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.
Theo worked with Owen Barder and the CGD Europe team. His work focuses on finance for development, with an emphasis on novel contracts and financing structures that enable development actors to deliver social returns by collaborating effectively with the private sector. Theo grew up in India and Ethiopia and has a Bachelor’s in economics, finance, and politics and an MSc and PhD in economics. During his PhD, he was an Overseas Development Institute (ODI) Fellow in the Pacific, where he worked as an economist for the Government of Vanuatu. Before joining CGD, he worked at the Ministry of Planning and Investment in Hanoi.
In his post, John Simon, a former CGD visiting fellow, politely disagreed with our suggestion that donors are mainly using the wrong instrument to support private-sector investment. John made some excellent points (which we urge you to read in full). And, as we stressed in our first post, we all agree that private investment is crucial for delivering social returns in developing countries.
However, we still believe that donors should work with the private sector by using contracts that reward success, instead of instruments like guarantees, soft loans, or equity stakes. Those instruments work best if DFIs have better information than the private sector about risks and rewards and can choke off competition by picking winners.
Below, we summarise what seem to be John’s main arguments (our takeaways) and then reply to them based on our paper.
Our takeaways from John’s post
DFIs have better information than the private sector because they get information from their government’s diplomatic corps or intelligence agencies, and — particularly in frontier markets — because they specialise in markets where the private sector has not made a similar effort to understand the terrain.
Most DFIs don’t get this kind of privileged information. In any case, there are many private sector firms that provide the same edge.
If donors really do know about profitable investment opportunities in emerging economies, wouldn’t publishing it be cheaper (and better for competition) than setting up a DFI to try and trade on it?
DFIs’ investments are ‘safer’. Because they’re attached to powerful governments, other parties involved in deals with DFIs are less likely to default or renege.
This may be true in some cases. But it isn’t a feature of DFIs — it’s a subsidy they get from the reputation of the governments that fund them.
More generally, governments often use their power on behalf of national firms to push their overseas partners to honour a contract, without participating in their deals. (We’re not saying this is good or fair).
Since it’s not something DFIs do themselves (or can bring to bear equally for all their deals), we don’t think it’s an argument for or against which kind of financial instruments donors should use.
Writing pay-for-success contracts is much more expensive for donors or DFIs than issuing loans and guarantees.
The costs of “paying for success” will go down as donors do it more often.
And since paying for success means that donors won’t need to evaluate business plans in advance, the total costs may well end up being lower than other instruments.
As the effort by donors shifts from rowing the boat to steering it, in the long run, the transactions costs could well turn out to be equal, or even lower.
As John says, OPIC in the United States and CDC in the United Kingdom both do well financially, so there is no question of needing a public subsidy. So too does the IFC, the World Bank’s private lending arm.
Since DFIs are generally profitable, not loss-making, the question of public subsidy does not arise. Moreover, their record of financial success in markets where private investors are scarce is proof they do add value in finding and promoting profitable investments in these tough environments — investments the private sector often misses.
We wish these organisations — and their investees — well.
But if DFIs were always profitable, it casts doubt on how much they are crowding in additional investment, so much as displacing private investment.
And if DFIs create additional investment by offering cheaper financing or better terms, they are subsidising the private sector.
We argue that ‘paying for success’ is a better way to allocate that subsidy.
* * *
There’s already long list of white elephant projects in development. Paying for success doesn’t eliminate the chance we’ll fund these kinds of projects, but it does mean taxpayers are less likely to bear the costs of donors’ irrational exuberance. It also makes it easier to use subsidies that encourage competition, rather than “picking winners.”
Our paper and blog post are not intended as an attack on DFIs. Far from it: we strongly believe that there should be more private investment in developing countries. But we continue to believe that donors must be thoughtful about how they do this: not only to get the best possible value for public investment, but also to have the biggest possible impact on growth and jobs in developing countries, and to do the best we can to help the world’s poor. John’s defence of DFIs does not shake our conviction that donors could be doing this better.
In an earlier blog post, we explained why we think that donors and development finance institutions are getting it wrong by issuing guarantees and cheap loans to the private sector. We argue they should instead be increasing the returns for firms when they succeed.
Today, a former CGD Visiting Fellow, John Simon, disagrees. Before coming to CGD Simon was the US Ambassador to the African Union and the Executive Vice President of the Overseas Private Investment Corporation (OPIC), where he led efforts to finance housing markets across Sub-Saharan Africa and SMEs in Liberia. He is a founding partner of Total Impact Capital.
Thank you, Owen and Theo, for the chance to comment on the paper.
Unfortunately, I’m generallyunpersuaded by your main argument. Having represented the American government as an Ambassador and at a senior level at OPIC, an American DFI, I’ve seen first hand many of the unique benefits of marrying development-focused investment with commercial discipline.
