 All right. Good evening, everyone. I'm pleased you could make it. Former UK Prime Minister Margaret Thatcher once said, you and I come by road or rail economists travel on infrastructure. So it is a great pleasure to be here today and to welcome Pierre Richard Aginor. Infrastructure, as we all know, whether it's energy or water, and telecommunications, it touches our lives every day. And well-designed infrastructure facilitates economies of scale, returns to trade, reduces costs of trade, and is therefore central to the specialization of modern economies and the efficient production. No doubt it is a vital ingredient to economic growth and development. Development for developing countries, poor countries, and it is absolutely key to raising living standards. Unfortunately, many issues surrounding the economics of infrastructure are still not very well understood. Enter Pierre Richard Aginor, who for many years now has been at the forefront of putting some clarity into the analysis. So it's a pleasure to welcome you, Pierre Richard. You will today provide an overview of your new book called Public Capital Growth and Welfare, which I've just learned has been endorsed by luminaries such as Jeff Sachs, Kostas Azariades, and Stephen Turnofsky. I think that says it all. The book addresses a series of issues pertaining to the impact of infrastructure on growth, education, health, innovation, and its implications, in particular for poverty traps, gender bias, and public debt dynamics, all highly relevant and important topics. Professor Aginor is the Hullsworth Professor of International Macroeconomics and Development Economics and co-director of the Center for Growth and Business Cycle Research at the University of Manchester. In the previous life, he was lead economist and director of the macroeconomics and policy assessment skills program at the World Bank and also at the International Monetary Fund. He also recently joined as a research associate at the Center for Applied Macroeconomic Analysis here at A&U, so we hope to forge and maintain closer ties in the future. He has visited and taught in countless academic departments, has given lectures, seminars, and short courses in universities and research centers in more than 40 countries. However, this is the first time in Australia, so you can add one more country to your list, and we hope it won't be the last time that you will grace us with your presence. Apart from the sheer number and volume of his articles and books and monographs, what I find truly amazing is the number of self-disciplines of macroeconomics that Pierre Richard straddles and specializes in development macroeconomics, monetary policy, international finance, labor economics, macroeconomics of poverty reduction. And the list pretty much goes on, so it's a rare breed of macroeconomists who seems to effortlessly straddle all these different self-disciplines. He holds a PhD in mathematical economics and econometrics from the University of Paris. I think that's all I have to say at the moment. Now the floor is yours. You will present for about half an hour, I understand, then we'll open it up to questions and answers, and then we can convene for some drinks afterwards. Thank you. Thank you, Timo, for this extensive presentation. Yes, this is my first visit here, but no regret, wonderful weather compared to where I was. I'm going to talk about public capital, but essentially public capital means core infrastructure here. Economists like you used to use the term public capital, and in general public capital could be brought, of course, through assets such as hospitals, schools, etc., but we're going to use a more narrow definition of public capital, central infrastructure, core infrastructure, energy, telecommunications, transportation, and water and sanitation. And as Timo said, the book is sort of an overview of my research over the past few years, but it's more than that. It's taking a series of topics that have been addressed in the literature before many of them, and try to provide a common thread, essentially an analytical framework that can be adapted to answer various questions. So the first chapter of the book starts with a workhorse model, if you want, and the model is adapted in subsequent chapters to address different issues as we will see, gender, innovation, etc. So what I'm going to do is not so much talk about the analytical framework itself. I don't want to get into any technical details. I don't think it's necessary, actually. I'm going to essentially give you an idea of the different channels that are discussed in the book and what are the policy implications. A broad policy implications, of course, given that this is not designed for a specific country, but I think that there are general lessons that can be drawn from this line of work. And I still have to write down one equation, though, because it's quite important. And this is the equation that economists typically use to define a public capital stock. So if we denote K as the capital stock, T plus 1 is time, so what this is saying is that the stock of capital at period T plus 1, I think I scared somebody already with that question, but the stock of capital at period T plus 1 is equal to the stock of capital in the previous period that's KT, multiplied by 1 minus a coefficient that we call delta. Delta is simply the depreciation rate and plus I, which is investment I at period T. So the capital stock at T plus 1 is the capital stock that you had at period T adjusted for depreciation plus the flow of investment. So this is the standard formula that we use. Well, to better think about what the difference between stock and flows, imagine that you multiply the coefficient here of the flow investment, the variable IIT, by a coefficient that we call alpha and alpha is between 0 and 1. And it's convenient to think of alpha as a measure of efficiency of