Transcript

Transcript of an IMF Economic Forum -- Globalization: North-South Linkages

April 11, 2002

View or listen to this press conference using Media Player

MR. STARRELS: My name is John Starrels and on behalf of the IMF's External Relations Department, where I work, welcome to this Economic Forum. I'm absolutely delighted you're here, as are the participants.

Just a couple of very quick announcements so we can get on with the presentations.

First of all, we strongly endorse engaged discussion, so at the end of these proceedings, you will notice at each of your seats there is a microphone. Please use it. We'd love you to both introduce yourself and, if appropriate, give your institutional affiliation. We want to stay in touch with you, and that is one way of doing so.

Secondly, there is some literature relevant to today's proceedings. Please avail yourself of it. You'll see it when you exit the two doors on your left and your right.

And on that note, I take great pride and great pleasure in introducing Carmen Reinhart, who will be handling these proceedings. Carmen, the floor is yours.

MS. REINHART: Okay. Welcome to our session on Globalization, North-South Linkages. I'm very happy to introduce David Dollar, who is Research Manager of the Development Research Group at the World Bank. David's work on trade has received considerable attention and so did—the Dollar-Rodrik debate is now very, very famous. He will speak first. And Carol Graham, she is Deputy Director of Economic Studies, Brookings Institution. David will talk about trade. Carol will also talk about inequality and income mobility.

I'm going to take an opportunity to advertise her excellent new book, "Happiness and Hardship: Opportunity and Insecurity in New Market Economies." I recommend you read it.

And, last, Graciela Kaminsky, who is my long-time co-author. She's a professor of economics at George Washington University, who is going to address not the trade side of the equation nor the inequality issues that arise in the context of globalization, but the financial markets, capital markets.

Also, those of you that came in early noticed that I had her name switched to me. This is because the authors are supposed to speak for only for ten minutes, and Graciela does have a tendency to run overboard. So if she sits next to me, I can at least kick her.

So without further ado, you know, once we hear from them, we will open the floor to questions.

MR. DOLLAR: Thank you very much, Carmen. It's really a great pleasure to be here. I'm going to talk about some of the results from our recent World Bank book "Globalization, Growth and Poverty" that I co-authored with Paul Collier, with contributions from lots of people inside and outside of the World Bank.

The topic of this forum is North-South Linkages, and I'm going to focus on the real side of that. I want to begin by pointing out that there has been really a dramatic change in the trade relations between North and South over the past 20 years, and we can characterize that in a number of ways. But here I'm pointing out that, 20 years ago, developing countries as a group mostly exported primary products—minerals, agriculture—and one of the big changes in the past 20 years is that developing countries have moved into manufactures, so that about 80 percent of their merchandise exports are manufactured products. There has also been a big increase in service exports from developing countries. So there's been a real change in the nature of North-South trade relations.

Globalization is obviously a big, complicated issue that involves many things. I'm going to be focusing particularly on this trade side.

Now, while that was an aggregate figure for the developing world, another interesting point is that there are a group of developing countries that are participating very actively in the trade side of globalization, and here are just a selected group of countries. This is showing the increase in trade to GDP, real trade at constant prices over real PPP GDP, the change over the past 20 years. And it's just trying to highlight that there's this great diversity of experiences in the South. So you have countries like China, where trade to GDP is actually—this is actually an understatement. It's more than tripled over the past few decades. Some of the other countries there you can see with very large increases in trade to GDP.

Some of these are the poorest countries in the world, so that's an interesting phenomenon. But you can also see there that a lot of developing countries trade less today than 20 years ago. So we have a diversity of experiences in the developing world, and that creates the potential for researchers to look at what's the impact of these differences.

The last kind of introductory point I want to make is that, while I was asked to focus on trade, I want to emphasize that that manufacturing trade that I just highlighted, in the real world that's very closely tied to direct foreign investment. So what you see is the same—pretty much the same countries that were at the top of that last list are the developing countries that are getting large flows of foreign investment, direct investment. This is just for one year, but if you did something similar for the 1990s, you would find there's a group of developing countries that are receiving a lot of direct investment, that are exporting, and are really involved in the global economy. And then there are a lot of locations in the developing world that are just not involved.

So our research is really aimed at trying to understand what's the impact of this integration or non-integration on developing countries.

I'd like to start with some ideas or hypotheses that I take from the policy debate, and this is a quote from about a year ago now from one of the leading British newspapers, from one of their columnists, one view on globalization and poor countries, poor people. He's talking about British political parties, and he says all the main British parties support non-stop expansion in world trade and services, although we all know it makes rich people richer and poor people poorer. And I think this was a fairly common view. A year, two years ago, you'd see a lot of comments along this line. You would also see a different kind of viewpoint, for example, coming from President Fox of Mexico, a somewhat more nuanced or balanced view of this integration. President Fox on a recent trip to Washington said, "We're convinced that globalization is good," but then he added, "It's good when you do your homework," and he then gives a list of some things that fall into developing countries' homework from his point of view.

So I think of these as different ideas or hypotheses about the effect of integration on poor countries and poor people, one view that it's largely a destructive force and another view that at least under some conditions it helps developing countries with their growth and their poverty reduction.

And so what we've done in our research is we've tried to look at the empirical evidence to see which of these ideas or hypotheses have more support.

So let me just quickly go through a couple of results that are relevant for this discussion here today.

One thing that we've done on the trade side, we identify the top one-third of developing countries in terms of increased trade to GDP over the past 20 years, and we call these the new globalizers. It includes China, India, Mexico, Brazil, Bangladesh, so some very low-income countries, some more middle-income countries. And that group pretty much by construction has had a very large increase in trade to GDP over the past 20 years, more than 100 percent on average.

The rich countries, the integration has increased by about 70 percent, and you can see that the rest of the developing world actually—as a whole actually trades less of its GDP today than 20 years ago. So there's this diversity in the developing world.

Now, to sort of give a punch line to a fairly long and complex research agenda, what we find in our work is that the countries that are integrating more with the global economy have generally seen accelerations in their growth rates over the past 20 years. This is population weighted, so China plays a big role. But if you take China out, you still find that most of these globalizing developing countries have seen accelerated growth during the past 20 years.

And to put that in context, in the 1990s, which were a pretty good decade in many ways, the rich countries were growing at 2 percent per capita on average, and these globalizing developing countries were growing much faster, at 5 percent per capita. If you take China out, it's 3.5 percent per capita, which is still almost twice as fast as the rich countries.