First, you assert that DFIs don’t have better information than private investors. I don’t agree. If this were true, DFIs who take risk in frontier markets that private investors are not willing to take — or take alone — would all be loss-making entities. In fact, OPIC and the IFC and almost all the other DFIs make healthy profits.
The reason is they have made the investment to understand these markets in a way private investors — unsure of the returns — do not, and hence they do have better information — some in the public domain, some in their intelligence bureaus — than the more risk averse private sector. The insider information comes partly in the form of reporting from our embassies and intelligence agencies. Since the apparatus would be there with or without OPIC, it is a costless subsidy to OPIC. Access to it gives DFIs an opportunity to avoid deals that have serious problems and de-risk deals that have minor ones. Because this information is sensitive, it cannot be made available to the broader public.
More importantly, DFIs do nothing but invest in emerging and frontier markets, allowing them to build the skills, knowledge, and networks to succeed where others fail. This includes knowing the good local lawyers and accountants who can bring a project to closure, legal templates that work in the environments in question, and an understanding of what risks really matter in these developing country contexts. For example, I have seen many experienced investors come to OPIC with a government off-take agreement they thought was rock solid based on their experience in developed countries, only to have to renegotiate it from the start once OPIC's lawyers and underwriters have a go at it.
Second, you dismiss some of DFIs unique comparative advantages, which make them better channels for private investing for social returns in many developing country environments.
DFIs are extensions of governments most developing countries — and individuals in those countries — do not want to upset. Therefore, while a government official or local businessman may not think twice about cheating a foreign investor, they will think at least twice before they do the same to the US government.
For these two reasons, DFIs can indeed perform on par with the private sector and identify deals that the private sector ignores.
This undermines your assumption that the cost of the OPIC loan and the pay for success contract are the same. That would only be true if the OPIC loan is concessional (I would dispute that point, again looking at OPIC's loss experience) and also that the transaction costs are similar. In fact, we know only too painfully, the transactions costs of setting up pay for success schemes are quite high, sometimes so high as to make the whole enterprise impractical, whereas the transactions costs for an OPIC loan are manageable.
Even if the transaction costs were the same, from a real world perspective, the pay for success mechanism will hopefully cost the Treasury something – i.e., we will have success and there will be a budgetary payout, whereas the OPIC loan, if all goes well, will actually make the Treasury money. Of course, the reverse is true in the case of failure, but if the OPIC folks have done their job right the successes will outnumber the failures by several times. Do you think 10 years from now the same will be true for PFS?
Finally, it is not a choice between private investors and DFIs. The DFIs can only operate with private investors. But I am confident very few of the billions of dollars in deals the DFIs support every year would happen without their participation. And that would be too bad, because the vast majority of those deals bring significant developmental and financial returns.
John makes some excellent points. We are certainly not arguing against the idea of DFIs in general, nor OPIC in particular. Even so, we continue to believe that pay for success is better than guarantees and loan subsidies — we explain why in our next blog post.
Tanzania’s Morogoro Shoe Factory: Underwritten by the World Bank in the 1970s to supply the entire domestic market and produce for export, it never produced more than 4 percent of its capacity before folding, leaving behind a trail of dodgy deals and unpaid debt.
In early 2014, USAID announced a $10 million facility to backstop loans by the Kenya Commercial Bank to local firms so that they could buy GE medical equipment, repaying half the face value of loans that default.
The Advance Market Commitment for pneumococcal vaccines: Because pharmaceutical firms cannot invest in R&D for vaccines against diseases that primarily affect poor people, donors offered vaccine makers a guaranteed price per dose if they could produce a vaccine that met effectiveness and safety criteria.
No prizes for identifying Number 3 as the odd one out. In the first two cases, the guarantees and loan subsidies were given to firms picked out by the donor. The donor support was not conditional on success. The first project turned out to be a famous boondoggle. The second uses the US aid budget to subsidize an American manufacturer by enabling a bank to make riskier loans. The AMC, by contrast, offered a reward for which any company that could compete. Firms would get paid only if they succeeded in developing and producing a new vaccines. This induced two competing pharmaceutical companies to develop and produce new vaccines. Projections through 2020 suggest that access to pneumococcal vaccines will prevent an estimated 1.5 million children’s deaths.
Of course, we have hand-picked these examples to make a point. We are enthusiastic about the growing interest in supporting private investment in developing countries, but it matters a lot how this is done. The sorry history of failed and distortionary partnerships should tell us something about how donor countries can do a better job of working with the private sector.
In a new paper, we argue that the tools that donor countries usually use to “crowd in” the private sector — guarantees and cheap loans — distort firms’ incentives by reducing their risks or increasing their rewards irrespective of how well they do. This can lead to poor choice of projects and poor management. Worst of all, it relies on donors to pick winners, instead of allowing the market to decide which firm succeeds. Anointing the successful firms in this way stifles innovation and can choke off the very market forces that would bring about a more successful economy in the long run. And at worst, it can descend into expensive government backstops for firms who do the best lobbying for support rather than the firms which are the best at innovating and serving their customers. Past experience of these kinds of partnerships suggest that these are not merely theoretical concerns.