investment or the quality of governance in general. What it means essentially when alpha is less than 1, it means that only a fraction of investment actually turns into capital. To give you a preview, empirical estimates of alpha range in recent study conducted by the IMF and the World Bank between 0.3 and 0.6. If you take an average value of that, it means that typically more than 50% of investment actually does not turn into capital. Whether it's waste, whether it's corruption, whatever it is, it's just a common fact and it does not depend on the level of income. You don't necessarily see a smaller number or a higher number rather in richer countries. So that's the only equation that you're going to see and it's very helpful because it allows us to make a distinction between flows and stocks even if you invest a lot. If alpha is very small, the capital that you're creating as a result is going to be relatively small. So that's the sense in which what matters in this analysis is not the flow of investment itself. It's the stock and more precisely the flow of services that you can derive from the stock of public capital. That's it in terms of equation. So let me first address the issue of why economists typically think about when you ask a question what does public capital do for growth? And typically the answer is something like this. The first point is that there is a distinction between flows and stocks. That's what we just saw. That's the efficiency factor that you saw earlier. The main effect, as Timo was saying earlier when we think about infrastructure is the impact on productivity and production costs. If flows are in bad shape, typically it's going to cost you more to operate trucks, to move goods around the country. Typically public capital being available in greater supply would mean that it has a positive effect on the productivity of private inputs, whether it's labor or capital. So that's the conventional view of looking at public capital. The second effect is a so-called complementarity effect. And it is the idea that if you are in an environment where public infrastructure is relatively good, you tend to have higher rates of private investment. Poor countries typically have where public capital stocks are relatively low have enormous problems in raising private rates of investment because you cannot really operate the firms in an environment where you don't have access to core infrastructure services. So the complementarity effect is actually quite important. It means that when you invest in infrastructure, it may have actually a benefit in terms of leading to higher private investment as well. Now, both effects are positive. The third effect that economists would typically talk about is a so-called crowding-out effect. It's essentially the idea that if, in order to invest, you have to raise taxes or you have to borrow from banks or you have to issue debt in order to finance that increase in public capital, so typically what you are going to see is that it's going to be more costly for the private sector to invest. So if you think of a given amount of resources available for investment, if the government uses more of it to invest in infrastructure, there will be less resources available for the private sector or it's going to cost the private sector more in order to borrow. So in a sense, if the crowding-out effect is very strong, it could even counteract the productivity and complementarity effects that you see here. These are the typical effects that are well known if you want for economists. The first chapter of the book actually goes through the evidence for all three actuals, how all three effects can be accounted for in the model. But most of the book actually is not about those conventional effects. It is about what I call here the new channels and it's important. The one thing that I do is actually take a lot of the literature that is available in professional journals, scattered in professional journals and essentially bring it together in a coherent manner. And there are also a number of publications by international organizations, especially the World Bank and the UN. Now the point is, it's not that all these channels were suddenly discovered. In fact, many of the channels that I'm going to talk about have been well documented for years. And if you think about two of the important channels that I'm going to describe, health and education, if you talk to experts in health and education, they've been aware of the infrastructure aspect of their work for a long time. But the problem is that at the macro level, there has been very little work on integrating those things into the type of macro models that economists use. So when you talk to health or education experts, you're very well aware of these things, but people working on infrastructure and macroeconomies, whose job it is to think about how to allocate resources at the national level, are actually not taking all those aspects into account. And I must say, myself, when I started to think about these issues, I started to look at them because I was working on poverty issues. And when I started to look at education and health and the role of public capital, I must say that it was very difficult for me to actually get into that literature because it's a completely separate literature, and macroeconomies typically would not go into that sort of professional journals and publications where these issues were raised. So one aspect of the book is actually service to both microeconomists and macroeconomies by trying to reconcile these two strengths of the literature. So let's see what these channels are. Well, the first is, again, we already talked about efficiency. Efficiency in the more recent literature has been related more explicitly to corruption, poor governance in general. There are now very explicit models that allow us to understand better the determinants