So this is an interesting new phenomenon that much of the fastest growth in the world is occurring in poor locations, particularly locations that are getting a lot of foreign investment, increasing their trade, and I emphasize that a lot of that turns out to be manufactured trade. You can see the rest of the world had negative per capita GDP growth in the 1990s.

So there's important diversity in the world, but I want to emphasize that it's not initially a rich-versus-poor diversity. This globalizing group of countries has about three billion people, and if you go back 20 years, the vast majority of the world's poor lived in these globalizing countries because it includes China, India, Bangladesh, those countries in particular with large numbers of poor people. And, on average, this less globalized group tend to be more lower-middle-income countries, though there are obviously some Sub-Saharan African countries in there with quite low income.

If these growth trends persist, of course, it will be a rich-versus-poor story, but looking at the last 20 years, it's interesting that a lot of the most rapid growth has been in countries that were very poor 20 years ago.

So that's really my first point, that there is a very strong correlation between increased participation in trade and growth, and it's hard to prove the causality. But in our serious econometric work, we control for other growth factors, we address endogeneity, and we find, I think, pretty solid evidence that this participation in direct foreign investment and trade is really good for developing country growth.

Now, my second point is going to address what I view as one of the big myths about globalization. You often hear it said that globalization is leading the growing inequality within countries, and this is just simply not true. We now have quite good household income distribution data from a large number of countries, and if you just look at the trends, you find that in some countries in the '90s inequality has gone up. In just as many countries, inequality has gone down. But actually in most countries the changes over time are rather small.

More important for this discussion, if you try to explain changes in inequality with measures related to globalization, like trade flows or trade policy or capital flows or capital controls, there's no systematic relationship between any of these globalization measures and changes in inequality.

So let me—and notice I didn't throw a lot of slides at you during that brief digression because I want to try to keep this brief and I want to kind of emphasize, well, what is the implication of that. And so here are five good examples of globalizing developing countries in the 1990s, where I'm focusing particularly on some very poor countries, and I said as a whole that growth rates are well above OECD growth rates. But also notice there is a lot of dispersion. You know, China's doing a lot better than Bangladesh. You know, that's interesting. We can talk about that.

The point I want to make here is that all of these countries have had successful poverty reduction during the 1990s, and, in general, there's a very close relationship between poverty reduction—here it's the percent rate of change, percent rate of reduction of the poverty rate—very close relationship between poverty reduction and growth.

China is actually one country where inequality clearly has gone up, so, you know, its rate of growth of poverty reduction, rate of poverty reduction is not—doesn't have quite the same relationship to growth as some of the others. But, of course, China has had the fastest growth in the world, so it's also had the fastest poverty reduction.

In Vietnam, there's been virtually no change in inequality during the reform and liberalization. In Uganda, inequality has shifted in favor of the poor, so you can see Uganda has had very impressive poverty reduction with just a pretty good growth rate of 3.8 percent per capita.

So among these globalizing developing countries, in some cases income distribution has shifted in favor of the poor. In some cases, inequality has gone up. In all of them, there's a pretty tight link between growth and poverty reduction, so the poor have clearly benefited from the growth that's going on in these countries.

Now, a couple more points about that inequality issue. We also have some really good data from the International Labor Organization that looks at wages at very specific occupational levels, like primary schoolteachers or garment workers. And so we can compare changes in wages properly measured, and what you find is that wages for formal sector workers tend to be going up everywhere in the world, but they've gone up a lot more rapidly over the past ten years in these globalizing developing countries compared to the rest of the world. So workers benefit, wages go up.

I also want to emphasize that the benefits are not just material. If you look at social indicators, infant mortality, school enrollment rates, you'll find that in these globalizing developing countries there's been very impressive social progress, and one nice example of this comes from a good study of child labor in Vietnam. Here this is from a good survey in 1993. This is basically lining up all the households in a big survey in Vietnam from poorest to richest and looking at the incidence of child labor, how many 6- to 15-year-olds are working more than 20 hours per week.

You can see, as you go left to right, what you're seeing is that child labor is much more common in poor households, which is what you'd expect, but the important point here is we went back to the same families five years later, and the 1998 line is looking at the same families and looking at the incidence of child labor, and you can see it's gone down for every income group. And this is partly because income has gone up very dramatically for all of these households, even the very poorest households in Vietnam. So poverty is one of the main determinants of child labor. But also Vietnam has done a very impressive job of building up secondary education, so what you see mirroring this is a big increase in secondary school enrollment rates throughout Vietnam. So child labor goes down, secondary school enrollments go up. So we see the social indicators generally going hand in hand with the poverty reduction in countries like Vietnam.

Now, my third point quickly that I want to emphasize is that this integration doesn't just depend on trade policies or investment policies, but it really depends on a wide range of policies in developing countries. And I want to give you a few examples of some of the policies that really seem to matter.

This is from an interesting database we purchased from the U.S. Customs, with all of these shipments into the United States from all ports in the world over the past ten years. So we have something like three million observations. And we can go down to a very detailed product level, like men's shirts, and we can look at the cost of sending a container from different ports in the world to, say, Baltimore.

And the point—I just used this in Bangladesh. The point here is that while Bangladesh has reduced its tariffs and kind of marginally makes it into our group of globalizers, you know, Bangladesh still has some very serious impediments, one of which is their main port is extremely inefficient and corrupt. And the result of that is the cost of shipping a container of men's shirts is about twice as high as from some of these nearby ports that have—you know, are just the same distance away from Baltimore, essentially. And this is basically like an 8 percent export tax. So you can reduce your formal taxes, your formal import tariffs, your formal export taxes. But if your port is really incompetent, then in this case it acts basically like an 8 percent tax. And in Mombasa, it's more like a 10 or 12 percent tax. So a lot of developing countries are not integrating very successfully, not so much because of their trade policies but because of these real bottlenecks that exist, for example, in the trans—in this case it's just one aspect of the transport network.

Another thing we're doing is supporting large surveys of private firms in developing countries.

MS. REINHART: Oh, David?

MR. DOLLAR: Am I out of time? Okay. Just one last interesting point. We're also looking at how long it takes to get goods through customs. And what you find is in Shanghai they've created a very nice production environment where you can get stuff in and out. And in India it just takes a lot longer to get goods through customs.

So the point here really is that integration requires not just trade policies, but really a whole set of policies that I would put under the framework of investment climate. So I'm just going to skip ahead to the punch line, final last slide. Bottom line, estimates from sources outside the World Bank of the number of poor people in the world, total number of people living on less than $1 a day. So if you want to look at kind of a bottom line, the number of poor in the world has been rising, you know, partly because of population growth up through 1980. Since 1980, world population has continued to increase. We've added another 1.6 billion people since 1980. But the number of poor people has gone down by an estimated 200 million because of this rapid growth in low-income countries. Most of this poverty reduction is in China, India, Bangladesh, Vietnam. Most of the poor lived in Asia. Most of the poverty reduction has been in Asia.

