59551 POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise FEBRUARY 2011 · Number 49 Delivering Aid Differently--Lessons from the Field Wolfgang Fengler and Homi Kharas The global aid environment has changed profoundly over the last decade. New official and nonstate players have emerged as prominent actors, new challenges like climate change and stabilizing fragile states have arisen, and new approaches to providing aid are being tried. This note summarizes findings presented in the authors' 2010 publication, Delivering Aid Differently--Lessons from the Field.1 We live in a new reality of aid. Rich countries delivered $3.2 have so many new countries, individuals, celebrities, founda- trillion of aid to poor countries between 1960 and 2008, and tions, and other groups joined in giving aid? aid is a $200 billion industry today. Despite disputes and up- Although all sides have some good arguments, we believe heavals, the core of the aid architecture--the set of rules that the developments of the last decade have radically re- governing aid flows--has changed little over the past few shaped the aid environment and that traditional arguments decades. Now the new dynamics of aid may be too strong to no longer hold. Today, the main paradox of aid is that despite resist fundamental change. increasing flows and more players, aid has declined in relative Delivering Aid Differently (Fengler and Kharas 2010) was importance in most countries. written at a time when the future of foreign aid is being This new aid environment is characterized by three im- fiercely debated. On one hand, aid experts (including Peter portant shifts that have emerged in the last decade: Singer [2009] and Jeffrey Sachs) support a huge increase in 1. Strong growth in many developing countries has rede- the size of foreign aid budgets, faulting a lack of donor gen- fined the role of aid. Several countries (notably in East erosity for the continued existence of severe poverty. Their Asia) have fulfilled the dream of development aid--to message: aid works; we just don't do enough of it. make it nearly obsolete. For many years, foreign direct In scathing critiques of aid, on the other hand, William investment has exceeded aid by a wide margin. Devel- Easterly (2006) and Dambisa Moyo (2009) argue that by oping countries' demand for aid is now increasingly dif- detaching recipient governments from accountability to ferentiated. Gone is the "third world" as a homogenous their citizens, aid has retarded the progress of poor countries. block of poor countries in the South. Even in Africa, Between these camps, aid practitioners argue that foreign aid many countries have been growing fast since 2000. works well when it is done right. The key is to learn lessons 2. The donor landscape has changed fundamentally over to improve aid effectiveness. the last decade, a trend that will likely accelerate in the What should we believe when there is such disagreement coming years. New private players--international non- among experts? If development aid is really so ineffectual, governmental organizations (NGOs), foundations, and why has it increased so rapidly over the last decade? Why corporations--are responsible for an ever-larger share 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise of aid volumes. Private philanthropy from developed Figure 1. Total Development Assistance, 1992 and 2008 countries to developing countries now tops $60 billion annually. These new players have brought fresh energy 1992: and approaches to the delivery of aid. But although the $92 billion volumes of aid are growing, the average project size-- NGOs even for traditional donors--is shrinking. Small proj- 5% ODA ects can deliver benefits to isolated communities, but 95% they also increase aid fragmentation that multiplies ad- ministrative costs and complicates donor coordination by recipient governments. In 2007, for example, offi- cial donors sent more than 30,000 missions to manage their aid projects.2 3. Innovation, especially in information technology, has 2008: started to reshape development aid. Knowledge trans- New $194 billion bilaterals fer has become as important as financial aid, and com- 10% bining the two can be remarkably transformative. We believe that information--the hardest currency in the 21st century--ought to be at the center of aid delivery NGOs models. Recipient countries could champion a "single- 27% ODA 63% window" approach to aid regulation, whereby all infor- mation can be found systematically. Such an approach would reduce the confusion and inefficiency of current delivery methods. These one-stop information systems would more readily point out the comparative advan- tages of different providers and make known what each player is doing. Source: Authors' calculations based on OECD/DAC (2008) and Hudson Institute (2009). The New Reality of Aid Over the last two decades, aid has been growing strongly: substantially lower (perhaps one tenth as much) because more than doubling from $92 billion in 1992 to around salaries of local experts are so much lower than those of in- $200 billion in 2008, and representing a 50 percent increase ternational experts. There are examples in which much of in real terms. Most aid still comes from the members of the the aid pays for multiple teams of foreign advisers who give Development Assistance Committee (DAC), a club of 22 sometimes contradictory advice--thus, that aid funds du- rich countries ($120 billion) providing