of efficiency and how to actually improve efficiency. But the more important ones in terms of new channels, if you want, is a rediscovery in a way of the importance of maintenance. Not so much for maintaining the quality of public assets per se, but the recognition that poor quality of public assets means actually extra costs for the private sector as well. And there are calculations showing that for every dollar that is not spent in maintenance of public assets, it does have a significant cost in terms of depreciation of private sector assets. There are studies for developing countries, quite a few, and there are also studies for the U.S. where they can calculate how much it costs an individual, on average, for maintenance of a car as a result of poor roads. But more importantly for our purpose here is the discovery that public capital matters a lot, actually, when it comes to the production of education services and the production of health services. In fact, it's none of the production but also the delivery of these services. And if you take some basic examples, if you want, what electricity is essential for both schools and hospitals to work, to operate, you can, if you have vaccines, they need to be refrigerated. If you don't have electricity, then this is hard to do. Roads are very important because in many countries, many developing countries, access to medical facilities is hampered by the fact that people have to walk several miles to get access to a medical facility. And of course, as you may know, sad stories of mothers dying on the side of the road because they don't have the time to reach a medical facility in order to deliver their babies. And roads are also important not only to get to those facilities but also to attract medical workers and teachers in particularly rural areas without access to roads. Many, it's difficult to attract qualified doctors into rural areas because people are simply not willing to live permanently in those areas. And water and sanitation are very important in schools. There are studies showing that in countries where schools have access to sanitation enrollment rates for girls increase very significantly. So the book actually goes through a series of examples of that sort. And of course, in terms of water and sanitation, I should add that there are very... There's a wide range of studies showing the impact of access to clean water on malnutrition and infant mortality. So all these things are well documented, but when you put them at the macro level the key concept then becomes the concept of effective labor. Effective labor is the view that what matters for production is not only raw labor, the amount of time basically that you spend into production, it's going to depend on the quality of education and on the quality of the labor force. It's actually the health status of workers. So effective labor is a concept that in a sense summarizes all those three aspects. And clearly public capital is going to matter to the extent that it affects effective labor through the production of education and health services and eventually it's going to matter for market production. There are other aspects as well and directly related to the production of health services is the fact that the production helps your people tend to think more about the longer run prospects in principle. So it does affect the rate of time preference, whether or not you want to... You're going to save, whether or not you want to invest in higher education if your horizon, time horizon, tends to increase then that tends to affect your decisions to save and consume. And those decisions to save and consume of course are going to affect your propensity to invest in physical and human capital. If life expectancy, the typical response when life expectancy increases is for people to save more. Either through personal accounts of your private pension system but typically people would tend to save more. And as a result, more savings in the economy to the extent that it translates into high investment is going to affect private capital and therefore market production but also the decision to invest in human capital which in turn of course is going to affect the production of education services. So these are really some of the key channels that matter. There are a couple more actually. One is so-called network effects. Network effect is not an independent effect. It's really a size effect and to give you a simple view of what the network effect is think of the stock of public capital here as the number of road segments and on the vertical axis here you have what I call the efficiency of public capital. And if you're building, you're starting to build a network of roads. If you're building roads between say points A and B and points C and D that's good for trade and business between those separate entities. But the moment you link B and C, you create the opportunity for trade between A and C and B and D and A and D. So in other words, by adding one segment to the network you've actually multiplied considerably the opportunities to trade and exchanges between those different points. So one way of capturing that is to simply assume that up to a certain point when you are building infrastructure up to a certain point you have some efficiency gains but those efficiency gains are constant. Beyond a certain point here between points A and B you see that the curve is increasing pretty fast. That's the idea that you've reached a stage where adding the marginal gain from adding one more connection to the network is going to expand considerably interactions and therefore increase productivity of the existing capital. And beyond a certain point, typically what you would have is that beyond point B, if you continue to add roads basically it's going to make a difference but not as much. A very good example for industrial countries is the completion