Food, Agriculture, and the Environment Discussion Paper 36 Governance and Food Security in an Age of Globalization Robert L. Paarlberg International Food Policy Research Institute 2033 K Street, NW Washington, DC 20006–1002 USA February 2002 Copyright © 2002 International Food Policy Research Institute All rights reserved. Sections of this report may be reproduced without the express permission of but with acknowledgment to the International Food Policy Research Institute. ISBN 0-89629-642-3 Contents Tables iv Foreword v Acknowledgments vi 1. Introduction 1 2. Dominance of National Governments Despite Globalization 3 3. Public Goods Provision by Government 12 4. Performance of Global Governance Institutions 15 5. Africa’s National Governance Crisis and Food Security 27 6. Strategies to Correct National Public Goods Deficits 38 7. The Role of NGOs in Supplying Missing Public Goods 47 8. Conclusion—Assigning Responsibilities 50 References 52 iii Tables 1. Institutional alternatives for governance of food and agriculture, by level and sector 4 2. FAO estimates of the incidence and prevalence of chronic malnutrition in developing countries and countries in transition, 1996–98 8 3. Prevalence of child malnutrition in developing countries, by region, 1975–95 9 4. Incidence of child malnutrition in developing countries, by region, 1975–95 9 5. Net private capital flows and foreign direct investment into selected low- and middle-income regions, 1990 and 1998 10 6. Poor country (GNP per capita less than $1,000) grain import dependence by region, 1973 and 1993 15 7. Consumption of all cereals—wheat, maize, rice, other course grains—by IFPRI IMPACT regions, 1971 and 1974 18 8. Global food aid deliveries, cereals in grain equivalent, 1990–99 22 9. Food aid deliveries to Sub-Saharan Africa, cereals in grain equivalent, 1990–99 22 10. Global distribution of good governance 31 iv Whose responsibility is it to assure food security in an age of globalization? Is improved governance at the international level our greatest need, or are governance deficits most severe at the national level? When national governments lag in assuring food security for their own citizens, can outsiders help make up the resulting governance deficit? What role can bilateral donors and international financial institutions, such as the World Bank, play? Is it possible for NGOs to step in to do the job? These and related pressing questions are addressed in this discussion paper by Robert Paarlberg. He argues that the problems of hunger and food insecurity urgently require a national, not global focus. Many national governments in developing countries still do not provide essential public goods, such as civil peace, rule of law, transport infrastructure, clean water, electrical power, and public research to generate new agricultural productivity— essential ingredients in the effort to boost incomes. For tackling hunger, the weak per- formance of nation-states remains most critical—and in most critical need of improvement. According to Paarlberg, the governance challenge as far as food security is concerned is to persuade sovereign governments to provide the necessary public goods that would ensure access to adequate food. This paper was commissioned for IFPRI’s 2020 Vision Initiative conference, “Sustainable Food Security for All by 2020,” held on September 4–6, 2001, in Bonn, Germany. A sum- mary version was presented at the session on “Whose Responsibility Is It To End Hunger?” The presentation sparked a long overdue discussion on who are the key actors in the effort to eliminate hunger, how their role has changed over time, and what their responsibilities are likely to be in the future. I hope that this paper contributes to continuing this much needed debate, so that we can work more effectively to assure a food-secure world. Per Pinstrup-Andersen Director General, IFPRI Foreword v The views expressed in this discussion paper are entirely my own, yet a number of colleagues provided supportive and corrective assistance along the way. Dana G. Dalrymple, David Orden, and Nora Ng gave useful reactions to early drafts of the paper. Per Pinstrup-Andersen read a draft and provided extensive comments as well. Raymond Hopkins and Michael Lipton, two formal reviewers suggested by me, offered sometimes sharp but always construc- tive comments and criticisms. I also thank my third formal reviewer from IFPRI, who remains anonymous to me. Uday Mohan at IFPRI did a prompt and professional job managing the editing and production process. My most important supporter in this effort has been IFPRI’s Rajul Pandya-Lorch, who commissioned the paper in the first place, then gave me wise and helpful substantive reactions all along the way. Working with IFPRI, and especially with Rajul, is always rewarding. Acknowledgments vi Globalization is said to be the “key governance challenge” for the twenty-first century (Reinicke and Deng 2000, vii). The forces of globalization, which include the spread of international markets for goods, services, capital, and labor and the emer- gence of new institutions and network organiza- tions that operate easily across borders, are said to diminish the capacity of sovereign nation-states to govern their own affairs. According to a report commissioned for heads of government prior to the 2000 U.N. Millennium Assembly, “One of the chief characteristics of these globalizing dynamics is that they overwhelm the attempts of states to manage globalization alone or control its effects” (Smith and Naim 2000, 11). A conclusion sometimes drawn is that gover- nance activities must now be shifted to a level above the nation-state, into the jurisdiction of more cosmo- politan or global governance institutions (Held 1996). Others argue for a shift in the opposite direc- tion, moving governance downward in the hope that “localization” of politics and policy will keep institu- tions accountable to communities (Hines 2000). The need to globalize institutions of governance has become obvious in some policy areas, such as international money and finance, climate change, and even public health. Is the same true in the area of hunger and food security? This discussion paper argues that the greatest governance deficits in the food security area are still at the national level, not the global level. Significant hunger persists in some regions largely because of governance deficits and failures at the national, not the global level. Too many national governments in the developing world fail to provide the essential domestic public goods— such as peace, rule of law, public research, and rural transport infrastructure—needed for sustained growth of farm productivity and rural incomes. Global governance institutions have at times tried to step in and fill these national governance deficits in the developing world. But most such attempts have ended in frustration since the traditional norm of state sovereignty continues to stand in the way. The governance challenge for food security is not so much to deliver more public goods at the global level, but instead to persuade existing sovereign governments to deliver the minimal public goods needed at the national level. This argument for the continued centrality of national government goes somewhat against the grain of recent opinion. Czech President Vaclav Havel has forecast that in this new century most states will evolve into “far simpler, less powerful administrative units,” while power will move “upward to regional, transnational and global organizations.” Former U.S. Deputy Secretary of State Strobe Talbott sees the same trend: “All coun- tries are basically social arrangements. [T]hey are all artificial and temporary. Within the next hundred years nationhood as we know it will be obsolete; all states will recognize a single global authority.” Leaders of intergovernmental institutions often wel- come and encourage this supposed trend. Speaking to the General Assembly in 2000, Secretary-General Kofi Annan asserted that the institutions of the United Nations ought to be viewed as having a direct mandate from the peoples of the earth, rather than a mandate derived from U.N. member governments (Thiessen 2001, 64). Such hopes and expectations of globalized governance are certainly appropriate and accurate in some policy areas (for example, trade regulation, the battle against HIV/AIDS, open-ocean fisheries protection). Yet in the area of hunger and food secu- rity it is the performance of separate sovereign 1. Introduction 1 nation-states that remains most critical—and in most critical need of improvement. This paper emphasizes the continuing centrality of governance at the nation-state level in several steps. First, it reviews the dominant role that national governments still play in most food production, trade, and consumption activities around the world. Global markets and interstate institutions may be spreading and proliferating overall, but in the poor- est countries where large numbers of people are still hungry, and particularly in the rural regions of those countries, international food markets and global institutions still tend to have weak influence relative to local or national food markets and local or national food governance institutions. Second, it offers a concept of adequate gover- nance for food security based on public goods pro- vision and asks which institutions are currently doing the job well and which are doing it poorly. At the global level a range of international institutions have now evolved to provide a substantial number of essential food and agricultural public goods. Public goods delivery at the national level in some developing-country regions has been far less ade- quate, particularly in Sub-Saharan Africa. It is not primarily a global governance deficit that leaves Africa struggling to improve its food-security per- formance. Within Africa too many national govern- ments still fail to provide their citizens with essential national or local public goods such as civil peace, rule of law, rural roads, clean water, electrical power, and public research to generate new agri- cultural productivity. Improved governance is need- ed at every level, but these governance failures at the national level now substantially outweigh gov- ernance failures at the international level. A final section of this paper looks at some cur- rent options for improving the performance of nation- al governments in countries where hunger remains a growing problem—again primarily in Africa. It reviews a number of supporting or gap-filling roles that governments from the industrial world or inter- national governance institutions might attempt to play, and it reviews a range of options available to nongovernmental organizations (NGOs). Yet the conclusion reached is that in regions where hunger is still a serious problem, national governments on the scene must take the largest responsibility for solv- ing the problem. Our fascination with the rapid pace of globalization and with the rapid evolution of glo- bal governance institutions in other policy areas should not distract us from the heavy responsibilities that traditional national governments continue to bear in the struggle to end hunger. 