American Enterprise Institute

April 10, 2008

[Edited transcript from audio tapes]

9:45 a.m.	Registration and Breakfast
10:00	Presenter:	Ann Owen, Hamilton College
	Discussants:	Paul Wolfowitz, AEI
		Jeremiah Norris, Hudson Institute
	Moderator:	Roger Bate, AEI
12:00 p.m.	Adjournment

Proceedings:

Roger Bate, AEI [Moderator]:  Some of you may well attend AEI's Election Watch series.  It's excellent, the monthly Election Watch series.  If you don't, I suggest you do come to it, because it's fascinating, especially for me, who can't vote in the election.  I can just watch all of the debates with a kind of mild enthusiasm and interest. 

One area where sentiment is certainly lacking at the moment is free trade.  NAFTA is being kicked with the discussion of possible timeouts and bilateral and multilateral trade deals.  And obviously while some of this is playing to the crowd, the vested interests at the time in the short run and weren't sustained, there is certainly a protectionist undercurrent at this moment in time.  Yet at the same time, we see President Bush and his administration spending vast amounts of money, tens of billions of dollars on health aid, much of it being very well spent, saving many, many lives; and the presidential candidates are all saying that they want to spend even more.

Now I'm speaking today without directly answering this question.  We'll probably put some information our way, which might explain why this is somewhat paradoxical.  It is my great pleasure to introduce our speaker today.  Ann Owen is an Associate Professor of Economics at Hamilton College.  She's a graduate of Boston University and has a Ph.D. from Brown University.  Her work has been published in numerous journals and you have the handout, the flyer, which has lots and lots of information about her in it. 

It was her paper in the Review of International Economics, from September, that drew her to my attention.  In fact, it was Foreign Policy Magazine who alerted me to the paper, because they thought it was particularly interesting, a friend at Foreign Policy.  The subject, the title of that paper is the subject of today's event, "Is Free Trade Good for Health?"  So without further ado, I'm going to let Ann take over.

 Ann Owen:  The paper I'm going to talk to you about was, as Roger mentioned, published in the September Review of International Economics.  It's a joint work with a colleague of mine at Hamilton, Steve Wu.  So what I'm going to try to do today is tell you a little bit about the background and very briefly about some related literature.  This paper, you'll find out quickly, is an empirical paper, and the strategy that we take is very important.  I'm going to talk a little bit about that.  I'll talk about this overall correlation that we find, and then some attempts that we made to try to understand the reasons for the correlation, and then we'll wrap it up.

 So what are we going to do in this paper?  Well, we're trying to answer the question, "What are the effects of openness to trade on health outcomes?"  And we're particularly interested in thinking about whether or not these effects are varied with the country's income.  As you probably already know, we do find effects.  Then we're trying to figure out, "Well, why does this happen?  What are the channels that help explain these linkages?"

 So I'll tell you right up front what we find.  Increased trade is associated with better health outcomes in developing countries.  Then we also, in trying to figure out why this is, we find that, "Well, policy and trade are related."  The overall policy environment and the amount of trade that a country does are related; and we also find some evidence consistent with there being knowledge spillovers.  Right?  And this is the idea that if you interact with people in other countries, you might learn something from them that might improve health outcomes.  Now you can't ever observe these things directly, so what we're going to look for is evidence that's consistent with that happening.

 Several different strands of the literature are related to what we do.  There are many papers on the issue of trade and just income growth.  We think what we're doing here is related to that literature for two reasons.  One, is that healthier people are more productive; so health, in a sense, is an input into production.  So there's going to be a relationship between health and income.  Then the other issue is that health, in and of itself, might be a legitimate measure of welfare for us to think about, in addition to income.

 There's also several papers on trade, institutions, geography and income growth.  What this literature is about is, it's really a debate about which is the most important.  We're going to have some of the same issues as this literature does, in the sense that it's really hard to attribute causality to anything in particular when they're all occurring at the same time.  We kind of have that issue in our paper too, when we try to figure out, "Why is this happening?"  And we try to do a little bit to try to disentangle different effects of these very factors.

 Also there's literature on trade and the environment.  And again, the environment may be a legitimate measure of welfare for us to be thinking about in addition to health.  Also, the environmental quality could be related to health.  And there's vast literature on economic activity and health, both at the microeconomic level, which looks at income and socioeconomic status and how that affects health; as well the macroeconomic level, where we're looking at relationships between income per capita and health incomes.  So there are many different strands of literature that our work is related to.

 One thing that's important to realize is that what I'm going to show you, controls for income.  So to the extent that trade causes higher income, and income causes better health, that's a channel that's over and above what we find.  Okay?  Because we're going to already control for income.  So we are really trying to get at the idea of this trade, of exchange.  How does this affect this interaction between countries?  How is this affecting health? 

And there's a couple of reasons why we might expect these kinds of interactions to be related to health.  First, knowledge spillovers.  Right?  This is what I mentioned earlier.  If we're interacting with other countries, then maybe we're going to learn something from that interaction that's going to help us improve health outcomes.  This ability to exchange with other countries might give some countries better access to medical resources -- vaccines, other medical supplies that would help countries be healthier.  Also, there could be a relationship between trade and the development of institutions, which provides a policy environment that might be better for health.

 And then finally, trade could actually be worse for health if it facilitates transfers of disease; is related to increased pollution; or maybe it allows the trading of unhealthy consumer goods.  So theoretically, we can tell the story either way, and that's why this paper focuses on the empirics.  Because this is primarily an empirical question.

 So what are we going to do?  Our main methodology is a fixed effects model.  A couple of things that are important about this.  One is that because we're using this method, we are already controlling for characteristics that don't change over time about the country.  So things that you can think of, characteristics of this country that don't change over time, that's not the reason for this correlation.  We are controlling for it.  So, for example, you might think that geography is related to how much you trade and geography might also be related to health.  Well, that doesn't change over time, so we're controlling for that.

 We're also controlling for effects of time periods with this method.  So to the extent that, "Oh, in 1985, there was a big flu pandemic and people got sick," well, we're controlling for that as well.  That's not the reason.  We're also controlling for income levels.  GDP per capita, so you can see this coefficient Beta 1.  We're actually expecting that to be positive.  Richer countries are healthier. 

 The two coefficients to be particularly concerned with are these coefficients, Beta 2 and Beta 3.  Beta 2 is the coefficient on openness, and I'll talk in a minute about how we're measuring openness, but we're using several different measures.  And then Beta 3 is the coefficient on an interaction of openness with GDP.  So if Beta 2 is positive, that means trade is associated, or openness, is associated with better health outcomes.  Beta 3, let's take into account the income of the country.  If it's a positive sign, that means that effect is going to be even bigger the more your income is.  If it's negative, that means that the bigger your income is, we're going to subtract something off that effect.  Right?  So if it's negative, that means the effect is stronger in the poorer countries.

 So we're going to look at several different measures of openness, two different kinds.  One is just purely trade volume related.  The first one is probably the one that's most commonly used in these kinds of studies.  Exports plus imports divided by GDP.  We also looked at imports divided by GDP and exports divided by GDP, but they're so highly correlated that you get exactly the same results, so we're only going to show imports divided by GDP.

