Your independent source for Harvard news since 1898 | SUBSCRIBE

Your independent source for Harvard news since 1898

Features | On the Agenda

The World's Poor

A Harvard Magazine Roundtable

11.1.00

Such distinguished recent speakers at Harvard as Kofi Annan, Nelson Mandela, and Amartya Sen have decried the desperate plight of the world's poorest people--the estimated billion-plus human beings for whom clean water, minimal nutrition, basic immunizations, and elementary education are luxuries. As part of its series of roundtable discussions on urgent social issues, Harvard Magazine asked six experts in economic development, international aid, public health, food policy, and the environment to talk about those left behind in a growing global economy. Participants in the conversation, which was held at the School of Public Health and moderated by the magazine, included:

Zanny Minton Beddoes, M.P.A. '92, Washington economics correspondent for the Economist;

Barry R. Bloom, dean, Harvard School of Public Health;

Michael R. Kremer '85, Ph.D. '92, professor of economics and founder and executive director (1986-1989) of WorldTeach, which places volunteer teachers in developing nations;

Jessica Tuchman Mathews '67, president, Carnegie Endowment for International Peace, Washington, D.C., and previously senior fellow at the Council on Foreign Relations (1993-1997) and founding vice president, World Resources Institute (1982-1993);

Dwight H. Perkins, Ph.D. '64, Burbank professor of political economy and director (1980-1995) of the Harvard Institute for International Development; and

C. Peter Timmer '63, Ph.D. '70, professor at and dean (1998-2000) of the Graduate School of International Relations and Pacific Studies, University of California at San Diego, and previously Cabot professor of development studies at Harvard.

They sketched conditions barely imaginable in most of the industrialized world--short life expectancy, pervasive illiteracy, and debilitating diseases--and critiqued past approaches to delivering aid and development expertise. Comparing the situations in much of Asia with most of sub-Saharan Africa, they moved from technical issues to the broadest concerns about social cohesion, institution-building, government competence, and the quality of political leadership. And they proposed innovative ways of encouraging the investments in technology and research needed to improve living conditions within the most disadvantaged nations. Edited excerpts of the discussion follow.

 

Moderator: Who are the world's poorest people? Where do they live? What are the conditions of their lives?

agenda-new-perkins
Dwight H. Perkins
Portrait by John Soares
ÒPeople assume that governments are interested i

Timmer: Drawing on my own background in nutrition and my experience in Indonesia, if you don't have a high enough income there to consume 2,100 calories a day on an adult-equivalent basis, then you're defined as poor. Even in the United States, poverty is defined in terms of food costs: a family that spends more than 30 percent of its income on food is eligible for food stamps.

Beddoes: The most widely used general definition is income of a dollar a day. That's what the World Bank uses, on a 1985 exchange-rate adjusted basis, to define extreme poverty. By that standard, according to the latest figures, there are around 1.2 billion poor people in the world--about the same number as there were a decade ago. That's about 24 percent of all people living in developing countries, down from about 29 percent in 1990. In absolute numbers, most of those people are in Asia, but in percentage terms, the highest poverty rate is in sub-Saharan Africa.

Mathews: I recently sat through a furious debate in which several economists and World Bank experts contended about the numbers Zanny just gave. It turns out they are, in fact, extremely uncertain. Whether the number of very poor people has in fact gone up or down over the past decade comes down to how you measure what's going on in India, where you get a swing of some 300 million people.

Bloom: I just came back from a meeting on immunization at the Gates Foundation. We talk about the vagary of numbers--they had just found 5 million kids no one ever knew existed in one state in India, and there are probably more than that unregistered in China. Clearly, the real numbers may not be the official ones.

Kremer: Looking at the numbers a different way gives a more encouraging picture. There are 3.5 billion people in countries where per capita income is less than $760 annually: 1.2 billion in China, 1 billion in India, and another 1.3 billion outside those two countries--about a third in each.

According to official statistics, China's economy has grown 7.7 percent annually from 1980 until now--doubling every 10 years, roughly. At that rate, income increases eightfold over 30 years, roughly a single generation. This growth has caused tremendous reduction in poverty in China. Per capita income in India is growing at 2.7 percent--not quite China's rate, but still enough to cause income to double in a generation.

So China, where one-third of the world's poor used to live, is doing extremely well. India, where another third of the poor live, is doing quite well, compared, for example, to the historical experience of England during the Industrial Revolution, where income grew at a rate of 0.5 percent annually for 50 years.

But the picture is very different in Africa, where the per capita income is growing at roughly 0.4 percent per year. At that rate, it takes 150 years for income to double.

Perkins: We all know the problems with the absolute numbers, but we also know that the percentage of people in poverty is way down. If you go back to 1945 or 1950, basically only Europe and North America had escaped poverty. In China, probably only 10 or 20 percent of the population was not in poverty. In India, the situation was the same, and on around the world. Now, the Chinese say

90 percent of their people are out of poverty, and the World Bank says 80 percent. In India, it's not as good, but there is still substantial progress. Latin America has made substantial progress, as well. The lack of progress is mostly in Africa.

Kremer: That economic progress has been turned into gains in health. In 1980, in low-income countries, 150 children out of 1,000 would die before age five. Now that's down to 97 per 1,000.

Perkins: Health is clearly improved, even without income rising, but the big exception is what's happening now in Africa.

Bloom: In terms of life expectancy, the 10 worst figures are all in sub-Saharan Africa. In those countries, life expectancy is 33 to 45 years, versus about 74 to 77 years in Japan and Canada.

Beddoes: It's worth looking beyond the national figures to see where the abject poor really are. You get country differences within regions, and regional differences within countries. China is the most striking example. There's been enormous poverty reduction, but there are huge diversities between the coast and inland China in where the benefits fall.

agenda-new-mathews
Jessica Tuchman Mathews
Portrait by John Soares
ÒSo often, the IMF has insisted on huge  cutback

 

Perkins: Last year, I went to the two poorest provinces in China, and the poorest counties within them. In Shanghai, you have life expectancy not much different from that in the U.S. From there, it goes down a good way--but it doesn't go anywhere near where a lot of the African countries are. In the coastal cities you have good clothes and plenty of food. In the remote areas, the really poor farmers often go barefoot; their homes have dirt floors; the hogs live inside the homes; and even though there's compulsory education, many of their children don't end up in school. By the age of 40, the people look like they are 70 or 80. But even in these poor provinces in the interior of China you have substantial income gains overall, and those people who have been able go to the coast to work and send money back--they have tile roofs, for example. The ones who stayed behind have thatched roofs.

Mathews: While the trend of overall growth has been absolutely clear, so has been this striking spread in inequality--between the developed and developing countries, clearly, but also between the rich and poor within developing countries. Over the last 30 years, the difference in average per capita income between the rich and poor countries has tripled, so you have this group of people who have been largely bypassed by growth.

