With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
Last year more than 83 million people in low and middle income countries were affected by natural disasters. We may not know when or where the next disaster will strike, but we know it will. So why do we still treat disasters like surprises?
International appeals are generous, but they are usually launched only after disaster strikes, and funds and supplies are slow to arrive. A new CGD report urges a different approach: make disasters predictable. The report looks at how we can pre-arrange disaster response funding using the principles and practices of insurance, so that countries get the money as soon as they need it and donors actually pay less in the long run.
Vice President, Director of CGD Europe, and Senior Fellow
We launched the report at a recent CGD event with working group co-chairs Stefan Dercon—Chief Econmist at DFID, professor of economics at Oxford Univeristy, and co-author of a book called Dull Disasters—and Owen Barder, CGD vice president and director of our Europe program. They were joined by two members of the working group: Alice Albright, head of the Global Partnership for Education, and Rowan Douglas of insurance group Willis Towers Watson.
Today's podcast brings you a flavor of that event.
With Hurricane Irma now pushing a devastating path through the Caribbean, USAID is gearing up to do what it does best. Its Disaster Assistance Response Teams (DARTs) do amazing work—deploying rapidly in the wake of natural hazards like hurricanes and often bringing the logistical might of the US military with them. These teams go in big and fast to save lives, distribute food, set up emergency shelter, and prevent secondary impacts like disease outbreaks.
But then things begin to falter.
When the effort pivots from disaster response to post-disaster recovery, USAID’s efforts are almost invariably understaffed, underfunded, and underwhelming. USAID development professionals are increasingly trying to take the baton when their disaster response colleagues depart; but invariably they find themselves hamstrung. This is because the powerful toolkit USAID has for disaster response abruptly turns into a pumpkin and disappears during the reconstruction and recovery phase.
With Irma raging ahead and other storms on the horizon, I fear the USAID is again poised to win the response and lose the recovery.
As the head of US foreign disaster relief during the Obama administration, I managed many of these mega-disasters. Typhoon Haiyan in the Philippines, the strongest storm to make landfall in recorded history (a record that Irma could eclipse). The massive 2015 earthquake in Nepal. The Ebola outbreak in West Africa. Hurricane Matthew, which ravaged Haiti last fall. In each of these disasters, USAID disaster teams deployed decisively, flexibly, and with ample resources—and their actions saved many lives.
But if the agency went four-for-four on those disaster responses, we went one-for-four on their recovery phases. Only the Ebola recovery was well resourced, through a special appropriation from Congress. The other three disasters were largely orphaned once the recovery phase began, because USAID’s ambitious plans for recovery support received only a fraction of what they required.
This is not a new problem. The post-disaster recovery gap has bedeviled the US government for decades. Over the past 25 years there have been countless initiatives and acronyms thrown at the problem, but little real progress. The core obstacle is that the USAID lacks the people, resources, and flexibility needed to ensure effective recovery.
This is a shame, because the real work of a natural disaster is in the long-term rebuilding, not the immediate lifesaving. Most of the people who die in a natural disaster do so in a short window of time as the event peaks: crushed in falling buildings as the ground trembles or swept away by torrential currents of water. Few of these people are plausibly savable once an event is underway. But nearly all of them are savable through good preparedness and early warning (which are chronically underfunded).
For those who do survive, the road ahead is difficult. International resources flood in for the immediate response and invariably falter beyond that. So families often find that the “temporary” shelters they receive in the immediate aftermath become semi-permanent due to lack of reconstruction assistance. And without support to re-start their livelihoods, they lack the resources to rebuild for themselves.
The jarring contrast in how USAID performs on relief vs. recovery mirrors a jarring contrast in how Congress allocates and oversees the resources it gives to USAID. For disaster response, Congress grants the agency broad flexibility to contract staff, allocate resources, and set response priorities. It has established a long-standing and well-funded disaster assistance account to support the agency’s disaster efforts. And it has also given USAID a “notwithstanding” authority for disaster response—meaning that USAID can (judiciously) choose to bypass normal legal requirements if they would obstruct a speedy response. This means that when a disaster strikes, USAID can rapidly mobilize reserve staff, assess needs, and focus resources on the areas of greatest need.