So the bottom line is that this integration strategy, participating in trade and investment, has been one part of a successful strategy for many low-income countries to grow faster and reduce poverty. And I'll stop there and just take one second to get out of this. I may not be smart enough to figure out how to get out of this.

MS. GRAHAM: Thank you, both John [Starrels] and Carmen. I want to talk today a bit about some research that has been trying to contribute to the very contentious debate on globalization and inequality. There is also one on globalization and poverty, but I think David has—among others, but I think David has done probably the most of anybody along these lines in really documenting the very clear evidence that countries that open up trade and grow reduce poverty. And I just don't think there is that much serious debate left on that issue. It seems to be the countries that don't trade and don't grow where the poor do the worst, and that's very clear. Again, a very big contribution on his part.

This is based on some research that's been trying to take a new focus on—rather than on static measures of inequality, to look at income mobility—in other words, who's moving up and down the income ladder—and also to look at public perceptions of those trends—in other words, how do people think they've done compared to how they actually did.

The mobility work was earlier a joint work with Nancy Birdsall, and the work on public perceptions is some work I've been doing with Stefano Pettinato, who's actually my co-author on this book, and he's here today in the middle of the room. So maybe he'll correct me if I make any large errors.

Anyway, the first point is why look at income mobility, and much of the work that's done trying to look at the effects of globalization on inequality basically looks at Gini coefficients, which are very static measures, snapshots in time of particular countries' income distributions and of the whole distribution. And what we know about Ginis is that they don't capture the whole story and they don't change very much. Chile three decades ago had approximately the same Gini coefficient as it does today, but for those of you who know Chile, it's a tremendously different society and economy than it was three decades ago.

So the reason for looking at mobility is that you can think about how people do over a particular period of time, ten years, say a period of market reforms or policy change or opening to globalization. You could look at people over their lifetime, or you could look at people between generations. And this slide is a good example.

A Gini—say this was a society. These are the earnings curves of lawyers versus bricklayers, and, you know, bricklayers start off higher than lawyers. Law students tend to have below zero earnings—this should probably go down lower. So bricklayers—at this point in time the bricklayers look better than the lawyers, and the Gini, if you were able to figure out where each one was at a particular moment in time captured by the Gini, you'd be much more worried about the lawyers than the bricklayers.

But if you look at the earnings curves over the lifetimes, obviously lawyers do a lot better, earn a lot more over their lifetime. And if you were worried about redistributing income from one to the other, you would presumably be worried about the bricklayers rather than the lawyers. The point here is that the Gini doesn't capture this.

So what do we know about mobility? Well, first of all, income mobility measures are hard to get because you need panel data, the same data over time for the same people. And that's rare, particularly for developing economies. But we do have some data, and it's—again, this is just—this is one example. The data we have is not only anecdotal, it's slightly better than anecdotal, but it's certainly illustrative.

This is a comparison in a ten-year period of mobility rates in the U.S., which is known to be the land of opportunity, and Peru, which is an emerging market economy, in a period that it implemented fairly dramatic market-oriented reforms and opened up its market to free trade, basically globalized.

Not to belabor this, if you look at this line, the vertical line is where people start off in the quintile distribution, and the horizontal line is where they end up. So in the U.S., 61 percent of the people that start in the bottom quintile are still there ten years later; 59 percent of the people that start off in the top quintile are still there ten years later. One percent of the people move all the way from the bottom to the top. That's the rags to riches story. Five percent move up to the fourth quintile.

There's also some downward mobility. Three percent of the people that started off in the top quintile fell all the way to the bottom in ten years, and those are the riches to rags story.

The people here on the bold diagonal are the people that basically start off and finish off in the same place, so that that's no mobility.

Now, here's Peru, and what's quite remarkable is that there's much more mobility in Peru than there is in the land of opportunity. So only 45 percent of the people that started off at the bottom were still there ten years later, and 5 percent of them moved all the way up to the top quintile. So that is indeed the rags to riches story.

There's another equally interesting story here, which is, if you look—so there's more upward mobility. There's also more downward mobility in the sense that—look at the fourth quintile. Only 32 percent of the people that start off in the fourth quintile are still there ten years later. But a lot of them moved down. So these are basically middle-class people who experienced falls into less better living standards.

One thing also is the U.S. data is income data, and the Peruvian data is expenditure data. So the Peruvian data really underestimates the amount of mobility there is in Peru.

Anyway, that is just an illustration, but what it shows it that there's a lot of movement up and down the income ladder, and Peru I think is not all that different from many other Latin American economies.

What explains this? Well, we don't have—we can't say that this is only the result of market reforms because we don't have comparable data from before. But what we do know is that during the period of opening to free trade, turn to—entering a globalized economy that there have been major changes in rewards to education, to different categories of education in Latin America, and that contrary to what the theory would have predicted, which is that unskilled labor would do the best, the highest rewards with the opening to free trade have been skilled labor and higher educated groups. And I think this is because—David would be able to talk about this probably better than I, but I think it's because the real rewards to unskilled labor have been in places with cheaper unskilled labor in Asia.

So in Latin America we've seen the rewards go to people with higher education. You see this goes way up. And another part of this story is that people with primary education are doing about the same as people with secondary education. So while before, having a secondary education made you quite a bit better off than those who only finished primary school, that difference is almost—that difference has really narrowed. And I think if you just think about this in terms of what's happening to people, prior to the sort of turn to free trade, what you got was a lot more—a lot of people with secondary education basically had stable middle-class lives. They weren't necessarily middle-class by developed economy standards, but they tended to have stable jobs in the public sector and things were okay. Now people with a secondary education are seeing very mixed rewards. Some of them do all right and others don't.

So just with this as a background, what do we know about how people think they're doing? And these are the results of a survey that I did with Richard Webb in Peru where we went back—we had panel data for these people, and we went back and asked people, How do you think your economic situation is today compared to ten years ago? And this is data for a ten-year period of time.

If you look at these axes, on this side we have what we call objective mobility. It's percent income change. So these are people that had losses, stayed about the same, people that had 30 to 99 percent gains, and people that had 100 percent gains or more, going up that way.