2 Traditional nation-state institutions continue to domi- nate in the area of food supply and food security, particularly in poor countries where hunger prob- lems are most acute. This dominance of nation-states is somewhat surprising, given the proliferation of so many powerful and influential public and private institutions both above and below the national level. In the public sector, the second half of the twen- tieth century saw a dramatic increase in the num- bers of intergovernmental organizations (IGOs) operating above the level of the nation-state. Some of these IGOs are regional (for example, the inter- governmental institutions of the European Union). But a number are genuinely global institutions (for example, the universal membership institutions of the U.N. system). In the private sector, two addi- tional kinds of institutions have proliferated above the level of the nation-state: private multinational corporations (MNCs) and not-for-profit private inter- national nongovernmental organizations (INGOs). This emergence of multiple institutions above the nation-state dates in the modern era from the late nineteenth century. It became pronounced after World War II, and grew into a virtual explosion in the last two decades of the twentieth century. The number of intergovernmental institutions recognized by the Union of International Associations (UIA) in Brussels nearly tripled during this period, from 1,039 in 1981 to 3,019 by 2001 (Yearbook of Inter- national Organizations 1981 and 2000/2001). MNCs have proliferated as well. The number of multinational business firms in the world’s 14 richest countries more than tripled, from 7,000 in 1969 to 24,000 in 1994. By 1992 the sales of each of the top-10 MNCs was more than the gross domestic product (GDP) of at least 100 nation-states. Measured in global terms, in the 1960s and 1970s foreign direct investment (FDI) by MNCs increased at roughly the same rate as world output and trade, but then between 1985 and 1995, FDI increased eight times faster than output and more than twice as fast as trade (Reinicke 1997). In con- sequence, the stock of FDI around the world increased over six times, to reach $3.2 trillion. A number of developing countries shared in this growth. In the single decade between 1985 and 1995 annual inflows of FDI into the developing world increased from $18 billion to $99.7 billion (Vernon 1998). Private international financial flows grew even more rapidly than direct investments. Cross-border equity flows initiated by private firms increased by 300 percent in the last five years of the twentieth century, growing from $268 billion in 1995 to an estimated $1.1 trillion by 2000 (Persaud 2001). INGOs also increased dramatically in number, activity, and visibility as the twentieth century ended. According to one count, the number of INGOs worldwide grew from fewer than 10,000 in 1978 to more than 40,000 by 1997 (Cusimano 2000). Growth in number and activity of INGOs was especially strong in the area of international development. During the 1980s, development assis- tance transferred through INGOs grew twice as fast as official development assistance (ODA) trans- ferred government to government. By the end of the 1980s, some 4,000 development NGOs estab- lished in the wealthier countries of the Organisation for Economic Co-operation and Development (OECD) were disbursing billions of dollars a year for development, working with 10,000 to 20,000 “southern” NGOs in the developing world, thereby 2. Dominance of National Governments Despite Globalization 3 providing assistance to an estimated 100–250 mil- lion individuals (UNDP 1993; Clark 1991). This late twentieth century increase in the visi- bility and activity of IGOs, MNCs, and INGOs is only part of the story. Even as these international public- and private-sector institutions proliferated, other new challenges to nation-state authority arose as well. Private markets expanded, sometimes at the expense of state-owned enterprise. Under a combination of political and market pressures, including pressure from IGOs such as the World Bank and the International Monetary Fund (IMF), state institutions yielded more economic control within their borders to private markets and local pri- vate corporations. Not-for-profit institutions at the national level also grew stronger and more numer- ous in developing countries. Within a short space of time, 10,000 national and community-level NGOs were established in Bangladesh, 21,000 in the Philippines, and 27,000 in Chile. Observers have called this a “global associational revolution” which could prove “as significant to the late twentieth cen- tury as the rise of the nation-state was to the late nineteenth” (Salamon 1994, 109). At the most local level as well, the traditional institutions of the nation-state seemed increasingly under challenge. Local government authorities have demanded greater control over public revenue and greater decentralization of government regulation. These newly invigorated and diverse local institutions have proven harder for central state institutions to regulate, and a profusion of grassroots organiza- tions, including community-based development or social service organizations have taken their own initiatives to balance or correct the perceived fail- ings of the nation-state. Table 1 maps this expanded institutional ter- rain, locating traditional nation-state institutions at the center of what is now a wide range of alterna- tive institutions capable of challenging the nation- state for dominance. This proliferation of institutional alternatives to the nation-state has visibly weakened the control of national authorities in many areas of contemporary political and economic life. Nation-states are finding it harder to act alone when they seek to govern inter- national investment and finance, global environ- mental issues such as climate change, global com- mons issues such as ocean fisheries, and even public health issues like HIV/AIDS. Trade, international com- munications, and hard-to-contain cultural industries such as entertainment also elude the nation-state’s grip. Yet traditional nation-state institutions continue to dominate in the less globalized policy areas of farm- ing and food security. National political dominance over farming is conspicuous in the industrial world, where so many producers work under inducements provided by lavish national farm subsidy programs. In poor states as well, the food and farm sector tends to remain under considerable national political con- trol. Even in “weak” developing-world states where national governance institutions lack key resources, they still tend to be stronger than any alternative insti- tutions within the food and farm sectors. 4 Table 1—Institutional alternatives for governance of food and agriculture, by level and sector For-profit Not-for-profit Level private sector Public sector private sector International Multinational corporations Intergovernmental International (MNCs) organizations (IGOs) nongovernmental organizations (INGOs) National National corporations National government National nongovernmental organizations (NGOs) Local Local private tradespersons Local authorities Grassroots organizations Source: Devised by author. For similar classification scheme, see Nye and Donahue 2000. Industrial Countries In the affluent industrial world, transnational and supranational globalization forces are strong, but traditional national authorities continue to dominate food and agricultural policy. Food production pat- terns and practices continue to be shaped by na- tional agricultural trade restrictions or by national farm price support and income subsidy policies. In the case of the European Union, these traditional farm subsidy and protection policies have been aggregated into a regional policy. Nonetheless, the governance institutions of the Common Agricultural Policy (CAP) remain dominated by a council of national ministers of agriculture, and ultimately deci- sions are made through bargaining among the heads of separate national governments. IGOs such as the World Trade Organization (WTO), the OECD, and the forum for Asia Pacific Economic Cooperation (APEC) have repeatedly attempted to impose restraints on farm subsidy and price support policies in industrial states, but politically organized associations of farmers within those states have exercised enough influence to keep the lucrative subsidy systems in place. Consider the relatively weak authority of the WTO in industrial country agricultural policy. The industrial country governments that have dominated multilateral trade negotiations in the WTO have not yet been willing to subject their national farm sup- port policies to any significant international disci- pline. The obligations to reduce farm supports and limit direct export subsidies that emerged from the 1986–93 Uruguay Round of multilateral trade negotiations were so weak that they forced neither the United States nor the European Union to under- take any reforms beyond those already being con- sidered for other reasons (such as budget con- straints). The 1993 Agreement on Agriculture did require industrial countries to convert nontariff agri- cultural border protections to tariffs. But even for this technical change some important exemptions were made (Japan and South Korea were permitted for the moment to avoid tariffication obligations for rice). Moreover, the new tariff bindings were set so high that in some cases they implied an increase, not a reduction, in permitted border protection (this practice came to be called “dirty tariffication”). At the insistence of the European Union, nations unwill- ing to allow additional imports were permitted to bundle together sensitive with less sensitive products when calculating their compliance with the market access provisions of the Agreement. The most im- portant cash income support payments to farmers in use at the time in both the United States and the European Union were exempted altogether from discipline, by placing them in a so-called “blue box.” A number of other subsidy instruments were also excluded from discipline, because they were said to have either minimal market-distorting conse- quences, such as payments decoupled from market prices and planting decisions, or a public goods dimension, such as public research programs and payments supposedly linked to environmental pro- tection (Orden, Paarlberg, and Roe 1999). The financial resources of national governments in rich countries continue to dominate agricultural development assistance policy as well, despite the dramatic increase in IGOs, INGOs, and NGOs working in this area. NGOs emerged into the field of international development assistance in the 1980s not so much as challengers to donor state governments, but as adjuncts. In Norway, for example, 22 out of 70 INGOs engaged in devel- opment work obtained more than 80 percent of their budget from the state, and 39 more were state- dependent for at least 60 percent of their budget (Tvedt 1998). Nation-state tax revenues, not private voluntary contributions, are the major source of INGO funding in many donor countries including Sweden (85 percent), Belgium (80 percent), Italy (77 percent), Canada (70 percent), and the United States (66 percent) (Smillie and Helmich 1993; Riddell, Bebbington, and Davis 1995). The recent NGO revolution in the area of international devel- opment is therefore not so much a challenge to tra- ditional nation-state dominance as it is an informal institutional extension of that dominance. Developing Countries Turning to poor countries, here as well food and farm production systems and development policies tend to be shaped by national government authori- 5 ties. In many cases this large role played by state institutions is a legacy of colonial rule. The public- sector export crop production and trade systems set up by colonizing powers in much of Africa and Asia did not disappear following independence. These national commodity production and market- ing systems, dominated by state-owned corpora- tions and state monopoly marketing boards, in most instances, were simply taken over by the newly independent national government and run for the purpose of generating state revenue. At times these developing-country governments taxed the farm sec- tor so heavily as to impair agricultural productivity. Maurice Schiff and Alberto Valdes calculate that between 1960 and 1984 the net effect of direct and indirect state policy interventions in 18 devel- oping countries was an enormous income transfer out of the sector, averaging 46 percent of agricul- tural GDP per year (Schiff and Valdes 1992). The newly independent governments of the developing world may indeed have been weak in some respects, but not in their ability to extract resources from their own farmers. The biggest contrast between rich and poor countries can usually be seen not in the relative strength of national food and farm policies, but rather in the pro-farmer versus antifarmer bias of those policies. In wealthy industrial countries nation- al policy has long tended to subsidize farming thus generating surplus production, whereas in poor countries governments have more often imposed explicit or implicit taxes on farming, causing a slow- down in productivity growth. It is a perverse irony that governments in rich industrial countries, where farmers are few in number and already productive, tend to support investments in farming more than governments in poor agricultural