 And then we do something a little different, in that we're looking at the imports of the country, but now we're looking specifically at who they're importing from.  And before we add them up, we're weighting it by the health of the trading partner.  And the results I'm going to show you, we're weighting these by one over the infant mortality rate.  Why we're doing this is that if it's the interaction -- right?  If there's knowledge spillovers, then the more we trade with healthy countries, the bigger the effect should be. 

In addition to volume measures, we're also looking at two policy-based measures.  We're looking at the Sachs-Warner index.  This is just, you're either a measure where Sachs and Warner either classified countries as "open" or "closed."  You're closed if any of these five things are true:  You have a high black market premium; you have export distortions resulting from an Export Marketing Board; you're socialist; you have extensive import quotas; or extensive tariffs.  So they make a judgment.  If you fail any of those tests, you're closed.  Right?  So it's one zero.

 Then, some people that have looked in this Sachs-Warner index, say, "Oh, all the action is coming from the black market premium, that's really what's going on,” so we're going to look at the black market premium separately.  So we start with data for 219 countries.  We're going to look every five years.  We don't actually think that if you trade more today, that tomorrow you get healthier.  Right?  There's going to be a time lag.  And we're going to lag our measures of openness.  "Oh, you trade more now; five years from now, we expect we're going to see this effect on health."  And we're going to use different measures of health -- infant mortality, life expectancy.  We have those for males and females. 

We're also going to kind of control for what we just call kind of human capital policy of the country, by looking at population growth and secondary schooling rates.  We get our data from a variety of sources:  The Penn World Tables; the World Bank Development indicators and the detailed trade data come from the U.N. 

 Some summary statistics from our data.  It's over the whole time period, 1960-1995.  The important thing here that I want to make is that there is variation in both health and trade measures.  So for example, the average infant mortality -- so this is number of infant deaths per 1,000 births is about 82.  But there's a very wide range.  The low is Sweden at 3.7; and the high goes to Sierra Leone at 180.  So there's a big variation in these health outcomes.  The same thing with life expectancy.  It ranges from 37 to 83 for female life expectancy.  And trade has a similar kind of variation in it.  So there's a big variation.  Singapore trades four times its GDP; Brazil, less than an eighth.

 So let me show you the main results.  There's a couple of numbers here, but focus on the first column.  So this is male life expectancy.  So this, with the little stars, that means it's statistically significant.  So positive number here on GDP means, "As we expect, the richer countries have higher male life expectancy."  A positive number on openness, so this is a measure of openness with exports, plus imports divided by GDP.  More openness is associated with higher male life expectancy, but this negative coefficient on the interaction means that the higher your income is, the smaller that effect is.  So we're finding that effect to be the biggest in developing countries. 

 So what the other columns in this table do is they look at different measures.  They look at female life expectancy and also infant mortality.  And when you look at infant mortality numbers, you should be expecting all the signs to be opposite, because higher infant mortality is worse health.  Right?  We add controls, we still get this effect.  We use this now with all our different measures of openness.  So this is trade volume.  This is the black market premium.  The Sachs-Warner index.  Imports divided by GDP.  And then our health adjusted imports.  I know I'm going through this really quickly.  You'll have to kind of trust me that, "Yes, we found the same thing." It's important.  Well, it's at least noticeable. 

One standard deviation increase in the log of openness for the poorest country results in a decrease in infant mortality of 7 deaths per 1,000.  What's interesting is when we use the health adjusted import measure; it's twice as big.  Okay?  So if you're trading with healthier countries, that effect is bigger.  The same kind of increase in the log of openness, increases female life expectancy 1.3 years.  And again, it's about twice as big, if we're using health adjusted imports; and male life expectancy, almost a year.  So these effects are noticeable.

 So now what we're worried about is, "Is this actually really true?  Right? Are we just identifying some kind of spurious correlation?"  So I'm going to go through a couple of things that we did to kind of convince ourselves that we didn’t.  For people who are really interested, you should read the paper, because I can't go through everything.  But just to give you an idea of the kinds of things that we did. 

 One thing that we're concerned about is that there's some omitted variable.  Even though we're controlling for lots of things, we are concerned that there's an omitted variable.  So we tried thinking about what that might be, to see whether or not, by including it, did we make this correlation go away?  We added income and equality, that didn't affect it. 

Female secondary school enrollment rates.  Right?  You might think that that actually could be related to trade, if trade makes countries be more, efficiently use their resources, they're going to educate women more.  That is also related to some of these health outcomes, in particular, infant mortality.  We put that in, that doesn't make it go away.

 Civil wars.  Right?  If you're in a civil war, you're not going to trade.  You're going to have bad health outcomes.  We put that in, it doesn't make it go away.  The other thing that we did was, "Well, we realized that our results implied is that the biggest effects are in developing countries, not in developed countries." 

So what we did was we split the sample and we say, "Okay, well, what we should see is very smaller, no effects in developed countries.  Then we should see the effects occur in developing countries."  And this is for developed counties now.  We're looking only at the openness coefficient, and you get here, nothing is significant.  So we find basically no effect of openness on health in developed countries.  But when you look in developing countries, you see the positive effects.  So that, again, makes us feel a little better that this is not just some artifact of our estimation strategy.

 What else did we do?  Well, we dropped countries with the largest change in health outcomes; dropped OPEC countries; dropped transition economies.  We even looked to see if we flipped the timing.  Regress openness on lagged health outcomes, is that going to get us anything?  Again, that would be some indication that it's just an artifact of our estimation strategy.  We didn't find anything there.

 So now what we do is we go to think about, "Well, why does this correlation exist?"  If we can find a reason for it, that actually makes us feel better that it actually exists.  So we looked at a couple of different things.  We tried to find some evidence of these knowledge spillovers.  We looked at the relationship between openness and policy regime.  We looked at this idea that more open economies can import more medical resources.  And then we also looked, again, at the relationship maybe between openness and development aid.

 Of these four, we find some evidence for three of them.  First of all, knowledge spillovers.  The first piece of evidence, I've already showed you; and that is that health adjusted imports are stronger.  But then people said to us, "Well, maybe that's -- You have to take into account the fact that countries, a lot of the trade that occurs are actually occurring between countries that are like each other.”  Right?  So it's rich countries trading with rich countries.  So what we did was we predicted, based on the characteristics of the country, what their trade should be with each country. 

So for those of you who are familiar with gravity models, that's what we used.  So we used a gravity model to predict trade.  Then we took the residual of that prediction.  So countries that are trading more than you might expect them to, based on their characteristics, they got a positive residual.  We took those residuals and we weighted them by the health of the trading partner and we get even stronger results.  I have to admit, we cannot prove that knowledge spillovers occur, because you can't see them.  But this evidence is definitely consistent with that idea.

 Then we looked at this idea that, "Well, maybe openness and policy regimes are related."  And here what we do is we exploit the idea that we actually have two different kinds of measures of trade.  We have measures of volume and also policy.  So we said, "Well, what happens if we put them both in?  Who wins?  In a sense, who wins that horserace?  Is it trade volume?  Are these overall measures of policy?" 