of the highway system in the U.S. where it's well documented that this led to a very significant increase in interstate trade. But beyond that, if you are adding more, the additions of additional roads if you want the highways did not have the same marginal benefit. It's still positive but not as strong. So that's one way of interpreting network effects and it tends to be quite significant if you're in a poor country starting with basically no roads. You don't gain much by adding a single road. You're going to gain a lot by connecting some important points and that's a consideration when you're devising a program of investment in transportation. A few other aspects if you want to be trans-structured that I did not capture in that graph. One is an impact on innovation. And that's quite significant because this refers to both the ability to innovate and the diffusion of innovation. And the key issue is really to what extent public capital matters when you are developing a growth strategy based on imitation. Essentially for a developing country, you're importing blueprints from the rest of the world and you're simply replicating those. So in a way what you're doing is imitation, which is a process through which many currently middle-income countries have gone through already. But imitation can take you only so far. At some point the critical issue is to go from imitation to what you may call true innovation. The difference is of course that true innovation is more demanding in terms of skills and also it turns out in terms of infrastructure. And the key issue here is access to broadband. And the study is showing that broadband begins to make a big difference in terms of growth when it reaches 8, 10% per year first human rights. So in that sense, why is broadband so important? Broadband is important because it allows fast access to information and the diffusion of ideas. And again, there are several studies looking at those issues. The key point here is that the type of infrastructure that you need to be successful with an imitation strategy is not the same type of infrastructure that you need to be successful for true innovation. This is actually one of the conclusions of a recent report on innovation by the OECD among other institutions. Now there's also an impact of infrastructure on income distribution. And there are a couple of studies showing that improved access to infrastructure may actually reduce inequality. Well, several reasons. One of them is that improved access benefits are poor more than the rich. And if you think of access to clean water, for instance, you might say that the marginal benefit for the poor can be considerably higher than it is for people who already have basic access. So if inequality is bad for growth, then you might say that if you're reducing inequality, inequality infrastructure can contribute positively to growth through that channel. My own take is that when you look at that part of the literature, it's not as strong as some of the rest because causality could go both ways. It could be that improved access to clean water gets you to be healthier and that might be in better health you actually work more. And therefore your productivity and wages improve. That's one way to look at it. The other way to look at it is that people who work more have higher wages are able to build themselves of cleaner sources of access to water. So causality is not quite clearly established. It's my own judgment on that point. But you have to keep in mind that there is a correlation between the two. Another important source of impact, if you want, of infrastructure is the impact on women's time allocation. Now I like what I mentioned is I like to say that it's really an important issue for developing countries but usually the audience says, well, it's very important for industrial countries as well. If you think of Germany in particular. And the idea is that women in developing countries can spend up to two hours or three hours a day collecting wood and water. And those constraints, those activities impose significant constraints on their time. And what improved access to infrastructure would do, of course, is to reallocate time to other activities. So let's take a more detailed view of this point because I think it is quite important. Let's think of women allocating their time to three activities. To home production. And home production is what I just said. Getting water, cooking, et cetera. All activities related to the home. There's also time that they allocate to their own health and time allocated to the child's delivery. So in that sense, time allocated to market activity is the residual. Given the total amount of time, you can calculate as a residual the time allocated to market production. I'm excluding leisure here. I assume that it's not an option for many women. Time allocated to market production is the residual. And clearly time allocated to market production has a benefit in terms of market activity. But time allocated to own health obviously affect women's health and their productivity and eventually market production as well. Time allocated to child rearing also has an impact on children's health and education. So think of child rearing essentially as taking not only cooking for children who are rather, which is in the first category, but rather taking children to medical facilities, et cetera. So that link essentially creates an intergenerational link between mothers and children. And that's the link between mothers' health and children. And you also have an intergenerational link which means that if children are healthy or in better health in childhood, eventually they will be more productive when they become adults. And that also has a benefit in terms of market activity. So this is important because when you think about women's time, very often in economic models, the time allocated to home