This is how people answered, and their possible responses range from very positive, positive, indifferent, to very negative. And if you'll see—and what surprised us was that there's quite a strong negative skew in these four boxes in the sense that these are people that had the most income gains in the sample. This is a sample of 500 people. And yet they answered that their situation was negative or very negative compared to before.

Now, when we first got these results, we thought, well, maybe it's the time period people were interviewed. We repeated this for three years in a row, and we got fairly consistent findings. So then we said, all right, maybe it's Peru, maybe it's Peruvians. I'm part Peruvian. I know we're a little weird, so maybe that's it. What explains this?

We then found some comparable data for Russia, a shorter time period, it's a five-year time period. But what we got was an even stronger negative skew in the responses of the people that had done the best during the ten-year period, and an even higher percent, well over 50 percent of them said that they had done very negatively or negatively.

So we started to worry about this. First of all, what explains it? Where are they? Is this the very poor? Well, for Peru, we find that it's not the very poor that are negative, poor people that moved up. It's people that are roughly in the middle of the income distribution.

Now, this chart is only people that had positive upward mobility, and you'll that—and here's people—negative answers, people that answered the same, and people that answered positively. And you'll see sort of the biggest block of negative answers are not among the poorest quintile, but right here in this middle group. And if you think about this story that it's probably people that—that the very poor probably do quite well with opening to free trade, and people with just primary education are doing better relative to secondary education completion than they were before. It's probably people in the middle that are more frustrated.

We then did a more detailed portrait of these frustrated achievers, as we call them in the book, to see what was the difference between them and their counterparts, people that had upward mobility that were not negative. And we found that they were—these little points on the right mean that the differences are statistically significant. What we found is that they were slightly older on average, that they were more urban, and I think this is important because urban people's reference norms are higher. They are comparing themselves to a much broader base of people than are rural people. There were no gender difference, no education difference, no difference in terms of household expenditure.

We then thought maybe its volatility. Maybe these people are having a more volatile income trajectory and that's explaining their frustrations. But instead, when we looked at the coefficient of variation for the period, we found that in Peru the frustrated achievers actually had less volatility and income than did the non-frustrated groups.

I still haven't fully explained that finding. But the other thing we found is that we had a whole other set of perceptions questions. In other words, how satisfied are you with your job? How do you think your economic situation will be in the future? Where do you place yourself on a notional income ladder if society is a nine-step ladder where the poor are one and the rich are nine? And we found that the frustrated achievers scored significantly lower on all of these things. So, for example, they placed themselves lower on a notional income ladder than do non-frustrated respondents. So you start to get a pattern in terms of how these people behave or how they assess their situation.

Very briefly, Russia, not that many differences except the Russian frustrated achievers actually had a higher coefficient of variation, so they had more volatility in income. We also had some additional questions. Russian frustrated achievers were less happy with their lives, and I'll get into happiness very briefly in a minute.

MS. REINHART: Very briefly.

MS. GRAHAM: Okay. But the more important finding for policy is that the frustrated achievers were less favorable about democracy and less favorable about markets and much more likely to say that society should restrict the income of the rich. So there's some beliefs here that come with these frustrations that would have real policy implications.

Very quickly, we just wanted to see if these things were behaviorally determined and if there were very big differences in terms of the way, say, Latin Americans and Russians assessed their well-being than other societies. We did the first study of happiness in Latin America to match what's been done in the developed economies.

Basically, at the country level, what the findings show is that as countries grow wealthier over time, mean happiness scores do not increase. I don't want to belabor this too much except that if you—these are the developed countries, and you'll see that there's sort of a mixed relationship. The U.S. is the wealthiest country; it's not the happiest. The Dutch are the happiest. My husband's Dutch, and it's freezing and rains a lot there, so I'm not sure I understand this finding. But the Dutch are the happiest population.

If you look at Latin America, you get the same thing. Venezuela, Brazil, Panama are the happiest countries. This could be cultural, whatever. The unhappiest countries are the Andean countries. Peru doesn't do very well at all. Neither does Russia.

But the point is that there's sort of no clear correlation just between income and happiness after a certain minimum level of income, and this holds at the individual level as well, which is in keeping with our frustrated achiever findings.

Give me time to put this one—these last two slides up, Carmen, because I think people will enjoy them.

This is at the individual level. Latin America looks very much like the U.S. or Europe in that there is an age relationship with happiness that holds across thousands and thousands of individuals. And even though everybody talks about their 20s and 30s with great nostalgia, it turns out that the lowest point on the happiness curve is mid-40s, and—

[Laughter.]

MS. GRAHAM: Now, in the U.S. it's a little bit earlier. In Latin America it's about 47. In Russia it's at about age 50, which isn't a good story. But the point is for those of you on this end of the curve, you have nowhere to go but up, as long as you're healthy.

[Laughter.]

MS. GRAHAM: And the last thing, just to relate this back to policy relevance, which is obviously what we're interested in, is that in Latin America—I could go on about happiness, and I won't because I'll get in trouble with the moderator. But what we find in Latin America, as well as in Russia—I'm just showing you Latin America now—is that happier people are on average more favorable towards democracy and they're more satisfied with democracy. They don't necessarily prefer democracy to any other system. That just was statistically insignificant. But they tend to be more favorable about the market, about democracy, not on this slide but also about free trade. And happier people are less likely to want to redistribute income, which is a whole other set of findings.

We don't know the direction of causality here, obviously. It could be that happier people are happier under any circumstance they are, which, you know, that's—I'm actually doing some more work in this area which establishes the direction of causality. But without getting into that, just to go back to the frustration of these achievers, in other words, what drives the frustration of these people in emerging market economies that are doing well but say they're doing badly?

Well, clearly, part of it could be behavioral, but I think there's much more, and I think there are two key policy issues. One is insecurity. None of these countries have—very few countries in Latin America have any kind of unemployment insurance or broadly available social insurance. There are targeted programs for the very poor, but if you're in the middle, you may be doing well today. But if you're not doing well tomorrow, you have no safety net to fall back on. I think that's a big issue. Insecurity is a big issue, and if we're interested in sustained public support for globalization and the kinds of policies that have people—that do reduce poverty and have countries grow over time, I think we should think about this. And, secondly, I think these frustrations are also driven by relative income differences, and that's what we find from the happiness literature. And if you think about Latin America and certainly Russia in recent years, there are very large gaps between how the very, very wealthy do and the rest of society. And they seem to be driving significant frustration—

MS. REINHART: Carol, I'm about to become very unhappy.

MS. GRAHAM: I'm stopping. I'm finished.

MS. REINHART: Okay.

Graciela Kaminsky?