official development plicative and unnecessary efforts and is wasted, with most of assistance (ODA). However, a large and increasing portion the money going to foreigners. (some $65 billion) comes from private NGOs, foundations, Beyond the problems in categorizing aid, however, the faith communities, and corporations. A further $15 billion more fundamental issue is that aid is coming from more or more comes through bilateral aid from non-DAC govern- places and being allocated through more channels. The ments. This means that two thirds of the growth in aid flows growth in aid volumes is paralleled by a growth in the num- since the early 1990s has been coming from nontraditional ber of aid agencies involved. In a recent count, some 263 sources, and a third has come from DAC members (see fig- multilateral aid agencies give money to promote develop- ure 1). However, pledges to dramatically increase aid made ment (Kharas 2010). "Vertical" multilateral agencies that fo- by G-8 countries whose representatives gathered at Glenea- cus on a single theme (such as the Global Fund for AIDS, TB gles have largely fallen short of their targets. and Malaria) are now large providers of aid. There are at Several studies show that there can be large differences in least 56 countries with aid agencies that provide bilateral aid, depending on the method of counting (Hudson Institute foreign assistance, and most of these have several agencies 2009). Think of two options to provide technical assistance: that undertake such programs. More and more countries are In the first case, the development agency uses predominantly opening up new aid agencies. In the past few years, econ- international consultants. In the second case, it uses mostly omies such as those of Brazil, China, India, the Republic of local expertise and networks. Even if the second case proved Korea, Turkey, and República Boliviariana de Venezuela have to be more effective, the amount of recorded aid would be developed large aid programs. New international aid players 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise include some countries that are still characterized as devel- ternational NGOs, foundations, and private corporations-- oping economies. channel significant volumes of large and small contributions. These flows are changing the aid architecture. Figure 2 Hudson Institute's (2009) recent estimate of private giving shows the old reality of aid flows, illustrating how a large for international development measured private philanthro- majority of aid was organized for much of the second half of py from developed to developing countries as an annual the 20th century. Taxpayers in rich countries provided mon- minimum of $49 billion and possibly as much as $70 billion. ey to their governments to channel, either directly through The aid landscape now includes thousands of privately fund- bilateral programs or indirectly through multilateral organ- ed international NGOs and perhaps hundreds of thousands izations, to governments in poor countries. The recipient of community-based and civil society organizations in devel- governments, in turn, used the money to deliver services to oping countries themselves. As a result, aid from non-DAC poor beneficiaries through development projects. bilateral donors and private sources of funding soon will rival However, that flow pattern no longer adequately de- in size the traditional aid from rich countries. The aid picture scribes the majority of aid flows. New players--including in- now looks more like what is depicted in figure 3. Figure 2. The Old Reality of Aid Multilateral institutions Rich governments Poor governments Rich individuals Poor individuals Source: Authors' illustration (Fengler and Kharas 2010). Figure 3. The New Reality of Aid Rich governments Poor governments Multilateral institutions Vertical funds Private NGOs Rich individuals Poor individuals Source: Authors' illustration (Fengler and Kharas 2010). 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Existing aid monitoring mechanisms tend to handle new key government positions. One evident cost of fragmenta- donors poorly. Many databases that track aid incorporate tion is the time government officials spend receiving high- only ODA funding data or offer spotty coverage of NGO level delegations from donor organizations. In Cambodia, and vertical fund projects. This means that they leave out senior government officials spend half their working hours some 40 percent of the aid flows. As a result, projects funded meeting with donors (Ek and Sok 2010, p. 72). Multiple by emerging donors are often tracked in disconnected sys- project implementation structures may also create systems tems that may fall into disarray when funding is interrupted, that bypass government structures and procedures, and or simply are not subject to any monitoring. small projects are less likely to be included in recipient coun- try budgets. Aid Fragmentation A survey to monitor implementation of the Paris Decla- The multiplication of donors has also led to greater fragmen- ration targets for improved aid effectiveness found 2,473 ac- tation of aid into ever-smaller activities. Globally, as the tive project implementation units parallel to government number of donors has increased, the number of new aid systems. Because the sample covered countries receiving just projects has skyrocketed and the average project size has over half of ODA, the total number of parallel