countries where hunger persists and productivity is lagging. Scholars studying this different policy bias in rich versus poor countries have been able to link it statistically to the process of industrial development itself. When the comparative advantage of the agri- cultural sector tends to weaken relative to industry, the “national political marketplace” tends to shift from supporting an urban-biased policy of taxing farmers and subsidizing consumers toward a rural- biased posture of subsidizing farmers at the expense of consumers and taxpayers. Where the industrial sector has become most highly advan- taged relative to agriculture, as in Japan or Europe, nominal rates of agricultural protection (measured as the internal-to-border price ratio) tends to be very high. Where the farming sector has not lost so much comparative advantage (as in Australia or New Zealand) nominal rates of farm protection still tend to be positive in the industrial world, but may actu- ally be quite low.1 These differing biases in national agricultural policy tend to determine not only commodity market outcomes (surpluses in rich countries versus lagging production in poor countries) but also rural environ- mental outcomes. The threat that agriculture pres- ents to the rural environment can depend on whether the sector is being taxed or subsidized by the state. In industrial regions where farmers are well-organized politically and where national gov- ernments tend to subsidize farming, the resulting inducement to boost crop yields often encourages excessive use of chemical fertilizers and pesticides. This is one reason why so many farmers in Europe, North America, and Japan overuse chemical inputs. The environmental outcome is chemical pol- lution of surface water and groundwater down- stream from farms. Meanwhile in nonindustrial regions where national policies impose heavy taxes on farming, a different kind of environmental damage occurs. Rather than applying too much chemicals in response to subsidy incentives, heavily taxed farm- ers in most poor countries do not use enough inputs and end up mining soil nutrients, so soil fertility declines. They may also underinvest in drainage and irrigation systems, leading to problems of waterlogging and soil salinity. When yields then 6 1 Those who study this industrial transformation model of policy outcomes have used quantified measures of industrial comparative advantage within countries to predict 60 to 70 percent of all variation in nominal rates of farm-sector protection across countries (Honma and Hayami 1986). start to lag the only way to boost production to feed a growing population is to expand irrigated or cropped area. This often leads farmers to plow and irrigate fragile grazing lands, to move onto poorly suited sloped lands, or to invade forest margins. The results are accelerating desertification, soil and forest destruction, and a rapidly shrinking habitat for native species (Paarlberg 1994). Developing-country governments have been under pressure to reduce their interventions in the food and farm sector. Poor countries that borrow from international financial institutions such as the IMF and the World Bank have been told to lift national controls on internal commodity and input supply markets, relax restrictions on foreign currency exchange, privatize state-owned enterprises, and reduce wasteful employment in state bureaucracies (including food and agricultural ministries). These powerful international pressures to weaken the role of the state over the food and farm sector have been exercised in part through “structural adjustment” lending programs and policy reform assistance projects. Yet many national governments in the developing world have shown a remarkable ability to resist such pressures. In 1994 the World Bank completed a compre- hensive study of 29 Sub-Saharan African countries that had undergone structural adjustment. The study revealed that 17 of the countries did reduce the overall tax burden on farming. But some—because of persistently overvalued exchange rates—actually increased that burden, and only four of the 29 countries had eliminated parastatal marketing boards for major export crops. Pressures on gov- ernments to reform macroeconomic policies pro- duced somewhat better results, but this study drew the telling conclusion that “no country [in Africa] has good macroeconomic policies and good agri- cultural policies” (World Bank 1994, 1–2; 76–88). This finding of incomplete reform was reinforced by the conclusion of an October 2000 IFPRI Food Policy Report on agricultural market reforms in Sub- Saharan Africa: The pace and extent of reforms have varied widely across countries and crop subsectors. For the most part, reforms were not fully implemented. For example, many governments liberalized internal trade but maintained a state monopoly over exter- nal trade. In other instances, although fixed prices were eliminated, price bands for food crops were imposed to limit market price fluctuations and pro- tect consumers and producers from the allegedly “exploitative” behavior of private traders. State- owned enterprises remain active in several com- modity subsectors, notably cotton in West Africa and maize in Kenya, Malawi, and Zimbabwe. Many countries reversed reforms as a result of external shocks or changing economic conditions (Kherallah et al. 2000, 9). From the vantage point of the rural poor, the tra- ditional powers of the nation-state remain surpris- ingly dominant in most developing countries. State powers continue to be exercised through a broad range of public-sector institutions: national or paras- tatal marketing boards that monopolize the pur- chase of commodities, national or parastatal seed and fertilizer companies that monopolize the supply of key inputs, nationally controlled co-ops and nationally managed agricultural credit institutions, national research and extension services, national commodity import or export authorities, national irrigation or land-titling agencies, national forest departments, centralized service delivery agencies in areas such as health and education, and nation- ally organized public works projects such as food for work and public relief. State power is also exercised in rural areas by local representa- tives of national ruling party organizations, by national taxation and revenue authorities, and of course, by national police and military forces. Even in supposedly weak states, such national governance institutions tend to dominate in the countryside. It is often where such national gover- nance institutions most dominate that hunger prob- lems are most severe. Hunger and National Dominance Poorly fed people are found in all regions of the world. But the greatest concentrations of hungry people today are in South Asia and Sub-Saharan Africa. Not coincidentally, traditional state institu- tions continue to be strong in the food and farm sec- tors of these two regions, particularly relative to global institutions. 7 South Asia and Sub-Saharan Africa stand out as the only two developing-country regions where both the incidence and prevalence of human mal- nutrition remain high and where trends toward hunger alleviation remain weak. Table 2 presents region-by-region estimates of levels of chronic mal- nutrition in the developing world by the Food and Agriculture Organization of the United Nations (FAO). The figures are based on highly aggregated data so they conceal some important internal and local differences. Yet only in South Asia and Sub- Saharan Africa do we still find more than 175 mil- lion hungry people, combined with a regional prevalence of malnutrition above 20 percent. Table 2 reveals a large number of people still undernourished in East Asia, as might be expected given the region’s large population and its still recent movement away from deep poverty. Yet the clear trend in East Asia is now dramatically away from hunger. Between 1980 and 1997 the per- centage of East Asians experiencing undernourish- ment declined from 29 percent to just 12 percent. The prevalence of hunger also declined in South Asia (from 38 percent to 23 percent) and in Sub- Saharan Africa (from 38 percent to 34 percent). But even with these declines the prevalence of hunger remains high in both South Asia and Sub-Saharan Africa (FAO 2000). Moreover, high rates of popu- lation growth in these two regions meant that the total number of hungry inhabitants did not decrease at all between 1980 and 1997. Persistent hunger in South Asia and Sub- Saharan Africa is again visible in parallel estimates (using different data sources) of child malnutrition by region in the developing world. As of 1995, the prevalence of child malnutrition was higher than 30 percent only in South Asia and Sub-Saharan Africa (Table 3). There was some decline in South Asia from an extremely high earlier level, but there was no decline at all in prevalence of hunger among children in Sub-Saharan Africa. When considering the actual incidence of child malnutrition, South Asia and Sub-Saharan Africa stand out even more as the two developing-country regions farthest from solving their hunger problems. Table 4 shows that South Asia has more than twice the number of malnourished children as East Asia. Sub-Saharan Africa has fewer malnourished because less people reside in the region, but in Sub-Saharan Africa the absolute number of malnourished children has recently been rising rather than falling. This prevalence of hunger and chronic malnutri- tion in South Asia and Sub-Saharan Africa cannot be explained through reference to rapid globaliza- tion or international governance failures because the most powerful forces of contemporary globalization have had only modest impacts on the food and farm- ing systems of these two regions. Most South Asian and Sub-Saharan African states have opted to shield their food and farm sectors from the forces of glob- alization. Both are postcolonial regions where gov- ernments remain strongly nationalistic and eager to 8 Table 2—FAO estimates of the incidence and prevalence of chronic malnutrition in developing countries and countries in transition, 1996–98 Undernourished Region Number of persons Share of population (millions) (percent) South Asia 294 23 Sub-Saharan Africa 186 34 East Asia 155 12 Near East and North Africa 36 10 Countries in transition (former USSR, Baltics, East Europe) 26 6 Latin America and Caribbean 55 11 Total 792 18 Source: FAO (2000, Table 1). keep external market influences at bay. National institutions in these two regions have remained strong enough, despite globalization, to resist many IMF and World Bank demands for market-oriented policy reforms. It is not because the forces of global- ization have remained ungoverned in South Asia and Africa that hunger has persisted; strong nation- al governments in these regions have substantially resisted the forces of globalization. We already noted the power of national gov- ernments in Africa to resist most IMF demands for durable liberal market reforms. The consequence of this resistance has been, for Sub-Saharan Africa, a growing disconnection from many international markets, including commodity markets. During the colonial period African agriculture was deeply inte- grated into the global commodity markets of that day, but more recently this deep integration has weakened. Africa’s volume of exported coffee, groundnuts, palm oil, and sugar has actually been shrinking. It was smaller in 1997 than in 1970. As with trade, so with international investment. National policy controls over investment in Africa are so strict that they help keep most private multi- national corporate investors away. After gaining their independence in the 1960s, most African states embraced tax, regulatory, and trade policies that proved highly discouraging to new private- sector FDI. New foreign investment was also dis- couraged by the failure of some African states to preserve internal peace, enforce private contracts, or invest adequately in power, transport, and com- munications infrastructure. Thus at a time when MNCs were assuming a larger role in the econ- omies of developing countries in East Asia and Latin America, their role in Africa scarcely grew at all. By 1991–94, when average annual FDI inflows into the developing world as a whole had reached $62 billion, total inflows into all of Sub-Saharan