MS. KAMINSKY: It's going to take some time. It's an antiquity, this computer.

[Pause.]

MS. KAMINSKY: Finally, we made it. Well, today I'm going to talk about the other aspect of globalization. I'm going to be talking about financial globalization.

Let me tell you a little bit about the facts. Here you have a graph with the amounts of capital flows to developing economies, and this is what we are talking about when we talk about globalization. You see that when we start this, it starts in 1970. The level of capital flows to developing countries is quite small, and as we move into the 1990s, capital flows surge, especially foreign direct investment, portfolio equities, and bonds.

There was in the 1970s also a period of financial globalization, in the late '70s, beginning of the 1980s, and basically the beneficiaries of that globalization were countries mainly in Latin America.

So as with trade, and globalization and trade, when we talk about globalization and finance, there are also two opposing views. Some argue that it's a blessing. Some argue that financial globalization is a curse. Let me start first with the blessing part.

In general, those that favor globalization, financial globalization, argued that this improves the functioning of financial markets. Markets become more liquid. There is more monetization in the economy, so firms have more access to credit and, therefore, they can produce more.

Also, it allows cross-country risk diversification, so a resident in Chile can buy assets in Chile and can also buy assets in Europe and, therefore, not suffer as dramatically when there is a recession in Chile.

There is also evidence that financial liberalization triggers growth, and at least some of the estimates in various pieces of research said that financial liberalization can trigger at least one percentage point higher growth rates in other developing economies. So this is the blessing part.

Just to give you an idea about the deepening of financial markets and more access to funds all over the world, here you have a graph that shows how firms in various countries obtain funds overseas. This is issuing of equity, so that when they list in different foreign asset markets, this is the amount that they raise. This is in billions of dollars. So you can see that as you move into the late 1990s, this has increased substantially.

Not only firms can issue equities all over the world. They can also have more access to lending through international bond issuance. And although after the Asian crisis bond issuance has declined—this is bond issuance in industrial countries from developing countries. Although it has declined, this still has increased compared to the numbers that we have in the early 1990s.

As everything in economics, we have the opposing view, and this view is that financial globalization is a curse. And, in general, what critics of financial liberalization point out is that most of the financial crises that we have seen in the 1990s, and also crises that happened in the 1980s, especially in Latin America, have been preceded by financial liberalization. And, in particular, everybody points to the excessive booms and busts in financial markets due to these excessive capital inflows following financial liberalization. As many of you know, this was a stylized feature in most of Asian countries that suffered a recession. All of them experienced excessive real estate market booms and busts and also stock market booms and busts.

In general, from various sources of research, you can also find that, following financial liberalization, the odds of having a banking crisis increase at least about 40 percent. So this is the arguments in favor of turning the clock back to the times of financial repression.

In some research that I'm doing jointly with Sergio Schmukler, we investigated the effects of financial liberalization and globalization on financial markets. And our idea was to try to make consistent the findings of two types of literature: one literature that looks at the short run, what happens in the immediate aftermath of financial liberalization, and in those cases you have crises; and the other literature that focuses on the short run and focuses on the effects on growth.

One literature finds that globalization is a curse; the other, the long-run literature, finds that it's a blessing. So we were trying to understand a little bit what was happening with financial markets in the immediate aftermath of financial liberalization, and if liberalization persists, what happens with financial markets. And what we have here is just the cycles, averages of stock market cycles in 28 different countries following liberalization. So the red lines are the short run after financial liberalization, and the green lines are the long-run behavior of stock markets. And what you do find is that in the immediate aftermath of financial liberalization, it's true, these booms and busts in financial cycles increase. I don't have here the size, the amplitude of these booms and busts during repression, but stands in the middle between the green line and the red line.

QUESTIONER: What are your two axes?

MS. KAMINSKY: My axis is the size and—the vertical axis is the size of the boom or the bust. In the horizontal axis, what you have is demands before the peak of the stock market boom.

So, for example, what you observe here is that in the immediate aftermath of financial liberalization, the 24 months before the peak of the financial cycle, you have that the stock market is about at the level 30, and then it increased to 100, and the amplitude is much smaller after—in the long run.

So we do observe that after you open up the capital account, after you liberalize financial markets, you see these excesses coming on. But as you move into—as this liberalization persists, financial markets become much more stable than in repression times. So this opens up an important question in economics, in international economics, about what is the proper sequencing of liberalization and reform.

Many argue that it's very risky to open up a financial system that is not prepared to cope with free capital movement, and in general, they argue that banks during repression times don't work very properly, are very inefficient. And when you open up the capital account, then you open up new sources of funding to these already bankrupt banks, and they bet, and you have this huge capital outflow—inflows in the economy that then trigger these financial crises. That's the idea of the supporters of this hypothesis, is that you first have to clean up the institutional system, the banking sector, and then open up the economy. And they argue there are many reforms that you have to make. You have to reform the government. The regulatory burden of the government has to decrease. Property rights have to improve. You have to modify commercial codes. Government accountability has to increase. Transparency of the government has to improve.

But the question is that many argue—and this is some of the evidence that we have—that all these changes in institutions never occur before financial liberalization, and perhaps it's because of several various political pressures. After financial transactions are free, they argued that a capital market itself fills the important supervisory function of financial institutions. If that's the case, if the changes in the institution is not an exogenous parameter but depends on financial liberalization, then sequencing may not be optimal.

And liberalization by itself can change institutions. So, for example, when firms become global, they list in foreign exchange markets. They are in the jurisdiction of superior legal systems. They have higher disclosure standards. They become more transparent. They are monitored by better investment bankers. Even when firms do not become global, they show themselves as adhering to a poorer standard and, therefore, there is also better monitoring. So by itself, financial liberalization may trigger an improvement in institutions.

It's also true that globalization may put also pressure on countries to improve their legal systems so they can better participate in the global economy. In preliminary results that I'm doing with Sergio Schmukler at the World Bank, what we find is that changes and improvement in indices of law and order occur in all the countries after financial liberalization occurs. We also find, for example, changes in laws such as passing laws on prosecution of insider trading, this happens after financial liberalization. So it does seem that the process of globalization by itself triggers changes in institutions.

So with this caveat in mind, I just want to bring the last transparency because Carmen otherwise is going to get very upset, and she's my co-author so I have to deal with her all the time. So we have to be very careful. It's true that financial liberalization has triggered in the past financial crises. Many countries after these financial crises have closed their economies.

For example, in the green line here, what I'm showing is an average of restrictions in the capital account of various regions in the world, and a big number three is when the economy is closed to capital account transactions with respect to the rest of the world, and a one means that it's fully liberalized.