And what we find is that trade volume -- so we're putting in this exports plus imports divided by GDP, plus the Sachs-Warner index, which is a broader measure or overall policy.  What we find is that the volume always comes in significantly and the policy one sometimes does, especially in the infant mortality equation.  So we interpret this as being, "Well, there's a little bit of both things going on here."  That volume, trade volume does matter, but also the overall kind of policy environment that you're in is mattering too. 

 So then we looked for evidence for this channel.  Maybe these open economies are importing more medical resources.  And when you look cross-sectionally, this makes quite a bit of sense.  Countries that immunize children have lower rates of infant mortality.  Right?  So you need to get those vaccines from somewhere.  Maybe they're getting them through trade.

 Well, what we did was we got some detailed trade data to look at the amount of these medical resources that countries were importing, both per person and also divided by GDP.  And when we put those into our estimation, unfortunately, we don’t find evidence that this is what's happening.  The only thing that we find evidence of is some reverse causality.  Countries that are less healthy import more of these medical resources.  So that makes some sense, but we can't find evidence that this is what's going on.

 And then finally, we looked at this idea that, "Well, maybe what's really happening is it's development aid.  Countries that are open are being rewarded with more development aid, particularly maybe bilateral aid.  And that it's really the development aid that's happening.  It's not the trade."  We looked at eight different kinds of aid, and of those different kinds of aid, we found one kind is actually associated with better health outcomes in our specification, and it's water resources aid.  And then we also actually found a correlation between water resources aid and the volume of trade.  So there's something there, but when we include this also in our regressions, the impact of trade is only reduced slightly by this channel.  So it's possible this is explaining a little bit, but it's not explaining everything.

 So overall, what we are finding is that more open developing economies exhibit better health outcomes.  We don't find these effects in developed economies.  Either they're going to be small or they're statistically insignificant, so we can't make a strong conclusion about that.  We think that we have some evidence that's consistent with knowledge spillovers.  And also, we think that policy may evolve as the country opens up to trade, both that country's policy environment as well as maybe other countries' policy, in terms of how much development aid it gets.  But overall, we do think that there's a strong correlation here.

 Roger Bate:  Thank you very much, Ann.  That was excellent and incredibly concise from what is a fantastic, but very complicated paper.  That was extremely good.  We probably need to load up the next PowerPoint presentation.  Karen, if you can help with that.  While we're doing that, I have a couple of questions.  I'm abusing my privilege as being Moderator.  Before, I said we wouldn't take any questions until the end, but I lied. 

 I sent this paper to a few health economists and macroeconomists who are always critical of large macro studies, because there was a fight between the macro and micro folks.  The main question that came across, which I wondered if you could deal with, was the health indicators.  Did you look at other health indicators?  Why did you pick the two that you chose?  What is the data reliability of those?  Because, of course, critics could say, "If the data quality of the health indicator is not very good, your time model collapses, because garbage in, garbage out."  I wondered if you could just deal with that question.

 Ann Owen:  Yeah.  The reason why we chose these is because these were collected by the World Bank.  Many people use these in studies, so in a sense, we thought that this was probably the most reliable source of cross-country data.  I, of course, in all kind of aggregate data, you always have to be aware of the fact that there's noise.  But I think that actually poor data quality, unless there's some systematic bias, in that more open economies are overstating their health outcomes or something like that, more noise in the data should actually make it harder to find something, I think.

 Roger Bate:  Well, that's good to hear, that you looked through these.  We may well have more questions related to that later on.  If we can pull up Paul's presentation and we'll hear some comments and response.  Thanks.

 Paul Wolfowitz:  Thank you.  First of all, I'd like to compliment Ann and her coauthor on what I think is not only a very interesting paper, but I think it's a very important paper.  I think it's distinguished by being very careful and precise, and perhaps it suffers from being very careful and precise.  Because I think, and I guess I'm going to err on the other side here, I think it has an impact on a very important public debate, where unfortunately, people are at times, I think, border on being inflammatory.  And we see that in various manifestations.  But in effect what you're saying is that openness and trade promote health.  In contrast, we have people like Nobel Prize Economist, Joe Stiglitz, talking about the devastating effect that globalization can have on developing countries, especially the poor in those countries. 

I'm just going to put up a slide here, if I can.  This is a relatively restrained version.  But it seems to me, underneath it all, it's of all things, a U.N. Trade and Development Report of 2004, suggesting that after all, trade may not be so good for development.  One of the world's richest men wrote, "Many people, particularly in developing counties, have been hurt by globalization without being supported by a social safety net."  In case you couldn't guess, that's George Soros, who goes on to talk about, "How markets are good at creating wealth, but are not designed to take care of other social needs.  As long as capitalism remains triumphant," he says, "The pursuit of money overrides all other social considerations." 

So the implication here is that we need government to look after the poor because markets devastate the poor.  I think perhaps this is the macro versus micro debate here.  There are many cases, I'm sure, where improving trade policies actually have had harmful affects.  But it seems to me that people who make statements, like the ones I've just quoted, really have the burden of proof against them.  In your paper, it seems to make it strongly, to show exactly where is it that they expect a better result from trade restrictions.  And in doing so, I think they have to confront the fact that, the implication that markets don't help the poor, but governments do, then has to confront the fact, "Well, what is it exactly the governments do?"

 Back, before the reaches of time, I was the Assistant Secretary of State for East Asian Affairs and dealt a lot with trade policy.  One of the wisest observations I came across was someone who said, "Well, trade policy is the policy of the weak."  You may notice we never heard from the American IT sector.  We never heard from the American biotech sector.  We heard from steel and textiles.  You have to qualify that though.  It's the policy of the economically weak, but the politically powerful.  And what's true in the United States, I think is true in developing countries, maybe even more so. 

So to the extent that government policy modifies trade, it’s not done in the interests of people who are earning a dollar a day or less.  It's done in the interests of people who are doing very well, thank you, from those trade restrictions.  So I think this is an important paper and an important conclusion.  Because I do think development policy should be concerned with the deserving weak, which is, I think, a reason why health and education properly get so much attention in the development community.

 Bringing me to the second point.  You say, and I guess I'm partly asking you to elaborate on a little bit -- but I'm going to venture out and elaborate before I let you answer the question -- where you say one of the reasons that trade and health are positively correlated is simply because good government provides policies that are conducive to both trade and better health outcomes. 

One of the things that came as a shock to me, when I went to Indonesia as Ambassador, again, a long time ago.  It would not have come as a shock to Roger Bate and his colleagues who have an excellent report -- Roger, I'm advertising for you -- from 2006, called, "Tariffs, Corruption and Other Impediments to Medicinal Access in Developing Countries."  The thing that came as a shock to me was when the deputy in my embassy, no economist, said, "Well, of course, everyone knows the Health Ministry in Indonesia is the most corrupt ministry in this rather corrupt country." 