Kremer: If you look at it on a population-weighted basis, poor countries have improved their relative position. In China, per capita income has grown tremendously, and India has grown at a reasonably high rate, so two countries that accounted for a large proportion of the world's poor are actually growing faster than the rich countries.

Timmer: But simply growing faster is not the same as Jessica's point. The gap in real dollars rises even if China grows at 8 percent, and the U.S. grows at only 2 percent. What's going to happen in the future, given those growth rates?

 

Is Growth Enough?

Moderator: Are growth-oriented policies sufficient to improve the lives of the very poor?

Timmer: It sounds like we all agree that growth is good for the poor--but is that enough for dealing with the problem of poverty? The gap between the rich and the poor has a political dimension, as well as an economic one, especially if it feeds back into asset formation. Serious questions grow out of that gap.

Kremer: Even if economic growth is very good for the poor in the long run, some children aren't getting immunizations or primary schooling, and we need to directly address those problems now.

Bloom: One small interjection: We're talking about the rest of the world, but--curiously--not the United States. I have to point out that the disparity in life expectancies in 1990 between Native American males in Bennett, Jackson, Shannon, and Todd counties in South Dakota and Asian females in Bergen, New Jersey, was a massive 41 years. That's almost as great as the gap between the world's richest and poorest countries. And it's rather sobering that male life expectancy in our nation's capital, at 62 years, is less than in some of the world's poorest countries, such as Iran, Tajikistan, or Vietnam.

Perkins: Returning to the rest of the world, there are two propositions that we'd probably agree on. One is that the engine that has the biggest impact on poverty is growth. Economic growth has by far the greatest ability to raise income 10 times, 15 times, 20 times. But that said, there are clearly many people who either don't benefit at all from growth, or don't benefit very much. The question is how you're going to alleviate human suffering for those people. We have to think of other methods besides growth if we're going to do anything about their condition.

Timmer: Is it a question of the nature of that growth?

Perkins: It's more than that. It involves how income is distributed. People involved in development tend to divide between those who are distribution (or redistribution) people and those who are basically growth people. It's unfortunate, because the two really ought to work together.

Beddoes: That's a real division within the World Bank.

Kremer: I would think that 90 percent of the policies that promote growth are the same policies that promote poverty alleviation--take education and immunization, for example. In contrast, having a state-run factory that produces automobiles is terrible for growth and terrible for poverty alleviation.

Beddoes: You're right, Peter, that there is a very clear division, certainly in academia, and beyond that in the policy arena. There seems to be a sentiment that you have to be either on the growth side or on the anti-poverty side of this debate--that it's no longer possible to straddle the two. But promoting growth doesn't necessarily have to come at the expense of either. As Michael said, economic growth can reinforce all these other priorities.

So I wonder whether this debate actually masks a growth versus anti-growth debate. Are there people in this debate who feel that growth is somehow not good?

 

Where Markets and Governments Don't Deliver

Timmer: I recently gave a presentation at the rural-society seminar at Yale. It's a wonderful group, but there aren't any economists--it's all sociologists and historians and anthropologists. I presented a paper called "How Well Do the Poor Connect to the Growth Process?" The other participants were flabbergasted that I was asking the question that way. The most pointed question I received was, "Can you give me a single example where growth has been good for the poor?" I was really startled. I
said, "Let me start counting for you: China, Indonesia," and--right down the list--growth is good for the poor, but not all growth is equally good, and not all the poor participate equally. Somehow, in that seminar, it was a yes or no proposition.

The debate is really over whether you trust markets to deliver. Will markets solve this problem by themselves? Or do you need sensible government intervention? Beyond low inflation, sound macroeconomic policy, open economies, and provision of the standard public goods, are there more active things governments should do? That is the nature of the debate within the World Bank: are those steps important, or is the most important thing just to get the government out of interventions in the market? The three economists in this room think most governments intervene in the economy too much, to the detriment of the poor.

Bloom: Even the World Bank, the bastion of market thinking, looks at the realm of health. A question: Who pays for prevention? People pay when they're sick, so there are probably more market failures in the realm of health than almost anywhere else. It is very difficult simply to allow health to depend on markets.

For example, there are four companies that make vaccines. Who is going to make a profit on vaccine for dysentery? For cholera? For typhoid fever? Or even for tuberculosis, AIDS, or malaria? At the moment, the president of the United States has had to offer $1 billion in tax write-offs, thanks to pressure from Michael [who has proposed public mechanisms to stimulate vaccine development], even to get them to think about that. Where's the market?

Beddoes: It seems to me that even as it reallocates its lending, the World Bank maintains its correct emphasis on the need for the right kind of national policies. The Bank continues to focus on government at the national level, and with good reason: there is a lot to do there.

In doing that, however, there's been an underemphasis on international public goods. There's been an underemphasis on the kinds of areas where rich countries' dollars can make a big difference--in vaccine research, providing money to give immunizations, in provision of primary education--funds for practical programs that go beyond policy at the national level. This is where we've not done enough, because we've concentrated so overwhelmingly on a model that said--properly--that countries had to get their overall government policies right.

Perkins: No economist thinks a market economy solves everything. It doesn't solve the income-distribution problem. No one believes that growth brings down inequality automatically, because the original distribution of assets seems to persist. Brazil, for example, has enormous inequality--and has had for a long way back in history; China has relatively little inequality--and Korea, Taiwan, and others are similar to China. The reasons for the lowered inequality in these latter cases have to do with war and revolution that fundamentally redistributed assets, not market forces.

The biggest problem in dealing with poverty is that it's much harder to deal with than growth. Concentrating on health and education, where you have clear market failures, is one area where non-market interventions can play a major role. But when you talk about direct interventions to help the poor, government often lacks the administrative capacity to do this well. The interventions are made in the name of alleviating poverty, but actually benefit the bureaucrats and the upper-income groups.

Timmer: Food subsidies are a key example. Food security is one of the defining features of people who have escaped from poverty, so you want to do something about hunger. That mobilizes public sentiment, it mobilizes governments. Yet there are a lot of very foolish, not to say rapacious, food policies in the world alongside some good ones. The difficulty is sorting out the ones that work and helping governments deliver on those policies while encouraging them not to do the silly things. That is where the "Washington consensus" [on how best to support sound economic policies and development] breaks down. It's not very discerning about stabilizing rice prices in Asia--which turned out to be really important for everybody--versus controls over imports or exports that tax the poor on behalf of a very small, rich elite...

Perkins: Or highly subsidized credits that end up going into the hands of people in political power.

Timmer: Exactly--rural credit schemes that end up going to the rich.

Mathews: Those problems have occurred even in Green Revolution policies [the investments in agricultural technology in the 1960s and 1970s aimed at boosting yields of rice and other staple crops].

Beddoes: In fact, it was 25 or 30 years of failed government interventions that finally, a decade ago, brought about the consensus that less government is better. Now we're shifting back tentatively toward the view that the right type of government action makes sense. But do these governments have the institutional capacity, for example, to do the right kind of food-security policy--or is it apt to turn into some kind of subsidy scheme?