What you see here is that Latin America, after the early 1990s—in the 1970s it started to liberalize, and then it had the debt crisis. And after the debt crisis, it introduced capital controls again. Capital controls, it regulated the banking industry, it regulated the stock market. And so we have to be very careful because if all the benefits of financial liberalization take place in the longer run, we cannot at the brink of one crisis reintroduce all the restrictions again to capital mobility.

MS. REINHART: She miraculously adhered to her time slot.

Now we have some time for questions. Please remember to turn on your mike and identify yourself.

QUESTIONER: My name is Dario Skuka (ph). I represent the Institute for Global Energy Affairs. I'd like to offer three remarks to deal with historical background and the political economy of what is globalization.

During the 1980s, the Cold War era, few analysts have noted the significance of fast-growing foreign trade between or among the LDCs, which I define as South, as opposed to trade from the OECD to LDCs. It's possible to explain why if anybody wants it.

The second observation is that globalization as we know it now has been imposed by the West, multinational corporations, with little assent or organizational input by the new LDCs or by the old LDCs, ergo, unrest manifested in various ways, and usually supported by Western-educated elites.

Between the two, what we now have and what was during the Cold War in the '80s, we have had experience of Iran. The Shah of Iran imposed modernization in a fashion which was too fast and too heavy, which then brought revolution. The Iranians simply were not able to absorb the significance, the importance, and possibly the benefits of what the Shah was trying to do, ergo, the revolt.

There is a parallel to what happened in Iran at that time and what seems to be brewing right now around the world.

Thank you.

MS. REINHART: Well, there is, I mean, within this institution, outside this institution, a continuing debate on the issue you raised of whether one does it gradually or whether one does a big bang. And I think that possibly I'll leave the rest of the panel to, you know, remark on your statement. But my sense is that prior to the Asian crises, the big-bang approach was much more the dominant thinking than it is today when it comes to liberalization, be it capital account, be it trade.

QUESTIONER: How about we collect a bunch of comments?

MS. REINHART: Yes, okay.

MR. LANKES: Hans Peter Lankes from the IMF. I have a question for Mr. Dollar.

A frequent criticism of the points that you made empirically about the relation between openness and growth is that the countries that did best on both scores are actually not the most open, that China, India, Vietnam, et cetera, were late liberalizers and haven't actually liberalized as much as many other countries that performed less well on these scores. How do you answer to that?

MR. O'FLAHERTY: I'm Dan O'Flaherty from the U.S. National Foreign Trade Council. I'd like to ask especially Mr. Dollar, but others as well, about the automaticity of the relationship between following correct macroeconomic policies and achieving high levels of growth, because there are countries in the world where that hasn't happened and where there's a sort of counter-reaction, South Africa being one. And the comments made by their Finance Minister in Monterrey reflected considerable disillusionment with what's known to be the Washington consensus and signals, I think, very negatively a trend that might be reflected in the Mexican Congress' recent restrictions on your President Fox.

MR. SAGAFI-NEJAD: This is Tagi Sagafi-nejad. I'm a professor at Loyola College. I enjoyed all the presentations, and I have a lot of questions. But with respect to Professor Graham's comment, the punch line that came at the very end regarding the relationship between happiness and income distribution and the relative deprivation idea, I would like to hear more of your comments with respect to the empirical evidence that exists in that.

I know that as early as the early '70s Richard Easterlin, who is the father of economics of happiness, had proposed that it's not the absolute levels but the disparity in income, and what Ted Gurr called relative deprivation, and you mentioned the Gini coefficient as perhaps one way of explaining that. That was contributing to the level of happiness. But, however we measure it, I think it's an important ingredient in the process of globalization because we need to know whether globalization in the end is increasing happiness or not. I think that's really sort of the bottom line.

And if I may while I have the mike make a small comment regarding the fast pace of reform in Iran, there is the misperception that the Shah introduced too much reform too fast. That is not quite true. In fact, it's the opposite. The wrong types of reform were introduced, and they were introduced in the wrong direction and the wrong way. It wasn't that the Iranian economy was not ready. It was quite ready. But the Shah did it the wrong way, and that's why he stumbled and fumbled.

Thank you.

MS. REINHART: We'll take one more question and then do a round and continue.

QUESTIONER: In fact, it's not a new question. It's a continuation of my neighbor speaking from the IMF to Mr. Dollar. To what extent—how do you define a globalizing country or a country that has an integrationist policy? Is it by the results of it having invested more, increased foreign investment and trade, or by a set of policies that do result in increased trade and investment? Because as my neighbor just outlined, the number of countries who succeeded in globalizing in terms of trade and investment did not have open policies as such, and other countries maybe have been a bit more successful in opening up but they don't have the same results.

Thank you.

MS. REINHART: Okay. David, we'll start with you.

MR. DOLLAR: Okay. Let's see. The way that I would approach this is to sort of start with kind of a factual answer. In my kind of grouping of globalizers that I showed you in these slides, those are defined on the basis of increases in constant price—trade relative to constant price, GDP over the past 20 years. You can do a grouping based on, you know, foreign direct investment. You'll get a similar list of countries. Population weighted, you'll get very similar results.

So the question then is why don't we try to define a group of globalizers based on policies, and my view that I've come to after a number of years is that it's very hard to measure the level of openness of trade policies. So different approaches have been tried, and I've contributed to that literature. But, in the end, personally I'm not satisfied with any of these different measures.

Countries have reported tariff lines. You can just take an average across all of these. But, of course, some products are just much more important in the economy than others. You can weight by actual trade, but then if you have extremely high tariffs against a product, you don't import it, therefore, you don't trade it, therefore, it doesn't come into the weight.

So there's a large literature on this, and the bottom line for me is if you look at the existing tariff data, there's very little relationship between that and trade. So that says to me that that's not a particularly good measure of trade policy. The same point can be made about trying to measure non-tariff barriers.

So in the end, when we get down to specific countries, I think we end up arguing a lot about, you know, who's open, who's not open. So what I've fallen back on is using the actual increases in trade. I would say that for a lot of this debate about globalization, that's exactly the right measure—right?—because some of the anti-globalization movement is saying very clearly trade and foreign investment—they don't put the word "policy" in there. They say trade and foreign investment are bad for poor countries, so to test that, using trade and foreign investment data is exactly the right data.

But then in my serious work with Art Cray (ph), what we do is we recognize that the increase in trade volume is endogenous, so we control for changes in the rule of law, changes in other policies. We treat this as endogenous. We instrument for it, and we show that it's an extremely persistent result in that dynamic cross-country literature, that instrumented increases in trade and instrumented increases in direct foreign investment are just very, very robustly related to increases in growth.