Africa (including South Africa) were still just $447 mil- lion, which was then less than 1 percent of the developing-country total (Cantwell 1997). All of Sub-Saharan Africa was taking in just $447 million in FDI annually at a time when China—nominally still a communist country—was taking in $125 mil- lion in FDI every day. Also, the scant MNC invest- ments that did go into Africa almost never went into the farming sectors of the poorest countries in the region. Nigeria alone got 44 percent of Africa’s the FDI total in 1991–94, mostly in its energy sector. In South Asia as well, national governmental institutions and policies have tended to keep the forces of globalization at a distance. In post- independence India, it was national policy for roughly four decades to pursue development essen- tially without foreign MNCs. A labyrinth of national policies barred foreign investors from some industries entirely, restricted them elsewhere to minority owner- ship, required extensive reviews and official 9 Table 3—Prevalence of child malnutrition in developing countries, by region, 1975–95 (percent) Region 1975 1985 1995 South Asia 67.7 61.1 49.3 Sub-Saharan Africa 31.4 29.9 31.1 East Asia 33.3 26.5 22.9 Near East and North Africa 19.8 15.1 14.6 Latin America and the Caribbean 17.0 10.6 9.5 Source: Smith and Haddad (2000). Table 4—Incidence of child malnutrition in developing countries, by region, 1975–95 (millions) Region 1975 1985 1995 South Asia 90.6 100.1 86.0 Sub-Saharan Africa 18.5 24.1 31.4 East Asia 45.1 42.8 38.2 Near East and North Africa 5.2 5.0 6.3 Latin America and the Caribbean 8.2 5.7 5.2 Source: Smith and Haddad (2000). approvals, placed tight controls on their currency transactions and distribution practices, and restricted use of foreign brand names. In consequence, as late as 1988 the total stock of FDI across all sectors in India was worth only $1.2 billion. Not until 1991, when the government of Prime Minister Narasimha Rao began a number of sweeping investment pol- icy reforms, did India’s national regulatory grip on MNC activities at last begin to weaken. South Asia’s negligible use of international food markets is another indicator of its weak connection to the modern forces of globalization. The poor countries of South Asia are home to 21 percent of the world’s population and an even larger share of those who are still hungry in the developing world (roughly 38 percent in 1997, according to FAO). Yet these South Asian countries together take in only 2 percent of the world’s grain imports. The region’s reluctance to use commercial international grain markets reflects a conscious policy choice by nation- al governments to promote “self sufficiency” in food grains rather than depend on international trade. India is again a case in point. India has recent- ly accounted for roughly 10 percent of total world agricultural production, but less than 1 percent of world commodity trade. Some 2.7 million children die in India every year, 60 percent of them from dis- eases linked to malnutrition (Sharma 1999), yet national authorities tightly restrict the movement of foreign grain into the economy. India does occa- sionally import small quantities of corn, but it strictly controls these imports with a tariff rate quota that places a 60 percent duty on above-quota imports. The Government of India recently imposed an 80 percent duty on rice to curb the influx of what it called “cheap grain.” For wheat, India allows imports only rarely, to offset specific internal trans- port cost problems (for example, to allow less expensive imported wheat to reach coastal flour mills in the southern part of the country). The coun- try also exports very little wheat, despite its occa- sionally large internal surplus stocks.2 Table 5 provides additional evidence of global- ization’s weak impact in South Asia and Sub- Saharan Africa. While net private capital flows into low- and middle-income nations in Europe, Central Asia, and Latin America were increasing sevenfold to tenfold from an already substantial level during the 1990s, and while net flows into East Asia (despite the 1997 financial crisis) were more than tripling from an already high level, private flows into Sub-Saharan Africa and South Asia increased very little from a low base. FDI in South Asia and Sub-Saharan Africa also increased only slightly from very low levels, while pri- vate FDI in Latin America, East Asia, Europe, and Central Asia was exploding upward. To summarize, globalization’s impact has been relatively weak in the two regions of the world where 10 2 In 2000, when India’s excess wheat stocks reached 27 million tons, efforts were finally made to clear the stocks through export. These were frustrated in part by the presence of a wheat crop disease—”Karnal Bunt” fungus—in some parts of India, which has put India’s wheat on the import ban list of some 30 countries (APBN 2000). Table 5—Net private capital flows and foreign direct investment into selected low- and middle-income regions, 1990 and 1998 (millions of dollars) Region Net Private Capital Flows Foreign Direct Investment 1990 1998 1990 1998 Sub-Saharan Africa 1,283 3,452 834 4,364 South Asia 2,174 7,581 464 3,659 Middle East and North Africa 369 9,223 2,458 5,054 Europe and Central Asia 7,649 53,342 1,051 24,350 East Asia and Pacific 18,720 67,249 11,135 64,162 Latin America and Caribbean 12,412 126,854 8,188 69,323 Source: World Bank (2000, Table 21, 315). food-security problems remain most conspicuous. Traditional nation-state institutions remain strong in these regions, particularly relative to global-age insti- tutions such as international markets, MNCs, IGOs, and INGOs. People remain hungry in these regions not because the traditional power of sovereign states has been undercut by global markets, but more often because the powers of traditional nation-states have not yet been properly used. How should the powers of the nation-state be employed? Here disagreements abound, but for- tunately the most important function of government is also the least controversial: to provide basic public goods such as national defense, social peace, rule of law, macroeconomic stability, public education, public health, a public infrastructure for power, trans- portation, and communication, and research. These are all goods that societies need to prosper, and they are goods that the private sector is ill equipped to provide. Where hunger is worsening today, it is usually because these basic public goods are not be- ing provided by still-dominant nation-state institutions. 11 The quality of governance institutions can be rated in many ways. Here we stress a minimal compo- nent of good governance that enjoys wide accept- ance. We assume that government’s first task is to provide the public goods needed by societies to remain peaceful and prosperous, goods that are unlikely to be produced in sufficient quan- tity by private markets alone or by nongovern- mental institutions. Defining Public Goods Economists define public goods as goods with ben- efits that are available to all (they are “nonexclud- able”) and which are not diminished in their avail- ability even when consumed (they are “nonrival” or “nonsubtractable”). World peace is an example of a pure public good. It is nonexcludable in the sense that all can enjoy the consumption of world peace once it is achieved; and it is nonsubtractable in that one person’s enjoyment does not reduce the total amount remaining for others to enjoy. Another ex- ample of a pure public good is a traffic light. The safety that traffic lights offer to drivers and pedestri- ans is available to all who drive or walk on public streets and sidewalks, so it is nonexcludable. It is nonsubtractable because the safety offered to one person does not diminish that provided to others crossing the same street or to drivers at the intersec- tion in question (Kaul, Grunberg, and Stern 1999). The provision of public goods such as these can be understood as the first task of government. At the national level, important public goods would include adequate national defense, a public infrastructure of roads, power, and communications, adequate pub- lic schools and public health services, a monetary system supplying a common currency of stable value, and a court system able to enforce laws and con- tracts and protect life and property. Some of these goods may be partly price-excludable or partly rival- rous (for example, public schools that charge fees or public health services with a limited budget), yet all must be provided primarily by government. The nature of these goods is such that they are unlikely to be provided by business firms or private voluntary associations. Profit-seeking firms lack the incentive to invest scarce resources in the production of goods that are nonexcludable, since they cannot earn profits from goods available to nonpaying cus- tomers. While private business firms are generally not suppliers of public goods, they are nonetheless among the most demanding consumers of such goods. Private companies usually hesitate to locate new investments in nations where governments fail to provide peace, rule of law, or an adequate infra- structure for power, communications, and transport. Voluntary agencies that do not work for profit (including NGOs) are often motivated to produce some nonexcludable public goods. But seldom will they have sufficient resources or authority to do so on their own. NGOs do many good things, but they do not build national power grids and trunk roads, create criminal justice systems (police, courts, pris- ons) to protect life and property, or establish the lab- oratories needed to carry out basic scientific and medical research. Public governmental institutions have traditionally held the role of provider of these more expensive public goods. Governmental insti- tutions are more likely to have the financial means to provide such goods within their jurisdictions because of their exclusive sovereign right to raise revenues through taxation. Governments will also have an incentive to spend public revenues for pub- lic goods, because only through the increasing prosperity of their domestic societies (made possible through public goods provision) will their tax base 12 3. Public Goods Provision by Government and hence their own revenue grow in the long run. Theorists of political economy argue that even non- democratic governments thus have an incentive to use revenues to provide public goods, in order to maximize tax revenues for the state over the long term (Olson 2000). Additional Components of Good Governance Provision of essential public goods is, admittedly, only the first task of good government. A more ambitious vision based on an expanded view of economic or social justice would have to include provision of some nonpublic (subtractable or exclud- able) goods as well. In the area of food security, one such good might be a supply of cheap food made available to the poor through a public food- distribution system. In other cases the pursuit of food security might even require that private goods (such as land) be taken from a traditionally privileged cat- egory of citizens, with or without compensation, for redistribution to disadvantaged citizens. In still other cases food security might require government action to reduce racial prejudice or gender inequity. These are important tasks in some cases, and a more complete review of governance would have to address them. Yet public goods delivery is an essential task in all cases. The public goods definition used here stops short of requiring governments to provide essential public goods through one particular kind of gov- ernmental system, for example, a democratic rather than an authoritarian system. There is evidence that democracies are more likely than authoritarian sys- tems to provide essential public goods related to food security due to their institutionalization of social accountability through regularly scheduled competitive elections under the scrutiny of a free press (Sen 1985). Yet the statistical link between democratization and hunger reduction is simply not strong enough to focus on this characteristic of gov- ernance alone. Of the several independent vari- ables offered and examined by Smith and Haddad in a multiple linear regression to explain reduced child malnutrition in developing countries