production and time allocated to child rearing is usually viewed as non-productive. But clearly that is not... If you think about child rearing, if it helps to improve child health and education, this would have significant effect in terms of market production. And the implication is quite important. When you look at the impact on growth of changes in access to infrastructure operating through this channel, it does not really matter that at the end of the day women end up increasing the time spent in market production. It could be that there is an indirect effect going through reallocation of time toward children. It's another feedback effect that is quite significant. Out of market production women generate wage income like men and within now family resources what we can see is changes in intra-household bubble. There are studies showing that in households where women have high levels of education their bargaining power in terms of allocating resources tend to be larger. I'm just stating a fact. I'm not totally convinced. But the study is showing that this is indeed a possible challenge. So what may happen indeed is that as women's time is reallocated to their own health, productivity improving, market activity generating high income you're going to see a reallocation of resources toward typically items that matter more for women which is again based on documented studies women tend to be more concerned about the health and education of their children. I'm talking about developing countries. So that's another possible channel through which at the end of the day infrastructure can matter. So providing better access to infrastructure means that time allocated to home production is going to fall. Then the issue is how women reallocate their time. Whether they reallocate it to their own health or to child rearing, you can have benefits at the end of the day in terms of market production and growth. Now this is a critical agenda. In fact, right now the World Bank they have asked all their economists any report, any country report has to have some discussion of the gender dimension of their economic programs if you want. And you can actually take all this and turn it into an actual model that you can use in order to do this kind of analysis. So the argument that gender is not important when it comes to growth or gender cannot really be put into the type of economic models that macroeconomists like to use is not true. This can be done and you can do it in a way that captures some of these elements. And you can make of course this graph a lot more complicated than it looks right now. So all I've said, as if it wasn't already complicated, but all I've said so far is really good things about infrastructure. And obviously there are other potential negative effects of what we call here negative externalities which are of course related to environmental damage and pollution. And those could create negative effects on growth when in terms of environmental degradation that would be a direct loss of physical assets to be involved. And that would certainly have an impact on production. But the effect could be also indirect when you think about the impact of pollution on health and productivity. If you think about the diagram that we showed earlier if life expectancy is adversely affected this could have a negative effect on savings and this could translate into an adverse effect on growth. Now does that imply that you have to completely disregard a few other externalities that we discussed? No. What it means simply is that when it comes to thinking about infrastructure you have to in a sense account or internalize those negative effects. So if there is a risk of pollution or the risk of environmental damage this has to be accounted for in the sense that the allocation of resources has to take those negative effects into account. It may well lead to even to a conclusion that the cost in terms of environmental damage or pollution may be so high that the investment is not worth it. But the reality for many developing countries is that growth is such an overwhelming objective environmental damage and environmental pollution unfortunately are not always given the consideration that they deserve when discussing large infrastructure projects. So to summarize if you want but in the form of a few policy implications the first is I think it was clear from the beginning the flow of investment is not a very good proxy for the accumulation of public productive assets. Remember the alpha that we talked about earlier it can be very significantly less than one for some countries it's only three. So even if you think about the stock the stock can have benefits you can have the flow having a negative effect a crowding out effect and this distinction between stock and flows is extremely important when you think about the sustainability of public debt. That's a topic that I'm going to address tomorrow in a seminar but the idea is simply that when you invest and you're issuing debt to finance the investment the debt obviously has to be repaid with interest and well depends on which debt you're talking about but in principle the debt has to be repaid and what it means essentially that you have a by issuing the debt you're in a way diverting savings from private capital formation to the benefit of the government so on the one hand you have a positive effect of the public investment for all the reasons that we discussed earlier but on the other you may have a negative effect coming from the fact that the government using more of the existing pool of savings it means that the private sector has less of it so the issue is in order to have a sustainable public debt the benefit that you get from the debt has to be in a sense large enough to compensate for the fact that you're going to get lower private capital accumulation the second is that I did not quite address that