But in the recovery phase, by contrast, USAID becomes straitjacketed by congressional requirements. Reconstruction needs are generally funded through other foreign aid accounts, which lack the flexibility the disaster account enjoys. To spend money from these accounts, USAID must go through arduous federally-mandated procurement procedures, which can take one to two years to complete. Needless to say, that sort of delay doesn’t much help a population that is living under plastic sheeting after losing their homes and livelihoods.
But it gets worse. Before USAID can even begin that arduous programming process, it first must find the money—and rarely does Congress appropriate any extra funds. While Congress issued special appropriations in the cases of Ebola and the 2010 Haiti earthquake, these are rare exceptions. More typically, USAID is forced to find resources within its existing funding—meaning cuts to other programs. USAID needs to be willing to make those hard trade-offs—but even when it is, it finds itself prevented from doing so by congressional spending directives. Each year, Congress mandates that USAID spend predetermined amounts on major sectors like education, health, or microfinance. These directives leave little flexibility for the agency to change course and reprioritize after a disaster strikes.
And then of course there is the problem of people—or rather the problem of too few people. Ramping up a large recovery effort takes bodies—people to assess needs, coordinate with partners, plan programs, and issue grants. At the height of the 2016 response to Hurricane Matthew in Haiti, USAID deployed more than 60 disaster professionals to the country. Once that team departed, the recovery planning fell to the local USAID development mission, which lacked the bandwidth to cover this role along with its day-to-day obligations. This is a common story, yet the agency lacks an organized roster to identify, hire, train, and deploy staff for this sort of function. Congressionally-imposed caps on staffing levels make it difficult for the agency to staff for these occasional contingencies without undermining other functions.
Fortunately, much of this can be fixed—if Congress and the Trump administration choose to do so. The administration is midway through a review and reorganization of foreign affairs and foreign aid tools. Congress has signaled that it wants to be involved—the foreign aid budget draft released by the Senate this week requires that Congress be consulted on any major changes. So the time is ripe to fix this long-running problem.
How to do so? As I wrote in a recent paper advising the administration’s reform effort, there are some concrete options available—right now—to give USAID the resources, people and flexibility to nimbly execute post-disaster recovery efforts.
On the resources front, Congress could help greatly by authorizing earmark relief to USAID in the wake of a declared disaster. Granting USAID’s country programs a two- to three-year waiver on fulfilling sector earmark directives would free up resources to be redirected toward reconstruction programs. Similarly, loosening the restrictions on hiring—particularly on the agency’s staffing cap—could enable USAID to build a deeper bench of transition and recovery experts to deploy after disasters, without undermining ongoing program obligations elsewhere. Finally, USAID should be more aggressive—with congressional backing—in waiving standard grant-making processes in post-disaster settings. Rather than reverting to the standard one- to two-year turnaround for developing and issuing new grants, USAID should exercise authorities to waive those requirements when exigent circumstances demand faster action.
The perpetual question, of course, is one of trust. Congress applies tight restrictions on staffing, funding, and procedures because it has not historically trusted the agency’s own judgment in these matters. Yet the agency’s consistently excellent performance on disaster response is a powerful counterpoint: when given greater leeway, USAID can use it responsibly to deliver powerful programming. It’s time for Congress, and the administration, to take a joint leap of faith and equip the agency to be as effective on disaster recovery as it is on disaster response.
HLM2 at Nairobi’s Kenyatta International Conference Center
The panel is part of CGD’s on-going Payouts for Perils working group, focused on pinning down technically specific, politically saleable, and financially scalable innovations that we can invest in to respond better to expensive but predictable perils. That includes events like storms (think Haiti), earthquakes (think Nepal), or droughts (think the ongoing risks across the Sahel, including Kenya and Ethiopia).
Our panel in Nairobi brought together two critical skills to improving emergency aid: development expertise and risk management nous. We left it with four strong takeaways.