be- tween1970 and 1995, democracy had the weak- est correlation to hunger reduction. Other underly- ing determinants, such as women’s education, per capita food availability, women’s status relative to men, improvements in the public health environment (such as access to safe water), and per capita national income, all emerged as more powerful explainers (Smith and Haddad 2000). We will argue later that promoting the democ- ratization of some political systems in the develop- ing world (in countries with minimal internal ethnic conflict or at a more advanced stage of urbaniza- tion) is both a worthy and realistic objective for food-security purposes, because a democratic rule of law is likely to be better for the poor than nonac- countable rule by the strong or corrupt. Nonethe- less, there is little evidence to suggest that democ- ratization by itself can bring economic prosperity to poor countries. According to a 1995 review of 20 separate empirical studies, half of the studies had found no significant relationship between democracy and economic growth. Three did find a positive relationship, and five found a conditional positive relationship, but two actually found a neg- ative relationship (Brunetti and Weder 1995). A subsequent review of 12 additional studies uncov- ered a slightly stronger link between democracy and growth, but only slightly stronger. Of these 12 more recent studies, only one found a negative correlation between democracy and growth, while seven found a positive relationship and the remain- ing four showed results that were either inconclusive or mixed (Goldsmith 2001). In Africa specifically, no studies have found the few emerging democra- cies in the region since the 1990s to be any more prone than their predecessors to adopt economic reform programs or do better than authoritarian regimes in the region in terms of economic growth, stable prices, or balanced budgets. African democ- racies in the 1990s on the whole did neither better nor worse than nondemocracies (Goldsmith 2001). The important case of China suggests that it is entirely possible to increase food security without moving all the way to democracy. Beginning in 1978 a nondemocratic Chinese regime led by Deng Xiaoping introduced new market incentives and individual household land contracts into the nation’s farming sector, thus giving farm families more secure 13 control over their land and labor. At the same time, it made substantial public investments in agricultural research and rural infrastructure, particularly roads. Over the next two decades China’s total grain out- put increased 65 percent, from 305 million tons to annual levels averaging 500 million tons by 1999. The Chinese farmers who participated in this impres- sive feat saw their incomes rise markedly as well. Annual per capita net income for rural people in China increased from a destitute level of only 134 yuan in 1978 to 2,210 yuan ($276) by 1999. As a result, the absolute number of Chinese people liv- ing in poverty—unable to feed, clothe, or house themselves adequately—fell from 250 million in 1978 to only 34 million by 1999 (Chen 2000). This sharp decline in absolute numbers of poor people was all the more impressive given China’s continued overall population growth. Never before in human history have so many people escaped deep poverty and food insecurity so quickly. China’s leaders were not providing their citizens with a competitive elec- toral democracy, but they were providing essential public goods such as road and power infrastructures in rural areas, property security (including household control over land), access to a system of market- based exchange, and public investment in research. National Public Goods versus Global Public Goods In many sectors in today’s age of globalization, the greatest public goods deficits are no longer at the national level, but rather, at the regional or global level (Kaul, Grunberg, and Stern 1999). In finance, trade, communications, transport, public health, mon- etary policy, and environmental protection globaliza- tion has increased the need for common interna- tional regulatory frameworks and mutual assistance schemes. The struggle has been to design and empower global institutions capable of delivering these global public goods. Yet in food supply and food security, today’s most conspicuous public goods deficits are not found at the global level. While the demand for such public goods at the regional or global level has certainly increased in recent decades, fortunately the supply of these goods has also increased. It is at the national level that public goods deficits remain most pronounced. In order to make this point with adequate precision, we must first examine at some length the relatively strong per- formance of international institutions in delivering the global public goods most important to food security and hunger reduction in poor countries. 14 Global governance institutions for food security and hunger reduction are poorly funded and far from per- fect. Nonetheless, they work well enough to provide an impressive range of tangible hunger-reducing global benefits. Function by function, the governance institutions currently in place at the international level have done a relatively good job of delivering the essential public goods they were designed to pro- vide. Consider three global food-security functions in particular: the regulation of international commodity markets, the international delivery of food aid (includ- ing the provision of famine early warning and famine relief), and the supply of internationally usable agri- cultural research. Regulation of International Food and Commodity Markets Today’s problems of malnutrition and food insecurity in Africa, and elsewhere in the developing world, are not strongly linked to the governance of global food markets. International food markets are not an especially important factor one way or the other. Local commodity markets are far more important than international markets in determining the nutri- tional circumstances of the poor. It is generally wealthy countries, not poor countries, that dominate international markets for food and animal feed, both as importers and as exporters. Genuinely poor coun- tries tend to be less prominent exporters into these international markets, and only a few are heavily reliant as importers. International food markets do tend to be heavily used by some upper- middle-income developing countries in the oil- producing regions of the Middle East and by rapidly industrializing countries in East Asia. Yet these coun- tries are not genuinely poor and their hunger prob- lems are far less severe than those of Sub-Saharan Africa, South Asia, or even Central America. If we consider only those developing countries that are genuinely poor (defined as those with a gross national product (GNP) per capita of $1,000 or less in constant 1987 dollars) and measure im- port dependence on world food markets as the ratio of annual cereal imports to annual national produc- tion, Table 6 shows average poor-county regional import dependence for 1973 and 1993. These data reveal that dependence on international cereal mar- kets is quite low for most genuinely poor countries. It has been extremely low and declining for the gen- uinely poor nations of East and South Asia (this includes China, Indonesia, India, and Bangladesh under the definition of poverty used here). This poor-country pattern of not relying heavily on world markets for grain imports goes against some of the expectations popularized by respected 15 4. Performance of Global Governance Institutions Table 6—Poor country (GNP per capita less than $1,000) grain import dependence by region, 1973 and 1993 (percent) Import Dependence Region 1973 1993 Sub-Saharan Africa 10.0 13.6 South Asia 5.5 2.0 East Asia and Pacific 5.3 3.8 Latin America and Caribbean 17.6 36.5 Middle East and North Africa 22.5 8.3 All poor countries 6.4 5.2 Source: FAO (1973, 1975, 1985, 1993, and 1995). analysts several decades ago. In 1977, IFPRI pro- jected that India’s food import dependence would increase rather than decline, and would reach 10–12 percent by 1990. IFPRI expected that by 1990 Bangladesh’s food import dependence would be as high as 30–35 percent (IFPRI 1977). We can see in retrospect that these projections were off by several orders of magnitude. In the case of India and Bangladesh, analysts two decades ago badly overestimated the willingness of poor-country governments in South Asia to depend on the world market for imports. Nations in this region do have unsatisfied food needs, yet their restrictive food importing policies have kept reliance on world mar- kets to a minimum. Perhaps restricting imports is an excusable pol- icy in countries where the prevalence of hunger is declining, as in South Asia. But it is difficult to understand how it could be acceptable in Sub- Saharan Africa. Table 6 shows that among the increasingly hungry countries of Sub-Saharan Africa, dependence on the world market for imports remains minimal. Sub-Saharan Africa imports a total quantity of cereals (combining both commer- cial purchases and food aid) equal to less than 15 percent of annual domestic production. By focusing only on grains, Table 6 actually overstates the import dependency of poor countries in Africa. Taking into account foods other than cereals (such as tubers and root crops, for example) Africa’s dependence on imports from the world market is lower than 15 percent. The World Bank estimates that while more than 10 percent of Africa’s total grain consumption may have been imported in 1988–92, only 6.5 percent of total calorie con- sumption in Africa came from imported grains (Ingco, Mitchell, and McCalla 1996). Such averaged or aggregated estimates con- ceal significant variations within the region of course. If data from regional grain exporters such as South Africa and Zimbabwe are excluded, the import dependence of the rest of Sub-Saharan Africa is somewhat higher in most years. Yet even so past expectations regarding Africa’s food import dependence have simply not come to pass. IFPRI in 1977 projected much higher import dependency ratios for all of Sub-Saharan Africa, including a 44–46 percent import dependency ratio by 1990 for the Sahelian countries in particular. Only in Latin America and the Caribbean do we find a grouping of genuinely poor countries where import dependence on international cereal markets has been relatively high, and climbing. The genuinely poor countries in this region—Bolivia, Dominican Republic, El Salvador, Guatemala, Guyana, Haiti, Honduras, Nicaragua, and Peru— imported 4.6 million tons of cereals in 1993, while producing only 8.0 million tons at home, giving them an import dependence ratio of 36.5 percent, well above the 17.6 percent ratio seen in 1973. These poor western hemisphere nations, rather than the poor nations of Africa, appear to have the largest interest in good governance of international grain markets, yet these are all relatively small nations with relatively small populations currently experiencing food deprivation. Together, these poor western hemisphere countries contain only 1.3 per- cent of the world’s citizens, and they take only 2 per- cent of world cereal imports. So even if their import needs were suddenly to double or triple, the world market would be able to accommodate the increase. Apart from the limited dependence of poor countries on global food markets, we must ask how well managed those markets are. Advocates of improved global governance might argue that dependence on world markets for imports is low among poor countries today precisely because of those markets’ substandard performance. Perhaps if world food markets were better managed and more dependable, poor countries would be willing to depend more upon them. While there is certainly room for improvement, from the vantage point of importers international food and commodity markets have in fact performed quite well, often much better than internal food markets in most poor countries. One way to judge the performance of inter- national food markets from the vantage point of poor- country importers is to look at the changing purchase price of basic staple grains in those mar- kets. Over the long term, these prices have fallen significantly. Adjusted for inflation, the price of wheat and corn available