And I guess it came as a shock, because one thinks of health as something that is generally recognized, important public good and attracts people who are public-spirited and care about the deserving poor.  It was explained to me, and I still think the explanation holds as a kind of iron triangle, if you like.  It sort of starts at the doctors who provide the care, or the lack of care; the pharmaceutical companies, which are not infrequently owned by the doctors; and the bureaucrats, who often, if not owned by the pharmaceutical companies, or at least rented by them -- if you understand that joke -- coming together to find a way to make money by restricting markets and restricting access to drugs.

 We've had, when I was at the World Bank, a couple of major problems.  One with, I have to say both with friends of the United States.  One with India and the other with Kenya, precisely over corruption in the health sector.  One of the arguments that I encountered -- you encountered also I think in the issue of corruption, "Well, what's a little waste between friends."  If 50% of the money goes into a Swiss bank account, at least 50% goes to fighting AIDS or fighting malaria.  Indeed, in a more inflammatory version, we were accused and we held up money for some of these programs, that we were actually killing people. 

It's a sort of strange argument in a way, because it kind of assumes that the 50% that goes into Swiss bank accounts isn't killing anybody, which would make the assumption, the further assumption, that there's such an abundance of foreign assistance that we really can afford to waste it.  My feeling is that when there are countries that are making good use of that assistance, that itself is a major problem.  But unfortunately, it doesn't stop there, because particularly in the health sector, the methods by which people steal are not simply to have a little bit of off balance sheet money.  But in fact to sell harmful drugs, to build hospitals that don't work, to possibly    in the case of the Indian experience    actually deliver blood-testing kits that fail to detect the HIV virus in blood supplies. 

 Transparency International, which is another very -- I think their 2006 Global Corruption Report focused on health with some very interesting information in it.  They did a study in Senegal, in this case with the cooperation of the Ministry of Health, which found, among other things, that -- They found vaccines and antibiotics that were diluted or replaced by water, which is obviously a good deal more than just stealing.  They saw medications sold to street vendors, who would then pass them out indiscriminately, without regard to what disease they were supposed to be treated for.  They saw people who had to be turned away because they couldn't pay the bribe that was necessary when they brought a sick child to the clinic, or as x ray machines and other medical devices that don't work. 

In the Senegal case, according to -- If you get Labelo [phonetic] TI, they were able to work with the government and with civil society to fix a lot of those problems, and it's a story with a good ending.  But I think these issues really have to be taken seriously.  And it seems to me -- and I'm sort of back to my question -- whether I'm over-generalizing from your study, in an anecdotal way.  But it does seem to me that it's part of the explanation of perhaps why more open economies may also be somewhat less corrupt and therefore producing results you have.

 One last general comment.  This is more in the form of a question.  But it has struck me that there's a wide variation in developing counties, in the degree of public versus private provision of health care.  A bright graduate student at Georgetown, who may or not want to be identified.  His name is Ryan Winger, who's done some work on this for me.  And the data are pretty sketchy and pretty hard to draw inferences from.  But it is striking.  Let me see if I've got the right slide up here.

 This is 10 African countries, and to the extent the data    I'm not going to get into the issues of the quality of the data.  It's pretty bad, I guess I would say.  But you can see, there's a very wide variation in the percentage of health care delivered from the private sector versus the public sector.  And it seems to me that also might have some correlation with the openness of economies.  Now he's going to get mad at me because he told me this isn't scientific and it isn't statistically significant. 

I said, "Could you do it anyway?"  And if you plot life expectancy versus the percentage of private provision of health care, you do get a -- He would say, "a statistically insignificant," but a somewhat interesting positive correlation.  And similarly, if you applied infant mortality with private provision of health care, you'd get a, again, "a statistically insignificant," but somewhat interesting negative correlation.

 I really raise this more by the way of asking a question, which is it seems to me that, and a very important area of, I guess, micro research, probably around macro research, is this issue of delivering mechanisms.  I don't think it should be approached in a doctrinaire way.  I've seen anecdotally some very different results.  For example, on private versus public provision of insecticide-treated bed nets.  It really strikes me that this is a very rich field for development impact research and for controlled experiments.  The consequences, I think, could be quite dramatic.  I guess I've sort of, the trade of bias toward some degree of market mechanism.  But I don't believe in uncontrolled markets either.  I'm not trying to do anything here other than to be a bit suggestive. 

 And then finally, I'd just like -- These are in the nature of really genuine questions, as opposed to questions that are biased behind them.  One is, I'm a little struck, if I understood you correctly, that you -- By controlling for income, you're talking about the affects of trade, regardless of its impact on income?  Yet, if there's in fact a strong positive correlation between openness and income economic growth, then in a way that's the most important thing. 

Another one of the world's richest men in the debate, Bill Easterly, said, "I don't care about GDP.  GDP never cured anyone of AIDS," which suggests that he better go learn some basic economics, it seems to me.  But in fact, I guess I'm just asking if I misunderstood you.  Because if I've understood you correctly, then it seems to me the impact of trade, when you take account of income affects, is even larger.

 Secondly, I was very interested in your results about foreign aid and would very much like to hear any elaboration of it; and struck that it seems that the one place where foreign aid makes the biggest difference is in water supplies, which intuitively sounds right, when one thinks of how much disease is spread by polluted water and whether we give    I think increasingly now there is attention being given to that issue.  It seems to me it's -- It's not very sexy.  It probably doesn't get as much attention as it should.

 And then finally, this is a question about the knowledge spillovers.  I guess I'm a little puzzled because, again, you can correct me.  If, for example, a country imports a great deal of automobiles from Sweden, let's say.  And Sweden has high health indicators, then it will turn up on your data as a knowledge spillover.  But it's hard to imagine that Swedish automobiles bring much more health benefits with them than let's say German or American.  So I'm wondering about that.  That's maybe down in the weeds.  But it did seem to me that the proxy you're using is an odd one; I guess is all I was saying.  Thank you.

 Roger Bate:  Thank you very much.  That was very interesting, and numerous questions that have come out from that, which I would like to throw to Ann now.  But I won't, because I think it's probably best that we get Jerry's comments, and then we have a much greater free-for-all with the questions.  So over to Jerry.

 Jeremiah Norris:  Thank you, Roger.  Ann, thank you very much for your presentation.  I want to provide some context to her position, that international trade has an effect on a basic national indicator such as health.  I won't challenge her very extensive study, hundreds of countries, a general study.  But what I would like to do is to talk about a specific country situation because it is so much in the news today, in terms of free trade, and that would be Colombia.

 So the issue I addressed is what improvements on basic national indicators have taken place in Colombia prior to the Free Trade Agreement, which came into effect in 1993 or so? 

I served in Colombia for three years as a Peace Corp. volunteer in the 1960's.  This was at a time when the civil war between the conservative and liberal factions had exhausted both sides to the point where they needed a timeout.  The modern history of Colombia is one marked with specific periods of violence, not only from drug lords, nor from the FARC, that is the armed insurgents, but also from civil unrest, which almost tore the nation apart in the 1950's. 