project im- shrunk drastically. In 2008, the DAC bilateral donors regis- plementation units in developing countries may be closer to tered 84,764 new aid activity commitments, up from 15,750 5,000 (OECD 2008). in 1996. Over the same period, the mean size of activities fell steadily, to a level of $1.35 million in 2008 from $2.97 Aid Volatility million in 1996. The median size of a new activity is now Net ODA disbursements, seen from the perspective of a re- only $87,499. It appears that the increase in total ODA is cipient country, are highly volatile; and that volatility dimin- the result of adding many small new projects rather than of ishes their value. From the early days of development assis- scaling up projects that work well (see figure 4). tance, the Resolution of the Common Aid Effort in 1961 Of course, small projects in themselves are not bad. They recognized that "assistance provided on an assured and con- are an important source of experimentation and innovation. tinuing basis would make the greatest contribution to sound They may be able to provide exactly what is needed in iso- economic growth in the less-developed countries" (OECD lated communities where small amounts of money can make 2006, p. 10). Unfortunately, that understanding has not been a significant difference in people's lives. But the fragmenta- put into practice. In terms of loss of national income, aid tion of aid comes at a heavy price. Each project must be pre- shocks in poor countries have been as large and as frequent pared, negotiated, supervised, and reported on. Many proj- as the major global economic shocks faced by rich countries ects create project implementation units and steering over the last century--namely, the Great Depression, the two committees entirely outside existing bureaucracies. These World Wars, the Spanish Civil War, and the recent global re- structures incur substantial administrative costs and may cession. The risk of an aid cutoff in aid-dependent economies weaken domestic institutions by poaching scarce staff from may have significant consequences for how aid is used. It is useful to focus on episodes of aid crises because there Figure 4. Trends in Project Size and Count, 1990­2008 is strong empirical evidence (World Bank 2005) that crises are what matter most in reducing growth. Aid crises have signif- 9 90 Number of commitments (thousands) icant disruptive effects on exchange rates, public investment, Size (constant 2008 US$, millions) 8 80 and inflation. Sound macroeconomic policy--universally ac- 7 70 knowledged as the foundation of growth and development-- 6 60 is impossible with high aid volatility, especially if such 5 50 volatility tends to compound the normal business cycle. At 4 40 a macroeconomic level, the volatility of country programma- ble aid is seven times as great as the volatility in GDP and 3 30 three and a half times as big as the volatility of exports.3 2 20 Volatility is higher in the poor and fragile countries and 1 10 in those where political or security concerns play into donor 0 0 funding decisions. Thus, Pakistan, despite its relatively robust 02 homegrown coordination mechanisms, has seen highly 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 01 20 3 20 4 20 5 20 6 20 7 08 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 20 19 20 Year volatile aid flows in recent years, driven largely by swings in U.S. grant aid. Short-term orientation in aid planning is also size projects a direct cause and consequence of high aid volatility. In Pak- Sources: U.S. Census Bureau; Banco de Mexico. istan, approximately 34 percent of overall disbursements 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Table 1. Old and New Models of Aid Management Aspect Old model New model Overall approach International coordination and burden sharing; In-country coordination and collaboration collaboration between headquarters dominates by recipient governments, differentiated by situational demands Sources and channels Official development aid focus Diverse multiplayer, multichannel aid Programming Annual Multiyear, dynamic Delivery Project-based, fragmented Programmatic, scalable Capacity building Capacity substitution, technical assistance Capacity development, local networks Information systems Global Global and local Source: Authors' analysis. from ongoing development projects in 2008 were from proj- A fix for the development aid system--by expanding the ects with durations of three years or less. Thirty percent of old model to include aid from large, new bilateral donors or all projects extended over a period of more than five years by pursuing local ownership through joint country strate- (Malik 2010, p. 139). gies--could help improve efficiency in some cases. Three Major donors are wary of trusting weak governments with principles could provide useful guidance for a new model of large amounts of money, and new donors have not yet stepped aid delivery: in to cushion in a coordinated fashion the swings in aid from 1. Aid should be differentiated by country circumstance. traditional donors. But the resulting instability in aid flows car- This may sound obvious but, in practice, there is a ten- ries its own costs and reduces the benefits of aid, even when dency for donors to lump countries into categories: recipient governments improve their performances. fragile states, low-income countries, small island The case of Kenya highlights the costs of volatility on the economies. Although