for export from the United States fell by 63 percent between 1910 and 1988. The price of wheat available for export fell 16 by 67 percent over this same period (Johnson 1991). Because ocean transport costs declined as well over these decades, the final import price for poor countries fell even more in real terms. Economic models suggest that the import price of cereals on the world market will remain low even if some large developing countries should abandon their current practice of avoiding use of those mar- kets. For example, if personal incomes in India were to grow rapidly, causing meat consumption there to double by 2020 over the currently projected level, and if India, in consequence, began importing much more meat and 26 million tons of cereals by 2020 (for livestock feed) rather than the currently projected 6 million tons, the impact on world cere- al prices would still be quite small. World maize prices might increase by 5 percent rather than decline by 1 percent, and wheat prices might decline by 3 percent rather than decline by 8 per- cent (Rosegrant et al. 2001). It could be argued that the low price of cereals on the world market is partly a consequence of sub- sidized food production in rich countries and that it indirectly weakens food security by allowing governments in poor countries to skimp on agricul- tural investments and instead rely on imports from abroad. This would be a stronger argument if the poor countries in South Asia and Africa—those with the most acute food-security problems today—had in fact allowed themselves to become significantly reliant on food imports. But as noted above, most of these very poor countries have decided not to rely on imports, in hopes of being able to claim “national self-sufficiency” in basic food supplies. The price of food on the world market has been low and declining overall, yet importers do suffer from occasional price spikes. One example was the relatively sudden increase in international wheat prices between 1994 and 1996, from an average $157 per ton in the 1994/95 season to a momen- tary high of $271 per ton in early May 1996. These price spikes have sometimes been caused by malfunctions from well beyond the food and farm sector, such as inflationary or deflationary macro- economic policies or adverse trends in international financial confidence. Yet sometimes world market price spikes do reflect the poor governance of inter- national food markets. One reason for the price increase between 1994 and 1996 was a destabi- lizing policy switch within the European Union away from subsidizing exports toward imposing duties on exports as international markets tightened. As noted, the disciplines of the WTO have yet been insufficient to block all such market-destabilizing rich country policies. Such flaws not-withstanding, inter- national markets provide important and mostly de- pendable options to importers, and when prices go up suddenly it usually does not take long for an off- setting global production response to follow. The 1996 price spike in international wheat markets trig- gered so much added global production that by 1998 the export price of wheat on the world market had fallen back down to below the 1994/95 level. There are reasons to believe that price stability in world food markets will in any case improve in the years ahead, as the volume of food traded on those markets continues to grow (international cereal markets today are already 50 percent larger than they were in the 1970s) and as the agricultural poli- cies of the nations that dominate those markets continue to move gradually away from the use of destabilizing illiberal practices. The industrial coun- tries in Europe, North America, and the Far East that have long been the shapers of world food mar- kets continue to subsidize their own farmers heavily. Yet they have now taken at least some steps to restrict the use of highly trade-destabilizing policy instruments, such as nontariff import restrictions and export subsidies. The international Agreement on Agriculture that emerged from the Uruguay Round negotiations is designed to add further stability to international prices by converting nontariff barriers to tariffs, and also by reducing permitted export sub- sidy use. Under the Agreement, permitted budget expenditures for export subsidies were reduced by 36 percent and the permitted volume of subsidized exports was reduced by 21 percent over a 10-year period (Dixit 1996). These disciplines are relatively weak, as noted earlier, but they do move the governance of international food markets in the right direction. International grain market price fluctuations receive considerable attention, and they are an im- portant issue for the heaviest users of international 17 grain markets (mostly the nonpoor). But these price fluctuations have never been the main source of food insecurity among genuinely poor developing coun- tries. This can be said even for the so-called “world food crisis” period of the 1970s, when world market conditions were badly disrupted and then widely blamed for a perceived increase in hunger. At that time when the price of internationally traded food rose sharply, it was simply assumed that hunger in poor countries probably would increase. Yet there was never much evidence to support this conclusion, and in most cases it was an erroneous inference to draw. Most genuinely poor countries relied so little on the international market, or their internal markets were so segmented by policy from the world market, that their own domestic food prices moved up much less than the world market prices. Also, while inter- national food prices did rise on this occasion, the principal reason was not a global food production failure. High-income growth around the world, due to easy credit and inflationary macroeconomic poli- cies, had driven the price increases, and these were macroeconomic circumstances under which most people actually found themselves better fed. Between 1971 and 1974 the real export price of U.S. wheat increased by 103 percent and the real export price of U.S. maize by 58 percent. World food reserves simultaneously declined from 71 days worth of grain consumption to just 33 days (Johnson 1991; Hopkins and Puchala 1978, 7). Many analysts assumed that under these tightened world market conditions only the rich would be able to sustain their accustomed consumption levels. In fact, it was the rich who cut back most during this crisis, by reducing their per capita meat consump- tion. Per capita food consumption in most genuine- ly poor countries did not decline. FAO estimates of 1971–74 per capita grain consumption levels by country and region (Table 7) show no overall pat- tern of decline. While per capita consumption did decline slightly in some nations or regions, else- where in the developing world per capita cereal consumption either remained steady or actually rose while the “food crisis” was at its worst. Consumption adjustments were small in poor countries in part because the world market worked well enough to trigger large adjustments in rich countries. In 1973–74 when grain prices rose, the feeding of grain to livestock declined in the United States by 37 million tons, or approximately 25 per- cent. Canada and Australia also cut feed use in response to high prices. Use of feed grains declined so much in key exporting states in 1973–75 that it was possible at the height of this so-called world food crisis for the rest of the world to continue increasing grain consumption, not only by people but also by animals (Johnson 1991). Reduced feed use of grains in wealthy exporting countries did not result in food insecurity among the wealthy, of course; it led to higher meat prices and reduced consumption of red meat, which was on balance a nutritional benefit. The later increase in world cereal export prices in 1995–96 also failed to produce any noticeable decline in per capita consumption in genuinely poor countries. Between 1994/95 and 1995/96 U.S. wheat export prices increased from $157 per ton to $216 per ton, and world cereal stocks 18 Table 7—Consumption of all cereals—wheat, maize, rice, other coarse grains— by IFPRI IMPACT regions, 1971 and 1974 (kilograms per capita) Region/Country 1971 1974 Latin America Mexico 167 168 Brazil 96 102 Argentina 131 127 Colombia 76 81 Other Latin America 108 107 Africa Nigeria 64 61 Central and West Africa 66 65 Southern Africa 115 117 East Africa 70 78 Egypt 165 174 Asia West Asia/North Africa 155 167 India 130 126 Pakistan 115 125 Other South Asia 96 99 Indonesia 125 135 Malaysia 157 160 The Philippines 114 119 Myanmar 176 175 Other Southeast Asia 161 168 Source: FAO (2001) compiled into regions used for IFPRI's IMPACT model. as a percentage of world consumption fell from 17.8 percent to just 14.1 percent, generating talk of another world food crisis. Yet the imports of most developing countries were sustained and average per capita food use of cereals in developing coun- tries overall continued to increase. Average annual per capita cereal consumption in the developing world as a whole actually increased from 170 kilo- grams in 1994/95 to 171 kilograms in 1995/96, and then to 172 kilograms in 1996/97, despite much higher world grain prices (FAO 1998). In many poor countries food consumption cir- cumstances were actually better in the mid-1970s and then again in the mid-1990s when grain export prices were high, than in the mid-1980s when grain export prices were low. Over the “food crisis” decade of the 1970s, the share of the popu- lation that was chronically malnourished signifi- cantly dropped in Latin America from 19 to 13 per- cent, in the Near East from 22 to 12 percent, and in Sub-Saharan Africa it remained steady (at rough- ly one-third) despite exceptionally rapid population growth in that region (USDA 1995, 46). During the 1980s, in contrast, when world grain markets were slack, export prices low, and world stocks abundant, food consumption circum- stances in many poor countries actually worsened. In Africa overall, rates of dietary improvement fell by two-thirds during the 1980s compared to the 1970s, and FAO estimated that the number of chronically undernourished people in Latin America and the Caribbean grew from 46 million around 1980 to over 60 million by the early 1990s, reach- ing roughly 14 percent of the population (Alexandratos 1995; FAO 1991). The 1980s were marked by low international grain prices (described as a world food glut at the time). Yet this decade was one of severe food crisis within both Africa and Latin America due to the onset of a world recession. High interest rates after 1980 and lower world demand brought reduced income and export earn- ings to these developing regions, unserviceable external debts, and almost no income growth. For Latin America and the Caribbean, real GDP growth rates fell from a 1970s annual average of 5.7 per- cent to just 1.2 percent in the 1980s. For Sub- Saharan Africa, real GDP growth fell from a 1970s annual average of 3.4 percent to a 1980s annual average of just 1.8 percent (Grindle 1996, 20). Under these circumstances hunger increased, despite the abundance of grain on the world mar- ket (Paarlberg 2000). Fluctuating world food market conditions are therefore not by themselves a reliable indicator of food insecurity or hunger in most poor countries. Internationalists have repeatedly sought to improve world food security by imposing tighter regulations on price movements in international markets or by creating grain reserves and compensatory finance mechanisms to assist importers. But the evidence suggests that little would be gained by adding such supplementary governance features to world food markets. International markets for other kinds of goods, such as currency exchange and finance, are clearly in need of improved global governance. The free international flow of capital promoted in the 1990s by institutions like the IMF has at times imposed needlessly harsh adjustment burdens on the poor, leading in East Asia after 1997 to a tem- porary increase in hunger. Yet by comparison, inter- national food and commodity markets have operat- ed remarkably well. The poor countries in the developing world that have been willing to use these markets, rather than impose arbitrary anti- trade restrictions in the name of self-sufficiency, have found them to be an affordable