 On a sunny day in 1948, in the nation's capital of Bogotá, the very popular liberal leader, Jorge Gaitan, was having lunch in a sidewalk café outside of the country's leading newspaper, El Tiempo.  He was assassinated at the table.  A crowd gathered immediately and before the police could intervene, they tore the assassin's body to shreds, to the point where they could never identify who he was.  By sundown, the military stood shoulder-to-shoulder on the steps of the Palace of Justice firing into the crowd.  They could not be contained.  By nightfall, the city of Bogotá had gone up into flames.  The next day, [audible] down through the rest of the country.

 A Colombian sociologist, by the name of Orlando Fals Borda recorded this period in a book called, La Violencia.  In a riveting opening sentence, he said, "With a single shot, the thin veneer of a society was stripped away."  What followed was years of civil violence between the liberal and conservative factions.  From a population base of only 14,000,000 at that time, the lowest estimate in mortality was 375,000 people, almost all of these were civilians from neither the liberal or conservative party.  You can compare this mortality to what the United States suffered in World War II, which is just about the same.  But we suffered that rate of mortality from a population base of 145,000,000 people.

 When I arrived in Colombia, in the 1960's, both sides were spent.  No family had been unaffected.  It was a time of an unspoken truce.  Then in the late 1970's and throughout the 1980's, violence again broke out, but this time it was narco- activity, and into this maelstrom, the FARC emerged, the armed insurgents emerged.

 Fast forward.  A few weeks ago, I attended in Cartagena, a reunion of former Colombian volunteers.  On the day before it started, a teenage girl, 17 years of age, had organized in Cartagena, what was called "a March Against FARC" - again, this is the armed insurgents - a March against FARC and for peace.  Arguably, it was the largest demonstration ever witnessed in Colombia.  She organized the entire event by YouTube and this March Against FARC and For Peace was a worldwide event on that day. 

On the first day of our reunion, President Uribe came down from Bogotá to speak with us.  He was in a suit with an open collar; and after finishing his presentation, he joined the former volunteers on the main floor, and without a platoon of bodyguards, stood among us, taking in questions we wished to asked.  No aide attempted to pull him away and get him on to his next appointment.  But what impressed us most was the number of his ministers that came to the reunion, how young they were and how emblematic of the new Colombia they seemed to be.

 The Trade Minister came down from the podium, and rather than addressing us from the podium, he came down to the floor.  He was dressed in a casual shirt, rolled up sleeves and rumpled slacks, shoes that needed a shining.  He had left Colombia in the 1980's during the main narco-wars; went to Harvard and got an MBA.  He went on from there to work with McKinsey in Japan.  He learned Japanese.  Then he went on to China and learned Chinese.  On Christmas leave, in Bogotá, he listened to a campaign speech by a young upstart politician by the name of Uribe.  At that time, he was at 5% in the polls, probably just slightly above the margin of error.  He was so impressed with what this man was saying, that the next day, he walked into his campaign headquarters and volunteered his services. 

The Director for Colombia's National Social Action Program addressed us.  She could not have been more than 30 years old.  On one of her color slides, she showed us the impressive gains that Colombia had made over the past few years and improvements in basic social indicators.  Then, not unkindly, she waived the slide away, saying, "This is from the World Bank.  Pay no attention to it.  We are not doing that well, but we're doing okay." 

I felt caught in a time warp.  Wow.  A 17-year-old girl organizes Colombia's largest civil demonstration; the President of the country feels comfortable enough to stand among us, take any questions we might want to ask; a man, fluent in five languages, who could be making a ton of money in international consulting, returns to his country and becomes a Trade Minister; and a young woman, who had one of the major social action agencies, "This is the World Bank."  I thought to myself, "This is a country that's going to make it."

 But how has Colombia fared on critical and social economic indicators, prior to the Free Trade Agreement in 1993 or so?  Let's look at some numbers.  In 1975, the per capita income was $630, that moved to $1,400 in 1993.  In 1975, the gross domestic product was $7.2 billion, moving to $54 billion in 1993.  Exports were $1.7 billion in '75, moving to $7 billion in '93.  The average annual growth of exports in 1975 was 2.3%, moving to 11% in 1993.  Crude birth rates per 1,000 were 45, moving to 24 in 1993.  Death rates dropped in 12 in 1975, to 6 in 1993.  The infant mortality rate in 1960 was 100; it dropped to 36 in 1993.  Life expectancy was 55 in 1960; it dropped to 70 in 1993.

 Before the Free Trade Agreement then, Colombia was making substantial progress on these basic indicators, and it made this progress despite a long and costly civil war, the narco-war and the ongoing guerilla activity.  It's made more progress since too.  For instance, the life expectancy is now 73 years of age and infant mortality is down to 18.  The gains made prior to 1993 cannot be attributed to international trade, and the gains after 1993 were undoubtedly due largely to the momentum of previous years, rather than to the Free Trade Agreement.

 I believe it doesn't really matter all that much to future improvements and basic national indicators that the Colombia Free Trade Agreement passes or not in the U.S. Congress.  It does though matter to the U.S. because it will expose us as a protectionist nation, willing to put this agreement aside in favor of political expediency.  Do the Colombians want it?  Sure.  But they also would like to have Panama back and that's not going to happen.  Since it's unlikely to pass and we don't have many countries in South America that wish us well, the U.S. has a strategic interest in finding other ways to engage Colombia and sustain its strongest partner south of the border.  In the end, we need Colombia more than Colombia needs us and the Free Trade Agreement notwithstanding.  Thank you.

 Roger Bate:  Thank you very much, Jerry, for a very interesting tangential, but very important contribution. 

 Roger Bate:  Before I go back to the floor, several questions were asked indirectly, and some very directly, such as, "Opening up trade, does that increase transparency?"  Actually maybe that wasn't Paul's question.  Maybe that's another question I want to ask, "and is the transparency key?" 

 Paul Wolfowitz:  Yeah. That's exactly right.

 Roger Bate:  And, "Why did water aid work?"  I mean, that may be a micro study that has to be undertaken, but if you could go into detail.  Plus all the other questions.

 Ann Owen:  Yeah.  So the issue about aid, I mean I guess    What we were doing here, was we were trying to see if we could explain the relationship between trade and health with aid, so we did not    I guess I want to be a little cautious about what we didn't find about aid.  Right?  Because there's lots of studies out there that are specifically focused on that question.  So I think, I don't want to push too hard on the aid results.  I think from the perspective of our conclusions, what we showed was that by incorporating aid, it doesn't take away the effect of trade on health.  Right?  So I think that's our kind of takeaway. 

I mean, but in our kind of specification, we tried -- I'm trying to remember all the different kinds of aid.  Water resources aid; food aid; technical aid.  I think we included like an overall of aid, and none of those were robustly correlated with these health outcomes, which I think is somewhat consistent with some of the findings in the literature.  There's a huge debate about this, and I'm not sure that what we did was robust enough to take away big conclusions from that.

 The question about policy, and I think that's your question.  The transparency and good policy and --

 Roger Bate:  Well, the fact that if there is a more open trade environment, which I guess is domestic policy to the extent that they have an open trade environment, that that so drives quite a transparency, which lowers corruption.

 Ann Owen:  Right.  We did not have a direct measure of that.  Okay?  What we did have was the black market premium, which is going to be very related.  I mean, black market premium, we used it because it's a distortion in the exchange market, but it's very related to other kind of economic conditions.  So I think that that's related to that issue, but it doesn't directly address it. 