here but it's important to account for the quality when it comes to stocks of infrastructure whether you're looking at the benefits in general if you're building roads one thing that I've seen in developing countries very often when you start building roads the number of users increases much faster than the ability to expand if you want the road network so what you get is congestion problems and those congestion problems are very significant simply because they reduce the benefit of having the roads of the infrastructure asset in the first place these points if you want both suggest that when you think about scaling up public investment increasing, massive increase in public investment an issue that has been at the forefront of the international debate on development and how to promote growth in pool countries well it's not again simply about spending more it is about improving your ability to select, implement and monitor investment projects so what you're really talking about here is you need to address not only the issue of the level of spending but also the institutional framework within which investment is going high investment is going to take place third point which I think is clear by now is that beyond the productivity and cost effects that you typically hear when people talk about infrastructure projects whether it's a high speed rail or a new road or any of that sort it's not only about productivity and cost you have a range of other potential benefits that you need to consider and in many countries what you see is that very often when you build building a new road actually attracts a number of activities and beyond those activities you have benefits in terms of the things that I've talked about in terms of health, in terms of education which can be equally significant and network externalities it's important to account for them as we said earlier and when you talk about that an investment program well if you're going to build a high speed rail system in a way it's a system you can't limit yourself to just doing a small segment you've got to think big in those things so scale matters essentially when it comes to infrastructure if you want really to reap the benefits of the investment fourth is actually more of a call to macroeconomists to really start introducing those channels into their models I'm quite amazed when I go to some places and I look at what type of macroeconomic models that people continue to use when they try to introduce infrastructure where all the benefits or all the effects that have been captured are those that we discussed earlier impact on complementarity on the crawling out effect it's all good but what this is showing is that the human development aspects can be equally if not more important so there is a challenge here for macroeconomists to think in broader terms not only about the growth per se but also try to integrate human development indicators I've been involved in some of these working and it's actually doable you can link models with a range of human development indicators because some of these indicators are related to variables that you can capture in macro models and let's be provocative and simply say that when you think about the investment in infrastructures it's not even as much about promoting markets than it is about promoting health, education and possibly gender equality and I'm pretty convinced of that and the usual the implication of that brother is that sometimes the best way to improve outcomes in education and health may actually be to invest in infrastructure I think the example that I gave earlier of improved access to sanitation in schools especially in Muslim countries tends to have a very big effect on girl enrollment rates in those schools there are several studies that have documented that fact so there is here something to think about when you design infrastructure programs it's not simply about production people who are experts in education and health usually jump and say you want to cut my budget you may have serious problems in education and in developing countries there are serious problems in terms of access to school materials paying teachers something I firmly believe in but at the same time you can't forget the fact that in some cases the constraint is not in getting and spending more on producing services but it's the delivery of the services that matter in order to get more qualified health workers to go into rural areas well you may need to provide better access to those areas in concrete terms and finally what this is saying is that we have to rethink the way infrastructure projects are selected the traditional way of looking at infrastructure projects is to think of IRRs internal rates of return basically you calculate how much it costs what is the expected revenue and you decide on the basis more or less whether the project should go forward typically there is some political economy aspect to big infrastructure projects but usually the type of benefits that I've just mentioned in terms of health and education are very rarely taken into account when these type of projects are discussed in developing countries where health is such an important constraint on growth and development it would be a mistake not to think about those aspects which are now well documented at the micro level and hopefully with this type of contributions macroeconomists would be more willing to integrate this kind of reasoning in their macro models but it has to be also at the micro level at the level of the selection of infrastructure projects now how to do it in practice what is the methodology this is a different story I'm not necessarily in a good position to do it but there are ways of I'm sure by putting together education, health and infrastructure specialists there must be a way to develop a methodology to do these things more systematically the fact that we've never