and mostly reliable supplementary source of food supplies. Yet developing countries with food-security prob- lems require more than well-functioning international markets for the import of food. They also need effec- tive international markets for their own farm com- modity exports, and here the international market does less well in providing for the needs of poor countries. Especially for developing countries able to export value-added products in competition with industrial country farmers, the international market- place remains marred by serious protection. Indus- trial countries have at times been able, despite WTO rules, to impose restrictions on commodity imports, reducing export earnings for the developing world and harming the income prospects of poor farmers. For example, in 1998 the European Union established a new regulation limiting the amount of aflatoxins in imported food. This new regulation set 19 an aflatoxin standard tighter than that suggested by several international food safety governance bod- ies, including the Codex Alimentarius Commission in Rome, FAO, and the World Health Organization. Moreover, the health benefits that E.U. citizens can gain from this higher standard are likely to be triv- ial. The World Bank estimates that the difference between the new E.U. standard and the Codex standard may help Europe to avoid only 1.4 deaths per year for every one billion consumers. Yet it could reduce cereal, dried fruit, and nut exports to Europe from nine African countries (Chad, Egypt, Gambia, Mali, Nigeria, Senegal, South Africa, Sudan, and Zimbabwe) by 64 percent, costing these exporters $700 million each year (Otsuki, Wilson, and Sewadeh 2000). Such restrictive measures by importers are seri- ous market malfunctions. Still, evidence suggests they are not the chief reason why most commodity producers in Africa have lost export sales. Africa’s agricultural exports have dwindled not so much because of import protection by rich countries as due to prejudicial sectoral and macroeconomic policies imposed by African governments at home. Between 1962–64 and 1991–93, Sub-Saharan Africa’s share of various agricultural commodity exports (such as vegetable oils, palm oil, palm nuts and kernels, and groundnuts) dropped 47–80 per- centage points below earlier levels. Between 1955 and 1990 Sub-Saharan Africa’s share of global exports of all products fell from 3.1 percent to just 1.2 percent, implying substantial annual trade loss- es. Yet the World Bank determined that this disap- pointing export performance could not be explained by industrialized country import policies. African exporters have tended to face lower aver- age tariffs than other exporters. Nontariff protection against African exports is also generally less restric- tive than that facing other developing countries. The overall external environment for exports facing Africa today (tariff and nontariff) is actually more favorable than that which today’s more wealthy East Asian economies previously faced and over- came (Yeats, Amjadi, and Reincke 1996). Africa’s damaging marginalization in world commodity trade more nearly reflects impediments within the region itself to efficient commodity pro- duction and export. Africa’s shrinking share of world trade is most accurately described by Jeffrey Sachs as a “self-imposed economic exile” (Sachs 1996). Most African states have not actively pur- sued a trade-linked growth strategy. Trade policies in most other regions have moved slowly toward greater liberalization within the WTO, but during the recent Uruguay Round of trade negotiations the African continent mostly sought exemptions from trade-liberalizing obligations (Hertel, Masters, and Elbehri 1998). Governance deficits at the international level are thus not the principal reason why some poor countries are failing to make gains from inter- national commodity markets. These markets offer importers, in particular, an abundant supply of food commodities at prices low and stable enough to make the risks associated with dependence on com- mercial imports acceptable. Some poor and food- insecure countries remain reluctant to engage in commercial imports but the markets are nonetheless available as a valuable global public good for those that opt to use them. International Food Aid The availability of sufficient international food assis- tance might be viewed as a second important inter- national public good. Arranging adequate food imports on commercial terms can be difficult for poor countries with large external debts and lagging for- eign exchange earnings. The poor citizens of such countries, particularly those living in urban areas, may require a well functioning global concessional food assistance system to supplement commercial food markets. Concessional food aid since the late 1980s has in fact provided more than 40 percent of total cereal imports for over 40 recipient countries, mostly in Africa (FAO 1996). Fortunately, food aid is another area where existing global governance institutions have generally performed well. The international food assistance system is still dominated by national governmental institutions at both the donor and the recipient end. Virtually all international food assistance is financed by indus- trial world governments. NGOs do play a visible role in channeling food aid, but 97 percent of the 20 food they deliver is financed either by the United States, the European Commission, or by other indi- vidual national governments in Europe, Canada, or Japan. Only 3 percent of food aid delivered by INGOs is actually financed by the INGOs them- selves (WFP 2001a). Governmental donors coordi- nate their efforts through a variety of international governance institutions, particularly the international Food Aid Convention (also called the London Convention). This Convention is a legal international agreement that lays down minimum annual food aid commitments, donor by donor, either in terms of total tonnage or market value. The 1999 version of the Food Aid Convention set an aggregate mini- mum annual commitment from donors of 4.895 mil- lion tons of food assistance, plus a total value com- mitment of €130 million. Commodities considered eligible for these commitments include grain, pulses, edible oil, root crops, skimmed milk powder, sugar, and seed for eligible commodities. The United States and the European Union dominate as food aid donors. Since 1995 these two have coordinat- ed their actions separately under a food-security coordination program, as part of the U.S.– E.U. Transatlantic Agenda consultation process (Christensen 1999). Food assistance is delivered through a wide variety of channels, but again national govern- mental institutions tend to dominate. More than half (55 percent) of all global food aid moves directly from donor institutions to recipient governments. Another 29 percent moves through multilateral public-sector channels (almost entirely through the World Food Programme), and 16 percent is chan- neled through INGOs. Some food aid channeled internationally via multilateral public-sector agen- cies such as the World Food Programme is sub- sequently distributed by INGOs or NGOs within recipient nations. Food assistance is delivered to recipient nations in three general forms: as emergency relief, as proj- ect assistance designed to improve nutrition and support development (project aid), or through a continuing government-to-government commodity transfer program (program food aid). In 1999, 32 percent of all global food aid deliveries were for emergency relief (almost half of that went to Asia, in particular North Korea), 17 percent were project food aid deliveries (40 percent of which went to Sub-Saharan Africa), and 51 percent were pro- gram food aid deliveries, where Russia was the main recipient. The adequacy of this global food aid delivery system was twice briefly called into question in the mid-1990s: following completion of the 1994 Uruguay Round Agreement on Agriculture in the WTO and following the 1996 enactment of new agricultural legislation in the United States. The 1994 Uruguay Round Agreement raised concern among some developing-country officials that reduced domestic support to farmers in exporting countries could result in lower surplus stocks, increased international price variability, and reduced incentives on the part of exporters to pro- vide food aid (Ballenger and Mabbs-Zeno 1992). These anxieties were aggravated when the U.S. Congress in 1996 enacted a new farm law designed to support agriculture with cash payments to grain farmers—payments substantially decou- pled from traditional production incentives (Orden, Paarlberg, and Roe 1999). Roughly half of all inter- national cereal food aid traditionally came from the United States, so this policy move away from strong direct production incentives to farmers suggested that a traditional foundation of food aid—surplus production in wealthy countries—might be eroding. From today’s vantage point, these concerns appear to have been exaggerated. The Uruguay Round Agreement on Agriculture actually took great care to protect the functioning of international food aid systems. First it restated donor countries’ obli- gations to set their commitment levels to the Food Aid Convention high enough to meet the reason- able needs of developing countries during the trade liberalization process. Second, it exempted food aid shipments from the tightened restrictions on export subsidies imposed by the Agreement. Even program food aid shipments arranged through long-term credit agreements, such as the substantial U.S. grain exports funded under Title I of Public Law 480, were exempt from any new restriction under the Agreement (Christensen 1999). In the case of the new 1996 U.S. farm law, while it did promise to contain the size of publicly held food stocks, in 21 some other respects the new law was actually good for international food abundance because it increased commercial production potential in the United States by eliminating the authority of the U.S. Department of Agriculture to impose annual acreage reduction requirements (called ARPs) on farmers receiving income support payments. Since the mid-1990s food aid shipments have in fact been sustained above World Food Programme minimums, despite the new WTO agreement and despite the new U.S. farm bill. World Food Programme data show that interna- tional food aid deliveries did dip briefly in 1996–97, at a time when momentarily higher world grain prices discouraged large donor contri- butions. But total donor contributions never fell below the annual minimum of 5.4 million tons of cereals then prevailing under the Food Aid Con- vention. By 1998–99, as world grain prices fell, donor contributions climbed once again. Table 8 shows the total tonnage of global food aid deliv- ered over 1990–99. Recent trends in Sub-Saharan Africa specifically provide additional reassurance that the international food aid system can continue to provide adequate concessional flows. Program food aid to African governments has generally declined over the past decade, but project food aid for nutrition and devel- opment purposes has substantially increased, and emergency relief has been able to increase when necessary, as it did following the severe southern African drought of 1991–92. Table 9 shows recent trends in food aid deliveries to Sub-Saharan Africa. The availability of international food aid for emergency relief has played a significant role in containing some kinds of hunger, particularly in Africa. Emergency food aid is not always able to contain famine in Africa, but when failures occur the international governance of food aid is usually not the problem. Ethiopia’s difficult experience helps put such issues in perspective. Food aid arrived too late in Ethiopia to prevent famine in 1984. Most PL-480 shipments from the United States arrived in 1985 and 1986, and by then the worst of the famine had passed and a recovery of local production was already under way (Barrett 2001). Yet in this case the tardy arrival of the food aid could be blamed mostly on reluctance by the 22 Table 8—Global food aid deliveries, cereals in grain equivalent, 1990–99 (million tons) Contribution 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Relief 2.0 3.4 5.0 4.2 4.5 3.5 2.7 3.3 3.0 4.7 Program 8.4 6.9 7.7 10.6 5.7 4.3 2.9 1.8 2.7 7.4 