 Then the Sachs-Warner index, which is a little bit broader measure of policy; transparency is not one of their criteria.  My guess is that it's very related to the criteria that they used.  I had a couple other questions that I wrote down.

 Yeah.  This question about how health care is provided?  We don't have information about that.  I think that that could be related actually to your other question about, "What are these knowledge spillovers?"  I think the story that we have to tell in order for us to make sense is that knowledge about how health care is provided might be the kind of thing that is learned.  It's not attached to the car, but it would be something that because you're interacting with other countries, you're going to learn about this.  I think that would actually be a really interesting kind of further study, as to delve more into exactly how theses countries that are trading with each other, how maybe some of these policies might have changed.  But we don't have a direct evidence on that.

 And then this issue about controlling for income?  We did that intentionally, because we thought that if we wrote a paper that said, "Oh, trade is good for health" and everybody's going to say, "Well, yeah, because it increases income."  So I think that, although I guess there's people that would disagree.  In the economics literature, I think the consensus is that trade improves income and income improves health.  So for us to make a contribution, we actually wanted to look at this nature of exchange.  So I think you're right.  I mean, that's very important, and that would be an effect that would be added on to the effect that we find.  I think that was it.

 Roger Bate:  Okay.  That's great.  Thank you very much.  Well, we have just under an hour.  We don't have to take all of that time, but let's start taking some questions.  I'm sure we have a roaming mike.  Let's start with the gentleman over there, at the back.  If you have an affiliation and you want to explain what it is, please give us your name and affiliation, because this is being recorded.  Thank you.

 Peter Whitney:  Peter Whitney, Duke University.  I really enjoyed your presentation, and I think your rigorous approach with the way you went about it, to examine possible order aggressive correlation was very good; and to dismiss various factors, which could distort the results was terrific.  I think it'll serve a very useful argument in favor of free trade and so forth.  My question is, "Do you have any anecdotal evidence, individual countries, where you could go in and see actually what happened?"

 The only one I can think of is just one I've done research on, which is Chile.  I would say that the free trade there, in the '70's, they opened that but they also simultaneously had an infant feeding program.  It was all the Chicago's boys school of thought, and the two went hand-in-hand.  While their economy boomed, the infant mortality rate plunged.  It was a series of policies that put it together and so that openness to trade reflected an openness of the mind to modernization in every sense.  I would love to see some anecdotal studies, besides Chile, that maybe supported your case.  Thank you.

 Ann Owen:  Yeah.  We didn't go that route, in the sense that that -- The method that we chose was to look on average, you know, what's going on?  Because I'm a little bit cautious about going that way.  In the sense that we can provide an anecdote that's consistent, I think we can also find ones that aren't.  I think the story that you're telling with Chile is consistent overall with these results, that it's not just the volume of trade that's increasing, but also some of these broader measures of overall policy that are related.  So I think that's consistent with the broad findings that we have.

 Roger Bate:  The next question.  No questions.  Richard?

 Richard Tren:  Thanks.  I'm Richard Tren with the group, Africa Fighting Malaria.  Much of the discussion over the last few years on international trade and health has been obsessed with TRIPS and drug patents.  I was wondering if you have any comment on that.  If you looked at all of the impact of intellectual property and health in your study, or if you have any commentary on it?

 Ann Owen:  We didn't look at that.  We did look at the importing these medical resources.  I guess I'm not sure how that kind of thing would show up in the data.  But no, we didn't look at like the laws about drug patents in these individual countries and how that might be related to trade.

 Roger Bate:  Further questions?  I find it quite surprising for some of the countries, or at least it maybe just in aggregate, it's surprising.  Given the kind of modern transportation, just as how beneficial globalization, if you like, has been.  Because modern transportation does massively increase the transmission of pathogens.  We've seen this with West Nile Virus arriving in the United States; dengue fever being transferred around the planet, again by people traveling in airplanes.  The robustness of your paper, I think, is because of all of the various variables that you look at, I think is startling. 

 I'm hoping we're going to get a few more questions from people because otherwise we can wrap up relatively soon.  We'll move into that.  But I was wondering if you could, given that you're the expert, that you did the study.  I mean, we have already alluded to some micro studies that could be done.  I can think of at least four Ph.D. research projects that could come out of looking at micro studies.  The gentleman from Duke has pointed to one area.  You could look at an individual country to see if it conforms to this or where it doesn't.  But I was wondering if you could elaborate on a few of the individual projects that have maybe sparked in your mind from this?

 Ann Owen:  I think that this paper, when you read it, in some sense, it's a paper about what it's not.  Right?  We find a correlation and then we rule out, "Well, it's not this.  It's not this.  It's not this.  It's not this."  So I think that maybe where the interesting question is, is, "Well, what is it?"  Right?  We're working with aggregate data.  It's an average effect, and there's going to be things that are going against it and things that are going for it.  Right?  So yes, now we're more open.  We're interacting with people, you know, SARS gets transmitted.  Right?  So all those things happen. 

On average, we're finding that the effect is positive, so on net, we're finding the effect is positive.  But I guess some of the questions that I think are still out there, that would help make this I think even more convincing.  If we could now identify, "Oh, this is what's happening."  Right?  "This is how knowledge is being spilled over."  And maybe the way to do that is with case studies.  Right?  It's to say, "Oh, in this country, after they opened up to trade, here's now what happened."  I think that's the direction that I would like to go in, in a sense.  I think that this paper, it's good for what it is.  It shows there's this correlation.  It shows it's there, but a lot of the paper is about what it's not.  I guess the next paper should be about what it is.

 Roger Bate:  The lady in the back?

 Female Voice:  I had a question, and part of my thinking, I’m thinking maybe this one of those things where the economy opens up to trade and then in addition, some of these other factors.  It has to do with   I work for a USAID contractor in Egypt.  I was actually the trailing spouse of a foreign service officer.  I was working for a consulting firm.  We initially were working actually on water treatment and waste water treatment.  Then we got another project, which was women's education, which is strongly correlated with increase in income in a society. 

So I'm wondering if you happen to have looked into some of the, when free trade occurs, whether the education of women in some of these countries, becomes a large priority simultaneously.  And I guess it's a similar question to what this Duke fellow raised, about how do you figure out if it was just the economy becoming more open and more forward-thinking, or whether it was the trade?

 Ann Owen:  Right.  Well, we actually did look specifically at women's education in the sense that we looked at female secondary school enrollment rates.  And you do see a positive effect on health of that, in particular of infant mortality.  And other people have found this, so that's not necessarily a new result.  I think that it could be related to trade.  There's other people that have done some work on women in the labor market and open economies.  The idea is that when countries open up to trade, they have to be more competitive.  And because they have to be more competitive, they have to use all their resources efficiently, including women.  So that there's some work -- the name right now escapes me -- that has shown that when countries, more open economies have better opportunities for women?  Their logic is exactly that, that more competitive economies, they're only going to be competitive if they're using all their resources efficiently.  So I think there is. 