1. Perils - planning = tradeoffs
Our panel benefited from a stellar cast. Sitting in for the UK’s delegation to the meetings was Pete Vowles, the head of DFID’s programme in Kenya, ultimately responsible for delivering more than £150 million in UK aid. Pete told the audience how DFID is investing in new approaches to aid, including supporting a national safety net—Kenya’s Hunger Safety Net Programme—capable of reaching households affected by natural disasters instead of requiring a full-scale mobilisation by donors.
From left: Ginger Turner, Alex Palacios, Pete Vowles, and Rowan Douglas (CGD friend Rupert Simons sneaks into the frame at bottom right)
But in an environment like the horn of Africa, the budgets of DFID and other donors face a host of threats. Let’s say we allocate $100 to deliver vaccines. If a drought arrives, we’re faced with stark choices: tackle it by taking some money from our immunization campaign, or use some flexible funding we’ve held back. We either diminish one programme to subsidise another, reduce our pot of savings to respond to a later crisis, or fail to respond to the drought. That’s a list of bad choices.
2. Donors can’t, don’t, and shouldn’t “invest and forget”
Success in these difficult circumstances depends on the ability to mobilise the right amount of money at the right times. But that means either holding funds back against unknown future risks—and not being able to spend them to tackle poverty here and now—or facing the risk of being underfunded when rare-but-expensive perils arrive (or both).
That was the risk raised by Alex Palacios, who joined Pete on our panel. Alex is a senior official with the Global Partnership for Education, which raises international financing from donors to invest alongside more than 60 national governments in education systems around the world.
Those investments are now worth more than $4 billion globally. In Alex’s words, this is a portfolio that blends education inputs—like teachers’ salaries and line ministries’ capacity—with a policy shift toward spending more, over longer horizons, to educate future generations. Left intact, that portfolio delivers social and economic returns from schooling.
But that success, as Alex succinctly put it, requires resilience. When storms destroy schools, entire cohorts are set back—in some cases permanently. So delivering on goal 4 of the SDGs means both building education systems, as GPE does, but also making sure we can get kids back into school quickly after disasters strike.
3. There’s an industry that provides capital on call
Pete and Alex were joined by two professionals from an industry that we think might help donors solve some of these thorny problems.
Ginger Turner, a senior economist at the (re)insurance firm Swiss Re, pointed out that many of the problems we think of as “budgeting” are actually precisely the problems the insurance sector has native expertise to help solve.
One approach for DFID to manage its commitment to the HSNP, for example, would be to <deep breath> estimate the chance of drought, guess at the additional cost that implies, hold all that money on its books, and take the chance of either underfunding its response by guessing wrong or having more money left at the end of the year (which, in turn, means underspending on other urgent challenges).
The alternative is to set up a contract that pays out in case drought is on the horizon, for example by crunching average surface temperature and rainfall through a simple equation. As Pete pointed out, that’s precisely what the African Risk Capacity—a mutual insurance scheme across several African governments that the UK helped to set up—does.
Insurance, in other words, is an entire sector devoted to charging the rest of us a bit of money at regular intervals to keep money on call for when things go wrong. For frontline agencies and vulnerable countries, using insurance thinking and contracts could unlock a cascade of benefits. They don’t need to stock up on money that leaves them over- or under-funded. They know how much they’ll get if things go wrong, enabling planning. And they can match the payout in the insurance contract to the set of risks they face. That can catalyse faster, better-organised response—and avoid waiting for things to go wrong, hoping that donors will step in adequately and quickly.
4. (And it’s a pretty competitive one)
There’s a healthy dose of skepticism and distrust when we in the public sector contemplate shaking hands with our colleagues in the private sector. How can NGOs and donor agencies be sure to get a good deal?
Certainly this matters both for politics and for prices. This week’s biggest development story in the UK isn’t that the Multilateral Aid Review showed that the vast majority of British development dollars go the highest performing agencies. It’s about bad behaviour by Adam Smith International, a deep-pocketed aid contractor that submitted fake testimonials about its performance to MPs and took advantage of private government business plans when organising supposedly-competitive bids.
Rowan Douglas joined Ginger as an insurance sector voice on our panel. In addition to working his role as a senior executive at London’s Willis Towers Watson, he co-Chairs the Insurance Development Forum, an initiative to help more aid agencies and governments take advantage of insurance tools.