Project 2.7 2.5 2.6 2.5 2.7 2.3 1.7 2.3 2.6 2.4 Total 13.1 12.8 15.3 17.3 12.9 10.1 7.3 7.4 8.3 14.5 Source: WFP (2000). Note: 1999 data are provisional. Table 9—Food aid deliveries to Sub-Saharan Africa, cereals in grain equivalent, 1990–99 (million tons) Contribution 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Relief 1.5 2.4 3.7 3.0 3.0 2.0 1.6 1.3 1.6 1.6 Program 0.9 1.0 1.9 1.2 0.8 0.7 0.5 0.5 0.6 0.3 Project 0.5 0.6 0.6 0.7 0.7 0.6 0.4 0.6 0.7 1.0 Total 2.9 4.0 6.2 4.9 4.5 3.3 2.5 2.4 2.9 2.9 Source: WFP (2000). Note: 1999 data are provisional. Ethiopian government, earlier in 1984, to allow any international reporting of the severity of its growing internal food crisis. In 1999–2000, inter- national food aid shipments again arrived too late in Ethiopia to prevent famine deaths amid a dam- aging drought. Significant international humanitar- ian intervention began only in April 2000, by which time more than 70,000 people had already died. Yet World Food Programme officials point out that while some lives were lost before April, many more lives were subsequently saved by food aid. The World Food Programme eventually provided food to 2.5 million people in Ethiopia’s Somali region. Had there not been civil unrest in this region, even more could have been reached. When recipient governments are prepared to cooperate and when there is no violent internal conflict underway to obstruct an international relief effort, the international food aid system is usually able to provide timely assistance. One recently developed international governance instrument that has helped speed food aid relief has been the World Food Programme’s $20 million Immediate Response Account (IRA) system. Since 1993, this funding mechanism has given program officials working in developing countries the option to launch food aid operations immediately, on their own initiative, at funding levels up to $200,000 each. In 2000, coun- try directors used this authority 11 times in countries such as Nicaragua, Zambia, Mozambique, and Viet Nam. IRA funds were also used to initiate larger emergency food assistance operations in Eritrea, Ethiopia, and Kenya (WFP 2001b). Improved international famine early warning systems have become another important feature strengthening international food aid governance. These systems use a combination of market-price and meteorological data monitoring, plus increas- ingly sophisticated remote sensing satellite informa- tion, to plan and mobilize responses to food emer- gencies before they become acute. To illustrate the potential of these systems, consider the effective inter- national response when widespread drought hit southern Africa in 1991–92. The drought cut aggre- gate cereal production in the region by more than 50 percent on average; and in Malawi, Namibia, Swaziland, and Zimbabwe cereal production actu- ally fell by 60–70 percent. Because of already depleted maize stocks in the region, the drought put 17–20 million people at risk of starvation. Yet famine deaths were reported only in Mozambique, where relief was politically and logistically impossi- ble because a civil war was still under way. Starvation was avoided in the rest of the region because per capita food aid increased quickly and dramatically, from an average of less than 10 kilo- grams per person in the 1980s to a peak of more than 25 kilograms per person in 1992 (Pinstrup- Andersen, Pandya-Lorch, and Babu 1997). Improved international early warning systems played a key role in facilitating timely delivery of this assistance. In December 1991, famine early warn- ing systems supported by FAO picked up the devel- oping drought, and by the end of February 1992 the systems confirmed the situation was critical. In March and April, FAO and the World Food Programme sent joint crop and food-supply assess- ment missions to the region, to judge food import and food aid needs for the coming year and to com- plete a comprehensive logistics assessment, includ- ing a review of port capacities in South Africa. FAO/World Food Programme coordination with national governments was accomplished through a regional institution, the Southern African Develop- ment Community (SADC). To ensure adequate finan- cial and commodity support from the international donor community, FAO’s Global Information and Early Warning System (GIEWS) issued a special alert in April 1992, which was followed up by a joint U.N.–SADC consolidated appeal for assis- tance. Donor response to this appeal was gratifying, as pledges received covered 82 percent of all tar- geted food aid requests and 89 percent of all pro- gramme food aid requests. To move the assistance, SADC formed six differ- ent “corridor groups” to handle port, rail, and road transport through the region. Contributions to trans- port and logistics were roughly twice the amount requested by SADC and came from a wide variety of donors including numerous NGOs and other con- cerned institutions that participated actively and effec- tively in the various relief activities (FAO 1996, Volume 3, 42–45). Donors then worked through the World Food Programme to create a logistic advisory 23 center to collect and share information related to potential port and transport bottlenecks. Both FAO’s GIEWS and the U.S. bilateral Famine Early Warning System Network (FEWS-Net) developed and made use of extensive networks of on-the-ground informants to gather and then disseminate information. The inter- national response in southern Africa was a remark- able achievement in providing the global public good of famine early warning and prevention.3 International food assistance efforts have been far less successful in cases where recipient govern- ments either deny information (as did Ethiopia in 1984) or block international access (as with the North Korean famine after 1995) or where violent internal conflicts prevent relief from reaching the individuals in need. Violent conflicts are not only a leading cause of short-term food emergencies in much of the developing world, they are also a lead- ing barrier to effective international relief. The tem- porary interruption of World Food Programme over- land relief to vulnerable populations in Afghanistan following the onset of a U.S. bombing campaign in October 2001 is the most recent case in point, but a number of African examples are also illustrative: • Widespread drought in eastern and western Sub-Saharan Africa in the mid-1980s led to failed harvests for three consecutive years in a number of countries, threatening the survival of vulnerable populations. More than 35 million people were directly affected, and some 10 mil- lion eventually left their homes in search of food and water. Yet in the affected countries where peaceful conditions prevailed, international food relief was provided with gratifying success (Deng and Minear 1992). As Jean Dreze later observed, “Though drought threatened a large number of African countries at that time, only some of them—notably war-torn ones—actually experienced large-scale famine” (Dreze 1995). • When northern Sudan faced a severe drought in the mid-1980s, it managed to avoid wide- spread starvation thanks in part to the accept- ance and distribution of $1 billion in external assistance. Yet when violent civil conflict later escalated in southern Sudan, relief could not be delivered to areas still being affected by drought so hundreds of thousands starved between 1986 and 1988. By 1988 roughly half of the population in southern Sudan had been displaced by fighting, and famine deaths in that year alone reached 250,000. A new international relief effort (Operation Lifeline Sudan) was mounted in response to this conflict- linked emergency. But it was far less successful than the earlier international drought relief effort in the north, due in part to armed attacks on food shipments by the warring parties (Deng and Minear 1992). • Somalia, Ethiopia, and northern Kenya were all devastated by the same widespread drought beginning in late 1991. Yet in the latter two countries there were few deaths because inter- national relief efforts were able to get food to those at risk. In Somalia, however, food relief shipments were blocked by armed subclan mili- tia groups engaged in a struggle for political control, leading to significant starvation. Minimum food security was temporarily restored in Somalia only after U.S. military intervention late in 1992 afforded protection to international food relief shipments (Natsios 1996). International food assistance can thus be viewed as an area where coordination and gover- nance mechanisms at the international level are quite advanced. The global governance achieved is certainly far from perfect. Food aid availability is still too closely tied to donor country agricultural sur- pluses, implying that too much is sometimes given 24 3 Africa’s food problems are sometimes blamed on the region’s weak capacity for regional cooperation. Yet for food aid, compe- tent regional governance institutions are already in place. In the eastern part of Africa and in the horn of Africa, food aid can now be coordinated either through a fifteen-country regional institution—the Greater Horn of Africa Initiative (GHAI)—managed by USAID, or through the seven-state regional Intergovernmental Authority on Development (IGAD), consisting of Djibouti, Eritrea, Ethiopia, Kenya, Somalia, Sudan, and Uganda (Christensen 1999). when it is least needed, depressing incomes of farmers in recipient countries. Donors should pur- chase more of their food aid supplies from farmers in the developing world (“triangular” food aid), thereby giving income to the producers in greatest need. Nonetheless, these flaws have not prevented the current international food aid system from offer- ing substantial benefits to poor countries, particu- larly in times of famine emergency. The current food aid system works surprisingly well, as long as local governance problems within individual recipient countries do not get in the way. International Agricultural Research Agricultural research is another area in which the governance deficits of concern to poor countries are less pronounced at the international level than at the national level. At the international level, an expanded and highly capable Consultative Group on International Agricultural Research (CGIAR) has been operating for several decades now to pro- vide the global public good of research avail- able for use by farmers in poor countries. Unfor- tunately the national agricultural research systems (NARS) of many poor countries are simultaneously deteriorating in terms of budget resources and useful research outputs. The emergence of the CGIAR is further testimony to the capacity of international public institutions to provide some important global public goods. When the CGIAR was officially formed in 1971, it brought together under World Bank leadership four interna- tional agricultural research centers that had originally been established by the Ford and Rockefeller foun- dations. Today it has evolved into a 16-center system that carries out technical and policy research relating primarily to production of the major food commodi- ties consumed by the world’s poor people, but now with an eye toward protecting rural natural resources and biological diversity as well. International coordi- nation for these 16 centers is provided through the World Bank by the CGIAR Secretariat, a chairper- son, and the FAO-staffed Technical Advisory Commit- tee (TAC). This coordination system is loosely knit and decentralized. The CGIAR as a whole has no consti- tution and no by-laws and it reaches decisions by consensus. Individual centers are autonomous organ- izations with independent legal status and finances. Their research programs are separately directed by each center’s board and management (Anderson and Dalrymple 1999). The research mission of the CGIAR is precisely and explicitly to create global public goods. The centers focus on problems that cut across national borders or which lend themselves to international solutions. Of the 16 international centers, 13 are located in the developing world, yet they are con- stituted explicitly as international centers with man- dates and programs intended to be independent from purely national or regional influences. The centers’ germplasm resources are internationally mobile and research results are ma