The Persistence of Poverty

Mathews: I want to come back to inequality. As long as I can remember, I've read that Latin America can never reach sustainable growth until it solves its inequality problem. Today the United States is more unequal than Latin America in terms of income distribution--a shocking number. After two decades of the government steadily getting out of the economic realm, which has clearly paid off in terms of growth, how much further in that direction can we go? Dwight said something profound a minute ago: doing something about poverty has turned out to be a lot harder than doing something about growth.

 

Timmer: Thirty years ago, the thinking was the opposite--that you could certainly solve the problem of poverty, even if you couldn't figure out how to grow, because you could redistribute resources from the rich to the poor. That turned out not to be possible.

Mathews: Is something new going on, resulting from economic globalization, that is increasing the inequality of incomes--something we don't yet understand that bears directly on this question of why there are so many people living in such abject conditions?

Bloom: One could argue they're not inside the global economy at all--that what you just described is a global economy in which a large percentage of humanity is left out; the markets don't deal with them. These are the people who don't participate and can't purchase vaccines.

Mathews: If that's correct, it would explain this controversy among development professionals. These are two different jobs: achieving growth and dealing with poverty.

Timmer: But the point we made earlier is you can't deal with the billion-plus people in poverty if you can't figure out a way to get their economies to grow. The places where there are growing numbers of people in poverty are exactly in the economies that aren't growing, that aren't hooked into globalization.

In absolute numbers, China, Indonesia, and India have perhaps 75 percent of the world's poor, and all three of those economies are potentially hooked into global trade and advances in technology. But the really sharp rising numbers of poor are in sub-Saharan Africa, which isn't hooked into the global economy in any significant way, other than as a raw-materials supplier. We haven't figured out how to get Africa to grow.

Bloom: Look at literacy. We recently held a conference on the Internet and public health, and learned that in the poor states of India and the poorest countries in sub-Saharan Africa, literacy may be on the order of 10 percent for women. How can you be part of the Internet world if you can't read?

Beddoes: As a speculative question, can we be more optimistic about technology leapfrogging problems in health and literacy and education? I have heard people talking about billions of hand-held Internet devices within five years, with the price coming down to practically zero. Is that a way of decreasing what people call the "digital divide"? Techie types are very optimistic about the new media closing the divide over the next 20 years.

Mathews: They're also unbelievably naive. Every technological revolution in history has widened the gap between the haves and the have-nots.

agenda-new-beddors
Zanny Minton Beddoes
Portrait by John Soares
ÒTo the extent that there are opportunities  for

Kremer: The key issue is that very different technologies are needed for different users, and unfortunately, most of the technological effort in the world is directed toward technologies that are needed by the rich, because they can pay the most for it. We were discussing medicines and vaccines earlier, so it's worth noting that 1,200 drugs were licensed during the last 25 years. Of those, only 13 were for tropical diseases. Of those, two were modifications of existing medicines; two were produced for the U.S. military; and only four were developed specifically for tropical diseases by commercial pharmaceutical firms.

Bloom: Eighty-five to 90 percent of the burden of premature death and disability falls on people in developing countries. If you ask where we spend medical-research dollars, 94 percent of that is spent on diseases that afflict primarily people who are in rich countries. The only good news is that by 2020, people in developing countries on average will have the same major causes of death and disability that we do.

Kremer: There still are enormous differences. Malaria kills a million people annually in developing countries; it is not killing large numbers of people in rich countries.

I think this relates to our earlier discussion about growth and poverty reduction. There's not a lot of money available in aid budgets, so it needs to be targeted well. It's very difficult to find ways to deliver that aid directly to the poor without having it eaten up by bureaucrats along the way. Targeting aid to scientific research on diseases such as malaria, tuberculosis, and HIV offers a very effective opportunity to help the poor.

Timmer: The same thing is true on the agriculture side.

Kremer: I was going say that, for crops like cassava.

Timmer: Agriculture is the example where we actually have some historical evidence. It's the one counter-example to Jessica's argument that all technological revolutions have widened the income gap. The first couple of years of the Green Revolution widened the gap, but since then its benefits have overwhelmingly helped the poor, by keeping the price of rice and wheat well below what it would have been otherwise.

Kremer: Many new medical technologies have reduced the gap in life expectancy between rich and poor.

 

Mobilizing Societies, Building Leaders

Perkins: Gains in life expectancy clearly are partly nutritional and partly medical. When you talk about using limited resources to develop a vaccine for malaria or AIDS, then you would have an enormous impact for what would be a relatively small amount of money. But if you don't have a vaccine--and none of us expect one soon for either of those diseases--then you're still back to the tough problem of poverty-reduction.

You've got to get the affected societies to mobilize themselves to do it. There are societies that have done so. Thailand had an enormous HIV problem because of its sex trade, but they got control of it because they really faced up to it: they had some of the most successful publicity campaigns in the world to get people to recognize the danger. In most of Africa, you haven't got the recognition that there is a real problem, or the ability to do much about it once the existence of the problem is accepted.

Mathews: Thailand has now launched an antismoking campaign using Buddhist monks. They seem to have a gift for finding ways to act.

Timmer: There's a set of institutions in Thailand, rooted in the monarchy, that holds that society together, so you can mount effective campaigns and get messages out. That's not true in Africa. How one builds that seems to me the really difficult question.

Moderator: What are the institutional needs for societies to fight HIV or sustain economic growth? It's a different challenge from erecting dams or improving rice yields. How should aid donors and policymakers support or build those capabilities?

Timmer: The popular term is "social capital"--the institutions and linkages and relationships, the degree of trust and sense of community that allows self-enforcement of rules and regulations and norms. Societies that have lots of social capital function smoothly. Trust allows market relationships to take place. If you had to have a regulator enforcing every market transaction, you wouldn't have very many. You have to build up this degree of trust. I don't feel that the aid community knows how to build it.

Mathews: It may not be buildable from outside.

Timmer: But the big question then is, what do we as outsiders do?

Mathews: How has Uganda managed to recover a degree of trust in its government after what it went through [under Idi Amin]?

Bloom: Leadership. Uganda's president was the first to recognize that AIDS was going to be devastating. Every other leader in Africa had the same epidemiological data--that at that time, perhaps 20 percent of their military was infected. He took a leadership position, and that paid off. He became a darling of the international donors, and Uganda is a model country for reducing the problem.

The question I raise, particularly for a great university like Harvard, is how do you foster leadership? It's hard to do from outside, but it's also hard to do from inside. How do you generate leadership? How can we help?

Perkins: I don't think that you can bring people to a university to teach them how to be leaders and then have them go back to their societies and become leaders. Leaders have to rise up naturally through some basically political process.

Timmer: But you can empower them when they have arisen.

Perkins: Yes. But if the society itself is not united in any way, that society won't produce governments that have coherent national policies.

