, which is better than the alternative system where none of them really know. Probably all think the government will have to come and help. And then it's very hard to prioritize at a government level. Of course you know who will then pay. So the Kenyan government paid his insurance premium and it's not so straightforward because they have competing needs. But the point is that especially governments in countries where regularly or occasionally, I should have said, some of these big things happen and then they suddenly eat into the budget in a dramatic sort of way. They may well actually prefer to say why don't we pay every year 20 million, 30 million rather than once in a while 200 million. Or having to recover the cost of 200 million to actually have a response. We pay regularly for a premium. But yes, I totally agree there are really big political pressures. There's plenty of evidence that says that governments or people in power, they can get electoral benefit from being seen to be the benefactor after a disaster. While there is also evidence that money that alternative was spent for preparedness doesn't really give you electoral gain for this kind of reasons you have. And that's in all kinds of countries. This is evidence of us from India, from Mexico. And you know, and intuitively it makes sense. It's a very hard thing.

[Rajesh Merchandani]

Governing is about what's in the long term interests of a country, but government is about the short term politics.

[Stefan Durkan]

Well, no, that's A really good point there. And it's actually up to those of us who are involved in preparing policy, as I'm an official in UK government, to prepare policy to actually bring that long term dimension. So it's one of the things that we really are keen on, that we actually encourage a dialogue with officials, with scientists, with implementers on the ground, with beneficiaries to actually say, okay, what are these risks that you actually are facing and what are these possible ways to do in the long run and make it actually politically attractive to governments to respond to it?

[Theo Talbot]

Just very briefly, we should also acknowledge that this isn't something that governments or vulnerable populations want or frontline agencies are talking about and no one will provide. The insurance industry is increasingly, I'd say more so than many other points in the past, ready to work on this platform. And with this program, whether it's through groupings like the Insurance Development Forum or whether it's through just bilateral conversations, it seems like insurance companies benefit from having these risks that don't co vary very much with other risks on their books, but also acknowledge that these are new and important markets to be developing and these are new and important things to be doing. So we're not talking about a moonshot, we're talking about helping gently these two sides of the bridge connect, which seems from my perspective, more feasible than it ever has been before.

[Stefan Durkan]

What we also need, and that's beyond governments and insurance companies and the space in between, is actually there is not going to be a perfect way to use this kind of insurance, insurance like instruments to do it. We'll still need a humanitarian system that actually can bail the whole system out, that can cover those risks that are not well covered by the policies, that are hard to plan for, that are hard to contract for, that don't pay out, even though it's very clear there is a disaster happening and so on, and we need a humanitarian system as the insurer of the last resort.

[Rajesh Merchandani]

As chief Economist at dfid, you have a new minister, Priti Patel, in the new government in the uk, a post Brexit government. How is she going to go to the taxpayers of Britain under scrutiny and say we're spending some of your tax pounds on insurance as opposed to stuff.

[Stefan Durkan]

I think it's less difficult than you may imply. As the UK we spent billions on humanitarian aid. It's been a commitment that the previous Secretary of State has made in Istanbul that we're actually going to look more for insurance and insurance like instruments in the way we're doing things. We're not going to quickly rush and change the whole system and say it's all simply going to be insurance thing, but it actually is the kind of policy, you know, we may have a post Brexit government, but it's still a government select elected on the same manifesto. So we are still there with some of these commitments of getting humanitarian system to function better. And we think as one part of making decisions system to function better is to actually getting these incentives in place as we are describing, to getting the links with social protection and cash transfer schemes in the humanitarian system, which is something we strongly support, you know, using more cash and in the form of the scalable social protection, provide that kind of link here as well, and then using instruments that give the right incentives to the system. So I don't think there's a very big problem to do it. And on top of that, as seen from a post Brexit government, the UK is the world center of insurance. Surely we won't have that much objection to it that we actually harness some of the expertise and the capital that is present in London to actually bring to bear on the world's problems. Actually, this is a nice thing to do that you can work with the private sector, learn with the private sector, in this case, work with the insurance industry, learn from the insurance industry to actually solve really important problems in developing countries.

[Rajesh Merchandani]

Stefan Durkan, Theo Talbot, thank you very much for joining me.

[Theo Talbot]

Thank you, Rajesh.

[Rajesh Merchandani]

Stefan's co author is Daniel Clarke from the World Bank. Their book is called Dull How Planning Ahead Will Make a Difference. Theo and my other colleague in the London office, Owen Barda, are leading a new TGD working group on catastrophe Insurance for Humanitarian and Emergency Assistance. They put out a kind of policy paper to kind of frame this issue. It's called Payouts for why Disaster Aid Is Broken and How Catastrophe Insurance Can Help to Fix It. That's available on our website, cgdev.org as well as everything else that we're working on currently. And don't forget, also joining me, Rajesh Merchandani for the next edition of the podcast from the center for Global Development. Thanks for listening.

[Stefan Durkan]

Sam.