Rowan had two pithy answers to Owen’s question about getting a fair deal. First, donors should absolutely invest in technical capacity that would help agencies and governments be informed buyers of insurance. There are already growing teams in agencies like USAID and DFID (not to mention first responders like Mercy Corps) with backgrounds in actuarial science and an understanding of the insurance industry.
Second, competition is one of the best guarantors of fair prices. An annual review for ARC notes, for example, that it was able to get extremely competitive quotes for its insurance contract, ultimately receiving $55 million worth of coverage from 12 different reinsurers for just over $5 million in premiums. Insurance, in other words, is like any other industry: more transparency about procurement and more competition amongst providers translates into a fairer shake for all of us.
* * *
As Pete noted in his own smart blog post after the event, “risk thinking” and (where appropriate) insurance contracts aren’t silver bullets. But that’s okay: we aren’t hunting werewolves. Smarter financing won’t solve the humanitarian funding challenge, but it seems like a good way to free up more money for emergency response, make our assistance cheaper and more predictable, and leverage the skill and expertise of a competitive and capital-rich industry into the bargain.
Officials hard at work during the HLM2’s opening ceremony
Put simply, how we spend our aid dollars affects the quality of the help we deliver. But this is a piece of our humanitarian spending runs on legacy software: we wait until things go wrong before soliciting funding from donors to meet their costs. As Stefan Dercon, co-chair of CGD’s working group on this topic, has put it, that’s a 12th century financing solution to 21st century problems.
We’re confronted with refugee crises and tragic and continuing conflict in the Middle East and Africa. Our front pages are filled with evidence of our creaking emergency aid system. In a world where predictable, insurable risks contribute to instability, drive poverty and cause displacement, rich countries don’t need to appeal to humanitarian principles to inspire change. Simple self-interest should also do the trick.
In Haiti, already the poorest country in the western hemisphere, Hurricane Matthew’s devastation is still being calculated. We know that hundreds of people have died, and the damage to Haiti’s already-fragile infrastructure is immense. So what can people in rich countries do to help? Based on the latest research on humanitarian disaster relief and on the lessons learned in the wake of the 2010 earthquake in Haiti, here are some do’s and some don’ts for policymakers and individuals.
Give money, not stuff
If you are looking to help as an individual, give money, not stuff. Sorting, shipping and distributing donated items is expensive. That means your donations won’t do anywhere near as much good as it might seem. Money is also much faster to arrive, helping people sooner and more flexibly than a donation. The most needed staples in hurricane-hit areas are already available in Haiti. What’s needed is timely funds with which to purchase and distribute them. In the worst cases, donations can be highly damaging to the local economy, as was the case after the 2010 earthquake, when the huge amounts of food received depressed local prices and hurt farmers. Your donated t-shirt might not do so much damage, but a small part of your pay check would be much more valuable.
Support local organizations
Haiti’s ambassador to the US, Paul Altidor, has called for those who wish to help to engage with local organizations and municipalities in order ‘to avoid mistakes from the past’. After the earthquake in 2010, the hundreds of international organizations on the ground were not at all transparent to the Haitian administration or to each other, meaning that efforts were duplicated and people ended up getting in each other’s way.
There are a number of excellent Haiti-based organizations, like Zanmi Lasante and TiKay Haiti, who are best placed to understand and address local needs, but who struggle for funding. Just 0.6 per cent of the $6.43 billion donated in the wake of the 2010 earthquake found its way to local organizations. Unlike many international organizations, these Haitian charities are known by, and work closely with, local municipalities. This is important to the long-term development of Haiti. As Ambassador Altidor points out: ‘It is imperative that we take caution when offering assistance not to contribute to the destruction of local institutions by bypassing or undermining them.’
Don’t get on a plane
Unless you’re asked to come to Haiti by a reputable organization with a long-term presence on the ground, you will probably just be in the way. The cost of your plane ticket will do more good as a donation.