People assume that governments are interested in growth and development. But many government leaders are interested mainly in personal aggrandizement and in helping their group over another group. Go through sub-Saharan Africa: yes, Uganda and Botswana have unified governments of the sort required (although Botswana has one of the highest rates of HIV infection in the subcontinent), but not most of the others.

Mathews: The question is a tough one. I spend a lot of time these days on Russia, and even if a great leader arose there now, he'd have a very hard time making a difference. So many of the channels you need to convey a message from the top down aren't working.

Perkins: In sub-Saharan Africa, they haven't yet created effective nation-states. That's something the World Bank and others can't deal with directly. If you have a system that produces leaders who see the economy as a fixed pie, where the only way to do better for yourself is to make someone else worse off, what do you do?

Timmer: At least we know the precursors, if we're going to help at all: preventive health and education, those ingredients of public-goods formation. The alternative is that we simply stand back until they get their acts together. I don't think that's going to work.

 

Strategies for Assistance

Moderator: Is aid being delivered effectively?

Beddoes: The latest research suggests that the carrot and stick of "conditional" assistance [where the World Bank or other aid providers offer financing subject to the recipient's undertaking reforms in economic policy, tax collection, tariffs, etc.] doesn't work if you've got a government that doesn't want economic reform and is for personal aggrandizement, or you've got one that isn't capable of generating public acceptance. So, the aid community is moving away from the belief that conditional support works.

But what should we do then? The current trend seems to be toward the "participatory" approach--that programs need support from the ground up, and you build up civil society that way. Can that work in an environment where you still have all of these divisions in society? Do you just give up on aid to national governments and direct all the assistance directly to the grassroots?

agenda-kemmer
Michael R. Kremer
Portrait by John Soares
ÒThatÕs the warm and fuzzy aspect of democracy.

Kremer: It's very important to have grassroots participation, but projects designed to assist local community organizations of the poor and disadvantaged may wind up transforming them in unintended ways. Together with Mary Kay Gugerty, a graduate student in political economy, I have been examining a project that provided funding to women's groups in Kenya. We compared groups that received funding to those that did not, and found that once groups received funding, men and younger, better-educated women joined and took over leadership positions. By providing funding, the donors changed the very characteristics of the groups that had attracted them in the first place.

Timmer: Let's come back to some things that the international aid community ought to be doing, first and foremost. That's providing international public goods: the vaccines, the research that leads to the vaccines, the agricultural research, the funding for that from the world donor community. The decline in relative share of international public goods being provided by the aid community is just shocking--in particular, for agricultural research for Africa, where they need it the most. So the international community ought first to make sure it gets its own act together.

Then we ought to help societies figure out how to get primary education done. An educated population will be key to any productivity growth. And they can't educate kids who aren't healthy, so you're going to have to integrate that with rural health schemes.

If those were the priorities--international public goods and education with a public-health component--we would at least know we were doing no harm.

Mathews: You'd still have to deal with gender inequality, and with high rates of population growth, and with environmental degradation, because these all hang together.

Bloom: But to do that, you'd have to overcome two major barriers that I find extraordinary. First, even with what little we in the rich countries do, we hear about "donor fatigue." There are more good causes now, and the same or less money from the rich countries to address them. Maternal mortality, for example, was not on anybody's agenda a decade ago. Now it's a priority. We are spreading limited resources around to the point where it's not clear how effective we are.

Second, I just learned a new term, "recipient fatigue." Every donor country has its pet project, and they all divide a developing country into tiny little sectors. They all require paperwork and personnel--the people managing them are scrambling from one project to the other. So now there is an experiment called sectoral development, where, for example, all the donors in Ghana pool their money in the health sector, and let the country sort out its priorities. That's a pretty radical experiment in the way aid is disbursed, and I hope it will turn out well.

Kremer: Returning to conditionality, I agree it may be very difficult to change the overall policy orientation of a country in return for aid. Even so, it may encourage some worthwhile changes in particular areas.

I'll give an example from Kenya. I lived there 15 years ago, and at that time bicycles were rare. When I returned recently, there were many more bicycles, and people were using them--as people do in developing areas everywhere--not just for personal transportation, but as trucks. If you need to get supplies out to the rural areas, you put them on the back of a bicycle and push it. Lots of people can afford a bicycle, and you can move a huge amount of material that way. In a town I visited, hundreds of people make a living as operators of bicycle taxis--people who would have little income otherwise.

Why the change? Bicycles used to have a high tariff on them. They couldn't be imported. This benefited a tiny number of people working in Kenyan bicycle factories, but hundreds of thousands of people in rural areas suffered as a result. Kenya got rid of that tariff. I don't know whether that was the Kenyan government's initiative or the result of some outside...

Timmer: That almost certainly would have been conditionality.

Perkins: The current fad is that conditionality doesn't work, but I think there's lots of evidence that under some circumstances it does. It doesn't work very well if there's no one in the government who believes in it, but it can support the work of groups in government that want to do something. Peter and I have had a long experience in Indonesia, where the so-called "Berkeley Mafia," the economics ministers, used the international system and pressures from outside to win battles against their cabinet colleagues who wanted to do things that would harm economic growth. Kenya has had a whole series of reforms that were partly World Bank-imposed, but there were also key permanent secretaries who wanted them to happen. Because they could use that pressure, they could get their political colleagues to go along.

Timmer: China has used the World Bank the same way.

Perkins: Look at China's decision to risk joining the World Trade Organization and accept all of the quite draconian U.S. terms--to allow foreign-owned insurance and banking to enter within three years, and so on. I'm completely convinced the rationale is 90 percent not to promote exports; it's mostly to have a weapon for beating the State's enterprises over the head and getting reform.

Kremer: Under the WTO, China has to let in foreign private insurance companies. People in China will ask why they can't start their own private insurance companies, if foreign ones are allowed. And that's a big part of the reason why reform-oriented Chinese policymakers want to join the international bodies.

Mathews: An important part of the debate is about how World Bank and International Monetary Fund policies have affected the poor. So often, the IMF has insisted on huge cutbacks in fiscal expenditures just when developing economies are hurting [before providing short-term assistance during financial crises]--the opposite of what any developed economy does. Those cutbacks have ended up being visited largely on the poor.

Also, the IMF went through two decades of urging troubled economies to push exports that were largely unsustainable: exploiting forests and fisheries, with all kinds of unfortunate environmental consequences. The policies often produced short-term benefits, but equal or greater long-term costs that have been borne very disproportionately.

Beddoes: That's a debatable point. Over the past decade and a half, the IMF put a premium on economic stability and on liberalization as the sine qua non for a stable, growing economy. But I'm unaware of systematic evidence that it was particularly bad for the poor, as opposed to the large groups of middle-income and urban people who were hurt by the removal of unsustainable subsidies for products they use. After all, the worst thing for the poor is the "inflation tax," which is essentially what the Fund is trying to reduce in a high-inflation economy. So the IMF policies were not manifestly bad for the poor--although I grant you that thinking about the distributional consequences of its conditions is not at the top of the IMF analysts' list.