Then when I look at individual countries, see, I think some of the countries that were mentioned—you know, China—India is a weaker case, but I think in the case of China and Vietnam, trade liberalization was an important part of their initial set of reforms, in China's case starting in 1979, in Vietnam the whole reform program started later, more like 1989. But in both cases, trade liberalization, some trade liberalization was an important part of this. You basically—in each country you created a situation where producers could import capital equipment and materials relatively easily, and that has been a key part of their success. And a lot of the rest of the developing world, it certainly wasn't true 20 years ago that you could easily import capital, equipment, and materials, and it remains in many countries in Africa not easy to import capital, equipment, and materials.

Now, I did signal that where I'm kind of going with my research is focusing on this transport cost issue, which I think is very important, and some of that's related to geography, but a lot of that's related to policies concerning the administration of ports and customs administration. So if I were going to try again to kind of build up an index of trade policy, I would put a lot of weight on the policies around ports and customs administration because I think in practice that's where a lot of the bottlenecks lie.

Let's see. Now, the other kind of question that seemed pretty clearly directed to me was really concerning this idea of automaticity, and I don't particularly like the phrase "Washington consensus," but let me not be scared to work with it. The phrase was developed by John Williamson. If you look at his list of what goes into the Washington consensus, the first item on the list is good rule of law. So then you have to ask yourself: Do you think South Africa is a great example of the Washington consensus where the very first item is property rights and rule of law? And I think a lot of the countries that are disappointed with reform are countries that have made some macroeconomic and trade reforms, but have a very poor rule of law property rights situation, and it's extremely difficult—there's very clear evidence that the direct foreign investment pays a lot of attention to that basic governance environment of rule of law and property rights.

China looks pretty good on these rule of law and property rights measures, as does Vietnam, it might surprise you, as does India. So I think some of the countries that are disappointed have done some of the macro level reforms but are missing that key ingredient, which, frankly, was always part of the Washington consensus.

Investing in education, by the way, was part of the original Washington consensus, too. If you actually go back and look at the list, I'm not sure that the current development debates are all that different from the list that was in John Williamson's list of what goes into the Washington consensus. But I do think it is a fair criticism—last sentence. It's a fair criticism to say that the World Bank and the IMF probably put too much weight on the stabilization and formal trade liberalization part of the Washington consensus agenda and not as much weight on the property rights rule of law and other aspects of the agenda, which were always there in John Williamson's list. So I wouldn't so much criticize the Washington consensus as the way—the weights that were put on different elements of it by the IMF and the World Bank over the last 20 years.

MS. REINHART: Carol?

MS. GRAHAM: Just briefly, to add to the Washington consensus question, then I'll focus more on the other ones.

The other thing—I agree with David's response, but I would add to it that the Washington consensus is sufficient but maybe not enough to achieve growth under every circumstance. Those include circumstances where external conditions are adverse, circumstances where there are very large geographic disadvantages. And it's also not enough to make up for additional policies, for example, countries where education is extremely unequally distributed, all kinds of other situations, where more is necessary to really get a country growing in the long term at a high rate of growth.

Saying that, that said, I think it's also very important to note that the real disaster cases are cases where we've really seen major increases in poverty and really sad stories for large percentages of the population are in countries that ignore the basic macro rules in the Washington consensus, that pursue very poor fiscal policies, that just sort of decide that following standard macroeconomic rules don't matter. And there you get huge, huge costs that are borne disproportionately by the poor.

So I would say maybe one way of answering the automaticity question is these policies are necessary. They may not be sufficient in every case to get sustained growth.

Now to the two questions pertaining to sort of expectations and relative deprivation. On Iran, which I don't know very much about so I'm treading on thin ice here, but I would say just from the little I know, one could certainly say it's probably a situation where there was substantial heightened expectations by modernization, and they may well not have been met fast enough by objective economic conditions. But also I think one of the things that seems to go along with opening up and globalizing eventually is pressure to liberalize politically. And here I think the Iran story was as much about regime legitimacy as it was about economic modernization.

Related to that, the excellent question about relative deprivation and do we need to pay attention to it, and certainly that's what our findings show. In the book, we talk both about Ted Gurr's thesis of relative deprivation and social unrest, and we also spend quite a lot of time on work by Albert Hirschmann (ph) and his so-called tunnel effect hypothesis, which basically says that initial inequality in the development process is the design of some initial progress and it's a good thing. And if you were in a traffic jam in a tunnel and one lane started to move faster, initially it would give you hope because it would mean that there was some progress and that would be a signal where you might go in the future. That's what initial progress in development does.

But if that inequality persists and your lane doesn't move, eventually you get very, very frustrated and you're tempted to do radical things like jump the median strip or join a protest or start throwing rocks.

The point is that some inequality is almost a necessary part of economic progress. But sustained and persistent and high levels of inequality tend to increase feelings of relative deprivation, and they are precisely the kinds of feelings that Ted Gurr suggests can lead to revolution.

Now, in terms of our own findings, our frustrated achievers, even though they're doing well by objective economic conditions, are obviously comparing themselves to people that are doing much, much better, and they don't see a lot of prospects of ever getting to that point. Our frustrated achievers place themselves lower on a notional income ladder than they actually are. They compare themselves much less favorably to others in their community and others in their country than do the non-frustrated respondents. They're very, very concerned about insecurity. They have much higher fears of reported fear of unemployment, for example, and as I mentioned, they're much more negative about markets and democracy. And so I do think there's some—I think this does at least call for some increased attention to what kinds of effects persistent or very high levels of inequality can have, but I think equally important is the whole issue of insecurity in countries with no safety nets.

Back to Easterlin's findings, our findings at the country level support his findings as well—

MS. REINHART: Carol?

MS. GRAHAM: Okay. I'll stop there.

MS. REINHART: Graciela?

MS. KAMINSKY: I would like to talk a little bit about the issue of financial liberalization—or liberalization and globalization in general and its relationship in this case that you were talking. Is it Washington, is it mature countries that are triggering that?

We have to pay—we don't have to look just at the process of globalization that we observe now. Globalization has grown and declined through the years, and there have been reversals of financial development. It's not a monotonic increasing all the time. We have seen a process of globalization at the turn of the century during the gold standard periods and then a reversal. So it's not just the Washington consensus and the international institutions that are triggering all this opening up. It's economic reasons that are doing that, and when there are large trade flows and firms need financing to compete with other firms in other countries, then you open up.