Policymakers: Be transparent
In the wake of the 2010 earthquake and its associated relief effort, it has been almost impossible to trace the final destination of humanitarian funds raised. We know, because we tried. A lack of transparency about how funds were used means that international NGOs and private contractors were not accountable to the people they were supposed to be helping, or to the Haitian government. How then can we learn any lessons about what was cost-effective? This time, we must do better. The International Aid Transparency Initiative (IATI) and UN OCHA’s Financial Tracking Service are mechanisms that facilitate transparency. Organizations receiving public funds should be obliged to publish adequate data to IATI.
Buy insurance for next time
Recent work from CGD shows that Catastrophe Insurance could radically improve the way that humanitarian relief is organized. Using insurance principles instead of the current ad hoc system would save lives, save money, and provide incentives to invest in disaster preparedness. Haiti doesn’t need to face a debacle like the response to the 2010 earthquake. Donor countries can fix this problem, if they decide they want to do so.
In case you missed it, catastrophe bonds recently made it to the cover of the Wall Street Journal: a once-sleepy, wonky corner of the insurance market is poised for disruptive growth. “Cat” bonds are effectively a cheaper source of large-scale insurance coverage against clearly measured risks like earthquakes, storms, or even disease outbreaks.
The insurance industry’s rich jargon aside, these are simple products. Like a bond, someone lends you money and you pay them an interest rate. The twist is that if a pre-agreed disaster or other event happens, you don’t have to repay. Setting out clear conditions for “triggering” the bond based on objective data like the violence of an earthquake reduces uncertainty about the value and timing of your post-disaster income. That, in turn, facilitates planning and preparation.
Generally, though, coverage hasn’t trickled down to the poorer and most at-risk countries—precisely those which are most vulnerable when aid fails to arrive or arrives piecemeal. Parsing data on cat bonds handily provided by Artemis, an industry observer, shows that less than a twentieth of the total value of these products issued since 1996 covers risks in countries that the World Bank classifies as upper middle-income or below (places where income per capita is $12,475 or lower). They have mainly covered risks in the US and other high income countries (HICs).
Notes: Catastrophe bond issuance data from artemis.bm. CGD analysis. Totals overstate issuance due to overcounting, for example allocating the full value of a contract to both categories if both US and Caribbean risks are covered.
Information, cost, and basis risk seem to be at least three constraints on scaling catastrophe bonds up and across:
Agreeing on these contracts requires technical nous
They’re bespoke, so you need to pay for advisors and catastrophe modeling experts
The contracts are (mainly) based on triggers rather than actual losses. That creates basis risk—a storm might cause damage, for example, but not invoke the bond’s trigger.
Examining the handful of transactions that do cover developing regions bears this out. There are cat bonds covering risks in Turkey, Mexico, and a group of Caribbean countries (through CCRIF, a regional mutual insurance facility). Those issuances depended on technical and policy support from the World Bank, without which it would have been difficult—probably impossible—to navigate these complex transactions.
One way forward is to rely on centres of technical excellence like the World Bank to facilitate risk transfer deals. And it makes sense for the Bank to deliver those transactions alongside its other “risk finance” tools, like emergency credit lines. But this deal-by-deal approach is unlikely to catalyse a broad market of catastrophe bonds covering a full range of developing countries’ risks.
We might be sanguine about the gap between insurance demand in insurance supply if aid were much more effective at tackling the consequences of natural disasters—in short, if aid were ‘as good as insurance.’ But the premise of a new CGD working group focused on improving emergency response is that ex-post aid is increasingly unfit for purpose because it is overstretched, small compared to needs, prone to arriving piecemeal and late, and able to distort incentives to invest in lowering losses.
As a result, one of the exciting questions that our working group is tackling is how donors can scale up access to these kinds of insurance instruments, perhaps by streamlining access to the necessary technical assistance, bearing some share of premiums, or commodifying them so that more institutions, agencies, and governments can more readily deploy them.
If ex-post aid is a bad substitute for insurance, there’s a compelling argument for substituting insurance for aid. Cat bonds are one part of the toolkit. Scaling up this market for lower-income countries would provide better shielding against many risks that undermine development overseas. It would leave more cash in donors’ pockets to tackle the kinds of emergencies that are hard or too expensive to insure against. And it would provide the kind of protection to poorer and vulnerable places that rich countries and large firms are choosing for themselves.