Timmer: To the extent they thought about such things, it was always in the "what if" context: if you continue on the path you're on without reforms, the poor are likely to be hurt even worse than if you get the reforms in place and growth starts. The difficulty is the connection in the transition from the reforms to growth--and an awful lot of the reforms, in fact, have not led to growth. Then the poor were hurt...

Mathews: Doubly so.

Perkins: We're talking about different kinds of conditionality. The IMF is typically involved when there's a major macro imbalance, like high inflation or a major balance-of-payments problem, where a country is living beyond its means and is forced to cut back. People are going to get hurt in that process. Sometimes it's the poor, sometimes it's more general.

The other kind of conditionality is the effort to create a market system, not in the context of a macroeconomic crisis. That is the standard situation where the World Bank is the actor. Now they don't always get the financial liberalization right: the banks might be freed, but then they tend to forget about prudential regulation. But when they do get it right, they often make a big difference--and without the conditionality and the technical assistance, it doesn't happen.

Bloom: But what often isn't considered is the subsistence survivor. If the middle class and rich lose from increased constraints and taxes on gasoline and whatever, they eat. It's the poorest citizens who have only a marginal ability to survive the imposed constraints. Nutrition and education are cut back, and they are subject to disease, illness, and family catastrophe for which there is no insurance.

Timmer: That's true. Yet there's a short-run, long-run distinction. In the short run, the poor starve--that's a real dilemma for us. But in the slightly longer run, getting food prices up--so that farmers can invest and grow and have income--benefits the rural poor, who are really the poorest people out there. They depend on a healthy agricultural and rural economy.

Somehow, we have to figure out this transition process. How do we help the subsistence poor into a market- and incentive-driven rural economy and economy overall? That is the real political and even technical difficulty, because having the institutional capacity to deliver food subsidies targeted specifically to poor people is a challenging task. We have a hard time targeting food stamps in this country. For Indonesia to target rice during the 1998-2000 crisis--that's a real challenge.

Bloom: It's more than just technical. I will reduce your argument to one that's quite famous: the Black Death, which killed 50 to 70 percent of the urban population in Europe in 1348. From an economic point of view, it killed off marginal labor, raised wages, and increased productivity. The lack of faith in the church at that time led to the formation of the great universities. The Western reinvention of the printing press is attributed to the shortage of scribes. In fact, it is argued the Black Death was the greatest thing ever to hit Europe. The Amartya Sen position would be that there is a value to life and health independent of whether it is good for economic growth. That's lost sight of in these transitions.

Kremer: Again, I do think that 90 percent of policy choices involve no tradeoff between the economy as a whole and the welfare of the poor. That may not be true for rich countries like the U.S., where we may face more tradeoffs. But a lot of poor countries, unfortunately, have policies that reduce economic efficiency and hurt the poor. Cutting back is good for the poor, and good for the economy as a whole.

Beddoes: I agree, but the much harder question is what priorities are you going to put in place instead?

Kremer: Primary education and primary healthcare. For example, in a lot of impoverished countries, teachers often have poor attendance. What about requiring headmasters to keep careful record of teacher attendance? The records would be subject to inspection to make sure they are up to date. If the teachers aren't showing up, they would be subject to some action.

Perkins: And make that the condition of aid?

Kremer: I wouldn't go so far as to suggest that right now, but perhaps something like that should be considered.

Perkins: You've got to have some kind of conditionality. Are you just going to give assistance to countries you deem to be effective? In that case, you've eliminated a large number of countries from all aid. That's conditionality of a certain kind.

Are you going to give aid to all countries equally, just on the basis of their per capita income or their population? Then you're going to waste huge amounts of money in countries that are unwilling to take steps to promote development.

Timmer: I thought Michael had it just about right: 90 percent of the policies are going to help both the poor and growth, but the real question is how to deal with the other 10 percent--especially if they're gut issues like food prices and subsistence levels of nutrition. Do all countries have the same conditions put on them? Or do teams from each country work with teams from the donors to negotiate individual packages for given sectors and countries?

My experience working with Indonesia is that the IMF and World Bank teams didn't know very much about Indonesia, and when they came in to negotiate a package, they had 98 percent of the language already drafted when they left Washington on the plane.

Mathews: And it had probably been written for Korea.

Timmer: Right. Now because of the Berkeley Mafia and the Harvard team working there, Indonesia had the capacity to sit down with the IMF and Bank staff and say, "You've got it wrong. This isn't going to work in this context. You need to redesign how you want these policies implemented." When the international agencies had sent a good team, they'd listen and learn, and come up with a set of policies that the Indonesian team could take to the president and the parliament and say, "Here's how we're going to use these resources." It took very intense analysis of poverty in the country, of how the economy worked, the sectoral linkages, the trade interactions, and so on.

Mathews: That's why I don't agree with the premise that 90 percent of the policies are probably going to benefit both growth and poverty alleviation. The record doesn't show that. How many unanticipated, perverse consequences have we been subjected to?

Timmer: If you've got a policy that promotes low inflation, that's good for both growth and the poor. Sound fiscal policies with very little government consumption of subsidies--that sort of tight budget is clearly good for both.

Mathews: You're missing the point Barry made a minute ago.

Timmer: What I'm saying is that the small fraction of policies that might be good for economic growth but hurt the poor, at least in the transition, involve some really gut-wrenching issues like hunger, poverty, starvation...

Mathews: And environmental degradation.

Timmer: Yes. In those cases, individual country action, vis-a-vis donor proposals, is essential.

Beddoes: Without being an apologist for these institutions, I think they are now very concerned that the countries drive the new strategies. With so-called "poverty-reduction strategy papers," they are attempting to have the countries play a much bigger role. The practical problem is incapacity, because not every country has a Berkeley Mafia, and often the aid institutions' teams have much more expertise.

Mathews: I don't feel this conversation is anti-World Bank or anti-IMF, so much as it is a reflection of how hard this is to do, and how uncertain what we think we know really is.

Timmer: I have worked in Indonesia on food policy for more than 30 years now. They have had remarkable success. In the first two decades, they had a rural growth strategy that brought poverty down from 75 percent of the population to 15 percent, and they did it through a combination of the right macro policies--openness, stability, and rapid growth--along with a concern for where the growth was and poverty a lleviation through rural investments in education and health.

I took that model on the road, especially through the Philippines and Vietnam and even a bit in China. But what was really difficult was when I was asked to take that model and talk about Africa. None of the preconditions for creating and implementing a coherent development strategy were in place.

Perkins: From my experience with the Harvard Institute for International Development through the years, Africa is very much in a long-term decline, in contrast with Asia. Take the ideal model for developing a set of priorities, like Indonesia or Korea. Korea used to invite the World Bank to come in and help develop a plan jointly. They'd have the World Bank teams come in, they'd work together, and by the time they were through they all agreed on what was important.