classification is useful, across the ground. Over the past decade, the government has refused universe of 152 countries that the DAC recognizes as to incorporate committed aid into the national budget be- eligible for ODA there are always many more cate- cause the funds too often fail to materialize. A former gories of differentiated country circumstances that Kenyan finance minister confronted the development part- should be reflected in the delivery of aid. ners, stating, "Although some programme grants and loans 2. The delivery system should build on the diversity of may materialize during the course of the year, if and when aid providers. Different providers have different com- such additional resources materialize, I will use them for re- parative advantages. To identify and focus on these ducing our outstanding domestic debt" (Mwega 2010, p. unique advantages, each provider must know what the 117). His point: aid in Kenya is too unreliable to be treated others are doing. Too often, this is not the case. The as a source of financing for expenditures, but should be number of non-DAC donors and their overall share of thought of as a windfall. If it arrives, it is saved (debt is re- aid is likely to continue growing over time, and mech- duced). If it does not arrive, expenditures can be retained at anisms must be found to enable all players to be well their budgeted level. informed. Field-based coordination meetings have not been very successful, except in some cases where a fo- A New Aid Model for the 21st Century cused sectoral approach is taken. Many donors have no significant field presence or do not delegate manage- Aid can work, but it needs to be delivered differently to cre- ment to the field, so decisions constantly must be re- ate lasting impact in the years to come. Today, aid needs to ferred to headquarters. As a result, there is significant leverage knowledge--the hardest currency of the 21st cen- coordination fatigue. tury--to evaluate programs, identify successes, and then 3. The focus should be on the dynamics of development. scale up. A new consensus can be built around the ideals of Individual project success does not always add up to information openness and decentralization of coordination systemic change. Projects that work well are not sys- efforts. Networks of aid coordinators and aid agencies would tematically scaled up. Donors have limited long-term share information around the world, driven by common engagement in or accountability for results in a given standards for data management. With these improvements, area. A focus on dynamics can mean changing the in- money that is intended to help poor people will reach them stitutional setup, with more aggressive monitoring, in more efficient and equitable ways, and it will help deliver evaluation, and assessment of development results. greater development results (table 1). Scalable and programmatic approaches commensurate 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise with country needs have to be encouraged. It is impor- into account what others are doing. The issue is to create a tant to identify needs, interventions, and gaps at a local forum where there is a balanced representation of stakehold- and sectoral level, and to monitor progress in these ar- ers to build voice and participation, but not so many as to re- eas systematically over time. duce effectiveness and decision making. Such a model would allow for better integration of new Coordination donors and would address the criticisms of those who worry The Paris Declaration on Aid Effectiveness attempted to that coordination may stifle effectiveness and innovation in counter the fragmentation of donor interventions. The dec- the absence of a more competitive, market-driven approach laration was based on the premise that aid can never be to development. We believe that aid coordination imple- more than a complement to a government's own efforts be- mented correctly can yield benefits of tens of billions of dol- cause aid flows hardly ever exceed a government's own lars by preventing waste and achieving sustainable and fair spending, even in the poorest countries. It encouraged development outcomes. donors to use country systems and align with country prior- A fix for the development aid system--expanding the old ities. Implementing this principle, however, has been slow model to include aid from large, new bilateral donors or pur- across the board: traditional donors still rely on their own suing local ownership through joint country strategies--can systems; emerging bilateral donors often tie their aid to spe- help improve efficiency in some cases. This is especially true cific projects; and private aid donors raise funds for specific in sectors such as infrastructure where large official loans purposes, usually for social projects. The result has been an from relatively few donors are still the dominant funding increasing imbalance in sectoral aid allocations.4 channel. In other cases, however, it is unlikely that an expan- New coordination strategies aim to fix or expand the cur- sion of the current aid coordination system will be sufficient. rent model of state-based aid, and they have had some mod- The number of players is too large and the diversity of their est success in improving alignment with recipient-country approaches too great to return to the central planning ap- programs and harmonization among official DAC donors. proach of the past. What is