Our paper suggests that for the same amount of public money, donors (and their development finance institutions or DFIs) would normally do much better by paying for success than by issuing guarantees or cheap loans. Donors should define what success looks like (and how it will be measured) and offer contracts that reward any successful private firm for achieving it. This approach promotes innovation, sharpens incentives to choose investments wisely, keeps the firm’s shoulder to the wheel, limits scope for crony capitalism, and protects the public purse from the excess optimism of civil servants and from lapses in the collective decision making of bureaucracies.
For a more thorough consideration of the advantages and disadvantages of these different approaches, you can read the full paper here.
Shortly, our friend and erstwhile CGD Visiting Fellow, John Simon, a former Executive Vice President of the Overseas Private Investment Corporation (OPIC) and US Ambassador, will respond in a blog post explaining why he disagrees with our analysis.
Governments, donors, and public sector agencies are seeking productive ways to ‘crowd in’ private sector involvement and capital to tackle international development challenges. The financial instruments that are used to create incentives for private sector involvement are typically those that lower an investment’s risk (such as credit guarantees) or those that lower the costs of various inputs (such as concessional loans, which subsidise borrowing).
The emerging consensus is that the response to Ebola is a test that most richcountries failed. Given that the next public health challenge is a ‘when’, not an ‘if’, what can we do to be more prepared for the next emergency?
Cat bonds might be part of the solution. For the uninitiated, this has nothing to do with securitising housepets: the ‘cat’ stands for ‘catastrophe.’ Cat bonds work like regular bonds, with the twist that if a pre-agreed ‘bad thing’ happens, investors forfeit their cash to the borrower, who can spend it right away — for example to tackle an emergency or pay for recovery.
The World Bank issued its first one in June last year to help cover the insurance costs for earthquake and cyclone risks of 16 Caribbean countries. Now, it wants to scale up and across. The Financial Times reported from Davos that the Bank wants “...to create a global fund that would issue bonds to finance pandemic fighting measures by governments and other bodies.”
It’s not alone. Richard Wilcox, interim director general of African Risk Capacity (ARC) — an underwriter created to help African governments insure themselves against infrequent-but-costly disasters like poor harvests — told reporters that his organisation wanted to offer pandemic insurance to African governments by 2017.
Lest a useful piece of financial innovation fall prey to the development hype cycle, it’s worth pinning down what these contracts can and can’t do.
Cat bonds’ promise lies in the insurance facility they provide. Governments benefit from assured payouts when hit by bad shocks; unlike donors’ unenforceable pledges, these bonds pay out specific amounts under unambiguous conditions.
Compare and contrast: as Amanda Glassman and Karen Grépin noted, it took the WHO five months to declare the Ebola outbreak an emergency. Even if the organisation had access to fully funded contingency fund, they’re not convinced that it would have been triggered — and certainly not with the predictability and immediacy that is a feature of catastrophe bonds.
The investors themselves still seem keen: if the ‘bad thing’ doesn’t happen, they earn a nice payday. Better yet, as one typically breathless industry briefing puts it, cat bonds “are almost entirely uncorrelated with macroeconomic variables.” This means that cat bonds pay out when other assets don’t, and vice versa, a property much-beloved of portfolio managers.
Live deal tracking shows that, as a result, demand for these has grown at a fast clip, totaling $23 billion today (of which nearly $1 billion was issued in just the first few weeks of 2015). Healthy demand for these products drives down yields, making cat bonds potentially a cheaper way to provide insurance against some types of risks.
Insurance contracts need triggers for payouts that are measurable and transparent: we all have access to the same seismological data, so we can all agree whether or not an earthquake of magnitude 7.3 hit Port Vila.
But many of the problems we’d like to insure against don’t benefit from the same data infrastructure. One of the constraints on mounting an early defense against Ebola, for example, was the lack of monitoring and early warning systems. The “pandemic bonds” proposed for dealing with these risks would need to be priced and triggered based on credible, transparent, and timely data about the spread of diseases— effectively presupposing the existence of institutions that, were they in place, might reduce the need for risk financing in the first place.
Even when we can agree on a so-called “parametric trigger”, it’s not clear that catastrophe bonds are the best way to insure countries against shocks. The Pacific Catastrophe Risk Insurance Pilot, for example, pools risks from five Pacific island countries for total coverage of $43 million against tropical storms, earthquakes and tsunamis. Rather than a catastrophe bond, it is a traditional insurance contract. Structuring the transaction as a catastrophe bond might lower the cost of the insurance, but only if enough investors can be enticed to take the other side of the transaction.