In the African countries where we worked, it's very hard to find governments that have a set of priorities beyond the immediate priorities of particular ministers and presidents to accomplish particular things, most of which have very little to do with growth. So what happens is that the aid agencies all come in with their own agendas and you get this overload of often conflicting donor demands on these societies.

Bloom: Dwight, if you were the leader of one of the 20 poorest countries in Africa, how would you set an agenda when you're entirely dependent for support on outside agendas?

Perkins: What Africa needs to do against this economic blight is no secret. We're not talking about poverty right now, we're just talking about growth. It involves getting the exchange rates so they're not wildly overvalued, and getting rid of state enterprises...

Kremer: Especially marketing enterprises.

Perkins: Especially marketing enterprises, but state enterprises in general--which is very hard to sell, because the Africans say if you privatize them that just means the Asians are going to run them, and you won't have the political patronage, and so on.

Those are the kinds of things that everybody would agree you have to do if you're going to have effective growth policies. They are published in the country's plan, they're put into conditions for aid--and then they're not implemented time and time and time again.

So you have to ask, "Why is it that the African governments won't do this?" The answer is they have a different agenda, or they don't have an agenda, because the political system hasn't created effective governments that can develop their own agendas--at least when it comes to growth and development.

Mathews: What about enforcement? As part of conditionality, shouldn't the agencies get much more serious about cutting off lending when conditions aren't met?

Perkins: That's very hard to do. The World Bank is under pressure to give a certain amount of funding even if the funds won't be used well, and then, of course, the U.S. cites security reasons for supporting many of the countries.

Downsizing the Aid "Industry"

Beddoes: You could also see it as the inevitable outcome of the aid industry in decline. Overall aid spending has fallen since the early 1990s. Today, aid flows are small compared to increasing flows of private capital to developing countries. In 1998, for instance, rich countries provided around $52 billion of aid to the developing world. In the same year, poor countries received about $268 billion of private capital.

On the one hand, you've got people who are interested in reforming the entire aid industry--you get radical ideas about what aid is and how it is delivered. On the other, if you're delivering the aid, you're terrified of advocating any reforms that could be used as an excuse by the opponents of aid who want to cut it even further. The aid community has no interest in giving them any ammunition. It's a real dilemma for anybody in Washington now. We have an administration that's basically quite keen on aid, but terrified of putting forward any radical proposals for change in case they support the agenda of eliminating any aid at all.

Perkins: You could make a case that aid probably ought to be phasing out of the countries that are getting capital flows from the rich nations, or could easily do so--the countries that have made enough progress to participate in the global system of private capital investments. Maybe the aid should be concentrated in those areas where there are market failures, or on poverty alleviation within areas that have grown, as in Asia.

When you talk about Africa, the only private capital available for most countries deals with natural resources--and the one thing we have learned about natural resources over the years is that having a big mine is often one of the worst things you can have. Pouring lots of money into big mines means it ends up in the wrong hands, which then use the money to undermine the integrity of the government.

Timmer: Barry's point remains awfully important. If we're going to focus aid on poverty in those countries that can't get foreign capital from private sources, we've got to contend with the debate that started our conversation: is growth the mechanism for dealing with poverty, or do you deal with it more directly?

Bloom: Just look at the international lending organizations. The vast majority of money goes to countries that have the highest probability of success, not to the very poorest ones. That's true both of the World Bank and the World Health Organization, I feel sad to say. So the issue is, how do you use public resources to create public goods for those countries, when there is virtually no capital coming in any other way? We're afraid to do that.

Timmer: We may have it backwards. In Asia, we're concentrating on growth, on the assumption that that will take care of poverty, and in Africa, we're concentrating on poverty--neglecting the fact that you have to get growth going or it's not going to be sustainable. We need to flip those around.

You're right: there are a number of areas in Asia where more attention to the underlying issues of poverty would be very effective. But Dwight's right, too: we need to figure out how to get growth going in Africa.

 

Asia versus Africa

Moderator: What can we learn from the relative economic success of Asia versus Africa's persistent poverty?

Mathews: Part of the problem is surely that the initial engine for growth is agriculture, and Africa has a uniquely inhospitable environment for agriculture. The soil is worse, the rainfall patterns are harder. There are far fewer opportunities for irrigation.

agenda-new-bloom
Barry R. Bloom
Portrait by John Soares
ÒEighty-five to 90 percent of the burden of  pre

Timmer: And the cropping systems are extraordinarily heterogeneous and complex. Designing research to raise the productivity of those complex cropping systems is way harder than raising the productivity of pure stands of rice or maize. But we spend virtually no resources developing technologies for that environment, compared to what we spent for Asia and Latin America.

Mathews: So alongside the overall institutional question about Africa that Dwight has suggested, it is, by every measure, the continent where there is the least to work with.

Beddoes: One of the lessons I take from the Asian experience is that you really have to get the policies right. The Asian growth-promoting policies carry you a long way toward where you want to go.

But is the emphasis we had on national policies in Asia correct for Africa--in a continent where agriculture is so uniquely disadvantaged, and where the health situation is catastrophic and getting worse? That's why you have to concentrate on international assistance for Africa much more than you needed to in Asia.

Mathews: The other part of the Asian experience was that the focus on economic policies was preceded by national decisions to slow population growth. Except for the Philippines [and Indonesia, which started in the late 1960s], those decisions were made in the early 1950s. Every one of those countries brought population growth rates way down, and that made an awful lot of the other investments--in education and so on--pay off.

Timmer: For all his flaws, it took extraordinary leadership for President Suharto of Indonesia, as head of the world's largest Islamic country, to go on television and award free condoms as prizes for participation in family-planning programs--to push family planning as healthy, desirable, and part of the society's moral fabric. That was an astonishing act of leadership.

Mathews: In Africa, there's a deep cultural bias in the other direction.

Kremer: Even in Africa, some countries, including Kenya, reduced their population growth.

Mathews: Almost all of them have.

Perkins: It's hard to do without very widespread education of women. Lowered fertility has traditionally been associated with industrialization and urbanization--all parts of what we call the demographic transition. Without growth, that's been hard to do.

Mathews: But Bangladesh has done it. Every single factor that should make population policies fail is in play in Bangladesh, and yet they have made it work. It can work if you design the program correctly, locally.

Bloom: The most dramatic example, which is poorly known in this country, is that when the shah of Iran was in power, although the distribution of resources was very unequal, there was a normal decline in fertility as the economy grew. When the Islamic revolution came in 1979, the country had a pro-birth policy. The birth rate went out of control until the new regime came in and in 1988 those policies were reversed. Now Iran has the most rapidly declining fertility rate of any country in history.

What that says to me, again, is that within all the cultural constraints, if there is a concerted commitment to change the basic policies and conditions--whether for nutrition or health or fertility--leadership makes one hell of a difference.