needed is an innovative aid mod- Nevertheless, in the absence of a new and better coordina- el that incorporates new donors and aligns with national de- tion mechanism that includes emerging country donors and velopment strategies. We suggest each country develop its private NGOs, the existing aid system will face two funda- own aid coordination agency. mental challenges: 1. The system risks reverting to a pre-Paris Declaration Information mode in which traditional players will be reasonably At the level of a recipient country, a single-window approach coordinated but the overall system will remain frag- to aid regulation would be most effective in reducing the mented, resulting in sectoral and geographic misallo- confusion and inefficiency that typifies the current system. cations of aid. Each country's aid coordination agency should provide three 2. There will be a risk of renewed "monument building" basic services to donors and implementing agencies (see fig- and greater emphasis on concrete outputs than on ure 5). First, it should be the source of reliable information broader development results. Although it is easiest to about planned and ongoing projects in the country, covering report tangible outputs in education, health, and infra- all existing projects financed by the government and by both structure, the main challenge in most developing coun- public and private foreign donors. It would provide data on tries is the improvement in service delivery systems. "who does what where" and should share lessons from proj- Support for more abstract activities, such as institu- ect and program implementations. Second, it should compile tional reform, is much more difficult to mobilize and analyze data from other organs of government to present among donors and taxpayers in rich countries. These a single list of needs and to alert donors to gaps. Third, it are long-term programs and they demand strong and should debate, adopt, and enforce a basic set of minimum trusted relationships with local stakeholders, both in- standards among a country's donors and implementing agen- side and outside government. cies. These standards would include measures to ensure trans- In the new aid architecture, the inherent challenge of co- parency and prevent corruption, protect against discrimina- ordinating aid has not changed--but the magnitudes have. tion, and discourage waste. They would include approaches In the past, a developing-country government could convene to mainstream gender and environment issues. the top 10 donors and cover more than 90 percent of the aid In the case of economic development and poverty reduc- flows. Today, the top 10 donors typically account for less tion, the goals of a coordination strategy are to achieve scal- than 60 percent of total aid. Obtaining information about ability of projects, predictability of aid flows, efficient divi- aid flows and development impact is key so that govern- sion of labor, and low transaction costs. Whereas the national ments and donors can adapt development strategies to take aid coordination agency will play only a small direct role in 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 5. Country-Level Aid Coordination Documents and implementers Projects and programs Information National aid coordination Needs agency Standards Data on inputs Data on outputs Source: Authors' illustration (Fengler and Kharas 2010). securing these benefits, many of the other necessary ele- accountable to their ultimate beneficiaries--the poor people ments of an aid coordination architecture will arise naturally of the world. in the presence of open information. If donors are given in- Those who seek to improve aid coordination recognize formation on what projects are achieving notable outcomes, flaws in the aid system, but also see the promises of aid. For- they will be able to pool their resources to scale up those eign assistance cannot replace private investment or recipi- projects to reach more people. Projects that duplicate each ent-government programs in achieving desired development other's efforts will be identified and adapted to reflect a outcomes. However, efficient and well-coordinated aid is more logical division of labor, either by splitting up different critical to meeting the urgent needs of the world's poor peo- stages of the work or by operating in different geographic ar- ple and to speeding them along the path toward prosperity. eas. More generally, a geographic division of labor opens up Coordination allows for aid that better meets needs on the the possibility for competition among aid agencies--a po- ground, avoids waste, and achieves fair outcomes. tentially powerful driver of efficiency--or for tournament- A new consensus can be built around the ideals of infor- based or cash-on-delivery grant programs that offer funding mation openness and decentralization of coordination ef- to the first organizations to meet certain benchmarks as forts. In place of disconnected sets of large bilateral donors, measured by the aid coordination agency.5 multilateral organizations, and private agencies, we encour- age locally controlled agencies that promote competition Conclusions among donors with high-quality information on needs, aid inputs, and aid outcomes. Networks of aid coordinators and We advocate two institutional changes. First, we encourage the aid agencies would share information around the world, development of one (or more) geographically based