done it for no reason not to try and I'll finish on that one thank you very much we will open it up now to some questions if there are any I know it's a small room so we don't really need a microphone but because the session is being recorded we would like to be able to record and properly capture the questions as well any takers the narrow definition of public capital we were using is basically economic infrastructure you then extend if you like through a lot of externalities or complementarities or crowding out effects to other areas particularly health and education the narrative you gave essentially sees the direction of causation from the hard infrastructure and the soft infrastructure it's possible easily to conceive of a lot of causation from soft infrastructure to hard infrastructure so if I'm writing a book on education and health and the building of hospitals and universities and schools I can see substantial implications for educated women educated folk whatever wanting raw material wanting roads so they can drive their cars as they get into the labour market because of their education I'm wondering if what we need as a general equilibrium model of public capital broadly defined which is not to say we don't need all the insights you've got that's the platform for the meta model if not all the way to a macro model and just what you think of the two way causations here and whether no doubt in the empirical work that maybe in the book there is some capacity to have reflected that I think you're right the the focus here is on core infrastructure but building schools and building hospitals is obviously an essential part I think in terms of school but what really strikes me here is that I remember going to this country in Niger where the government the president decided he's going to build I don't know hundreds of schools so they stopped building those schools and then they said whatever that's relatively low cost to build a school how do you bring electricity how do you create give access to water and sanitation in those schools so there's an issue of scale that makes a big difference I think when it comes to building individual schools and when you talk about giving access to water and sanitation the scale of the problem is quite different you can't build those schools but to have to reap the benefits if you want of that you're going to need to provide access to infrastructure so that's the focus if you want of this book but clearly you could when you talk about allocation of public spending you could think of several types of public capital where you have not only core infrastructure which could be as I just said specified in such a way that it plays the role of a constraining factor on the other types of capital you need in order to produce a capital stock that can be used in education you need to combine core infrastructure and public capital in education the school, the school buildings and the core infrastructure you can do that and then discuss the allocation of resources among those different levels now you said something else which is also that it goes both ways so you could build the public capital that is complementary to the existing capital in education and health the schools etc but the very fact that you provide that is going to create a demand perhaps for a different type of infrastructure right? that's absolutely correct and the only question is that how does this change in demand translate into actual changes in policies so now you need in a way you need to go beyond what this book does and provide a framework in which you can think of the allocation of public expenditure being decided not to maximize growth not to maximize outcomes in education and health per se but how the political system in a way translates those demands for public goods right? into actual decisions and that's another that's another issue and the issue is in the political economy sense if you want how much power those who want this type of services are going to be able to affect the decisions of the governor and that's not something that I discuss here but certainly matters can be addressed in another book if you Australia infrastructure reform usually requires a number of governments and one of the techniques that we use to break deadlock is to ask body light productivity commission to work out the benefits of the proposed reform is the take home message from the seminar that we've been underestimating the benefits or is it the case in countries like Australia the health and education effects are pretty marginal well I don't know enough about Australia to say any statement that would be strong I can give you just two reactions to that one is the fact that usually when people think about you ask for analysis of impact usually people take a static view the static view is to say well this is the environment and this is what we believe is going to happen if you make the investment given the environment the thing is the infrastructure itself is going to change the environment and I think that's the important point when you create in Brazil for instance they are building a high speed rail between Sao Paulo and Rio it's not only about what this is going to do for improving traffic exchanges between Sao Paulo and Rio it's going to have an effect on all the cities in between those two big cities and with the proper design if you think about and that's what they are doing they are creating feeder services essentially access to that high speed rail from different cities along that road what would happen is that you are going to bring now more of those cities or areas if you want into the picture so you have to think in a dynamic sense not in a static sense you have to think to anticipate in a way how this is going to affect behavior but it could be that given the cost of living in Rio I may decide to go and work 60 miles from Rio for instance