Perkins: Leadership is obviously very important, but being an effective leader depends on a population that's willing to be led. Think of Robert Mugabe of Zimbabwe. When he came to power, you would have said he was just such a leader. This is not somebody who happened to be given the job as the head of his country--he fought and negotiated for it, and he represented the largest group at the time. He didn't just throw the whites out of the country--he recognized that Zimbabwe needed them and their productivity. Now, 20 years later, he's in trouble and he's playing to the worst instincts of his supporters in an effort to stay in power--he's playing the race card the way some of our southern governors used to.

Bloom: I accept that as a powerful and cogent argument. I would ask then, what are the alternatives? What do you support? Let me suggest one category that has been badly neglected: institutions. In Africa, virtually no new health institutions have been built in the last 25 years except private hospitals. As an international community, we have not supported great universities in Africa. There were great universities, but they were allowed to decay.

Perkins: Uganda destroyed a wonderful university. You can't build universities when the government is trying to close them down and beat up the students every time they demonstrate. The universities in Africa are closed almost as much as they are open.

Kremer: One key institution is government--without that, it's hard to get other institutions. A big part of Africa's problem is that its governments aren't working very well. That comes down to security issues: how do you get in power and stay in power? Africans may need to rethink some of those issues. In a place like Sierra Leone, there's no hope of any development until the security situation is worked out.

Mathews: Is the mismatch between national borders and ethnic communities worse in Africa than, say, in eastern Europe?

Perkins: I don't think there's anything comparable to the African problem in that regard.

Beddoes: Isn't that exacerbated by a unique natural resources blight? To the extent that there are opportunities for quick wealth in Africa that rely on natural resources within these very weird national boundaries, politics is about grabbing control of the natural wealth. Until you either lessen the reliance on that, or find a way of insuring that the wealth is distributed, I can't see a way out.

Mathews: That's also true of the Middle East and Russia.

Timmer: And of Indonesia now, too.

Perkins: Certainly in Sierra Leone and Liberia that's the case. Basically the motives are to use force to grab the diamonds.

Kremer: In those situations, maybe there's a role for the international community in supporting democratically elected governments to preserve security. It would be wonderful if we didn't have to rely on mobilizing a new force each time--if there were some sort of standing international force that could be deployed quickly.

Mathews: That's doable. You don't even have to create a standing force. Create a standing command structure with real professionals and it will work.

Kremer: It doesn't have to be U.S. forces at all, and it could be small: 10,000 soldiers in Sierra Leone or Liberia would make a huge difference.

Timmer: In East Timor, the U.S. didn't send any combat troops, but it did provide the cargo aircraft and all of the telecommunications equipment that allowed the Australians to get in there quickly, and then to mobilize the rest of the forces and bring them in. That's the sort of thing the U.S. could put in place easily and quickly.

Perkins: These things help, and they're clearly justified in some areas--the devastation in Sierra Leone is horrendous--but basically the people in these countries themselves are going to have to be mobilized in order to prepare to be led.

It's easy to be very pessimistic about Africa now. But I have spent my life studying Asia, and if I had been talking about these issues and about China in the 1920s, I would have had an equally hopeless view about what was going on. China finally did figure out how to organize itself: it took a revolution that cost millions and millions of lives; it took a long period of misrule by the Kuomintang and the Communists. But the fact is they got unity and eventually stability and growth.

The process whereby that happened in parts of Asia is not going to be the same in Africa. I don't know how Africa is going to do it, but my guess is that over time, people don't like living in societies that are riven as these are, and they will bring about change.

Timmer: This is very much an historical process. It takes time. You cannot create institutions overnight. Societies have to invent them, build them, and trust them--that's inherently time-consuming.

Perkins: It took China a solid hundred years to do this.

Mathews: Is the situation in Africa made almost impossible by the hollowing out of the population because of AIDS?

Bloom: I chair the UN AIDS Vaccine Advisory Committee. Even those of us in the business didn't know what the numbers were until recently. None of us really understands what it means to take out a fifth of a population between the ages of 15 and 59, to completely change the age and demographic structure, to produce 13 million orphans. How do you have a structure for society? What does it mean when there are more 60-year-olds than 40-year-olds? Or when a kid at age 15 has only a one in two chance of reaching age 40?

Almost nobody is doing anything. Total international AIDS resources are about $300 million. We're talking about a problem that is devastating Africa now. India is number two. There, 0.7 percent of the population is HIV-infected--that's 3.7 million people, and the total is shooting up.

Mathews: Africa has about 70 percent of the global total of 33 million HIV cases?

Bloom: That's right.

Mathews: The Black Death presumably killed across the age distribution?

Bloom: It killed almost everyone.

Mathews: But AIDS kills off the working population...

Bloom: Yes, but it kills a certain number of kids, too. The dilemma there--and part of the debate in South Africa--is that if you treat pregnant women to prevent the transmission of infection to their infants, you create orphans, because the mothers are all going to die. You could provide enough drugs perhaps to keep their babies from being infected, but they cannot afford the expensive drugs to treat the mothers for a lifetime. So the dilemma was whether to allow children and infants to die or, in fact, to create orphans? That is both a terrible moral dilemma, and an issue of resource allocation.

Perkins: In Zambia a large portion of the government staff either is HIV-positive or soon will be. The civil service will be wiped out, so their attitude toward governance is shaped by short-term behavior. When they surveyed three battalions of their army in the early 1990s, I believe that nearly 100 percent of the men were HIV-positive. How does that change discipline when they know they're going to be dead in 10 years no matter what?

Bloom: What does AIDS do to an education system? Teachers are a very high target group for HIV. They're losing about 1 percent of teachers per year, and there aren't that many to begin with.

 

Assisting Success

Moderator: We have talked about adverse natural conditions within countries that hold back development. We also talked about the role of countries' institutions in creating the conditions for sustaining growth and alleviating poverty, and about outside agencies' role in investing in public goods--vaccines or agricultural technology. So what is the strategy for the poorest countries in the world? What are the most important decisions they can make? What would you recommend, what tools and resources would you request, to help improve the lives of the people whose prospects currently seem worst?

agenda-new-timner
C. Peter Timmer
Portrait by John Soares
ÒEighty-five to 90 percent of the burden of  pre

Timmer: We all agreed on the provision of international public goods, those things that the donor community has a direct ability to fund and implement for the global community at large: health research; agricultural research; and maybe a peace-keeping capacity that could keep societies from tearing themselves apart over resource conflicts and similar things.

Our conversation has been much more pessimistic about what the outside community can do in individual-country circumstances. Dwight stressed more firmly than the rest of us that there, things really have to happen within those societies, and we should stand ready to help when the time comes. I think of Indonesia's circumstances in 1966, when Gunnar Myrdal is writing that the international community holds out no hope for Indonesia--it's a basket case. When it did turn around, the international community responded with a substantial amount of appropriately targeted aid. It made a big difference.