So it's not just you open up because there are benefits to be made, and that's why there are some reversals in the aftermath of crises and when there is a decline in inflows.

MS. REINHART: I just want to underscore briefly that what David Dollar mentioned on the trade side, there's an analogue on the financial side, namely, liberalizing your capital account, liberalizing capital flows. It may not be necessary to trigger a large increase in capital flows if you don't have an equity market that's developed, if you don't have firms that are big enough to borrow abroad. So just as high transport costs can make trade liberalization relatively ineffective, so to speak, all these other features can also make financial liberalization ineffective as well.

We have time to take a couple more.

QUESTIONER: Alexander Sarley (ph), School of International Service graduate student. My question is directed toward Mr. Dollar, but I'd appreciate the input of Ms. Kaminsky as well.

I understand that increased FDI has led to increased levels of manufacturing trade for less developed countries. My question is: Considering the recent efforts of the industrial North to limit laws on local sourcing, outflow of profits, and the recent WTO ministerial agenda to implement more modalities regarding the protection of individual investors and intellectual property rights, what is the effect of this largely multinational corporation FDI on the actual development of these countries? And what role do sovereign governments play in their own national development strategies?

Thank you.

QUESTIONER: Hi. Sean Sunderland (ph) with the Embassy of Canada. My question is for whoever would like to answer, but I'm curious if in the research that the panelists have done, if they have found any correlation between the relative amount of development assistance going to developing countries and their relative propensity to liberalize and/or the rates of economic growth that these countries have undergone.

Thank you.

MS. REINHART: Okay. One last one.

MR. ROWDEN: I'm Rick Rowden from Results Educational Fund here in Washington. Mr. Dollar, just to build on what this first man's question was, isn't it true, though, that although Vietnam and China and India had begun their liberalization, they actually had, in fact, achieved many years of high growth rates well before they undertook their substantial levels of trade liberalization?

And, Ms. Kaminsky, regarding foreign direct investment, isn't it true that the overwhelming bulk, like 95 percent, really goes to about 15 or 20 emerging market economies and that leaves about 5 percent to the other 150 or so? I just would like you to address that.

And then one last thing, Mr. Dollar. The anti-globalization protesters that I'm familiar with certainly don't call for autarchy. There's issues of justice and other matters at hand, so it's not like they're saying countries should not trade at all.

Thank you.

MS. REINHART: Okay. Let's all be brief.

MR. DOLLAR: Okay, but let me start with this last point, because, I mean, obviously the anti-globalization movement is diverse and I have lots of friends in it who work on particular issues that I sympathize with. But some of the direct objectives of the anti-globalization movement were, say, to keep China out of the WTO or prevent Latin American countries from having a free trade agreement with the United States. So some of the direct objectives of the anti-globalization movement have been to limit trade opportunities for poor countries. So a lot of my work—I view myself as an advocate for the poor, and I think, you know, a lot of the poor countries would like to have better access to the U.S. market and Canada market. And I see myself as an advocate for that.

I want to kind of come back to one of the first questions asked. I just don't see this agenda driven by multinational corporations at all, and I think it's insulting to China or India or Vietnam or Mexico to suggest that their policy is driven by multinational corporations. These are countries with strong civil societies. You know, India and Mexico are democratic. China and Vietnam aren't formally democratic, but they have very strong civil societies, in my view. So I think it's insulting to countries in the South to suggest that their policies are driven by multinational corporations. I just simply don't see the evidence for that.

But to kind of answer some of the specific questions, in the case of—I know Vietnam really well because I was an adviser to Vietnam starting in 1989. Trade liberalization was part of the initial reform. And, you know, I don't want to overemphasize the importance of one thing compared to others, but Vietnam is a nice example of how the different reforms interact, because Vietnam stabilized high inflation and it gave land to peasant families and it liberalized rice trade all at the same time.

And what happened in the summer of '89 is the rice crop increased by 20 percent just like that. Had it been a closed economy, the price of rice would have collapsed. But, instead, Vietnam became the third largest exporter of rice in the world immediately, and that supports the rice price. It started importing fertilizer at a much lower price than it had been producing it.

And if you look at what's happened to the poor in Vietnam, their income has gone up because they are getting a better price for their rice output, and they're buying their fertilizer at a much lower price. So you can relate the trade liberalization very directly to the improvement in people's lives in Vietnam starting in the first year of reform.

In China, in the second year of reform, 1979, they opened up special zones. The whole city of Shanghai is a special zone. So I would say there were discrete trade liberalization moves linked to attracting foreign direct investment that were part of a broad strategy, much broader than—yeah, you could say for China the external part was a small part. Maybe it was a small part. But, you know, I just don't see—ask yourself the counterfactual. If China had done everything else—if China had done everything else and it was one of the most closed economies of the world in '78, if China had done all the domestic reforms and remained completely closed to foreign trade and investment, do you think it could have grown as fast and had as much poverty reduction? And I'm not aware of any serious person in development who would argue that China could have done as well if it had remained a closed economy to trade and foreign investment. And that's what a lot of—that's what the part of the debate that I'm engaged in is really focusing on.

Then for the Canadian gentleman, we have a whole strand of literature on this aid effectiveness, which Carmen is not going to let me go into, but, you know, there's quite a bit of evidence that aid does matter, but it really does depend on the quality of policies in the recipient country. And we'll have to save that for another forum.

MS. REINHART: Carol, do you want to add anything?

MS. GRAHAM: Well, I'll try and add something for David since he ran out of time, which is exactly that, that aid seems to foster growth when policies are good, when institutions are strong, when there is a rule of law.

The more difficult question to answer is the one you posed about is there a correlation between aid and liberalization. And I think because of the mixed motive for which aid is given by different donors, I think that correlation probably doesn't exist, and David can correct me if I'm wrong. I think if it existed more strongly, we might see a better relationship between aid and growth.

MS. REINHART: Graciela?

MS. KAMINSKY: It's true that foreign direct investment, most of the capital flows have been to 12 larger developing countries, the largest countries in Latin America, China, some East Asian countries. That doesn't mean there are good opportunities in those countries. And it's true that capital doesn't flow to the poorest countries. Then we have to talk a little bit about what should you do with poorer countries. But that doesn't mean that these flows of capital to some emerging economies have led to larger growth in those economies and in that sense have been beneficial. There is a large literature that looks at these correlations between liberalization and growth, and it's very positive.

MS. REINHART: Okay. Well, thank you, everyone. We enjoyed having you.

[Applause.]

[Whereupon, the Economic Forum was concluded.]