There are independent studies that kind of confirm what you're thinking about.  If you have to then say, "Oh, as a policymaker now, what am I going to do?  Am I going to do water resources aid or am I going to do female education?"  That again, that's going to be an empirical question about, "Well, which is going to give you the biggest bang for the buck?"  I don't know the answer to that.  But that's a great question to ask.

 Roger Bate:  Bob?

 Robert Helms:  There is a question here, but I wanted to see if I could get somebody to talk about your health measurements a little bit.  You've used infant mortality, which you have statistics for from the World Bank.  Of course those get criticized a lot of times as being very inconsistent from one country to another.  But I wonder if you could say something about what the variable "life expectancy" means?  My concern is a lot of people sort of hear that term and they say, "Well, somebody's going out and doing a public opinion poll." You know?  But that's not what it is.  It's based on, as I understand it, hard demographic numbers, and then projected out in terms of what you would expect, based on the death rates and how long people are living and that sort of thing.  It's based on sort of hard numbers.  I'm sure some of you know a lot more about that than I do.

 Then my other question is, have you looked at some of the other WHO outcome measurements having to do with disability-adjusted life, your dailies and some of those?  Is that just the case?  They're used in developed country studies, but is it you just don't have those for the underdeveloped countries, or just don't think they're good enough yet or what?

 Ann Owen:  So the first question about life expectancy, what does it mean?  It's not an opinion.  It's based on the current conditions in the country; how long people are living now.  So it's how long you would expect somebody born, like in this year, to live.  Not based on opinion, but really based on the demographics of that country. 

 We didn't actually look at other health outcomes.  There is an issue, I think, with any of this kind of cross-country data, with data quality and also data availability.  But that would be a great way to actually also kind of extend this, what we're doing, and to see whether or not you find the same effects.  We should.  If our conclusions are correct, we should find the same thing.  But that would be a good idea.

 Roger Bate:  The next question, the gentleman at the back.  And from my limited input to this, I was at a Center for Global Development working group on health indicators.  I was not an expert really.  I joined into it and learned a lot.  The data in many countries, which is really far too poor to do any kind of daily calculations.  I mean, maybe 20 years from now.  But at the moment, you have to work on very crude basic indicators to get any kind of measurement.

 Marco Mostolleri [phonetic]:  Thank you.  My name is Marco Mostolleri from the Embassy of Panama.  My question is in regard to the knowledge spillover, vis-à-vis government policies.  The recent development and slight improvement in the health care in developing countries, particularly in Latin America, and specifically in the transfer of technology and knowledge.  With this, I mean the affiliation of national medical institutions with foreign medical instructions.  What do you think are stronger inputs into better health care?  Would it be stronger institutions and a bias to the rule of law, or trade openness?

 Ann Owen:  That's a good question.  I think that what our study is able to say about -- Because we're looking at five-year intervals?  My guess is that when you're looking at a short time period, that it's actually going to be the interaction and the volume that's going to dominate institutions.  When you're looking at a longer time period, probably the institutions are going to play a more important role. 

There's a paper by Dollar and Kraay, a couple of years ago    2002, I think -- where they look at this exact question.  But instead of looking at health outcomes, they're looking at income growth.  That's the conclusion that they arrive at.  That if we look at short intervals, then it's trade that's really causing these effects.  But when they look at longer time periods, they're drawn more to the institutional measure.  So I think that they don't have to be  - I don't think those stories are mutually exclusive.  They can both be working together, but maybe they're working together in different timeframes.  I think the timeframe of our study is more consistent with the shorter time period.

 Roger Bate:  Bill?

 Bill Levy:  Bill Levy, American Enterprise Institute.  I want to commend you on the paper.  This is very interesting and it looks very carefully done.  Given all the interest in trying to find case studies and antidotes, to back up, you mentioned that when you were doing your robustness checks, you excluded outliers.  That might be a very interesting way to go forward.  It's a good thing to check.  You confirm that this holds and it's not just one or two countries driving this.  But then when one's looking for, which are the case studies that might be the most fruitful?  Identifying those outliers would be very interesting.  I don't know if you have that off the top of your head, but otherwise it would be an interesting -- No?

 Ann Owen:  I don't know what those countries are off the top of my head, but that's a great idea.

 Roger Bate:  The gentleman in the back?

 Brockwel Heylin:  I'm Brock Heylin with Civic Service.  I just wondered if there was any link between a country's tolerance or intolerance of corruption in the health system and health outcomes?

 Ann Owen:  The paper that we did, didn't look at that.  The only measures that we have that are probably related to corruption are these kind of general policy measures, like this black market premium and the Sachs-Warner index.  But we didn't directly look for corruption in the health sector.  But I think that's a good point.

 Roger Bate:  You could look at the work of Maureen Lewis.  She did a chapter within the Transparency International's 2006,  as Paul Wolfowitz mentioned, the 2006 Report.  Also, she's done several papers, which will be published in journals.  Also I think you can pull one down from the Center for Global Development site, which discuses a lot of detail about corruption.  It doesn't specifically answer your question, but I imagine in the references, that you will find a good starting point for the literature.  I haven't seen, myself, much data on that association.  In terms of anecdotal stories about how bad corruption may be in particular countries, there is a lot of that in those papers, and that'll probably drive in some direction, I would think.  And the gentleman from Duke again?

 Peter Whitney:  If I may be permitted a second question.  Several things you all have said seem to tie together.  Maybe it suggests a possible paper.  Paul Wolfowitz said that, Roger, you've done a study on tariffs and corruption.  Certainly in the countries that I've worked in, quite a few countries, corruption was most manifest at an import policy.  Whether it was import licenses; corruption at the ports.  I could think of a number of countries -- Argentina, Mexico, Brazil -- where I've worked, where the corruption in the trade policy was tied to the protectionist policies.  So you might have something to answer that gentleman's question.

 And regarding Jerry's paper, in Colombia, Colombia -- along with Chile -- has the best reputation over the long haul for being the most ethical country.  They were the only two countries in the 1980's that continued to pay interest on their debt.  They postponed principal.  Believe me, the banks in New York know that very, very well and always have treated them with a special respect because they were ahead in that, in terms of economic thinking.  So it seems to me, you might have a whole bunch of ideas that could be tied together in a subsequent paper.  Any comments?

 Ann Owen:  Yeah.  I mean, that's a good thing to look at.  To the extent that things like Export Marketing Boards are related to corruption, then to some extent that's kind of in our study, but it's not there directly.

 Paul Wolfowitz:  Just to pick -- Corruption is a very precise term actually when you get into it, and it's a problem.  I suppose in a broader sense, it's abuse of power rather than stealing.  At the risk of losing any future donations I'm going to get from Carlos Slim, I think the way that he's made a fortune in Mexico at the expense of Mexican consumers is an abuse of power.  A lot of what I saw in Indonesia was perfectly legal, but it was legally created monopolies. 

Then you can get into the question of how they came about.  That’s why trade policy can be, in fact, such a politically biased instrument.  As I tried to say, it's not biased for the weak, it's biased toward the powerful.  It seems to me if one's going to look at the questions you raise, on the one hand, it will be useful to be precise about corruption, which I admit I wasn't.  But at the same time, not to be so narrow on how you define it, that you exclude what may, in fact, be the most pernicious effects.