When that window opens--a new government comes in, a leader takes over--you've got to be ready to move in a hurry. The UN in East Timor wasn't ready. And you've got to know what to do. A lot of academic research has to be done. We can't rely on the World Bank to do half of that; the universities are going to have to be engaged as well. And the research has to be country-specific because societies are different from each other.

Bloom: That implies there is a global willingness to be ready. But there is a cognitive dissonance in this country; on a number of surveys of American public opinion, foreign aid comes out as the least-supported activity every time. But "helping the world's hungry" or "helping the world's sick" scores high. Provided that the public has a sense that the humanitarian assistance will go to the people who need it, two-thirds of the people polled are supportive. Yet virtually no congressmen agree with that position.

Mathews: What makes things even worse is if you ask people, "What portion of the federal budget ought we spend on foreign aid?" they say 5 percent.

Bloom: And what do we spend? About 0.7 percent if you include military assistance.

Mathews: About half that if you exclude it.

Bloom: That's the lowest of any OECD [Organization for Economic Cooperation and Development] nation.

Mathews: In fact, the public--if it believes what it is saying--supports a tenfold increase in what we're spending on aid.

Bloom: So we need leadership here.

Timmer: This really is a dissonance between what the population supports now and what they would support if they were sure the funds were really aiding the needy. But aid gets caught up in all these fiascos.

Mathews: This continuing gap between public opinion and Congress's views ought not to exist in a democracy, so, frankly, I don't believe we have any satisfactory explanation for it.

Beddoes: One reason it exists is that, unlike any domestic priority, there's no constituency here that calls on Congress for countries to have that aid--at least not like the lobbying for your pet pork project in Louisiana.

Mathews: There is such a constituency--environmental groups, women's groups, development groups...

Bloom: You don't want to be called what we're called when we visit with congressional staff. We're called the "Misery Lobby." We're in a box, and we're put on the shelf.

Kremer: The public is skeptical about foreign aid. It's therefore important to try to design aid policies so the public in rich countries can feel that their money is going where they want it to go.

We discussed vaccine research. One approach worth considering is for rich countries' governments, including the U.S., to announce they'll participate in providing a share of the funds to buy vaccines if they are developed for malaria or tuberculosis or AIDS. With that sort of program, nothing would be spent unless the vaccines are actually developed. And the funds used to purchase vaccines in a crisis would be very modest relative to the human or economic cost of the disease, but would still be enough to motivate some biotechnology and pharmaceutical firms to start doing research on these diseases.

Bloom: A couple of years ago, NBC went to Africa. I dealt with Bob Bazell, their chief science reporter. He wanted to show success stories, and so they told the wonderful story of the elimination of river blindness, a worm transmitted by black flies that in some villages made 25 percent of the population blind. With efforts from the World Bank, which paid for a spraying program, and dosage with a single pill given free by Merck [that cured those already affected by the worms], the effect was extraordinary. It was a very positive week--except in the ratings, which were the lowest that NBC had ever received.

You don't get the media for much in this world that's positive; it's only the crises and the catastrophes that get covered, so it's very hard for the public to get a sense that American assistance accomplishes anything.

Kremer: Much of this conversation has been pessimistic, particularly when we were discussing terrible problems like AIDS, but it's important to remember some of the facts we started out with. Per capita, India's economy is now growing at five times the rate Britain achieved during the Industrial Revolution. There has been tremendous progress recently in much of the world.

Even in Africa, although income levels may not have improved, life expectancy went up until AIDS. At independence, there were only a handful of university graduates in some countries. Now there are education systems throughout these countries, and they have many university graduates. Birth rates are starting to fall. It may take a while, but in the long run I'd be optimistic even about Africa.

Timmer: When I first arrived in Indonesia in 1970, there were two Ph.D.'s in agricultural economics in the entire country. Now there are probably five faculties of agricultural economics with the capacity to do modest types of Ph.D. training. There is a real institutional capacity to think about their agriculture and food policies. Thirty years ago, it seemed that everything was wrong with that society--chaos, the lack of institutions, poor leadership. Now, it is democratic--even if stumbling--and it has made progress.

That brings us back to the interaction between the economics of growth and the politics of growth, and how poverty and inequality connect there. As economists, we don't see a lot of statistical evidence that inequality is in any way harmful to growth. But when you put the political connections in, and the inability to sustain growth processes in highly unequal societies, those two go together.

If we were remiss in our discussion, it was in not focusing on building political systems that are capable of sustaining growth. There's no clear evidence that new democracies do that any better than other forms of government--in newly democratic poor countries, that transition process is ugly no matter what.

Mathews: That form of government can even be counterproductive for a lot of years. It's a whole lot harder than it was in Korea.

Beddoes: You're overlooking some characteristics of democracy. It's not necessarily the one person-one vote aspect of governance that affects the economy. It's more the transparency: the press freedom, the sense that you can have nongovernmental voices. Those aspects seem to be much more directly linked to the economy than voting. It's hard to prove this, but intuitively you can see the linkage between a vibrant nongovernmental sector and women's groups and poverty alleviation. That's not the way we use the term "democracy", but it's actually being "liberal" in the English sense that is the underpinning of society or individuals.

Timmer: That's the sense that what really matters in sustaining democratic processes is "voice"--that people participate and have some say in the nature of their lives, whether through community-based organizations or an electoral process that produces a prime minister.

Kremer: That's the warm and fuzzy aspect of democracy. A complementary view is that democracy helps get rid of some really terrible things. Leaders who behave like Idi Amin would not stay in power if they had to face elections. Maybe democracy doesn't do much for growth overall, but it helps get rid of some of the worst disasters in the world...

Perkins: Like famines, which typically happen when there is a lack of information, where information is suppressed and the free press is repressed.

Kremer: In that sense, Kenya has changed tremendously over the time that I've been visiting the country. Now you see columns in the newspapers where people write in to complain that, for example, traffic police on a particular road are demanding bribes. Publicity may not always work, but surely sometimes it embarrasses people into quitting that behavior.

Mathews: The real puzzle is the interaction of economic freedom and political liberalization. Our whole debate over trade with China was about how closely political liberalization would be linked to economic growth. Clearly, you can get an awful lot of economic growth for a long time before you get any political liberalization.

Bloom: Jeff Sachs [Stone professor of international trade] moderated a debate at the Kennedy School last spring about debt reduction and how to deal with governments that are unresponsive to their people. Jeff's argument is that you target what you can do, in health and education, and work through nongovernmental organizations. You can be remarkably effective in both those areas, and in fertility and agricultural programs, I'm sure. You get right to the farmers, right to the children. The World Bank and International Monetary Fund representatives at the debate didn't disagree, but they said their role was to create sustainable civil societies, which depend on putting money through and to governments, to improve government. The meeting adjourned without a resolution, but it's a fundamental dilemma: where do you put your marbles?