develop- buoyed by common standards for data management. At the ment authorities within poor countries, with a focus on pro- end of the day, the money that is intended for poor people viding the information that is needed to run an efficient aid will reach them in more efficient and more equitable ways. program. Aid agency behavior needs to change. Individual aid and government agencies have no inherent incentive to coor- Notes dinate or share information. Their primary focus is on showing their stakeholders tangible results for their specific investments 1. The book includes an overview; case studies of and on mobilizing new funding. This has resulted in a system Aceh/Indonesia, Cambodia, Ethiopia, Kenya, Pakistan, and where the whole is smaller than the sum of its parts. Tajikistan; and thematic chapters on joint assistance strate- Second, we argue for an international body of national gies, information systems, and humanitarian aid. development aid agencies to deliberate, share best practices, 2. The Survey on Monitoring the Paris Declaration (OECD and provide an informal mechanism for holding aid agencies 2008) finds 14,000 missions in the 55 countries surveyed (p. 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise 15). Extrapolating to all aid yields an estimate of 30,000 ------. 2008. 2008 Survey on Monitoring the Paris Declaration: Making Aid missions globally. More Effective by 2010. Paris, France. http://www.oecd.org/dataoecd/58/ 41/41202121.pdf. 3. Table 1 of Kharas (2008, p. 8) compares the coefficient Singer, Peter. 2009. The Life You Can Save: Acting Now to End World Poverty. of variation of detrended country programmable aid with New York: Random House. those of GDP and exports. World Bank. 2005. Economic Growth in the 1990s: Learning from a Decade 4. At the aggregate level, there has also been a substantial of Reform. Washington, DC. http://www1.worldbank.org/prem/lesso increase in ODA allocations to social sectors, whereas aid for ns1990s/. Zinnes, Clifford F. 2009. Tournament Approaches to Policy Reform: Making infrastructure and production have declined from 59 percent Development Assistance More Effective. Washington, DC: Brookings Insti- to 38 percent between 2001 and 2004 (IDA 2008, p. i). tution. 5. For further discussion of how tournaments and com- petitive structures can contribute to aid efforts, see Zinnes About the Authors (2009). Wolfgang Fengler is the lead economist in the Nairobi office of References the World Bank, where he covers Eritrea, Kenya, Rwanda, and Somalia. He joined the World Bank in 2000 and worked in Easterly, William. 2006. The White Man's Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and So Little Good. New York: Pen- Africa and Asia. From 2004 to 2009, he was a senior econo- guin. mist in the Jakarta office and managed the Public Finance and Ek, Chanboreth, and Hach Sok. 2010. "Cambodia." In Delivering Aid Dif- Regional Development team. There he published extensively on ferently--Lessons from the Field, Wolfgang Fengler and Homi Kharas, eds., Indonesia's public finances, decentralization, and the post- 63­84. Washington, DC: Brookings Institution. tsunami reconstruction process. Homi Kharas is a senior fellow Fengler, Wolfgang, and Homi Kharas, eds. 2010. Delivering Aid Differently-- and deputy director of the Global Economy and Development Lessons from the Field. Washington, DC: Brookings Institution. Hudson Institute. 2009. Index of Global Philanthropy and Remittances. program at the Brookings Institution. He is a nonresident fellow Washington, DC. of the OECD Development Center and a member of the Na- IDA (International Development Association). 2008. "Aid Architecture: An tional Economic Advisory Council to the Malaysian prime min- Overview of the Main Trends in Official Development Assistance ister. He was a member of the Working Group for the Commis- Flows." Washington, DC. sion on Growth and Development, chaired by A. Michael Kharas, Homi. 2008. "Measuring the Cost of Aid Volatility." Working Paper 3, Wolfensohn Center for Development, Brookings Institution, Wash- Spence. Previously, Kharas served as chief economist for the ington, DC. World Bank's East Asia and Pacific region and as director for Malik, Abdul. 2010. "Pakistan." In Delivering Aid Differently--Lessons from poverty reduction and economic management, finance and pri- the Field, Wolfgang Fengler and Homi Kharas, eds., 129­64. Washington, vate sector development. DC: Brookings Institution. The authors acknowledge the contributions of those authors Moyo, Dambisa. 2009. Dead Aid: Why Aid Is Not Working and How There who provided chapters for the book on which this note is based; Is a Better Way for Africa. New York: Farrar, Straus and Giroux. Mwega, Francis. 2010. "Kenya." In Delivering Aid Differently--Lessons from and those of Jonathan Adams, Vincent da Cruz, Wren Elhai, the Field, Wolfgang Fengler and Homi Kharas, eds., 107­28. Washington, Joshua Hermias, and Abdul Malik who provided critical input DC: Brookings Institution. to this summary. They also appreciate the comments received OECD (Organisation for Economic Co-operation and Development). from Shanta Devarajan and Scott Guggenheim. 2006. DAC in Dates: The History of the OECD's Development Assistance Committee. Paris, France. http://www.oecd.org/dataoecd/3/38/18968 08.pdf. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. It is produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at www.worldbank.org/economicpremise.