These contracts could also create new risks if they make policymakers spend less on prevention, a problem economists call moral hazard. Imagine an overworked civil servant at the end of long budget meeting with her Minister, faced with a choice between, say, training additional midwives or investing in frontline monitoring for an unknown, future outbreak. I know that I’d pay for the immediate need — especially if I had some type of fiscal cover if I needed it. It might sound deeply uncharitable to claim that insuring against these risks affects our incentives; then again, I would think you uncharitable if you told me offering insurance refunds to drivers for not crashing would reduce accidents or that financial incentives reduced drug use or convinced people to stop smoking– but both do.
Solve some of the problems, some of the time
The excitement about finding new uses for catastrophe bonds is healthy. Cat bonds are perfectly suited for some jobs. And if investors remain enthusiastic about piling into frontier market cat bonds, they could provide a cheaper alternative to traditional insurance contracts.
Where they are appropriate, these instruments can bring market rigour to bear on the longstanding problems of how governments with stretched budgets pay for emergencies, and make sure those funds are mobilised when they’re needed. They also provide a clear framework for donors to choose how and how much to chip in, for example by covering interest payments or by guaranteeing the outstanding debt. Compared to unenforceable pledges to buy unknowable amounts of development impact, that’s progress.
But however sexy financial innovation for development might sound, it shouldn’t distract us from paying for and helping to build boring but essential public goods, including better data for development. Those investments are enabling technologies that would make insurance contracts and catastrophe bonds feasible— and perhaps even reduce the need for them in the first place.
Britain’s highly charged debate about immigration means that migration systems and policies are potentially in flux—a chance, perhaps, for innovation. We believe there are opportunities to tweak these policies so that they deliver big benefits for poor people, avoid the most harmful unintended consequences, and make British people better-off. Granted, development is unlikely to be at the front of politicians’ minds as they weigh up the options for migration policy, but now is exactly the right time for a discussion about how to shape immigration policy for development impact within the bounds of the current political agenda.
We put our heads together to come up with thirteen innovations for immigration policy to deliver meaningful benefits for international development. (They are offered as suggestions, about which further analysis and research would be necessary). They fall into three broad categories: capturing gains from immigration, for example by training and hiring more nurses from developing countries, rationalizing our rules about immigration, for example rolling back the nonsensical limits on the number of overseas students who can come to the UK temporarily to study, and innovating, by re-jigging global rules and domestic policies to help developing countries capture more of the benefits of sending their workers to the UK, however briefly.
Capture gains: make the UK a leader in Global Skill Partnerships (GSPs)
Set up Global Skills Partnerships that help pay for training of key workers (like nurses) overseas in exchange for the right to hire them in the UK. This is good for the nurses (who earn higher wages and gain valuable skills), the sending country (which has a lower cost of training), and of course patients and hospitals that get the staff they need.
Rationalize: shelter student visas from the migration debate and use them to promote development
Current rules focus on reducing the number of non-EU students studying in the UK. This damages higher education with no offsetting gain and deprives students of the chance to gain valuable skills by studying in the UK. A study by Universities UK showed that universities contributed over £36.4 billion to the economy (2.8% of GDP) in 2011, a fifth of which is due to non-EU students (£13.9 billion). Almost 59% people responding to a poll say we should not reduce the number of international students, even if this makes it harder to reduce immigration numbers.
Innovate: make the UK a (supply-side) leader on remittances
The government can be a leader in tackling constraints that keep remittance transaction costs high. Key agencies in the UK government (including the development agency, DFID) could work with the Financial Conduct Authority to investigate competition in banking services that serve immigrant communities so that these workers lose less in fees and charges and can ultimately send more of their salaries home.
Back on the front foot
Reasonable people can disagree about some aspects of freer movement, but some things are difficult to argue with: for example, clamping down on students and productive workers who pay more into the public purse than they take out makes us— and the people who want to work and study here—worse off.
Much of the British political class would prefer not to have to discuss immigration policies. But people concerned about international development should not see this as a time to move the conversation on, but rather to make the case for immigration policies which are politically realistic, good for Britain, and which are better for development too.
When a disaster strikes, we are urged to send money, and many people do—but is there a better way to fund the relief effort? My guests this week, DFID chief economist Stefan Dercon and CGD senior analyst Theo Talbot, believe that insurance can help.
“Private sector” appears 18 times in the outcome document from last year’s UN financing for development conference in Addis Ababa—exactly the same number of times as “international cooperation.” In part, this is driven by the financial shortfall traditional donors face in delivering this ambitious agenda, and partly it reflects the different skills our public and private sectors possess. Now, one year into the SDGs, where are those ideas that bring private sector ingenuity and capital into achieving the development goals? In this edition of the CGD Podcast, we'll introduce you to one of them.