  Jeremiah Norris:  I would like to go back to a question that Richard had asked, because I think we haven't addressed it fully, and ask Ann if she would comment.  It perhaps is a subject of future studies, Ann.  But Richard's point is that the TRIPS agreement, the World Health Organization and the World Bank contend that by subordinating intellectual property rights, our countries could produce their own drugs and this would give more people access to better health care at cheaper prices.  It's a contention rather than empirical study.  But it is being pushed by several countries today. 

Thailand is now doing what they call compulsory licenses, not only for AIDS drugs but also for cancer and cardiovascular diseases, contending again, that they can produce it without having to worry about patent rights, on the patent holder.  More people will have access at a cheaper price.  But the drugs they produce, especially in Thailand, are so bad that even the World Health Organization refuses to pre-qualify them, yet they're being used.  So it may have a negative effect, rather than a positive effect.  Thank you.

 Ann Owen:  Yeah.  We didn't look at that at all, so I have no evidence one way or the other about that.  But it seems clear, like it is an empirical question.  Right?  You can make the argument either way.  It's good; it's bad.  It seems like that's just a study waiting to be done.

 Philip Coticelli:  Philip Coticelli, Africa Fighting Malaria.  I'd like to know, Professor, on what feedback you've gotten, if any, on the paper thus far?  Particularly from globalization skeptics.  I don't know how many that you're in correspondence with. 

 I also had a question for Roger.  I don't imagine you sent it to Joe Stiglitz.  If they were to see the paper and comment freely, what do you anticipate they would say?  In the same turn, how would you recommend that health advocates -- like Richard and myself -- play the strengths of this paper to a community who sees malaria control as an inherent failure of the market and something that is very much a public good and a public health imperative?

 Ann Owen:  Actually most of the people that have actually contacted me have actually said mostly positive things about the paper. 

 Roger Bate:  It's early days.

 Sue Owen:  Yeah.  It's early.  Right.  We've had a few people actually, since the AEI sent out their announcement and asked us for the data.  So I suspect that criticism is coming.  I think that the criticism that you can throw at any macro economic analysis with the data as it exists, is, "This isn’t really true.  You're picking up something else.  There's an omitted variable out there that is really the cause and it's just related to trade."  So there's this third thing.  I guess we tried to think about what that was.  We tried really hard to think about what that was.  That's why there's so many different things that we do in this paper.  I guess maybe if there's any critics, you can send this criticism to me later today.  But that would be, I think, the criticism is, "Well, it's not really this.  It must be something else."  And then we can go and get out, and say, "Oh, well, is it that or not?"  Right?

 Gee, I don't know.  I'm not really sure that I know enough about what the objections are of this particular program to say how you could use this.  I think that the overall conclusion is that trade is -- I mean, it's certainly not harming the health outcomes.  That does not mean that there's not a particular situation and a particular country, and a particular time period, where you can't find evidence of the opposite.  Right?  "The results are, on average, in this group of countries, this is what we find." 

 Roger Bate:  I did not send it to Joe Stiglitz.  I suggest that Paul Wolfowitz send it to Joe Stiglitz, because I think you probably have a much better chance of correspondence with him than I will.  I would just echo those points.  One of the reasons I wanted to hold this event, I think the paper is very, very interesting.  As in my introduction, an hour and a half ago, I do find it very paradoxical that we have presidential candidates that are least -- I would say all three, frankly, who are not vehemently free trade.  As we have seen over the last 10-15 years, an increasing rhetorical support for protectionism, and yet at the same time, massive increases and a desire to spend foreign aid on health.  To me, the results of this paper show that there should be some stepping back.

 Now there are obvious vested interest as to why people don't want to make those assessments and don’t want to deal with that paradox, but I think it is extremely interesting.  There are obvious specific examples, and I think Ann has explained it better than I probably can.  It will be easy to criticize the paper based on the data, especially the health outcome data, which you have to go with the best data you have and you've used the best data, from my reading of the paper.  I read it carefully, because I reviewed for the next month's Foreign Policy Magazine.  Hopefully, you'll start getting more critics coming through to read the paper.  I looked at it very carefully.  You would no doubt be criticized for the data.

 I think that you will find anecdotal examples of countries, which do not comport with this.  It's very easy to be critical.  What is exciting about this, and why I hope, as I said, the event today, and why I hope we can generate some interest around this topic, is that it's quite likely -- at least from these data -- that it's quite possible if the U.S. government had done a better job on the farm bill and some other areas of protectionism, it could have done a lot more than a lot of the aid packages.  To me, that's an interesting discussion, and one that frankly I don't see getting -- It is not questions that would be asked of Senators Obama, McCain and Clinton.  But in fact it should be, as far as I’m concerned, which is why I was trying to elevate the discussion of the paper today.  Do we have any final, final questions?

 Jeremiah Norris:  If I could just make a comment here?

 Roger Bate:  Sure.

 Paul Wolfowitz:  I presume you people might criticize the quality of the data, but it seems to me all that’s actually does is say, "Well, maybe the conclusion isn't as strong as it might appear to be."  But unless someone has an opposite conclusion based on some other data, at least -- It seems to me the going in proposition is that this extreme criticism of trade is hurting the poor is, at least in this case, is very much open to question. 

 And to some extent, also for the anecdotal examples, I wouldn't be surprised, and I'd be very interested.  There are probably some countries that are fairly protectionist that maybe, because they do half the Chilean policy and not other half.  I mean, there are these interesting examples in South Asia, both in parts of India and Bangladesh, where you get health outcomes that are not correlated to what we think of as sort of the usual development indicators.  And that good policy probably can, in certain instances, have a powerful effect in spite of whatever the trade environment is.  But then it seems to me, you've got to start getting precise about that and not make a generalization that because maybe in South India they have these results, that therefore that is a justification for all the protectionism that takes place in the rest of India or in Africa.

 Jeremiah Norris:  And this is an issue maybe for your future work.  The three largest health programs being funded over the past five years, and certainly into the next five, are AIDS, TB and malaria.  Normally in medical care, the outcome is validated in the presence of a medical record.  There are very few medical records being kept on this treatment for AIDS, TB and malaria.  So what we have is reported outcome is basically the disbursement of the money, rather than the measurement of say the viral suppression rate over time.  This is going to corrupt the data to a very large extent.  It will be difficult to determine the relationship between the money going into health and the reported outcome, because the outcome is basically just the numbers being treated without any reference to what's happening to the patient.  Thank you.

 Roger Bate:  The metric is always a problem.  I recall reading a paper that said that until, I think 1974, all deaths in France were from heart failure, because that was what was recorded on most of the death certificates.  In France, until they can get this right, we're going to continue to have very poor data, I think, internationally for a long time to come.  My criticism of the data, it's none of your fault.  You took the best that you could find.  Do we have any final questions?  If not, I'm going to wrap this up.  Okay.  Thank you very much indeed for coming.

[End of file]

[End of Transcript]