well that is going to change the labor market and the property prices what are the repercussions firms may find it easier to move away from congested city areas into areas that have more where they are going to have to create more infrastructure themselves so all those are what you may call dynamic effects you may have accounted for them you have to in a sense anticipate and guess to some extent what those effects are but I can tell you for a fact that there is a series of studies by the European Union there is a report that just came out by the Lincoln Institute in the US where they looked at the impact of high speed rail in Europe particularly France and Germany where they looked at the impact that the high speed rail system had on the cities that were linked to the new rail lines and the benefits are quite significant they are more than in a sense were expected at the beginning simply because it creates some sort of a dynamic process and it leads to changes in the way people work in the way people live and the way people firms set their locations it's not necessarily all good there is a study in India where they argued when they built a major highway what happened is that it contributed to the spread of AIDS because it attracted a lot of prostitutes if you want along the highway not necessarily all that positive so that's the first point that I wanted to make and the second is to think in terms of network effects that perhaps if you are thinking about doing it for one particular area maybe you would be getting benefits that are more important if you were to consider linking a group of cities as opposed to linking only two of them and that again requires making judgment as to what this externality would be and maybe it's not in the nature of I don't know technicians to be to think in those terms I don't know because if you are being asked to do a task in a particular task you know what I think that tomorrow we are going to get 20 new firms sitting in that particular area you may not be able to make that kind of judgment simply because you don't have a strong basis but the fact is when you look at empirical studies and you may look at the ones that talked about earlier there are significant benefits that go well beyond what was anticipated initially okay Any other questions? Can I just add one quick one if you use my power to me a lot of it does come down to implementation which you brushed aside a little bit I think most people would accept that there are network effects to infrastructure for example but it's devilishly difficult to put a number on these network effects and because it's so difficult that's why infrastructure projects easily become a political football you have one side saying the network effects are going to be very large and the other side doesn't think so and who's to say who's right because it's so difficult to put a number on it anyway so do you think we'll eventually be able to arrive at a methodology where we can prove the methodology suspects it doesn't become a political football we can't put a reasonably accurate number on it that to me still seems to be a million dollar question well there's always a political economy aspect to infrastructure whether it's in the selection of projects, the implementation the location of infrastructure projects so I'm not sure that you're ever going to get a situation where you can take out all the politics out entirely of those things I'm not too sure about that but in terms of the precise point of estimating what are the gains and the fact that it's difficult doesn't mean that you shouldn't try and I think it's quite important and it really has to come from what is your question that you're asking if you're asking infrastructure specialists to just go about their business and calculate internal rates of return and tell me well that's your best guess of the flow that's your best guess of the flow of income they're going to limit themselves to their own that's what you need to do is to force those guys to work systematically with people in education and in health and people who are interested in gender so that they can force them to think about those issues and as usual there has to be a clear message from the top in order to get these things out and interestingly enough I give a presentation of your most of the media at the request of the World Bank where they were discussing all of those very important and even at the level of the World Bank now they decided they were going to do a pilot project where they're going to put together experts in all those different areas to sit down together and think about all those aspects of infrastructure systematic and that's what you are not saying that you're going to get a perfect methodology and an indication expert they can calculate what is going to be the impact in their view on the number of graduates the impact that this is going to have on health outcomes and health experts in all those things the question is to get these guys to work together that I think is the problem and the truth is because macroeconomists typically have not paid any attention to these issues infrastructure specialists have been doing their work but they haven't been asked necessarily to think about other aspects no wonder that there is no methodology that has been developed but if this serves any purpose is to say that well you know what you should do now let me add that the impact on education and health are clearly very very important for developing countries it's not clear that this is going to matter a lot for industrial countries it could well be that the other aspects are more important but that's a judgment that you have to make whether improved access to better health facilities is going to make that big of a difference in a country like Australia or the US is not clear but there are other aspects gender etc that could be equally important